Between paradox and prescribing: Understanding anticipatory prescribing at the end of life in UK care homes Megha Majumder Darwin College June 2023 This thesis is submitted for the degree of Doctor of Philosophy 1 Declaration This thesis is the result of my own work and includes nothing which is the outcome of work done in collaboration except as declared in the Preface and specified in the text. I further state that no substantial part of my thesis has already been submitted, or, is being concurrently submitted for any such degree, diploma or other qualification at the University of Cambridge or any other University or similar institution except as declared in the Preface and specified in the text. It does not exceed the prescribed word limit for the Degree Committee of the faculties of Clinical Medicine and Veterinary Medicine. 2 Abstract Between paradox and prescribing: Understanding anticipatory prescribing at the end of life in UK care homes By Megha Majumder Background: In the United Kingdom, there has been a widespread introduction of anticipatory prescribing, especially in care home-based palliative care. Early and appropriate prescribing and administration of anticipatory medicines for patients who require end-of-life symptom control has become established good practice in palliative and end-of-life care policy. In care homes, the intervention aims to ensure timely symptom control, reassure residents and family members, and avoid emergency hospital admissions. However, there is limited evidence about how the anticipatory prescribing system—including its bureaucracy, regulatory apparatus, and the processes involved including the prescribing, dispensing, procurement, administration, monitoring, storage, and disposal of anticipatory medicines—unfolds in these contexts and stakeholder perspectives have not been sufficiently investigated. In 2015, the National Institute for Health and Care Excellence issued guidelines on the care of dying adults and specifically highlighted the need for empirical research into the use of anticipatory medicines in practice. Despite substantial developments in the literature since then, the evidence base supporting the implementation of anticipatory prescribing remains limited, particularly in care home settings. This research responds to this call. Aims: The study aims to expand understanding of the anticipatory prescribing system in care homes and the experiences and challenges of health and social care professionals within it, specifically by exploring: The underlying systemic tensions and wider interprofessional contexts which underpin issues related to the wastage and shortage of anticipatory medicines in care homes; the algorithmic and discursive structures which shape the tensions between individualisation and standardisation in end-of-life care; and the structural logics underpinning prescribing paperwork and how they contribute to the tension between navigating bureaucratic regulations and providing good, patient- centred end-of-life care in care homes. Methods: This thesis was underpinned by a constructivist, relativist approach and paradox perspective. A systematic literature review was conducted to synthesise understanding of the prescribing, storage, administration, monitoring, and disposal of controlled drugs for end-of-life care for care home residents in the United Kingdom. The findings from the review informed the design of the qualitative interview study. In total, 42 interviews were conducted with health and social care 3 professionals working in 29 care homes. Reflexive Thematic Analysis of the qualitative dataset was informed by Braun and Clarke (2006, 2019, & 2021). All participants provided informed consent. Ethical approval was obtained from the University of Cambridge. Findings: Analysis revealed that the paradoxical occurrence of simultaneous anticipatory medicines wastage and shortage in care homes was rooted in poor interprofessional relationships and a contradiction between the mechanism of prescribing and the inherent uncertainty of the end of life; that care home residents going through the inherently individualized process of dying were often paradoxically treated in a standardised fashion via the use of prescribing protocols, standard anticipatory medicines packs, and nonspecific clinical language; and that anticipatory prescribing paperwork was constitutionally predisposed towards duplication, the multiplication of error categories, and the usurpation of time which could have been spent providing high-quality care to residents, something which it also paradoxically demanded. Conclusion: The study highlights the complexity of and key paradoxes within the anticipatory prescribing system in care homes. Robust professional relationships can provide a platform for effective communication, interprofessional collaboration, collective reflection on its various processes, its bureaucracy, and its role in end-of-life care. Moreover, such relationships can play a crucial role in establishing opportunities for training, securing sufficient resources, and offering support for professionals involved in anticipatory prescribing. The findings can inform future strategies aimed at improving the quality of end-of-life care within care homes in the United Kingdom. 4 Acknowledgements Like the subject in question, this dissertation is the culmination of much collaboration and care. In writing it, I found, in the way that scholars often do, that a delicate synchronicity intertwined my challenges with those of my participants. If, as people, we live in worlds of tangled contradiction, with bodies that resist and rebel, relationships that fuel and frustrate, and daily life that demands (relentlessly) and distracts (salutarily), how do we produce work that remains both good and caring? How does this work embody and reflect the worlds that give it form? This thesis was written amidst a whole lot of world—in college halls and hospital rooms, rushing trains and homebound planes, and even atop the quiet promise of new life. Its completion is a testament to the courage and generosity of those who shared their stories with me, the support of my supervisors, Professors Stephen Barclay, Kristian Pollock, and Claire Goodman, and to the thoughtful engagement of my examiners, Professors Catherine Walshe and Martin Roland. Their guidance has been an exceptional example of how to carefully interlace the rigour of qualitative inquiry with a deep- seated empathy for the human experience. For the cups of tea, biscuits, and warm congeniality that sustained me through this research journey, I am grateful to the whole Cambridge Palliative & End-of-Life Care group. I wish to recognise in particular, Drs Ben Bowers, Richella Ryan, Sarah Hopkins, and Tessa Morgan, as well as Isaac Winterburn, Lizzie Mann, Joshua Gallagher, and Susannah Browne for their invaluable contributions as my earliest sounding boards, and Isla Kuhn, for her expert tutelage in conducting a robust systematic review. I would also like to acknowledge the Abbeyfield Research Foundation for their generous funding without which this research would not have been possible (although the views and conclusions presented in this thesis are entirely my own). My academic career began with a stroke of luck and fortuitous connections that taught me the special sort of thinking that can happen when bright, big-hearted students engage sincerely with one other's scholarship. There is a magic in these circles that I try to capture in all my work and writing. Though we’re all a bit scattered now, those little groups comprised of friends remain my favourite people to think out loud with: Drs Isabel Esain, Arthur Davis, Matt Cairns, and Jumana Esau at Cambridge; Drs Rennie Burke, Jane Zhu, Clio Sophia Koller, and Casey Rojas at Harvard; Drs Smitha Milli, Guy Wilson, Joshua Price, and Alankrita Dayal at Berkeley; and from my hometown, Drs Rebecca Leibowitz, Joon Lee, Kimia Ziadkhanpour, and Mrs. O’Donnell. 5 For guiding my earliest forays into academia, I am profoundly indebted to many of my past professors and mentors. I wish to acknowledge Professors Riley Bove, Jennifer Doudna, Marilyn Wellcome McEntyre, and Monica Albe at UCSF and UC Berkeley; Professors Becca Brendel, Barbara Bierer, Rebecca Li, and Atul Gawande at Harvard Medical School; and most vitally, Dr. MR Rajagopal at Pallium India. Their model of melding exceptional clinical care with meaningful research, teaching, and policymaking has set the mark for all I hope to accomplish in my own career. I would also like to credit my role model, Paul Kalanithi, with my decision to study at Darwin College and centre my research on palliative and end-of-life care. As a writer, physician, and social scientist, I strive to embody the same caring qualities that he exhibited throughout his brief but beautiful life. For their many forms of support through my academic journey, I am thankful to my parents, Llony and Kaiser Majumder. Everything that I have done and will do is a feat of their nurture, not my nature. My second set of parents, Dr Tumpa and Prof Prithwish Banerjee, grounded me through a difficult writing process. Their nourishing meals, profound wisdom, and mentorship in academia and medicine have provided me the comfort and care needed to complete this work. Finally, I am most grateful to their son and the little family we’ve built together. I have been in graduate school for the whole of my marriage to Spandan Banerjee. He has patiently carved out the time and space for me to think and write, firmly anchoring me to the value of my research. His belief in it has sustained my own. Our dog, Baloo, grew up alongside this dissertation. Her two o’clock whines were my cues to pause, pop out for a walk, and delight in the long sweep of sweet smelling plants along the River Cam. Being in the world with her and Spandan lent a sense of balance to my life and this work. Lastly, I thank our little boy, Ray, who has given me one very good reason to finish. Your father and I can’t wait to meet you. 6 Abbreviations AP Anticipatory Prescribing AM Anticipatory Medicine BMA British Medical Association CD Controlled Drug CHM Care Home Manager CHN Care Home Nurse CCG Clinical Commissioning Group CN Community Nurse CQC Care Quality Commission DN District Nurse DOH Department of Health EHCH Enhanced Health in Care Homes EoL End of Life EoLC End-of-life care GP General Practitioner GSF Gold Standards Framework LCP Liverpool Care Pathway MDT Multidisciplinary Team Meeting NHM Nursing Home Manager NHS National Health Service NICE National Institute for Health and Care Excellence NIHR National Institute for Health Research NMC Nursing and Midwifery Council ONS Office for National Statistics PHE Public Health England PPI Patient and Public Involvement PRN Pro re nata QNI Queen’s Nursing Institute UK United Kingdom USA United States of America WHO World Health Organisation 7 8 Contents Contents Page 8 Figures 13 Tables 14 Boxes 15 Part 1: Chapter 1 Background End-of-life anticipatory prescribing in UK care homes 16 1.1 1.2 1.3 Introduction Background: Death and dying in UK care homes Defining care homes 16 16 16 1.4 1.5 1.6 Death trends in UK care homes The role of anticipatory prescribing in a ‘good death’ The anticipatory prescribing system at the end of life 17 18 19 1.7 Health and social care professionals supporting the AP process 27 1.8 Personal motivations 29 1.9 Impact of the COVID-19 pandemic 29 1.10 Rationale 30 1.11 Challenges, tension, and paradox 32 1.12 Core objective and research questions 35 1.13 A brief overview of the thesis 35 Chapter 2 Reviewing the evidence regarding End-of-Life Controlled Drugs in UK care homes 38 2.1 Introduction 38 2.2 Aim 40 2.3 Review questions 41 2.4 Methods 41 2.5 Results 46 2.6 Discussion and narrative synthesis of content 56 2.6.1 Patchwork prescribing 57 9 2.6.2 Challenges of clinical uncertainty 57 2.6.3 Challenges in interprofessional working 58 2.6.4 Implications for research, policy, and practice 58 2.6.5 Limitations of the review 59 2.7 Conclusion 59 Part 2: Chapter 3 Methodology & Methods Methodology and methods 61 3.1 Introduction and research questions 61 3.2 Ontology & epistemology 62 3.3 Theoretical perspective 63 3.4 Methods: Research Design 67 3.4.1 Introduction 67 3.4.2 Public and Patient Involvement 68 3.4.3 Setting 69 3.4.4 Population 69 3.4.5 Sampling 69 3.4.6 Recruitment 70 3.5 Methods: Data collection 71 3.5.1 Interviews 71 3.5.2 Reflections on Zoom interviewing 73 3.5.3 Recording 74 3.5.4 Transcription 74 3.5.5 Summary 75 3.6 Methods: Data analysis 75 3.6.1 Reading 75 3.6.2 Timeline 75 3.6.3 Reflexive Thematic Analysis: An introduction & rationale 76 3.6.4 Doing RTA 79 3.6.5 The paradox lens applied 80 3.6.6 Quality of the research 87 3.7 Research ethics 88 3.7.1 Recruitment in a pandemic context 88 10 3.7.2 Informed consent 89 3.7.3 Interviews in a pandemic context 90 3.7.4 Socially desirable responses 90 3.7.5 Confidentiality and anonymity 92 3.7.6 Protecting research participants from harm 93 3.7.7 Minimising harm to the researcher 94 3.7.8 Reporting procedures 94 3.8 Final remarks 98 Part 3: Chapter 4 Findings & Discussion Outcomes of recruitment 99 4.1 Participant recruitment strategy 99 4.2 Recruitment outcomes 101 4.2.1 Reflecting on recruitment outcomes 102 4.3 Participant and care home characteristics 103 4.4 Conclusion 105 Chapter 5 The Anticipatory Prescription Paradox 106 5.1 Background 106 5.2 Introduction 107 5.3 Wastage vs. Shortage 109 5.4 Nurse-boundary vs. Doctor-boundary 115 5.5 Training vs. Support 121 5.6 Anticipation vs. Reaction 128 5.7 Uncertainty vs. Certainty 133 5.8 Holding on vs. Pushing away 136 5.9 Chapter summary 141 5.10 Discussion 141 5.11 Conclusion 146 Chapter 6 The Protocol Paradox 147 6.1 Background 147 11 6.2 Introduction 148 6.3 Individualization vs. Standardisation 149 6.4 Micro-decisions vs. Algorithmic decisions 154 6.5 Individualized language vs. Jargon oblivion 166 6.6 Complex cases vs. Bureaucratic shortcuts 177 6.7 Chapter summary 181 6.8 Discussion 182 6.9 Conclusion 186 Chapter 7 The Paperwork Paradox 187 7.1 Background 187 7.2 Introduction 188 7.3 Clock-based time vs. Task-based time 190 7.4 Granular Certainty vs. Structural Stupidity 194 7.5 Risk-Management vs. Error Multiplication 199 7.6 Tick-Box Compliance vs. Duty of Candour 203 7.7 Guidelines vs. Orders 211 7.8 Chapter summary 218 7.9 Discussion 218 7.10 Conclusion 222 Part 4: Reflections & Conclusion Chapter 8 Conclusion 223 8.1 Story of the thesis 223 8.2 Recommendations for practice and policy 226 8.3 Reflexivity statement 229 8.4 Quality, strengths, and limitations 233 8.5 Future research directions 238 8.6 Final remarks 239 References 241 12 Appendices Appendix 1 Published systematic review 256 Appendix 2 Systematic review search strategy 264 Appendix 3 Systematic review data extraction template 267 Appendix 4 Email text to potential participants for the interview study 269 Appendix 5 Participant information sheet 270 Appendix 6 Study protocol 273 Appendix 7 Ethics approval 283 Appendix 8 Interview topic guide 284 Appendix 9 Reflexive journal extracts 287 Appendix 10 ‘Twenty Questions to evaluate the quality of Thematic 290 Analysis’ by Braun & Clarke (2020) 13 Figures Figure 2.1. PRISMA flow diagram in published review from January 2000 to January 2021 42 Figure 3.1. Example of coding 79 Figure 6.1. Compatibility chart for two drugs in water for injections 158 Figure 6.2. GSF-provisioned symptom control guidelines 159 Figure 7.1. A comparison of two Controlled Drugs logs 195 Figure 7.2. Liverpool Care Pathway variance document 203 14 Tables Table 1.1. Common symptoms and anticipatory medications 20 Table 1.2. Roles and responsibilities of health and social care professionals in AP 27 Table 2.1. Review specific Gough’s Weight of Evidence criteria 43 Table 2.2. Weight of Evidence Appraisal Matrix 43 Table 2.3. Summary of included studies in published review from January 2000 to January 2021 47 Table 2.4. Number of papers included in the published synthesis 52 Table 3.1. PhD tasks by date 76 Table 3.2. Phases of data analysis 83 Table 3.3. List of themes 86 Table 4.1. Participant numbers by professional role 101 Table 4.2. Aggregate table of participant characteristics 103 Table 4.3. Individual participant and care home characteristics 103 Table 8.1. Strengths and weaknesses of the qualitative research with Tracy’s (2010) criteria 233 15 Boxes Box 1.1. An overview of the AP system at the end of life 21 Box 2.1 Review inclusion and exclusion criteria 41 Box 2.2. Narrative synthesis strategy 44 Box 2.3. Addressing the review questions 52 Box 3.1. Reflections on my role as interviewer 94 Box 8.1. Recommendations for practice and policy 226 Box 8.2. Personal reflexivity 230 Box 8.3. Functional reflexivity 231 Box 8.4. Future research directions 238 16 Part one: Background Chapter 1 End-of-life anticipatory prescribing in UK care homes In this chapter I explore end-of-life anticipatory prescribing in UK care homes in the wider context of health and social care in the UK. I examine its roles in end-of-life care and in facilitating a 'good death’, and conclude by introducing my research program. 1.1 Introduction In the United Kingdom (UK), there has been a widespread introduction of anticipatory prescribing (AP), especially in care home-based palliative care. Early and appropriate prescribing of anticipatory medicines (AMs) for patients who may potentially require end-of-life symptom control has been a key aim in palliative and end-of-life care policy and practice. Anticipatory prescribing involves general practitioners (GPs) writing prescriptions in anticipation of them being needed by a certain resident and the development of systems to ensure that medicines can be dispensed and administered both in-hours and out-of-hours (Wilson et al., 2015). Anticipatory medicines confer considerable responsibility upon visiting and care home nurses, as they frequently include controlled drugs (CDs), which require additional legal and regulatory considerations, such as appropriate storage and documentation, as well as strict controls over their administration and disposal. However, there is limited evidence about how these processes unfold and how these decisions are made by practitioners. In this thesis, I set out to expand understanding of health and social care professionals’ experiences and challenges navigating, arranging, and delivering AP in UK care homes. In order to better understand AP as it operates in the care home context, I conducted and analysed interviews of health and social care staff involved in the process. In doing so, I foreground the working relationships between these professionals, and contrast the portrayal of AP by policy and clinical paperwork with the views and experiences of practicing professionals ‘on the ground’. AP is an inherently relationship-driven process involving the cooperation, coordination, and communication of health and social care professionals, family carers, and residents. This chapter is divided into three parts. The first provides a background on death and dying in UK care homes (1.2-1.4) and defines care homes (1.3). The second focuses on end-of-life AP in UK care homes (1.5-1.7). This includes the role of AP in achieving a 'good death' (1.5), the AP process at the end-of-life (1.6), the overall context of AP and care homes with regard to legislation (1.6.1), 17 geography (1.6.2), resource pressures and financing (1.6.3), descriptions of the health and social care professionals involved in the AP process (1.7), and their respective roles and responsibilities (1.7.1- 1.7.6). The third outlines my personal motivations regarding the selection of this research subject (1.8), the impact of the COVID-19 pandemic on this study (1.9), the study rationale (1.10), leading into the problem statement and research questions (1.11) and a brief overview of the thesis (1.12). 1.2 Background: Death and dying in UK care homes Care homes provide care for over 430,000 elderly and vulnerable people in the UK (NHS, 2016). According to Public Health England, in 2016, 21.8% of older people died in care homes (Bone et al., 2018). The COVID-19 pandemic dramatically shifted the health profile of England’s population by becoming the primary cause of death in 2020 and 2021 (Raleigh et al., 2020). The first wave in 2020 involved a surge in the number of residents dying in care homes, where there were approximately 27,000 excess deaths in comparison to previous years, accounting for approximately half of all excess deaths in England and Wales (Raleigh et al., 2020). Numerous factors, some of which will be examined in this study, had an impact on how death and dying were managed in care homes before, during, and after this period. Care home residents in the final phase of their lives have different clinical care and psychosocial needs than other residents, including the possibility of being prescribed a suite of anticipatory medicines (AMs). Rather than an exclusionary focus on restoration or on prevention of decline, care of residents involves a focus on palliation—on the management of the terminal decline and its accompanying symptoms. Care home staff, particularly managers and nurses, play a critical role in determining residents’ needs for pain and symptom control and enabling timely access to appropriate medicines (Wilson et al., 2014). 1.3 Defining care homes Care homes are always classified as either ‘residential homes’ (full-time care with no requirement for qualified nurses) or ‘nursing homes’ (full-time care with qualified nurses on duty) providing full- time care for residents, most commonly older people. In this thesis, both nursing homes and residential homes are collectively referred to as ‘care homes’, with the specific terms of ‘nursing home’ or ‘residential home’ used when discussing either in particular. Care home staff, including care home managers (CHMs), care home nurses (CHNs), and care assistants, provide end-of-life care to residents alongside a multidisciplinary team comprised of General Practitioners (GPs), District Nurses (DNs), pharmacists, and ambulance personnel (Kinley et al., 2014). These visiting professionals provide guidance, medicine, and treatment to residents; community nurses, including DNs, are the primary providers of medical and nursing care in residential care homes (Davies et al., 2011; Handley et al., 2014). 18 1.4 Death trends in UK care homes Populations are aging in many countries, with longer life expectancies, including the UK (Johns et al., 2020; O’Brien, 2016). The majority of older people (aged 85 and over) in care homes are affected by complex illness profiles characterised by multiple comorbidities, often including several chronic diseases, depression, pain, and sleeping difficulties (Kingston et al., 2018). It has been estimated that 80% of UK care home residents have dementia (Alzheimer’s Association, 2013). Close attention to symptom management and care are required for residents: the average care home resident is prescribed eight medications (Schneider et al., 2019; Gordon et al., 2014). The number of older residents dying in care homes is projected to increase considerably in the future (Bone et al., 2018). According to 2019 mortality data, 604,707 people died in the UK (ONS, 2019a); approximately 50% of these deaths occurred in hospitals, while 25% occurred in care homes, 20% in individual’s own homes, and 5% in hospices (ONS, 2019b). Hospital deaths have decreased by 11% since 2004, while care home deaths have increased by 5.3% (PHE, 2018). If these trends continue, care home deaths are predicted to rise by 108% by 2040, leading to care homes becoming the most common place of death in the UK, with care home staff predicted to become the majority providers of end-of-life care (Bone et al., 2018). Since the National End-of-Life Care Programme (2010), NHS England has viewed the provision of end-of-life care to individuals in their preferred place of care as a major policy objective (Neuberger, 2004; NHSE, 2010). 1.5 The role of AP in a ‘good death’ The concept of a ‘good death’ has been championed by the hospice movement as a fundamental human right for dying individuals, and the term holds significant import for professionals working in health and social care (Emanuel & Emanuel, 1998). According to Neuberger, health and social care professionals must be aware of patients' preferences and knowledgeable about how to provide a ‘good death’ in order to help them experience one (Neuberger, 2004). The 2008 Department of Health End-of-life Care Strategy outlined four factors foundational to the construction of a ‘good death: ‘being treated as an individual, with dignity and respect; being without pain and other symptoms; being in familiar surroundings; and being in the company of close family and/or friends (p. 9). The element of symptom control requires a great deal of coordination between different professionals, particularly for care home residents situated in the community. The inclusion of symptom control in advance care plans is viewed in policy as an essential component of end-of-life care. Advance care plans are documents that outline a person's preferences and wishes for medical treatment and care in the event that they become unable to make decisions for themselves (DOH, 19 2008). In the context of AP, advance care plans facilitate coordination among different care teams involved in resident care, including GPs, community nurses, out-of-hours services, and care home staff (NICE, 2015; HIS, 2014). Bradbury argues that the concept of a ‘good death’ exists for the benefit of the ‘living,’ including care home staff whose day-to-day work is at the threshold between life and death (Bradbury, 1996). In this thesis, I consider the ways in which AP supports ‘good deaths’ for care home residents. However, the ‘good death’ is frequently simplified to a binary of ‘good’ and ‘bad’ in care home policy and clinical paperwork, without detailing the care and coordination requisite to its provision. This issue will be explored in Part 3 of the thesis, where the implications of AP policy language will be examined in detail. 1.6 The AP system at the end of life Anticipatory prescribing is the prescribing and dispensing of injectable medicines to a given patient, in advance of clinical need, for administration by appropriately trained individuals if symptoms arise in the last days of life and they are unable to take medicines orally (Wilson, 2014; Twycross et al., 2017). In this thesis, AP is explored within the broader system that supports and facilitates it. This system includes several interrelated processes — the prescribing, dispensing, procurement, administration, monitoring, storage, and disposal of anticipatory medicines — all of which contribute to the functionality of AP in care homes. This idea is further explored and expanded on in section 1.6.1. Although interconnected, each process is distinct, and it is important to both understand them as such and also to understand how they influence one other. Recognition of the dying phase is fraught with uncertainty and it is challenging to predict how and when death will occur; however, it is important for clinicians to anticipate and prepare for symptoms that may arise at the end of life to ensure that the dying individual, as well as their family and caregivers, are not unduly distressed. To palliate the four common end-of-life symptoms (pain, nausea and vomiting, agitation, and respiratory secretions), anticipatory medicines are usually prescribed by a clinician familiar with the dying person, with authorisation for administration (NICE, 2015; Russell et al., 2018; Twycross et al., 2017). Clinicians often prescribe four or five injectable AMs, because oral medication delivery may not be possible as patients near death. Anticipatory medicines are most commonly prescribed to be administered as individual doses as needed (PRN) or occasionally as 24-hour continuous subcutaneous infusions via syringe pumps once symptoms have become established (Bowers et al., 2019; Wilson et al., 2015). The medicines commonly prescribed include an analgesic such as morphine or oxycodone, an antiemetic such as levomepromazine or haloperidol, an anti-anxiety drug such as midazolam, and an anti-secretory drug such as glycopyrronium or hyoscine butylbromide (Bowers, 2019; Amass & Allen, 2005). In care homes, these medicines are typically dispensed and stored as a ‘just in case box’ within a CD cabinet 20 (Seymour et al., 2011; NHSE, 2019). Trained professionals can then readily administer them to residents if and when distressing symptoms arise at the end of life, without the need for a new prescription to be dispensed from a pharmacy (Seymour et al., 2011; Wilson & Seymour, 2017). While AP may provide these potential benefits, its challenges must be acknowledged. There is uncertainty for the prescriber as they do not know in advance which of the prescribed medicines will be needed and administered to the patient, and at what doses, although a prescription always requires a dose to be stated (Bowers et al., 2019; NICE, 2015; Penge & Peter, 2014). Anticipatory medicines may at times be prescribed in a standardised fashion on pre-printed proformas and drug charts for the sake of simplicity and efficiency, rather than being individually tailored to a patient’s specific anticipated needs and preferences (NICE, 2015). Inexperienced doctors or nurses may not accurately diagnose whether and when a resident is dying and may prescribe or administer inappropriate medicines and doses (NICE, 2015; Penge & Peter, 2014). If anticipatory medicines are under- prescribed or under-administered, distressing symptoms may go untreated; on the other hand, over- prescribing or over-administration may involve harmful side effects including potentially the hastening of death (NICE, 2015; Wilson et al., 2015; Bowers et al., 2019). The storage of anticipatory medicines also presents challenges: many are controlled drugs that need to be stored in dedicated locked cabinets given the risks of theft, unauthorised access, and misuse (CQC, 2022; Teggi, 2023). After a resident’s death, the prescribed anticipatory medicines may not have been used and need to be disposed of in ways that adhere to national legislation for the disposal of CDs (CQC, 2022; Teggi, 2023). Though my literature review primarily focused on Schedule 2 and 3 Controlled Drugs including strong opioids and midazolam, in my fieldwork I included all of the commonly prescribed anticipatory medicines, not all of which are controlled drugs. Table 1.1 details commonly prescribed anticipatory medicines, alongside their drug schedule and class. Table 1.1 Common symptoms and anticipatory medicines Source: The Controlled Drugs (Supervision of Management and Use of) (NHSE, 2013) Symptom Common anticipatory medicines that are prescribed* Controlled Drug status and schedule Drug class Pain / breathlessness Morphine sulfate Controlled Drug, Schedule 2 Strong opioid; prescription only Diamorphine Controlled Drug, Schedule 2 Strong opioid; prescription only Oxycodone Controlled Drug, Schedule 2 Strong opioid; prescription only Nausea and vomiting Cyclizine No Controlled Drug status Anti-emetic Haloperidol No Controlled Drug status Antipsychotic, used as an anti-emetic in low doses; prescription only Levomepromazine No Controlled Drug status Antipsychotic, used as an anti-emetic in 21 low doses; prescription only Agitation / restlessness Midazolam Controlled Drug, Schedule 2 Benzodiazepine Levomepromazine No Controlled Drug status Antipsychotic, used as an anti-emetic in low doses; prescription only Haloperidol No Controlled Drug status Antipsychotic, used as an anti-emetic in low doses; prescription only Noisy respiratory secretions Glycopyrronium bromide No Controlled Drug status Antimuscarinic; prescription only Hyoscine butylbromide (BuscopanTM) No Controlled Drug status Antimuscarinic; prescription only Hyoscine hydrobromide No Controlled Drug status Antimuscarinic *Notably, participants from the empirical research used the terms ‘anticipatory medicines’ and ‘controlled drugs’ interchangeably, possibly because several commonly prescribed anticipatory medicines are classified as controlled drugs and require careful consideration and monitoring to ensure safe and appropriate use. This thesis distinguishes between the terms and only includes discussion of controlled drugs that are anticipatory medicines. AP is widespread practice across the UK, but the evidence base to support practice is currently limited (Amass & Allen, 2005; Bowers et al., 2019; Carney & MacRobbie, 2008; Lawton et al., 2012; Scott-Aiton, 2009). NICE guidelines on the care of dying adults called for further research in this area (NICE, 2015). The independent inquiry into the deaths at Gosport War Memorial Hospital revealed that over 450 patients had their lives shortened by being administered large doses of opioids that were not medically indicated. The investigation found ‘an institutionalised practice of the shortening of lives through prescribing and administering opioids without medical justification’ at Gosport from 1989 to 2000 (Dyer, 2018). The Liverpool Care Pathway (LCP) was a tool intended to transfer best hospice practice to hospital and community contexts and advocated as best practice in end-of-life care (DOH, 2008). The LCP emphasised ensuring that dying patients had access to AMs for pain, agitation, and respiratory secretions. However, the guidance was used in a formulaic way, with inadequate communication with patients and families at times (Neuberger, 2013). Furthermore, the LCP was not always accompanied by appropriate staff education and was at times misunderstood and misused by care home staff (Neuberger, 2013). Box 1.1 An overview of the anticipatory prescribing system at the end of life ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ o Anticipatory prescribing as a system The concept of a ‘system’ has been widely studied and recognised in the literature, defined by Von Bertalanffy as ‘elements in relationship’, or a set of interconnected components or processes that work together towards a common goal (Von Bertalanffy, 1968). Drawing on Bertalanffy’s systems theory, the inherent value of studying a phenomenon as a system lies in its ability to capture its holistic nature, moving beyond isolated components to understand how interactions contribute to the behaviour of the whole (Von Bertalanffy, 1968; Skyttner, 2006). AP and its related aspects 22 exemplify systemic characteristics: issues related to the prescribing of AMs cannot be meaningfully understood independently of its bureaucracy, the regulatory apparatus which sustains its bureaucracy, and the web of interconnected processes that together support timely and appropriate AP in end-of-life care for care home residents (NHS, 2019). In this thesis, AP bureaucracy refers to the wide range of administrative tasks and paperwork that govern AP processes, in line with Weber’s concept of bureaucracy wherein authority derives from established rules and is characterised by a complex division of labour, hierarchy of authority, technical competence, rules of procedure and behaviour, and an emphasis on written communications (Weber, 1946, p. 198). Within this understanding, tasks such as form-filling, record-keeping, reading policy, implementing guidance, and liaising with multiple care teams and agencies are in place to support the safe and appropriate use of anticipatory medicines, in compliance with internal care home policies and broader regional and national regulations (CQC, 2022; NHS, 2019; NICE, 2015). The regulatory apparatus for the AP system in care homes comprises multiple organisations, including NICE, NHS, DHSC, and regional CCGs, detailed further in section 1.6.2. These organisations define the rules and regulations for AP processes, and the CQC monitors and enforces them (Malley, 2014; Warmington, 2014). Together, they dictate the complex landscape of AP in care homes through paperwork, policy, and frequent inspections (Malley, 2014; Warmington, 2014). While these structures are in place with the intent to maintain quality and consistency, navigating them can often pose challenges for care homes in their daily operations (Warmington, 2014). The resultant landscape, shaped by these regulatory dynamics, actively impacts the ways in which AP processes are carried out by health and social care professionals. These processes include: • Prescribing: The initial step in the process is the identification of end-of-life symptoms in care home patients. Once these symptoms are identified, clinicians then move towards determining the need for medication (Goodman et al., 2010). Factors that influence the decision from the patient's end may include symptoms, existing medication regimen, and their specific needs and preferences. From the prescriber's end, factors that might influence the decision include experience and training, medication preferences, interprofessional collaboration, and prescribing guidelines (Majumder et al., 2022; Wilson, 2015). • Procuring: Once prescribed, the relevant anticipatory medicines are sourced from the pharmacy. Procurement ensures that prescriptions are filled and available for care home patients when needed. This involves balancing factors such as medication stock levels, potential supply chain disruptions, cost considerations, and communication with suppliers (Johnstone, 2017). • Dispensing: Pharmacists dispense medicines to procuring individuals, such as nurses or senior care home staff members. This involves verifying patient details, ensuring correct 23 labelling and dosage, and checking for potential side effects or interactions that may be more prevalent in older populations (Alldred et al., 2009; Becker et al., 2005). • Storage: In care homes, medicines must be stored carefully in accordance with regulations (including, but not limited to, NICE Quality Standard 85: Medicines management in care homes, and NICE Guidance SC1: Managing medicines in care homes). This involves ensuring medicines are stored securely to prevent unauthorised access. Temperature- and light-controlled storage may be necessary, and a designated locked controlled drugs cabinet, regularly inspected, is often a requirement in care home settings (Allitt et al., 2017; Royal Pharmaceutical Society, 2015). • Monitoring: Monitoring in care homes involves careful observation of a patient’s condition, prior to, during, and following anticipatory medicines administration to ensure the appropriateness of the medicines and to remain vigilant to signs of over- or under-medication in residents (Allitt et al., 2017). • Administering: Administering medicines to care home residents requires that the administering individual understands each resident's medication regime, considers issues of capacity or communication difficulties (NICE, 2018), and can manage PRN (pro re nata) medicines, which are given ‘as needed’, rather than at scheduled times. PRN administration involves understanding the patients’ immediate needs and potential side effects of the medicines (Barber et al., 2009). Administering clinicians are also expected to communicate with managers and prescribers to ensure that residents receive the right medication under optimal conditions, while also documenting all administered medicines to ensure regulatory compliance (The Royal Pharmaceutical Society, 2007 & 2015; CQC, 2020). • Disposal: In care homes, the disposal of anticipatory medicines, particularly those that are controlled drugs, requires adherence to a range of guidelines. Controlled drugs, due to their potential for misuse, have specific disposal protocols (2001 Misuse of Drugs Regulations) which stipulate that they must be denatured prior to disposal, rendering them irretrievable (DOH, 2001). In care homes, this process involves two individuals: one responsible for the act of disposal and another to witness and verify the process (CQC, 2022). Care homes are required to maintain a controlled drugs disposal book, which records details of the drug, the quantity, and the signatures of both the person disposing of the drug and the witness (CQC, 2022; DOH, 2007). The interrelatedness and interdependence of these processes render AP a complex system. Any disruption or inefficiency in one process can ripple through and affect the others, underscoring the importance of a systems-thinking approach. As Senge emphasised in his work on systems thinking, understanding these interconnections is crucial for ensuring the optimal functioning of the whole (Senge, 1990). o The legislative landscape of AP in UK care homes 24 AP practice began to spread across the UK in 2003, when published as a recommendation in the Gold Standards Framework (GSF) (Collis & Al-Qurainy, 2013; GSF, 2006; Roberts & Green, 2021); it was endorsed as an indicator of best practice in the 2008 Department of Health End-of-Life Care Strategy (NICE, 2008). The GSF offers a nationally recognised quality improvement training programme and resources for health and social care staff caring for patients in their final years of life at home and in care homes (GSF, 2006; GSF, 2021). The GSF was developed in the early 2000s by Dr. Keri Thomas, supported by Macmillan Cancer Support, with the objective to improve community-based palliative care provision by GPs (GSF, 2006; Roberts & Green, 2021; Thomas, 2001). Presently, the GSF is a community interest company that markets its training programmes and incentivises care homes to be re-trained and re-assessed triennially through an accreditation process in which homes are rewarded with bronze, silver, gold, and platinum status (GSF, 2019, 2020 & 2021). Out of the 17,000 care homes in the UK, approximately 3,200 (19%) have completed GSF training, though not all of these care homes have sought complete accreditation, which involves a rigorous process of self-assessment and external evaluation (GSF, 2019 & 2020). Additionally, the distribution of GSF-accredited care homes varies according to population demographics, local priorities for end-of-life care, and funding capacities (Hall et al., 2011). Obtaining GSF training and accreditation requires funding and time for GP practices and care home organisations to invest and engage in quality improvement and staff development. Many care homes have limited finances for such optional programmes (Allen et al., 2022). The rationale underpinning AP is that residents’ distressing end-of-life symptoms can be treated safely and without delay within the care home, avoiding unnecessary suffering and potentially distressing hospitalisations (NICE, 2015). This rationale and the widespread adoption of AP were supported by early evaluations in which it was determined to be a low-cost and valuable intervention, despite there being limited evidence of its impact (Thomas, 2003). In one early project in 2005, Amass & Allen reported data on a community AP pilot intervention: of 23 patients for whom anticipatory medicines were prescribed, 16 had them administered, none of whom were admitted to hospital or hospice (Amass & Allen, 2005). Since then, AP has been enacted extensively across the UK, Australia, and New Zealand, and is regarded as an established and key component of good end- of-life care (Antunes et al., 2020; Russell et al., 2018; Ministry of Health, 2017; Victoria State Government, 2017). National guidance has repeatedly endorsed AP as best practice for end-of-life care in hospice, hospital, home, and care homes: NICE (2004) ‘Improving supportive and palliative care for adults with cancer’; Department of Health (DOH) (2004) guidance concerning access to medicines; DOH (2006) ‘End-of-Life Care Programme’; DOH (2008) ‘National End-of-Life Care Strategy’ (which has since been replaced by the Ambitions for Palliative and End-of-Life Care: A National Framework (2015 & 2021), where notably, AP is not mentioned); and NICE (2015) guidance ‘Care 25 of dying adults in the last days of life’. NICE emphasised the importance of using ‘an individualized approach’ to prescribing anticipatory medicines, and the importance of reviewing anticipatory medicines as the dying person’s needs change, evaluating the person’s individual symptoms prior to administration, and monitoring for benefits and side effects daily (NICE, 2015). However, these guidelines lack detail on what ‘an individualized approach’ entails, how to determine when anticipatory medicines are needed, or who should be involved in making these decisions. This NICE guidance was based on expert opinion and just two qualitative studies. However, the guidance acknowledges the limitations of the current evidence base and encourages empirical research into the use of AMs in practice (NICE, 2015). The literature has developed considerably since 2015, particularly given the changes to guidance and in practice during the COVID-19 pandemic: this research seeks to further build the evidence base. Several developments in AP regulation and legislation occurred during the course of this study, impacting professionals’ perceptions and practice of AP. Prior to the COVID-19 pandemic in early 2020, NHS England published the ‘Enhanced Health in Care Homes’ (EHCH) framework, which presented strategies aimed at strengthening relationships between Primary Care Networks (PCNs), GP practices, and care homes on a wide range of care provision, including medicines management and end-of-life care (NHSE, 2020). The EHCH framework supports several key aims in the NHS “Five-year long-term plan” that includes a focus on the implementation of anticipatory care across the whole NHS, including patients with frailty, dementia, and palliative care needs in order to achieve ‘a reduction in avoidable emergency admissions and more people being able to die in a place they have chosen’ (NHSE, 2019). Importantly, this research took place during the COVID-19 pandemic. At its peak, the number of people dying in UK care homes rose sharply (Comas-Herrera et al., 2020; ONS, 2020; ONS, 2021). In the light of concerns about potential drug shortages and a lack of available professionals to administer injections in a timely fashion, rapid changes to drug legislation were introduced, accompanied by widespread changes in AP practice (Antunes et al., 2020; DHSC, 2020). Prior to the pandemic, national legislation stipulated that medicines prescribed for individual residents could only be administered to them exclusively. It was not permissible to ‘repurpose’ prescribed medication in order to administer it to another resident who may need it urgently. This was a source of considerable concern to many clinicians working in care homes, should end-of-life symptoms arise for a resident for whom anticipatory medicines had not yet been prescribed (Seymour et al., 2011). On 28 April 2020, in response to the challenges posed by the pandemic, new national guidance was issued permitting the repurposing of medicines, including anticipatory medicines, in care homes and hospices (DHSC, 2020). This legislation has since been withdrawn. While AP is widely recommended and practiced in care homes, the AP system in care homes prior to, during, and following these changes has been little investigated to date. My research seeks to 26 expand our understanding of the system in practice and address these knowledge gaps with a view to understanding and redefining best practices in care home end-of-life care. o Geographic variation across UK care homes It is widely recognised there are disparities in social care funding levels between urban and rural regions in the UK (Seymour et al., 2011; Glasgow & Doebler, 2020). Rural areas receive less funding per head, notwithstanding the additional costs involved in delivering care (Hardhill & Dwyer, 2011). There is limited research to date investigating the consequences of this geographical funding disparity and the implication for care home end-of-life care. My research considers these issues in relation to the AP system. o Changes in UK care home finances Peter Townsend’s seminal 1962 publication, ‘The Last Refuge’, found major inadequacies in the 173 public, voluntary, and private UK care homes studied (Townsend, 1962). Since this period, there have been major changes in the constitution of the care sector, including a shift from public to private sector provision, with voluntary, NHS and local authority care homes declining significantly from the mid-1980s onwards, and much greater private care provision (Lievesley et al., 2011; Harrington et al., 2017). This transformation towards a market economy in adult social care has progressed steadily; conversations about a ‘mixed economy of care’ with local authorities, private companies, and the voluntary sector competing on a ‘level playing field’ have quickly dissipated (p. 42, Doty, 1986). In 1979, 64% of care home beds were still provided by local authorities or the NHS; by 2012, the local authority share decreased to 6% (CMA, 2012; Stern et al., 1993). Private provision is currently at approximately 84%, the voluntary sector at 13%, and the public sector just 3% (CMA, 2022). There are over 17,000 independent care homes ranging from corporate chains to independent, family-run care home businesses (CMA, 2022). While the care home industry has thousands of small providers, the five leading operators (HC-One, Four Seasons, Care UK, Barchester and BUPA) provide about 50,000 beds in total (SRD, 2019). Care UK and HC-One are financed by private equity firms (Harrington et al., 2017). Four Seasons was previously owned by a private equity company until 2019, when it was acquired by US hedge fund, H/2 Capital (Bayliss & Gideon, 2020). Barchester is controlled by a group of fifty investors, although the company is majority-owned by three individuals (Bayliss & Gideon, 2020). An international body of research spanning the UK and US has shown that for-profit, corporatised care homes deliver lower quality care than do not-for- profit, independent, voluntary, or individual- and family-run care homes (Barrn & West, 2017; Comondore et al., 2009). During the COVID-19 pandemic, care homes run by private equity firms experienced over 150% of the death rate of independent care homes (Morris & Wu, 2022). It was 27 thus important for my research to consider the relationship between care home financing and the delivery of AP. o Care home staff recruitment challenges The whole UK care home sector has experienced staffing shortages for many years, particularly in rural areas (CQC, 2022; Green et al., 2018). A 2022 National Care Forum survey revealed that not- for-profit care homes were lacking, on average, one-third of the workforce required (NCF, 2021). A 2021 Association of Directors of Adult Social Services survey reported that 70% of social services directors found that care providers were taking drastic steps such as pausing new resident admissions or closing their care homes due to staff shortages (ADASS, 2021). Staffing shortages can jeopardise residents' safety and dignity, compromising resident care (UNISON, 2014). End-of-life care in general and AP in particular involves the coordination of several care home staff members and community-based professionals and is compromised by staffing shortages (Bowers et al., 2019; Wilson et al., 2017). ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 1.7 Health and social care professionals supporting the AP system Anticipatory prescribing in care homes is a complex intervention involving significant cross-sector planning and working between several professionals, including: General Practitioners (GPs), Consultants in Palliative Medicine, specialist community Palliative Care teams, community and District Nurses (CNs and DNs), community pharmacists, care home managers (CHMs), care home nurses (CHNs), ambulance staff, care home residents, and family carers (CQC, 2012; Wilson & Seymour, 2017). The roles and responsibilities of these professional groups are outlined in Table 1.2 below, providing context for their accounts in later chapters. Table 1.2 Roles and responsibilities of health and social care professionals in the AP system Type of professional Roles and responsibilities in the anticipatory prescribing system Care home manager (CHM) • CHMs of residential and nursing homes are responsible for overseeing and ensuring that the quality of the service and care offered meet the required standards (NCS, 2020; NHSE, 2008). • Their responsibilities are detailed in the Health and Social Care Act 2008 (Regulated Activities) and include: complying with regulations, ensuring the safety and suitability of premises; managing complaints; maintaining records; assessing the safety, availability and suitability of equipment; safeguarding service users from abuse; ensuring that an adequate number of appropriately qualified, skilled and experienced workers are employed; respecting and involving service users; cooperating with other providers; monitoring the quality of service provision; caring for and ensuring the welfare of service users; maintaining cleanliness and infection control; meeting nutritional needs of service users; and managing medicines (NHSE, 2008). • CHMs’ responsibilities have evolved considerably over recent years, with an increased emphasis on meeting regulatory requirements and profitability (Matosevic et al., 2007; Orellana, 2014). On average, CHMs spend more than half 28 of their time on administration and bureaucracy: 51% reported doing so in privately owned homes, 59% in local authority homes, and 60% in voluntary sector homes (Orellana, 2014). Care home nurse (CHN) • In the UK, all CHNs must be registered with the Nursing and Midwifery Council (NMC) and meet its standards for education, training, and practice. Most are registered nurses (RNs) who have completed a nursing degree and are registered with the NMC. Nurses in dementia care homes typically receive further specialised training in dementia care (NCS, 2020). • Care home nurses are responsible for the health care needs of residents, providing nursing care under the employment of a care home provider. Their work involves coordinating with health and social care colleagues, particularly visiting GPs, in order to procure and administer AMs (Cooper et al., 2017; Spilsbury et al., 2015). District Nurse (DN) • DNs work in community and primary care settings, mainly patients’ own homes and care homes, providing nursing care and support for patients, families, and care home staff (McCrory, 2019). • DNs are community nurses with further specialist education and often lead a team of CNs and health care assistants. Their work includes case management of people living with multiple long-term conditions and complex needs, including end-of-life care and medication administration (McCrory, 2019; Goodman et al., 2003). • They work with GPs and other community-based professionals as part of primary health care teams to provide home and residential care home nursing care (McCrory, 2019; Goodman et al., 2003). Community pharmacist • Community pharmacists’ role in relation to care homes is primarily one of dispensing prescribed medicines (Mossialos et al., 2013). • The NICE 2014 guideline ‘Managing Medicines in Care Homes (SC1)’ widened the responsibilities of community pharmacists to include: conducting medication reviews; optimising treatment schedules; ensuring medication supply; facilitating medication storage; reviewing administration and disposal records; developing standard operating controls; supporting and advising care home staff; and working as part of a multidisciplinary team to promote appropriate, high-quality and safe care for patients (NICE, 2014). Ambulance paramedic • Ambulance paramedics are employed by NHS Ambulance Trusts, trained to respond to unexpected and emergency events, conduct thorough assessments, initiate management plans and determine the optimal location for patient care (commonly involving a decision to transfer to hospital or remain at home) (Eaton et al., 2021; HEE, 2023; Schofield & McClean, 2022). • Residential care homes lack on-site nursing and frequently call on ambulance staff for medical support during urgent end-of-life care situations (Amador et al., 2014; Wagstaff & Mistry, 2020). General Practitioner (GP) • The role of the GP is defined by The Royal College of General Practitioners as follows: ‘GPs assess, diagnose, treat and manage illness. GPs act as a patient's advocate, supporting and representing a patient's best interests to ensure they receive the best and most appropriate health and/or social care. GPs also provide the link to further health services and work closely with other health care colleagues to help develop those services’ (RCGP, 2021). • Although a GP’s primary relationship is with their patient, the resident, rather than with the care home organisation, working in coordination with care homes is essential for the provision of high-quality care to residents, including prescribing (NCS, 2020; NHS, 2019) Consultant in Palliative Medicine • While GPs are typically the primary health care providers and prescribers for care homes, specialised Consultants in Palliative Medicine and their associated community Palliative Care nursing teams may be involved in the care of residents with complex needs, working closely with primary care teams (NCS, 2020). At times, particularly during the COVID-19 pandemic, Palliative Medicine Consultants and Palliative Care Nurse Prescribers, including those who participated in my study, also support GPs by prescribing end-of-life care medicines (Spilsbury et al., 2020). 29 1.8 Personal motivations My decision to pursue a PhD on the topic of AP in UK care homes stemmed from personal experiences with my grandparents' end-of-life care. Bearing witness to their journeys, marked by the absence of adequate medication and a depersonalised care system in under-resourced care contexts, emphasised the need for in-depth research in this field. Importantly, my academic trajectory has consistently bridged multiple disciplines. My undergraduate studies in chemistry and sociology were complemented by experiences in the Bove Lab at the University of California at San Francisco. Here, I began my journey in qualitative research, conducting interviews with patients, family carers, and their health care providers seeking to understand and improve the care of individuals living with advanced neurodegenerative diseases. This experience further contributed to my academic interest in palliative and end-of-life care. Later, while studying medical ethics and philosophy in my Masters, I was afforded the chance to participate in hospital ethics committee meetings. There, intense deliberations about end-of-life medicines were framed by the broader context of the United States’ opioid epidemic. Yet, to my surprise, similar committee meetings or multidisciplinary team discussions were conspicuously absent in my grandparents’ care. To explore this further, I immersed myself in a local care home environment during the early stages of my PhD, and also shadowed end-of-life care specialist community nurses. These experiences suggested that end-of-life anticipatory medicines posed many frustrating challenges to both patients and their care providers, particularly within oft-neglected care home settings. When the COVID-19 pandemic struck, I was no longer able to volunteer in care homes or shadow professionals, but personal caregiving roles during this period reaffirmed the relevance of my research. These experiences led me to commit to furthering our understanding of end-of-life medicines and the systems within which they’ve embedded. My hope is that this body of work adds something of value to an end-of-life care system that respects patients and upholds their dignity and autonomy. 1.9 Impact of the COVID-19 pandemic The COVID-19 pandemic profoundly stressed the AP system and end-of-life care in care homes across the UK, exposing vulnerabilities in AP and highlighting both the strengths and weaknesses of the system. The lack of effective pandemic planning had adverse impacts on care home residents, families, and staff including managers and nurses, as well as the wider health care community including community nurses, ambulance staff, GPs, and palliative care specialists (Antunes et al., 2022; Spilsbury et al., 2020). The pandemic exposed that fragmentation, privatisation, workforce shortages, and inadequate planning created unsafe and unsuitable environments for end-of-life care (Daly, 2020; PGHELC, 2023; Anand et al., 2022), poorly suited to a resilient or well-connected system that would support a collective response to the pandemic and share guidance (Bayliss & Gideon, 2020). These issues were compounded by pre-existing challenges in care homes, including 30 understaffing, low wages, and burnout, which threaten the retention of qualified and experienced staff (Bayliss & Gideon, 2020; Franklin, 2014). Batkovskyte et al. provided a lens into its impact on care home staff in the UK: an alarming 51% of the study’s survey respondents cited staff shortages as a significant issue during the pandemic (Batkovskyte et al., 2021). Respondents also pinpointed staff exhaustion and a conspicuous absence of support from other health care services as formidable barriers to quality care. Similarly, a survey study by Birt et al. reported care home nurses’ responses to the pandemic, revealing their feelings of unpreparedness and ensuing moral distress (Birt et al., 2022). Moral distress, as defined by Hossain & Clatty, can arise when an individual acts in a manner in conflict with their moral or ethical values, often due to external pressures (Hossain & Clatty, 2021). Major factors underlying this distress, identified by Birt et al., were professional isolation and a lack of pandemic preparedness. This data dovetails with the issue of the broader medical community, particularly GPs, experiencing moral distress during the pandemic. A 2021 British Medical Association (BMA) report, ‘Moral distress and moral injury’ highlighted that over half (51%) of GPs surveyed experienced moral distress due to time constraints preventing them from offering adequate emotional support to patients, including those in care homes (BMA, 2021). Furthermore, approximately half of these GPs (48%) cited limited staff and the necessity for remote consultations, stemming from guidelines to limit interpersonal interactions for infection control purposes, as causes underlying their moral distress (BMA, 2021; BMA, 2022). This is particularly concerning as remote AM prescribing by GPs is reported to have become more commonplace during the pandemic (Antunes eta al., 2020; Mitchell et al., 2021; Mitchell et al., 2022): face-to-face consultations carried substantial infection risk and were therefore reserved only when judged to be clinically necessary (BMA, 2022). Thus, recent studies support that both care home staff and GPs experienced significant distress during the pandemic. Given these data, there is a pressing need to critically examine the current structures of the AP system and end-of-life care in the UK, not only to address the immediate shortcomings but also to develop a system that is more resilient and compassionate. An improved system could better support care home residents, families, and health and social care professionals, mitigating their moral distress, thus supporting the delivery of high-quality end-of-life care, even in challenging times. 1.10 Rationale This research aims to address a gap in knowledge by expanding understanding of the AP system, particularly the experiences of and challenges that health and social care professionals encounter in their day-to-day practice within it. An important part of this study involves identifying strategies that can help to mitigate or address those challenges. 31 Therefore, I conducted a review of qualitative research to synthesise understanding of the AP system in UK care homes; to date, no other systematic reviews in this area have been conducted (Chapter 2). The findings from my literature review were subsequently used to inform the qualitative study design (Chapter 3), exploring the nature of the challenges, tensions, and paradoxes that health and social care professionals navigate within the AP system in care homes. Doing so enabled the identification of key recommendations for sustainable and patient-focused AP for care home residents. Existing research has not holistically explored AP as a system inclusive of its various aspects in UK care homes. Several studies such as Bowers et al. (2022), Kinley et al. (2016), Finucane et al. (2014), and Kinley et al. (2010) aimed to investigate the extent to which patients in the community, including those in care homes, had AMs prescribed prior to death and the circumstances around those prescriptions (Bowers et al., 2022; Finucane et al., 2014; Kinley et al., 2010; Kinley et al., 2016). This differs from the objective of this research to understand health and social care professionals’ experiences and challenges within the AP system in care homes, beyond the number of residents with anticipatory medicines prescribed or frequency of prescriptions. This thesis explores the hows and whys underpinning prescribing and other processes of the AP system. Teggi aimed to evaluate the use of injectable controlled drugs and their impact on the place of death among care home residents (Teggi, 2023). However, Teggi did not examine critical aspects of the AP system such as the storage, monitoring, and disposal of anticipatory medicines or controlled drugs in care homes, and calls for further research in these processes as they are carried out by health and social care professionals. This research responds to that call and seeks to expand understanding of AP as a system in care homes, inclusive of those processes, in relation to each other, and the challenges faced by professionals navigating them. Wilson et al. used questionnaires to gain insight into the roles and experiences of community nurses regarding end-of-life medication decision making, focusing on community nurses' views and experiences with: GP prescribing, procuring medicines, and medication efficacy (Wilson et al., 2016). However, the empirical study employs in-depth interviews to investigate and compare the diverse perspectives and experiences of various health and social care professionals involved in the AP system, including but not limited to those of community nurses. In addition, the scope of this investigation specifically explores the AP system within care homes, rather than a more general community setting. Morris et al. aimed to assess the feasibility of GSF- accredited nursing homes maintaining a 'stock' of end-of-life care medicines, including anticipatory medicines (Morris et al., 2013), which differs from an aim of this research in that it does not relate to the implementation of an intervention, but sets out to explore the underlying dynamics of anticipatory medicines supply and disposal in care homes – the factors which determine the necessity and appropriateness of maintaining a stock of anticipatory medicines in these settings. Therefore, qualitative research that expands our understanding of AP in care homes is needed. This involves studying AP as a system, in light of the myriad challenges that health and social care 32 professionals encounter while doing the work of anticipatory medicines prescribing, dispensing, procurement, administration, storage, monitoring, and disposal. Such research is integral to the development and refinement of AP policy and guidance, as AP is a prevalent practice across the UK, yet the evidence base to support its implementation is limited (Bowers, 2019; NICE, 2015). NICE guidelines on the Care of Dying Adults urge further empirical research into the use of anticipatory medicines in practice, recognising that the guidelines were primarily based on expert opinion supported by only two UK-based qualitative studies (NICE, 2015). Since 2015 and the COVID-19 pandemic, there has been contribution to the literature, and this study aims to further contribute to this evidence base, ultimately expanding our understanding of the AP system in care homes and the challenges that health and social care professionals encounter within it. Additionally, while the primary focus of this research has been on the practice of AP in UK care homes, scholars internationally have investigated this dimension of palliative care—end-of-life prescribing in the community, with a recent surge in anticipatory medicines management research (Ferrell et al., 2021; Zimmermann et al., 2016). This study builds on this body of work and responds to global calls for more detailed explorations into the operationalisation of prescribing systems in community care contexts (Connor et al., 2017), including the WHO’s 2020 palliative care initiative emphasising the need for comprehensive approaches for timely and effective end-of-life symptom management (WHO, 2020b). Importantly, the core issues explored in this study—such as bureaucracy in end-of-life care, medication management in care homes, and cross-team coordination in the community—are well-founded internationally (WHO, 2014 & 2020), albeit within varying healthcare infrastructures and cultural milieus. Countries with well-established palliative care frameworks, such as Canada, Australia, and some nations within the European Union, share commonalities with the UK's approach to end-of-life care (Wright et al., 2008; Harding et al., 2019). These nations, which have integrated palliative care into their community health services, may find the results of this study useful for refining their own practices (Lynn & Adamson, 2003; WHO, 2014). Conversely, in regions where AP is less common or nascent, the findings can serve as a cautionary tale or blueprint for these developing palliative care systems (Gómez-Batiste et al., 2020; Stjernswärd et al., 2007). 1.11 Challenges, tension, and paradox End-of-life care in UK care homes is a complex arena, marked by difficulties that can profoundly influence the provision of individualized, patient-centred care. AP is a key aspect of end-of-life care in care home settings, and its implementation can have important implications on the dying process (Majumder et al., 2022; Perkins et al., 2016; Wilson et al., 2015). However, health and social care professionals involved in the AP system often encounter considerable challenges while navigating its various processes, including the prescribing, dispensing, procuring, administering, storing, monitoring, and disposing of anticipatory medicines (Finucane et al., 2014; Kinley et al., 2016; 33 Wilson et al., 2015). Expanding understanding of these challenges can help to expand understanding of the AP system in operation. Central to this investigation is the nature of challenges within the context of tensions and paradox. The term ‘challenge’ is commonly used in academic literature and by policymakers (Kuhn, 2015; Avelsson & Sandberg, 2013). Kuhn (2015) observed that terms like ‘challenge,’ ‘issue,’ and ‘problem’ have become buzzwords that researchers and practitioners frequently use, often interchangeably, without specific delineation. This overreliance on broad terms can dilute the specificity required for meaningful discourse. Alvesson & Sandberg (2013) found that the general vagueness of buzzwords in policy and literature, including ‘challenge’, limits our capacity to understand the root causes, dimensions, and implications of the subjects under study. To understand and address the challenges faced by individuals working in a given system, it is thus important to understand the specific elements which constitute them. To do so, my study will explore their underlying causes. Smith and Lewis argued that the concept of paradox and its resultant tensions allow for meaningful exploration of the persistent contradictions and interdependencies which underlie the challenges experienced by workers in complex systems (Smith & Lewis, 2011). By understanding challenges as such, one can begin to address their root causes and the dynamics that keep them in play. The advantage of identifying tensions and paradoxes is that they compel individuals to look beyond surface-level issues. According to Jarzabkowski et al., addressing challenges without understanding their underlying causes is akin to treating symptoms without diagnosing the disease (Jarzabkowski et al., 2013). By understanding the tensions between contradictory elements and their paradoxical sources, it becomes possible to explore root causes rather than symptoms (Jarzabkowski et al., 2013; Lewis, 2000). As such, recognising systemic paradoxes and tensions can lead to targeted and effective interventions (Lewis, 2000; Smith & Lewis, 2011). As Lewis (2000) proposed, understanding paradox-induced tensions enables practitioners to strategise in ways that harness these tensions for organisational benefit, rather than being perpetually reactive to emergent challenges (Lewis, 2000). Thus, while ‘challenge’ remains a prevalent term, its broad and often vague use necessitates a clearer focus on the foundational elements that may give rise to these challenges (Kuhn, 2015; Lewis, 2000; Avelsson & Sandberg, 2013). The paradox lens and tension concept thus provide the clarity and depth needed to guide both academic exploration and practical management of the challenges faced by health and social care professionals in the AP system in care home settings. Smith and Lewis (2011) define a paradox as ‘a persistent contradiction between interdependent elements’ within a given system. These paradoxes can emerge in complex systems where conflicting demands coexist, thereby producing tensions. Tensions, the authors propose, are between two contradictory elements in a system that simultaneously pull workers in different and often opposing directions, which result in felt challenges, stressors, or strains (Smith & Lewis, 2011). Given its 34 complexity, this thesis argues that the reasons underlying health and social care professionals’ challenges working with AP include the coexistence of contradictory elements within the AP system, both rooted in and continually reinforcing systemic paradoxes. Literature has consistently supported the idea that such tensions can manifest in the form of challenges that individuals face on a day-to- day basis (Smith & Lewis, 2011) and as ‘stress, anxiety, discomfort, or tightness in making choices, responding to, and moving forward in organisational situations’ (p. 68, Putnam et al., 2016). Farjoun (2010), for instance, posits that when workers experience tensions in response to paradoxical systems, it frequently culminates in challenges in their daily operations, necessitating adaptive strategies (Farjoun, 2010). Similarly, Andriopoulos and Lewis assert that tensions both produce challenges and are challenging when they disrupt the equilibrium of a system, demanding resolution or at least some form of navigation (Andriopoulos & Lewis, 2009). From this understanding, the tensions rooted in paradoxes within a complex system such as AP may manifest as challenges for health and social care professionals. To this end, the empirical research analysis in this thesis is guided by a paradox lens which aims to provide a deeper understanding of professionals' challenges and experiences related to working within the AP system in UK care home settings. The use of paradox within this study is justified by empirical data which indicate the presence of persistent and simultaneously occurring contradictions in the AP system. For example, care professionals often contend with the paradoxical simultaneous occurrence of excess wastage and shortages of anticipatory medicines, further detailed in Chapter 5. Such empirical instances exemplify the concept of paradox as defined by Smith and Lewis (2011) and support its application in this research. While the paradox lens offers a way to conceptualise and address the persistent contradictions, it is not without alternatives. A dilemma-focused approach, for example, may consider these contradictions as mutually exclusive choices that exist within a binary, potentially obscuring the reality that choices in care work are rarely absolute and often involve a spectrum of considerations that extend beyond binaries (Cameron & Quinn, 1988; Clegg, 2002). A dichotomy perspective may simplistically classify challenges into two opposing groups, which also risks creating false binaries and does not accommodate the complex, often concurrent roles care professionals must navigate, such as being both caregivers and administrators (Smith & Tracey, 2016). Additionally, frameworks such as phenomenology, focusing on individual experiences and meanings, may not fully capture the systemic and collective nature of contradictions in working with medications in care homes. Phenomenological methods do not attempt to explain the ‘hows’ or ‘whys’ underlying these experiences (Tuffour, 2017), and the layered nature of the AP system requires a paradox framework to distil its tensions. Interactionism could provide insights into the social dynamics of this work, but similarly it may not explore the inherent contradictions of the role as paradox theory does (p. 176, Willis, 2007). Critical theory, while it can offer a critique of power structures and inequities, may not explicitly address or centre the simultaneously occurring contradictions which characterise the day-to-day experiences of health and social care professionals (p. 87, Willis, 2007). In contrast, paradox theory uniquely allows for the recognition that some 35 challenges in the AP system do not have clear resolutions, and that attempting to address one side of a paradox can exacerbate the other. For example, efforts to standardise AP to ensure efficiency can simultaneously undermine individualized care, as detailed in Chapter 6. The paradox concept embraces such complexities and seeks to understand and balance, rather than oversimplify or eliminate them. A paradox lens is particularly suitable then for this study as it aligns closely with the complex and often contradictory aspects of the AP system. 1.12 Core objective & research questions The core objective of this thesis is to: • Expand understanding of the AP system in UK care home settings by exploring the experiences and challenges of health and social care professionals working within it In doing so, this research seeks to contribute to the growing evidence base on AP in UK care home settings and identify strategies that can target health and social care professionals’ challenges to better support them in their work, ensuring that patient-centred care remains at the heart of the AP system. I aim to achieve the core objective of this thesis by addressing the following research questions (RQs): 1. What underlying systemic tensions and what wider interprofessional contexts underpin issues related to the wastage and shortage of anticipatory medicines in care homes? 2. What role do the algorithmic and discursive structures associated with care home-based AP play in shaping the tensions between individualization and standardisation in end-of-life care? 3. How do the structural logics underpinning AP paperwork contribute to the tensions between navigating bureaucratic regulations and providing good, patient-centred end-of-life care in care homes? 1.13 A brief overview of the thesis This thesis consists of eight chapters organised in four parts. Part one of the thesis provides background information regarding the research topic, including this introductory Chapter 1, and Chapter 2, where I review the published literature concerning end-of-life CDs in UK care homes and the experiences of health and social care professionals related to them. Given the limited evidence concerning AP practice in care homes, I broadened the review to include all end-of-life CDs, a literature that is still limited. The systematic review (section 2.1) has been published in BMJ Supportive & Palliative Care. This review synthesised the limited available research on this topic and described several potential areas for future research, including the need to explore the underlying 36 reasons of the ‘challenges’ reported by health and social care professionals working within the AP system. Part two of the thesis provides information about the methodology and methods of the empirical research program. Specifically, in Chapter 3, I detail the methodology, data collection and data analysis methods of the research: a qualitative interview study of health and social care professionals working within the AP system in care homes. Part three of the thesis presents the results of my analysis of the empirical qualitative dataset and analysis-enriching discussions. In Chapter 4, I provide contextual information, describing the outcomes of my recruitment strategy and participant characteristics. In Chapters 5, 6, and 7, I present the findings of my analysis of the qualitative dataset. In Chapter 5, I explore the underlying systemic tensions and wider interprofessional contexts that underpin issues related to the wastage and shortage of anticipatory medicines in care homes (RQ1). This chapter introduces the anticipatory prescription paradox theme, that anticipatory medicines wastage and shortage can and do occur simultaneously – in the same care home and at the same time. Building on the accounts of participants and contextualising them within salient policy and theory, I demonstrate the nuances of this paradox and its constitutive tensions. Emphasising the urgency of the issue in care homes, the chapter underscores the need for targeted policy and practice reforms to navigate this paradox effectively. In Chapter 6, I consider the role of the algorithmic and discursive structures associated with care home-based AP in shaping the tensions between individualization and standardisation in end-of-life care (RQ2). These tensions, I demonstrate, are rooted in the protocol paradox: though dying is by nature an individualized process, end-of-life care, and specifically anticipatory care, is a field beset by standardised models of health care provision. Grounding the analysis in participants’ experience and juxtaposing it with relevant policy, theory, and literature, this chapter underscores the implications of this paradox and its tensions on end-of-life care quality and emphasises the need to re-evaluate current practices in the interest of personalised care. Chapter 7 examines how the structural logics underpinning AP paperwork contribute to the tensions between navigating bureaucratic regulations and providing good, patient-centred end-of-life care in care home settings. The chapter introduces the paperwork paradox theme: that the excessive bureaucratic documentation imposed on care homes detracts from the time needed for health and social care professionals to provide good and patient-centred care, while simultaneously demanding that care. Grounded in participant accounts and drawing from relevant paper, policy, literature, and theory, I dissect the paperwork that professionals do, and the ways in which the demands posed by 37 excessive paperwork detract from the time needed for the patient-centred care also demanded by that very paperwork. Part four of this thesis entails a conclusion in Chapter 8, in which I integrate and summarise the thesis by addressing the research questions and the key findings. I then detail the thesis’ contribution to knowledge and the literature, outline its impact on policy and practice, offer a reflexivity statement, explore the strengths and limitations of the study, propose directions for future research, and conclude by sharing some final thoughts on this body of work. 38 Chapter 2: Reviewing the evidence regarding End-of-Life Controlled Drugs in UK care homes In this chapter I present the findings of my 2022 published systematic literature review and narrative synthesis of End-of-Life Controlled Drugs in UK care homes from January 2000 to January 2001. This section is primarily comprised of content from the published manuscript: Majumder M, Bowers B, Pollock K, Goodman C, Kuhn I, Barclay S. End-of-Life care in UK care homes–controlled drugs: systematic review and narrative synthesis. BMJ Supportive & Palliative Care 2022; 12(3): 253-261 (Appendix 1) 2.1 Introduction Care homes are an important location for end-of-life care in the United Kingdom (UK). Over 430,000 older people and vulnerable people in the UK live in care homes: of the 539,738 deaths in England and Wales in 2019, 137,998 (26%) occurred in care homes (ONS, 2019a). During the COVID-19 pandemic, the number of deaths within care homes increased by 134% in the first wave and 10% in the second wave (O’Donnell et al., 2021). By 2040, it is projected that care homes will be the most common place of death in the UK (Bone et al., 2018). Care homes are categorised as either ‘residential homes’ (social care staff but no on-site qualified nurses) or ‘nursing homes’ (social care staff and qualified nurses on site at all times): many are dual- registered (NHSE, 2016). For this review we use the term ‘care home’ for all types of homes. Care home residents are often prescribed numerous medicines, with additional drugs prescribed for the management of physical symptoms and distress as the end of life approaches (Goddard et al., 2013; Hall et al., 2014). In the UK these include ‘Schedule 2 and 3 Controlled Drugs’ (CDs) such as morphine and related opioids and the sedative medication midazolam. CDs are subject to regulations additional to those for other prescription-only medications, as they are potentially open to misuse. National legislation controls the prescribing, dispensing, administration, storage and disposal of medicines and CDs and is detailed in guidelines of the National Institute for Health and Social Care Excellence (NICE) (NICE, 2017), reinforced by the regulatory inspector of care homes the Care Quality Commission (CQC) (CQC, 2019), and guided by principles outlined in the Enhanced Health in Care Homes framework of NHS England (NHSE, 2020). 39 Controlled drugs and other end-of-life care medicines are commonly prescribed in the community in advance of need in a widespread practice known as ‘anticipatory prescribing’, to ensure that appropriate medicines are in place if needed during the dying phase. These ‘anticipatory medicines’ (AMs) are stored in the home or care home ‘just in case’ they are needed at the end of life (Bowers et al., 2018; Bowers et al. 2019). Care home residents are registered with one or more local General Practitioner (GP) practice(s) who are responsible for providing and coordinating their medical care, in partnership with care home staff and other members of local health and social care services. GPs are responsible for prescribing residents’ medication, which is supplied to the care home by a local community pharmacist, where staff store the medication according to the home’s medicines management processes and are responsible for ensuring that residents receive their medicines as prescribed. CDs are stored in separate locked cabinets in care homes, with strict regulations over their documentation and use (CQC, 2019). The potential for errors in the storage and administration of CDs is a major concern for care home staff and CQC inspectors (Ellis & Abbott, 2016; NICE, 2015). The storage capacity in care homes is often limited. Thus, the storage of such anticipatory medicines, prescribed in advance of possible need, may present challenges for care homes where they may be prescribed for several residents considered to be close to the end of life. Towards the end of life, people can often no longer take oral medicines for symptoms such as pain and agitation: a trained nurse may be needed to administer drugs by injection, either a visiting Community Nurse in a residential home or an in-house nurse in a nursing home. This requires the appropriate injectable drugs, frequently CDs, to have been prescribed for the person, which national guidance recommends should be in a personalised manner rather than a ‘blanket-like fashion’ (NICE, 2015). The decision to administer anticipatory CDs depends on nursing staff assessing that there are no reversible causes of a resident’s deterioration (Faull et al., 2013), that the end of life is approaching (Bowers et al., 2017; Faull et al., 2013; Mitchell et al., 2016), and that symptomatic management and comfort care is appropriate (Faull et al., 2013; Wilson et al., 2015). These judgements are frequently challenging for nursing home nurses who may have limited training in end-of-life care and limited access to colleagues such as General Practitioners and specialist nurses, particularly overnight and at weekends (Bowers & Redsell, 2017). In residential care homes, these decisions are made by visiting community nurses who may be more experienced in end-of-life care. The use of CDs such as morphine and midazolam at the end of life adds a further layer of concern for managers and nursing staff in both residential and nursing homes, with concerns over the appropriate use of 40 such powerful drugs, their potential for hastening the end of life, and the legal and regulatory frameworks for their storage, use and disposal (Goddard et al., 2013). Alongside the marked increase in care home deaths during the COVID-19 pandemic, changes were made in UK legislation which permitted the repurposing of end-of-life care drugs in care homes. During the COVID-19 pandemic in England, drugs prescribed and stored for one resident could be repurposed for and administered to another resident if urgently needed to address issues regarding the timely supply of drugs in care homes (DHSC, 2020). The extent to which this legislation was used in practice is unclear at present and is under investigation by the authors. Given the increasingly important role of care homes in end-of-life care in the UK, it was timely to review the published evidence concerning the prescribing, administration, storage, monitoring, and disposal of end-of-life CDs in UK care homes. The focus is on oral, injectable, and transdermal ‘Schedule 2 and 3 CDs’, primarily anticipatory medicines that are strong opioids, including morphine, oxycodone, fentanyl (excluding codeine, dihydrocodeine, and tramadol) and midazolam. These medicines are commonly prescribed for end-of-life symptom control, are subject to strict regulation due in part to their potential for misuse and abuse, and require specific training in their use. There was considerable concern during the COVID-19 pandemic that they might be in short supply (Faull et al., 2013; NICE, 2015). The decision to focus the literature review on Schedule 2 and 3 CDs was both a strategic adaptation to the heightened role of these medicines during the pandemic and a direct response to a gap in the literature on anticipatory prescribing. The decision to broaden the scope of the review was further reinforced by the recognition that commonly prescribed AMs, including morphine, diamorphine, oxycodone, and midazolam, are typically included within the ambit of controlled drugs studies, given that these medicines are crucial for managing end-of-life symptoms such as pain and agitation. The insights obtained from investigating the use of CDs thus fed directly into a broader conversation about AP. By examining the specific challenges posed by CDs within care homes, the review set the stage for a deeper exploration of AP as a holistic practice that encompasses these medicines in care home settings. The expansion of the review’s parameters allowed for a comprehensive collection of the AP literature, ensuring that the review maintained its rigour and relevance within the dynamic pandemic context. This approach addressed the gap in the existing literature and provided a substantive foundation for exploring the experiences and challenges of health and social care professionals working within the AP system. 2.2 Aim 41 This review builds on a previous review exploring AP in the community (Bowers et al., 2019), with a specific focus on end-of-life care CDs in the care home setting. 2.3 Review questions With regard to End-of-Life Controlled Drugs (CDs) in care homes in the United Kingdom: 1) For whom and by whom are they prescribed? 2) How are they stored and monitored? 3) For whom and by whom are they administered? 4) How are they disposed of after a resident’s death? 5) What are the views of care home staff and other health and social care professionals concerning these issues? 6) What are the views of residents and their family members concerning these issues? 2.4 Methods A search strategy, developed with a specialist information technologist (I.K.), was developed for CINAHL (Appendix 2) and adapted for six subsequent databases (Medline, Embase, Web of Science, PsycINFO, Social Care Online, and Cochrane Library). All databases were searched from January 2000 to January 2021: the start date was the selected as the year of publication of the Care Standards Act (2000) which covered controlled drug usage, storage, and administration in care homes. Hand-searches of two journals, Palliative Medicine and British Medical Journal Supportive & Palliative Care, from January 2000 to March 2021 were conducted, as well as reference and citation searches of all included papers. Papers were included if they presented empirical data on controlled drugs for UK adult care home residents. Due to the diversity of CD regulations internationally, only papers concerning UK care homes were included. Box 2.1 summarises inclusion and exclusion criteria. The review questions were developed iteratively through discussions with the research team, with clinical and lay stakeholders, and from initial readings of the policy and research literature. Box 2.1 Review inclusion and exclusion criteria Inclusion criteria: ● Published papers presenting empirical research on the prescribing, storage, administration, and disposal of these oral and injectable Schedule 2 and 3 controlled drugs: strong opioids (excluding codeine, dihydrocodeine, and tramadol) and midazolam primarily (oral or injectable) ● Studies published between January 2000 and January 2021 42 ● English language full text ● Study setting: UK care home, residential and/or nursing homes ● Peer-reviewed quantitative and qualitative studies, case studies ● Key areas for data extraction: 1. For whom and by whom end-of-life CDs are prescribed. 2. How CDs are stored and monitored. 3. For whom and by whom CDs are administered. 4. How are CDs disposed of following a resident's death. 5. The views of care home staff and other health and social care professionals concerning these issues. 6. The views of residents and their family members concerning these issues. Exclusion criteria: ● Controlled drugs in non-terminal situations ● Controlled drugs not relevant to end-of-life symptom control ● Children (aged under 18 years) ● Prescribing in hospital, hospice, home or prisons ● Papers with no new empirical data, for example, editorials ● Research examining assisted dying, euthanasia, or continuous sedation until death ● Grey literature The PRISMA diagram (Figure 2.1) summarises paper selection. Following de-duplication and title screening by MM, 81 abstracts were independently screened by MM and BB. Full text papers were assessed for eligibility by MM with a second review by BB where eligibility was uncertain, leading to identification of 14 papers for data extraction. Disagreements were resolved by consensus. Figure 2.1 PRISMA flow diagram A review-specific data extraction form (Appendix 3) was used by MM and BB to extract publication details, study aims, participants, methods and results relevant to the review questions. Titles from database searches (n=1279) De-duplicated titles screened (n=1205) Abstracts screened (n=125) Full-text articles assessed (n=42) Studies included in synthesis (n=14) Titles excluded as irrelevant (n=1080) Abstracts excluded: did not meet inclusion criteria (n=83) Full-text articles excluded: did not meet inclusion criteria (n=33) Reference and citation searching (n=3) Journal hand searching (n=2) Id e n ti fi ca ti o n Sc re en in g El ig ib il it y In cl u d ed 43 Two reviewers (MM and either BB or SB) independently assessed the quality and relevance of each study using Gough’s ‘Weight of Evidence’ (WoE) framework (Table 2.1). Each article received a rating – high, medium, or low – indicative of its quality, with disagreements resolved by consensus (Gough, 2007). Gough defines the Weight of Evidence approach as a tool used in the review process, emphasising its purpose: ‘considering how to make separate judgements on different generic and review-specific criteria and then to combine them to make an overall judgement of what a study contributes to answering a review question’ (p. 224, Gough, 2007). This approach offers a nuanced appraisal of studies, moving beyond broad-stroke generalisations and towards close examination of specific details to determine the extent to which a study supports or detracts from the research questions. Gough’s WoE breaks down the quality and relevance of included studies across four key domains (as outlined in Table 2.2): (1) the internal validity of the study (WoE A), (2) the appropriateness of the study’s design to the review’s aims (WoE B), (3) the relevance of the study’s content to the review’s aims (WoE C), and (4) a composite rating combining the three previous domains to provide an overall judgement of the quality and relevance of the study (see Table 2.3). Table 2.1 Review-specific Gough’s ‘Weight of Evidence’ criteria WoE A: judged against internal validity, whether study design was rigorous, whether this could be adequately assessed from a transparent, comprehensive, and replicable method; accurate and understandable presentation and analysis; if samples and data collection tools were appropriate to the method. Papers scored as high / medium / low. WoE B: relates to the appropriateness of the study design to the six review-specific questions. Papers were scored as high / medium / low. Review questions 1, 2, 3, and 4: the fitness for the purpose of the study design in answering the questions were made on a paper-by-paper basis. Review questions 5 and 6: inductive research designs interpreting the views directly reported by patients / carers / health care professions = high. Deductive research designs interpreting these views directly reported by patients / family carers / health care professionals = low. WoE C: relates to detailed judgements about each study relating to the relevance of the focus of the evidence for answering the review questions. This includes: consideration of any sampling issues relating to the interpretation of the data; whether the study was undertaken in an appropriate context from which results can be generalised to answer the relevant review-specific questions. Papers were scored as high / medium / low. WoE D: the above three sets of judgement scores are then combined to give the overall ‘weight of evidence’ as high / medium / low. Criteria adapted Gough (2007), table contents adapted Bowers et al. (2019) Table 2.2. Weight of Evidence Appraisal Matrix Weight of Evidence Description High Medium Low A Quality of study execution; clarity of purpose: accuracy; understandability; method-specific quality. 3 2 1 B Appropriateness relates to the review question; fit for methodological purpose. 3 2 1 C Provides a relevant answer to the research question. Has utility and value; the findings were generated ethically, appropriately, and generalizable 3 2 1 D Overall assessment of weight of evidence findings. A mathematical sum of A through C above. 8-9 5-7 3-4 Adapted from Gough (2007) 44 To capture the range of qualitative and quantitative evidence, a narrative synthesis was undertaken, involving three iterative phases (Petticrew, 2006; Popay et al., 2006). This approach was selected due to its ability to both support an empirical ‘data-driven’ approach to theme-identification (Bharmal et al., 2022) and to integrate findings from heterogeneous data sources, thereby offering a comprehensive view of the evidence (Dixon-Woods et al., 2005) required to understand the practices and the perceptions surrounding the use of end-of-life CDs in UK care homes. Narrative synthesis is particularly useful for analysing data from studies which cannot be statistically synthesised due to heterogeneity in methods or reported outcomes, and in facilitating the identification of patterns and themes across studies (Dixon-Woods et al., 2005). Its flexibility and capacity to holistically combine diverse data types thus allow for the interpretation of nuanced findings that arise from complex datasets, which is beneficial for this review, where both qualitative and quantitative findings are important. Alternative methods to narrative synthesis, such as meta-analysis and a realist review, were considered; however, these methods were deemed less appropriate due to their constraints in supporting the expected methodological and clinical diversity of the studies. Meta-analysis, for example, requires a level of homogeneity in the data that was not anticipated, given the varied practices across care homes and different types of CDs under consideration (Esterhuizen et al., 2016). Moreover, meta-analysis is optimal for combining results from similar quantitative studies to calculate an overall effect size and is less adaptable to qualitative research (Esterhuizen et al., 2016). Realist reviews offer a theory-driven analysis of the literature which seeks to understand the mechanisms by which interventions work or fail within particular contexts (Rycroft-Malone et al., 2012). However, as the literature on end-of-life CDs in care homes involves a variety of contexts, interventions, outcomes, and study designs, a realist review may be too narrowly focused on mechanism-context-outcome configurations to capture the breadth of evidence available (Rycroft- Malone et al., 2012). Thus, neither meta-analysis nor realist reviews align as closely with the objectives of this review as narrative synthesis, a structured yet flexible method capable of synthesising various types of evidence to provide insights into the practical and experiential aspects of end-of-life CD use in UK care homes. It supports, for example, the integration of both the quantitative data regarding CD usage patterns and qualitative data reflecting the perspectives of health and social care professionals. The narrative synthesis involved three progressive iterative stages described in Box 2.2. Box 2.2 Narrative synthesis strategy ___________________________________________________________________________________________________________________________________________________________________________________________________________________________________ The first stage involved the development of a preliminary synthesis. I started by creating textual descriptions of each study. These included key details from the data extraction forms, such as 45 methods, study populations, and main findings. I then organised the study descriptions based on specific criteria including the sample population and research questions addressed by the studies’ results (Petticrew, 2006; Popay et al., 2006). An initial inductive thematic analysis was conducted to identify the ‘main, recurrent and/or most important (based on the review question) themes and/or concepts across multiple studies’ (p. 18, Popay et al. 2006). The thematic analysis involved the following processes: data collection (identifying the studies through the systematic search strategy described in Appendix 2 and extracting data from the studies’ results relevant to my review questions); familiarisation with the data, in which BB and I reviewed and immersed ourselves in the collected data, making initial notes and observations (for instance, the repeated mention of GPs as primary prescribers or the challenges faced by care home staff in obtaining end-of-life CDs from out-of-hours GPs were initial observations); generating initial codes, which involved open coding in order to condense the data into analysable units; searching for themes, in which the codes were grouped into initial themes (for example, ‘difficulties procuring end-of-life CDs’ or 'impact of the Reuse of Medicines guidance'), then into overarching latent themes that represented patterns in the data and considered the underlying concepts that connected the themes (for example, ‘Challenges in care coordination' and ‘Challenges in crisis-driven care’); reviewing and refining themes, where the identified themes were reviewed against the coded data and the entire dataset to ensure that they accurately represented the nuances of the studies; and finally, reporting, in which the results were written up as ‘discussion’, integrating the findings to present a narrative of the dataset, discussing their implications, highlighting key themes, presenting limitations, and suggesting areas for future research. The second stage involved exploring the data. In this stage, key individuals – BB, a clinical academic nurse, SB, a clinical academic palliative care doctor and I – developed the interpretive synthesis by reviewing the thematic analysis independently at various stages (Petticrew, 2006; Popay et al., 2006). It also involved exploring heterogeneity across studies – for instance, examining similarities and differences between studies’ methodological approaches, research methods, contexts, populations under study, and findings (Petticrew, 2006; Popay et al., 2006). To ensure a comprehensive understanding of the evidence, the reviewing team discussed and deliberated on results from studies led by researchers from different disciplines and epistemological backgrounds (Popay et al., 2006). Consensus on the synthesis was achieved through these discussions. The synthesis was further refined by engaging in discussions on the review results and their broader implications with a range of interdisciplinary peers and my supervisory team, including SB (GP and palliative care clinical academic), KP (a medical sociologist), and CG (nurse and clinical academic). In the third stage, the robustness of the synthesis was assessed. Each stage of the synthesis was guided by the quality and relevance criteria set out by Gough’s WoE Framework (Table 2.4) 46 (Gough, 2007). Papers that received a ‘high’ rating for quality and relevance (WoE D) were regarded as more credible and significant than those with a ‘medium’ rating during data synthesis (Gough, 2007; Popay et al., 2006). Results from papers with a ‘low’ rating for quality and relevance (WoE D) were cautiously integrated, and only when they supported the findings reported in ‘high’ or ‘medium’ quality papers (Popay, 2006). The inclusion of studies with a low weight of evidence in the synthesis was intentional, demonstrating current reliance on low and medium-quality evidence in anticipatory prescribing practice and policy, underscoring knowledge gaps and areas in need of further investigation (see Table 2.4). Notably, Popay et al. highlight the discretionary use of Weiss’ 'theory of change' in systematic reviews as a potential early step in the development of a synthesis (p. 12, Popay et al., 2006). The theory describes ‘the chain of causal assumption that link programme resources, activities, intermediate outcomes and ultimate goals’ (Weiss, 1998). While the use of theory approach may clarify the 'how' and 'why' of an intervention's impact, its inclusion is not always necessary or suitable (Popay et al., 2006). The topic under study, the CD use at the end of life in UK care homes, is grounded in well-established palliative care principles with clear objectives: to alleviate pain and provide comfort to patients at the end of life (NHSE, 2020). This review is thus evaluative of existing practices rather than exploratory or developmental of a new intervention, which does not necessitate the modelling of a ‘theory of change’. Specifically, the review aims to synthesise empirical data on current practices surrounding end-of-life CD use in UK care homes across several domains: prescribing, storage, administration, monitoring, and disposal. The narrative synthesis method adapted for this review is adept at integrating various types of evidence and exploring heterogeneity without the constraints of a predefined theoretical model, which may be more appropriate for context-dependent practices. Therefore, the review does not employ a ‘theory of change’ as it focuses on synthesising evidence on existing practices rather than developing or testing a theoretical model of intervention. ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ The review protocol was registered with PROSPERO (reg. no. 42020173014) in June 2020. 2.5 Results Fourteen full papers, reporting on twelve studies, were included in the synthesis (Table 2.2). Methods included: interviews or surveys of health care professionals or care home staff (n = 11), retrospective care home resident notes reviews (n = 8), ethnographic observations of health care professionals across care homes (n = 3), and documentary analyses of provider End-of-Life Care policy documents (n = 2). One study was reported in three papers (Wilson et al., 2015; Wilson et al., 47 2016; Wilson & Seymour, 2017). Each paper presented different findings and was therefore treated as an individual unit in the synthesis. Several papers answered more than one review question. In addressing review questions 1 and 3, studies included surveys, analyses of patient records, and policy documents. In addressing review question 5, included studies drew on qualitative methods, namely interviewing. Studies that used more than one method contributed to more than one section of the synthesis. Table 2.3 summarises the included papers and their “Weight of Evidence”; two were rated high, nine medium and three low WoE. Table 2.3 Summary of included studies Author, Year Title Reference Aims Research methods Key Findings Gough’s WoE A+B+C = D Finucane, 2014 Anticipatory prescribing at the end of life in Lothian care homes Finucane, Anne M., et al. "Anticipatory prescribing at the end of life in Lothian care homes." British journal of community nursing 19.11 (2014): 544- 547. To investigate the extent to which residents in eight Lothian care homes had AMs prescribed prior to death. AP intervention (training, multidisciplinary meetings, and new AM stocks) implemented in 8 care homes. Data collected in each care home at baseline, and once the intervention started. Quantitative analysis (counts, percentages). RQ1: "Many care home residents do not have the recommended AMs in place in the last days of life and thus may experience inadequate symptom control. Anticipatory prescriptions (Aps) for analgesia were available to just over half of all deceased residents; however, less than one quarter had APs for nausea and respiratory secretions. There was a wide variety of AP across care homes, with all residents dying with AMs prescribed in one care home, compared to only 13% of residents dying with AMs prescribed in another care home. Of the 77 residents who died in the care homes, 54% had AMs prescribed. Only 15% had prescriptions for all four nationally recommended AMs. Many care home residents do not have the recommended AMs in place in the last days of life and thus may experience inadequate symptom control. Interventions that increase the availability of AMs to manage common symptoms at the end of life for care home residents are necessary." M M M = M Kinley, 2016 Achieving symptom control not medication burden at the end of life Kinley, J., et al. (2016). "Achieving symptom control not medication burden at the end of life." Palliative Medicine 30 (6): NP331- NP332. To identify the prescribing practice for symptom control in the last month of life for residents dying in nursing homes. Thirty eight nursing homes implementing an end of life programme (GSFCH) took part in a trial. In order to complete this programme, the care home staff needed to evidence symptom control at the end of life. The data presented is from all items prescribed in the last month of life from all deceased residents’ notes during the 2nd year of the GSFCH programme. RQ1: "The medical administration charts were available for 50% of the 638 residents who died in the nursing home during the data collection period. In the last month of life the mean number of oral medication prescribed for residents was 10 (range 0-27). The residents who died had been prescribed oral medication for symptom control in their last month of life. For example 84% of residents had been prescribed oral and/or transdermal analgesia." M L L = L Morris, 2013 Implementing stock end-of- life medication in UK nursing homes Morris, Kathy, and Jo Hockley. “Implementing stock end-of-life medication in UK nursing homes.” End of life journal 3.3 (2013). To assess the feasibility of nursing homes with Gold Standards Framework in Care Homes (GSFCH) accreditation keeping end-of- life care medications ‘as stock’, rather than having the medication prescribed and dispensed for individual residents when residents became symptomatic. Qualitative interviews with nurse managers of four nursing homes with GSFCH accreditation, local pharmacists and the hospice pharmacist. A strategy to implement the stock medication was prepared. A baseline review of the notes of deceased residents was undertaken and information regarding symptoms and end-of-life care medication extracted. The review continued for 6 months after the stock end-of-life care medication was obtained. RQ1: “Once the End-of-lifeC medication ‘as stock’ had been obtained, the number of deceased residents with a written order from the GP for anticipatory medication increased in each NCH.The number of orders increased in NCH 1 to 50%, in NCH 2 to 63% and in NCH 3 to 94%.” RQ1: "Up to 53% of the residents were symptomatic in their last days of life, meaning that access to end-of-life medication is important. A total of 47% of frail older people were asymptomatic in the last days of their lives. If end-of-life care medications are stocked by nursing homes with end-of-life care training, there is a potential for dying residents to receive timely symptom control." RQ3: “As a result of anticipatory medication being immediately available once the stock had been obtained, more residents who were symptomatic had their symptoms managed promptly.” RQ3: “Only four of the 92 residents reviewed required a syringe driver.” RQ5: “Although during the study staff in NCH 1 did not use the end-of-life care ‘as stock’ drugs as much as the nurse manager had hoped, the number of residents having access to medication did increase significantly.” RQ5: The staff in NCH 2 “did feel ‘safer’ in the knowledge that they had a supply of stock drugs that could be accessed in emergencies for residents who deteriorated unexpectedly.” RQ5: “In NCH 3, the nurse manager reported that having access to stock medications had ‘transformed’ the practice of her nurses. They now felt comfortable requesting prescriptions for every dying resident, as they were more confident that there would be no wastage. She also reported an example of how having the stock medication had helped a busy GP.” M M M = M 48 Owen, 2016 Pharmacy team led anticipatory prescribing in nursing homes: Increasing proportion of deaths in usual place of residence Owen J, Down A, Dhillon A, et al. "Pharmacy team led anticipatory prescribing in nursing homes: Increasing proportion of deaths in usual place of residence." International Journal of Pharmacy Practice 24 (Supplement 3): 12 (2016). To determine if pharmacy team led end of life anticipatory prescribing for nursing homes decreases hospital admissions at end of life. Data was assessed relating to 550 patients who died in 19 nursing homes between July 2014 and July 2015 and compiled into tables. Death certificates were accessed to determine place and reason of death. Supply of AMs was confirmed from prescribing software, and involvement of palliative care teams documented. Archived medicines administration records (MAR) and controlled drug (CD) registers were accessed to determine whether AMs were administered. RQ1: "Embedding a pharmacy team managing prescribing in GP led nursing home service contributed to an increase in the proportion of deaths in nursing homes, avoiding hospital admission, from <40% to 75% in 2 years. The embedded nursing home pharmacy team worked alongside the local palliative care team to implement end of life initiatives such as teaching and support of nursing staff, development of a formulary of AMs and ensuring urgent supply from community pharmacies." M M M = M Perkins, 2016 The care of dying people in nursing homes and intensive care units: a qualitative mixed-methods study Perkins, Elizabeth, et al. "The care of dying people in nursing homes and intensive care units: a qualitative mixed- methods study." (2016). To gain an insight into how care in different settings was documented and described in the very last hours or days of life. Data were collected from 12 ICUs and 11 nursing homes in England, though 3 of the 11 nursing homes did not participate in the observational aspect of the study: (1) documentary analysis of provider end-of-life care policy documents; (2) retrospective analysis of 10 deaths in each location; (3) interviews with staff about end-of-life care; (4) observation of the care of dying patients; (5) analysis of the case notes of the observed patient’s death; (6) interview with a member of staff providing care during the observed period; (7) interview with a bereaved relative present during the observation; (8) economic analysis focused on the observed patients; and (9) strict inclusion and selection criteria for nursing homes and ICUs applied to match sites on LCP use/non-LCP use. RQ1, RQ3: “Where the LCP was used to support care, variances were recorded for agitation, respiratory tract secretions or pain. The most common action taken in response to these variances was the use of medication, namely midazolam, glycopyrronium (or hyoscine) and morphine (or oxycodone or diamorphine), respectively. Nurses administered AMs as advised and prescribed by GPs and in accordance with the LCP, even against family members’ wishes.” RQ3: “There was a strong emphasis in the nursing homes on being prepared for a patient’s death. This did result in the prescribing of anticipatory drugs, just in case they were required, despite the fact that they were rarely used. Issues in achieving a timely visit from the general practitioner also created problems for the nursing homes and often meant that decisions relating to end-of-life care were taken by nurses.” RQ3: Data from observations of last days of life care in nursing homes: “In the majority of cases where it was established, through either questioning or observation, that there was a symptom that needed relieving, the patient was given medication, with morphine and midazolam being the most common. The route of administration of these drugs was usually via a subcutaneous injection or a continuous subcutaneous infusion via a syringe driver. A small number of patients who were able to swallow were given oral medication. Over the course of the observations the route of administration was tailored to meet the patient’s requirements; tablets were replaced with liquids, subcutaneous injection or continuous subcutaneous infusion via a syringe driver.” RQ3: “A number of nurses also reported that very few of the deaths in their nursing home had required the administration of the drugs that had been prescribed.” RQ5: “Respondents reported that GPs’ attitudes towards anticipatory prescribing were influenced by the costs associated with throwing away the drugs that were not used.” RQ5: “With the focus on avoiding the emergency hospitalisation of dying patients, and the importance of good pain and symptom control in these accounts of a ‘good death’, it was not surprising that many respondents in this study viewed anticipatory prescribing as essential in order to have the drugs available for use ‘as needed.’” RQ5: “All nurses talked about a pain-free and peaceful death as the goal of care, and the ability to administer pain relief was seen as an important element in achieving this. However, the administration of drugs often left nurses feeling uncomfortable, particularly if the patient died soon after their administration.” M M M = M 49 Spilsbury, 2020 LESS COVID- 19: Learning by Experience and Supporting the Care Home Sector during the COVID-19 pandemic: key lessons learnt, so far, by frontline care home and NHS staff Spilsbury, K., Devi, R., Daffu-O’Reilly, A., Griffiths, A., Haunch, K., Jones, L., Meyer J. “LESS COVID-19: Learning by Experience and Supporting the Care Home Sector during the COVID-19 pandemic: key lessons learnt, so far, by frontline care home and NHS staff.” Coventry, National Care Forum (2020). To capture the experiences of frontline care home and NHS staff caring for older people with COVID-19 and to share the lessons learnt about the presentation, trajectories, and management of the infection with care homes that have and have not yet experienced the virus. Phase 1: Interviews with frontline staff (objectives 1 to 3) in June and July to gather in- depth understanding of: • the clinical presentation and illness trajectory of COVID-19 in this population; • what had worked well, or what more was needed, for care and treatment; and • lessons learnt for supporting infected older people to recover or die well. Phase 2: Consultations with senior operational and quality managers in care homes (objectives 4-5) in September to establish: • the resonance, relevance, and any gaps in relation to Phase 1 findings; and • strategies for managing COVID- 19 at an organisational level within the home for the mutual benefit of residents, relatives and staff. RQ1: “The range of interventions used by frontline staff (in both care home and NHS settings) for symptom management included: Pain: small doses of lorazepam and/or oral morphine were prescribed (for some individuals) for pain relief. Frontline staff reported the variable effectiveness of these medicines for managing pain for older people.” RQ5: “Peaks in the virus outbreak can create localised shortages of EOL medicines for care home residents.” RQ3, RQ5: “Care home staff described that, where possible, they had in place anticipatory medicines (prescribed by a GP) for residents, where this was considered appropriate by staff, residents and, when relevant, family members. When these medicines were not in place then this created difficulties of access to, and timely administration of, medicines to manage and control symptoms at EOL. Participants described increased pressure placed on the medicines supply chain during the peak of the COVID-19 pandemic. When medicines were unavailable, then this caused distress for the older person and the staff caring for them. In the early stages of the pandemic, staff were not able to re-purpose medicines or to request a supply of medicines for storage in the care home (in case required). This caused frustration and upset for care home staff.” RQ5: “Frontline staff working in, or with, care homes without nursing (also known as residential homes) described situations where an older person with COVID-19 deteriorated at a rapid rate (covered in the illness trajectory section). Delays in access to a health care professional (for example a community nurse, or a palliative care specialist) often delayed the timely administration of EOL medicines. Some participants reported delays of up to 6 hours.” RQ5: “Participants highlighted the importance of cross sector working to ensure older people received appropriate interventions and care in a timely manner. For example, relationships between a care home and primary and community services enabled access to a wider range of medicines to manage symptoms.” M M M = M Tanghe, 2020 Opioid underuse in terminal care of long-term care facility residents with pain and/or dyspnoea: A cross-sectional PACE-survey in six European countries Tanghe, Marc, et al. "Opioid underuse in terminal care of long-term care facility residents with pain and/or dyspnoea: A cross- sectional PACE- survey in six European countries." Palliative Medicine 34.6 (2020): 784- 794. To examine the perception of barriers and their impact concerning opioid medicines, comparing policy makers, health care professionals working in the field of pain management, palliative care or harm reduction and other stakeholders. A cross-sectional survey describing the dying phase in long- term care facility residents in six European countries. Nurses assessed pain/dyspnoea with Comfort Assessment in Dying with Dementia scale and checked opioid prescription by chart review. Opioid underuse per country and per symptom estimated. Associations of opioid underuse calculated by multilevel, multivariable analysis. RQ3: Estimated percentage of overall opioid underuse by country: 29.3% (95% CI: 16.9–45.8) in England. Calculated as the estimated prevalence of diseased residents with pain and/or dyspnoea in the last week of life. The estimated prevalence of opioid underuse in the last 3 days of life was higher in residents with dyspnoea than in residents with pain. RQ5: Opioid underuse is strongly associated with the absence of symptom assessment. Opioid underuse estimated to be (in England) 33.7% (26.2–42.2) in long-term care facilities (opioid underuse defined as residents without opioid prescription despite pain and/or dyspnoea, 3 days prior to death). Implementing palliative care in long-term care facilities with robust symptom assessment programs and clearer indications for opioid use can contribute to improving symptom management in the dying phase for patients at these facilities. M M M = M Wilson, 2015 Administering anticipatory medications in end-of-life care: a qualitative study of nursing practice in the community and in nursing homes Wilson, Eleanor, et al. "Administering anticipatory medications in end- of-life care: a qualitative study of nursing practice in the community and in nursing homes." Palliative Medicine 29.1 (2015): 60-70. To examine nurses’ decisions, aims and concerns when using anticipatory medications. An ethnographic study in two UK regions, using observations and interviews with nurses working in community and nursing home teams (n = 8). Observations (n = 83) and interviews (n = 61) with community nurses. Coding of field notes and thematic analysis of interviews. RQ5: “Throughout this study, nurses reported working in pairs or liaising with colleagues during the decision-making and administration process. Education, training, experience, and clear access to and good communication with nursing colleagues/managers, GPs, disease-specific and specialist palliative care services are important for building nurses’ confidence and ensuring a safe, comprehensive AM culture in nursing homes. However, a fear about hastening death and a lack of consensus about what the ‘end-of-life’ phase means may restrict the use of AMs in groups of patients with uncertain prognoses such as heart failure or COPD, potentially leaving some with their pain and symptoms under- or untreated.” RQ5: “In accounts reported here, some nurses advocated the use of prescriptions that allowed drugs to be administered within a small range. Yet many did not want or feel it was appropriate for them to have decision-making responsibilities about dose ranges. Some considered such decisions to be emotionally burdensome, especially those with less training and experience. A number of other UK studies have illustrated the emotional strain felt by district nurses providing palliative and end-of-life care. This burden can be heightened by communication issues, limited knowledge, experience of some conditions, a lack of support and time pressures. Nurses working in care homes can often face similar barriers to providing good quality end-of-life care. There is now increasing evidence that symptoms are often undertreated, especially in older patients and in those with conditions other than cancer. The effects of the murders committed by Shipman continue to influence public and professional attitudes.” RQ5: “Administering the medication raised a number of concerns for nurses. First, they were keen to distinguish between pain and H H H = H 50 agitation so as to administer the most appropriate drug but sometimes found this difficult.” RQ5: “Despite all participants stating that they would always start on the lowest prescribed dose of the medication, some raised a fourth area of concern about over medicating patients and this resulting in unwanted side effects.” RQ5: “Nurses acknowledged that although relatives often provided the majority of personal care to patients and had considerable insight into their needs, they took great care not to be ‘unduly’ swayed by relatives’ judgements or requests.” Wilson, 2017 The importance of interdisciplinar y communication in the process of anticipatory prescribing Wilson, E. and Seymour, J. “The importance of interdisciplinary communication in the process of anticipatory prescribing.” International Journal of Palliative Nursing, 23 (3) (2017): 129- 135. ISSN 1357- 6321 To determine the nature and importance of interdisciplinary communication is to the process of anticipatory prescribing and end-of-life care. Ethnography of health care professionals across four care homes and four community sites in England (83 episodes of observation and 72 interviews). RQ5: “When interdisciplinary communication worked well, the AP process was smooth and optimised patient care. However, when communication was impaired between nurses and GPs, care home nurses reported a reluctance on behalf of the GPs to prescribe anticipatory medications and attributed it to a lack of experience and/or training.” RQ5: “Trust, access and clarification of responsibilities were considered to be interlinked and comprise a central component of successful anticipatory prescribing.” RQ5: “Good interdisciplinary communication has been shown to support all elements of the process of anticipatory prescribing.” RQ5: “GPs seemed to be willing to work within a small level of risk and reported being frustrated by nurses continually bringing prescriptions back to be rewritten because of small errors.” RQ5: “Having a clear understanding of each other’s responsibilities in the anticipatory prescribing process was perceived to be an important factor in good interdisciplinary communication.” M L L = L Wilson, 2016 Anticipatory prescribing for end-of-life care in the community: a survey of community nurses in England Wilson, Eleanor, Jane Seymour, and Clive Seale. "Anticipatory prescribing for end- of-life care in the community: a survey of community nurses in England." Primary Health Care 26.9 (2016): 22-27. To gain insight into the roles and experiences of a wide range of community nurses in end of life medication decisions. Postal questionnaires to 1,558 English nursing home, community/district, and palliative care nurses. Collected data subject to quantitative and qualitative analysis. Descriptive statistics and thematic analysis to analyse questionnaire responses. RQ1, RQ5: “Nurses reported working well with GPs and perceived that they had good access to the medications needed. Figure 1 shows that 79.2% (n=427/539) of nurses reported that they ‘infrequently or never’ found doctors reluctant to prescribe anticipatory medication, and 11.1% (n=60/539) reported this as ‘neither frequent nor infrequent’. But it was evident there was some variation. A small proportion, 9.6% (n=52/539), agreed that some doctors were reluctant to provide anticipatory prescriptions...” RQ1, RQ5: “Not all practices in my location are happy to prescribe anticipatory medications, as they feel this is a waste of their budget and that out-of-hours doctors are there to be called upon for the prescribing of anticipatory drugs. In the main, the reluctance of GPs in prescribing is improving.” Manager in a nursing home. RQ5: “Similarly, a few nurses (8.6%; n=45/525) said they ‘always or frequently’ experienced significant difficulties in obtaining the drugs specified in anticipatory prescriptions used in end-of-life care. For a further 11.2% (n=59/525) this was reported as ‘neither frequent nor infrequent’. However, the majority of nurses (80.2%; n=421/525) said they encountered these difficulties ‘infrequently or never’.” RQ5: “I have real problems with GPs prescribing [end of life] drugs, they are usually unable to estimate doses correctly, have no idea how to prescribe when a patient has a fentanyl patch, or is a slow release morphine. This weekend, in 1 day, two prescriptions were written incorrectly.” Palliative care nurse. RQ5: “Nurses reported that the anticipatory medications successfully controlled those symptoms they were intended to relieve in 89.6% of the patient cases they recalled. In a more general sense, most respondents (96.0%; n=504/525) agreed that when anticipatory prescriptions were in place they ‘always or frequently’ enabled respondents to improve the quality of end-of- life care they provided.” M M M = M Seymour, Jane E., 2011 Do nursing homes for older people have the support they need to provide end-of-life care? A mixed methods enquiry in England Seymour, Jane E., Arun Kumar, and Katherine Froggatt. "Do nursing homes for older people have the support they need to provide end-of- life care? A mixed methods enquiry in England." Palliative Medicine 25.2 (2011): 125-138. To identify key factors in the larger health and social care system impacting the quality of end- of- life care provided in English nursing homes. Mixed methods: 2 in-depth qualitative studies of nursing homes; a postal survey of 180 nursing homes surrounding the case study sites. RQ1: “Rural home also reported that accessing GP support, prescribed medication and transferring a resident to the hospital during out of hours was very difficult. Some GPs endeavoured to overcome problems by ensuring medication was pre-emptively prescribed for individuals prior to the weekends, but this did not always provide the solution to unexpected problems among residents.” RQ5: “Participants emphasized the importance of mutual trust between nursing staff and prescribers.” RQ5: “GPs not always keen to issue end-of-life drugs or to visit promptly.” Care home manager RQ5: “And she begged me not to send her into hospital and she ended up going to the hospice because we hadn’t a clue where to get a syringe driver from. The doctor didn’t know anything about the drugs or what we should be using.” Care home manager RQ5: “One or two GPs I think sometimes may be less geared up to end-of-life care than others. So I think that’s a challenge for the staff and we are working on that. I mean, generally, I think things have improved a great deal but there are just a few GPs that do hold back probably more than others . . . I think whether they [don’t] know the patient very well, whether they’re aware of the drugs they need to be prescribing, time probably as well, you know.” Macmillan Nurse RQ5: “. . .Very often the question of end-of-life pathway drugs has been brought to us by the nursing staff [here] and always appropriately, as far as I’m concerned . . . I mean we’re probably more familiar to them here than any other practices because we spend so much time here. So I think that helps really because the more you know people the more you come to trust them, or you could put it the other way, I suppose.” GP RQ5: “Problems for nursing home personnel in accessing support for end-of-life care included variable support by general H H H = H 51 practitioners (GPs), reluctance among GPs to prescribe appropriate medication, lack of out of hours support, cost of syringe drivers and lack of access to training.” Watson, 2006 Barriers to implementing an integrated care pathway for the last days of life in nursing homes Watson, Julie, Jo Hockley, and Belinda Dewar. "Barriers to implementing an integrated care pathway for the last days of life in nursing homes." International journal of palliative nursing 12.5 (2006): 234- 240. To explore the barriers that needed to be overcome during the process of implementing an integrated care pathway for the last days of life as a way of developing quality end-of-life care in nursing homes. Qualitative and quantitative data were collected in eight nursing homes before, during and after the implementation of a care pathway. Documentary analysis, baseline audit of notes around the last days of life; group interviews; field notes; participant observation; and interviews were conducted. RQ1: “When the notes of the last five residents to die in each nursing home before the study were reviewed, it was clear that the loss of the swallowing reflex in a dying resident was seldom anticipated and thus, subcutaneous or rectal medication rarely prescribed. Also the prescribing of prn medication in anticipation of symptoms was infrequent.” RQ1: “Nursing homes are not permitted to hold stock drugs unless they are for a named resident; it was often felt that prescribing prn medication was potentially a waste of money as drugs were destroyed if not used.” RQ1: “The only drug that was occasionally prescribed prn across the nursing homes in the last days of life before the study was oral morphine, whether or not the residents were on a previous opiate analgesic.” RQ5: "Nursing homes are not permitted to hold stock drugs unless they are for a named resident; it was often felt that prescribing prn medication was potentially a waste of money as drugs were destroyed if not used. However, the financial implications of this are negligible compared to the unnecessary length of time a resident has to wait if prn medication is not available. The only drug that was occasionally prescribed prn across the nursing homes in the last days of life before the study was oral morphine, whether or not the residents were on a previous opiate analgesic." M L L = L Kinley, 2010 A baseline review of medication provided to older people in nursing care homes in the last month of life Kinley, Julie, and Jo Hockley. "A baseline review of medication provided to older people in nursing care homes in the last month of life." International journal of palliative nursing 16.5 (2010): 216- 223. To understand the medication needs of the very old and frail in the last weeks of life and how they might differ from a model of care developed for people dying from cancer Questionnaires to 67 trained nurses, medication reviews of 48 deceased residents’ notes in the last month of life in seven nursing care homes, and coding. Quantitative and thematic analysis. RQ1: Regarding end-of-life care symptom management requirements of frail older people in care homes: "Syringe drivers were used in 23% of cases; however, only three residents required a syringe driver for more than a day and a half and nurses’ confidence and competence in setting up syringe drivers varied. The use of syringe drivers may not be the most suitable way of managing end of life symptoms in very frail and old people." RQ1: "Out of the 11 residents who had a syringe driver in the last days of life, eight of the syringe drivers were in place for less than 1.5 days. That indicates that in the last days of life, symptom control needs of older people may be more appropriately managed through the use of bolus subcutaneous medication or rectal suppositories." RQ1: "As syringe drivers are rarely used in nursing homes, nurses working in these settings lack competence in setting up a syringe driver." RQ1: “Provision and administration of medication in NCHs varied considerably. In NCH-A, for example, the GP reviewed residents regularly, there was recognition when residents were dying, information was well recorded, residents’ relatives were contacted and the situation discussed with them and the full nursing team. Even though preparations were changed to liquid, no injectable medication was prescribed in anticipation of the resident becoming unable to swallow. In contrast, in NCH-D, nine out of 10 residents were prescribed injectable medication for symptom control when they became unable to swallow their oral medication.” RQ1: “There was evidence that variation occurred by prescriber preference. For example, some residents with no previous analgesic history were prescribed PRN morphine sulphate 1.25 mg as a starting dose and others 2.5 mg. Glycopyrronium and morphine sulphate were the most frequently used subcutaneous medication.” RQ3: “The use of injectable medication by residents with non- malignant disease was small (n=7).” RQ3: “Eleven residents additionally had parenteral medication administered as a continuous 24-hour infusion via a syringe driver. What is of note is the length of time residents received medication via the syringe driver. Eight of the 11 residents (73%) received medication via this route for 1.5 days or less. These figures suggest that for the majority of NCH residents the use of PRN medication to control residents’ symptoms in the last few days may have sufficed.” RQ3: “The use of injectable medication most often occurred where the NCH was supported by a specialist palliative care service.” RQ5: “The majority of nurses stated that they did not have difficulties in obtaining medication for residents in the last days of life. However, nurses from two of the homes (NCH-D and NCH-E) found that prescriptions for medication got delayed. These delays were at different levels: delay with prescribing, lack of immediate availability at pharmacy, and difficulty with delivering medication to the NCH.” RQ5: “There was also considerable variety in the RGNs’ recent experience of caring for a resident receiving medication via a syringe driver and in education regarding its use. While 11 nurses had looked after a resident with a syringe driver recently (within the last 3 months) for 11 others it had been over 1 year ago.” M M M = M 52 Breen, 2018 DOOP Kit, Domestic Bin Or Watery Grave? A Study Investigating Disposal Practices Of Transdermal Drug Delivery Products In Care Homes Breen, Liz, et al. "DOOP Kit, Domestic Bin Or Watery Grave? A Study Investigating Disposal Practices Of Transdermal Drug Delivery Products In Care Homes." (2018). To 1) gain insight into current practices of health care professionals regarding opioid transdermal patches (OTPs) (fentanyl and buprenorphine) disposal practices and 2) identify knowledge and system awareness surrounding the disposal of these products in care home settings. Google forms / paper questionnaires sent out to the nurses of 85 care homes; 36 online responses, 20 paper responses, for a total of 56 responses. Descriptive statistics and thematic analysis to analyse questionnaire responses. RQ4: "More education and training is necessary regarding safe disposal practices of OTPs, informed by pharmacist-led interventions, in order to minimise confusion, reinforce safe disposal practices and support the reduction of unsafe disposal practices (domestic waste or flushing. Study results indicate that 52.8% of care staff were unaware of the recent safety update issued in July 2014 by the MHRA with regard to fentanyl patches, which suggests a gap in communication between health care professionals and external governing bodies, a lack of effective intra-company communication strategies, or both." M M M = M Table 2.4 Number of papers included in the published synthesis With regard to End-of-Life Controlled Drugs (CDs) in care homes in the United Kingdom: Review Question Number of papers answering each review question For whom and by whom are end-of-life CDs prescribed? How are CDs stored and monitored? For whom and by whom are CDs administered? How are CDs disposed of after a resident’s death? What are the views of care home staff and other health and social care professionals concerning these issues? What are the views of residents and their family members concerning these issues? 10 papers: 1 high, 7 medium, 2 low WoE No papers identified 5 papers of medium WoE 1 paper of medium WoE 9 papers: 2 high, 5 medium, 2 low WoE No papers identified Box 2.3 Addressing the review questions ________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 1) For whom and by whom are End-of-Life Controlled Drugs prescribed? Ten studies addressed this review question (Finucane et al., 2014; Kinley & Hockley, 2010; Kinley et al., 2016; Morris & Hockley, 2013; Owen et al., 2016; Perkins et al., 2016; Seymour et al., 2011; Spilsbury et al., 2020; Watson et al., 2006; Wilson et al., 2016; 111-119). One study, a review of eighty deceased resident notes from eight care homes, found up to 53% of residents in three nursing homes were symptomatic in the final days of life (pain, agitation, secretions, dyspnoea, nausea and vomiting) and required access to end-of-life medicines including midazolam and morphine. In one nursing home, of those who had anticipatory medicines, only 36% of residents had them prescribed prior to the study’s anticipatory medicines stocking intervention: rather than doctors prescribing then pharmacists dispensing end-of-life medicines to residents when they became symptomatic, Morris 53 and Hockley (2013) worked alongside nurse managers and local pharmacists on a strategy to implement tailored stocks of end-of-life medicines in the participating homes. If prescribed for a symptomatic patient by an out-of-hours GP, care home staff experienced difficulty obtaining the medication (Wilson et al., 2016). Some GPs tried to overcome this difficulty by ensuring that medicines were pre-emptively prescribed for individuals prior to weekends, but this did not provide a solution for residents who deteriorated unexpectedly (Seymour et al., 2011). Another study found that of 77 residents who died in eight care homes, 54% had anticipatory medicines prescribed and 15% had prescriptions for all four nationally recommended anticipatory medicines. The study noted considerable variation in frequency of AP practice: 100% of residents in one care home and 13% in another (Finucane et al., 2014). Prescribing practices also varied regarding the frequency of prescriptions and the doses of anticipatory morphine for residents with no prior analgesic history (Kinley et al., 2010). During the last month of life, residents were primarily prescribed oral medication for symptom control; for example, oral and/or transdermal analgesia were prescribed for 84% of residents, while only 37% of residents had anticipatory PRN injectable medication prescribed (Kinley et al., 2016). During the last days of life, the only oral medication that was occasionally prescribed to be administered as required (PRN) across nursing homes in one study was oral morphine, whether or not residents were on a previous opiate analgesic (Watson et al., 2006). Syringe pumps have been reported to be used in around a quarter of dying nursing home residents, usually for less than the final 24 hours of life (Kinley & Hockley, 2010). In all ten studies the prescriber was a General Practitioner; occasionally anticipatory medicines were also prescribed by hospital teams on discharge or palliative care community specialists (Kinley & Hockley, 2010). While most nurses felt that GPs were in general willing to prescribe anticipatory medicines, one large-scale survey found 20% of nurses reported some doctors to be reluctant to prescribe anticipatory medicines, leading to difficulty in obtaining these medicines in a timely manner (Wilson et al., 2016). 2) How are End-of-Life Controlled Drugs stored and monitored in UK care homes? No studies were identified that addressed this review question. 3) For whom and by whom are End-of-Life Controlled Drugs administered? Five studies addressed this question: in each, care home nurses administered the CDs (Kinley & Hockley, 2010; Kinley et al., 2016; Morris & Hockley, 2013; Perkins et al., 2016; Spilsbury et al., 2020). Kinley et al. (2016) found that while up to 53% of the residents in three nursing homes were 54 symptomatic at the end of life, not all symptomatic residents were prescribed or administered the medicines that the researchers assessed were needed from a post-death notes review. In one care home, 16% of symptomatic residents died without anticipatory medicines either prescribed or administered (Kinley et al., 2016). In a cross-sectional survey study assessing pain and dyspnoea in long-term care facility residents, nurses reported opioid underuse in the last three days of life to be higher in residents with dyspnoea than those with pain (Kinley & Hockley, 2010). In another study of 11 nursing homes, nurses commonly administered CDs including midazolam (for agitation), and morphine, oxycodone or diamorphine (for pain); residents were reported to have been ‘more comfortable and settled’ (Perkins et al., 2016). Kinley and Hockley (2010) found that the majority of residents required three or fewer injectable medications for symptom control in their last few days of life. Injectable medications were more likely to be used when the nursing homes were supported by a specialist palliative care service (Kinley & Hockley, 2010). A report from an ongoing project found that during the initial stages of the COVID-19 pandemic, care home staff experienced difficulties in accessing medicines to manage and control symptoms at end-of-life for residents. ‘Timely administration’ was especially difficult during the pandemic peaks, creating localised shortages of end-of-life medicines for residents (Spilsbury et al., 2020). 4) How are End-of-Life Controlled Drugs disposed of after a resident’s death? Only one study addressed this question (Breen et al., 2018). A survey of 56 care homes concerning the disposal practices of fentanyl patches found over half of staff (53%) to be unaware of regulations concerning appropriate disposal of fentanyl patches (Breen et al., 2018; NHSE, 2013). The authors suggested this indicates gaps in communicating changes in regulations to care homes, deficits of in- company communication strategies, or both. 5) What are the views of care home staff and other health and social care professionals concerning these issues? Nine studies addressed this question (Kinley & Hockley, 2010; Morris & Hockley, 2013; Perkins et al., 2016; Seymour et al., 2011; Spilsbury et al., 2020; Wilson et al., 2015; Wilson et al., 2016; Wilson & Seymour, 2017). One survey of 180 care home managers found consensus that morphine, midazolam, and haloperidol are essential drugs for symptom alleviation in the last 48 hours of life (Seymour et al., 2011). A survey of nurses found strong support for the view that AP was essential in both avoiding emergency hospitalisation of dying patients and effective symptom management (Perkins et al., 2016). Despite this recognition of their importance, two studies reported persistent challenges in securing 55 the necessary prescriptions and support from GPs, obtaining medicines both at short notice and out of hours, and accessing syringe pumps (Seymour et al., 2011; Wilson & Seymour, 2017). Problems included access to these medicines out of hours, prescribing delays, medication availability at pharmacy, access to syringe pumps, variable GP support and reluctance of some GPs to prescribe (Seymour et al., 2011; Wilson & Seymour, 2017). In a study of eight nursing homes, staff reported inadequate end-of-life symptom management when anticipatory medicines had not been prescribed, lack of knowledge of end-of-life symptoms and medicines, and difficulties recognising the dying phase (Watson et al., 2006). Care home nurses and visiting community nurses report that while they did not want to leave symptoms under-treated, they also worried that anticipatory medicines might cause over-sedation, and frequently started residents on the lowest prescribed dose (Wilson et al., 2015). If the resident died soon after administration of these medicines, less experienced community nurses and care home staff were left uncomfortable and some worried that the medicines may have hastened death (Wilson et al., 2015). Care home staff view training from local palliative care teams as essential in preparing frontline staff and promoting their confidence, particularly in medication administration (Watson et al., 2006). The importance of close team-working relationships between care home staff and GPs in end-of-life care is repeatedly emphasised (Kinley & Hockley, 2010; Morris & Hockley, 2013; Perkins et al., 2016; Seymour et al., 2011; Spilsbury et al., 2020; Watson et al., 2006; Wilson et al., 2015; Wilson & Seymour, 2016; Wilson & Seymour, 2017). Watson et al. (2006) identified the lack of multidisciplinary team working in nursing homes as a barrier to implementing integrated care pathways for the last days of life in nursing homes. Spilsbury et al. (2020) emphasised the importance of close working relationships between care homes and community services which enabled access to wider ranges of end-of-life care medicines. Reluctance by some GPs to prescribe anticipatory medicines is a cause of frustration for care home staff (Seymour et al., 2011; Wilson et al., 2016; Wilson & Seymour, 2017). Nurses reported that GPs’ attitudes towards prescribing anticipatory medicines were influenced by the potential wastage costs of unused medicines (Watson et al., 2006). Interviews with care home staff during the early stages of the COVID-19 pandemic revealed frustrations that end-of-life medicines could not be accessed or re-purposed (Spilsbury et al., 2020). Pressures on the medicines supply chain during the pandemic caused distress for both residents and staff members (Spilsbury et al., 2020). The introduction of changes in legislation in April 2020, permitting medicine repurposing in England, sought to address these concerns, although it is unknown to what extent this new process has been used (DHSC, 2020). 56 In a study of four care homes that implemented an intervention to keep a stock of end-of-life care medicines in-home, staff reported feeling safer knowing that they could access the stock in the event of an emergency (Morris & Hockley, 2013). One nurse manager reported that being able to access the stock ‘transformed’ the practice of her nursing staff, who grew comfortable requesting prescriptions for dying residents knowing that the medicines would not be wasted (Morris & Hockley, 2013). 6) What are the views of residents and their family members concerning these issues? No studies have investigated this review question. ________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 2.6 Discussion and narrative synthesis of content This is the first review to systematically identify and synthesise the literature concerning the use of End-of-Life Controlled Drugs in UK care homes. Searches of nine databases, developed with a professional medical librarian (I.K.), and supplemented by journal hand-searches robustly identified the published literature. The limited literature does not address all the review questions and is of mixed Weight of Evidence. Notably, there were no studies which investigated the storage and monitoring of End-of-Life CDs in care homes, or the views of residents and their family members concerning these issues. Only one study addressed the questions regarding the disposal of End-of- Life CDs following a resident’s death. At times it proved difficult to separate Schedule 2 and 3 CDs such as morphine and midazolam from other analgesics such as codeine, dihydrocodeine, and tramadol in papers. Further, it proved challenging to separate evidence pertaining to research questions 1 and 3: for whom End-of-Life CDs were prescribed and to whom they were administered. For example, one study in which nurses reported opioid underuse did not specify whether the underuse was a result of under-prescribing or under-administration of End-of-Life CDs (Tanghe et al., 2020). It was noted that most included studies did not distinguish between End-of-Life CDs and anticipatory medicines. In practice, this is not problematic because the majority of prescriptions for anticipatory medicines will include Schedule 2 CDs. Due to the diversity of CD regulations internationally the review findings are limited to the United Kingdom. Care homes are an increasingly important setting for palliative and end-of-life care; they are the location of approximately a quarter of all deaths in England and Wales, a proportion that increased markedly during the first wave of the COVID-19 pandemic (ONS, 2019a; ONS, 2019b). They provide care for many of the most vulnerable members of the community, often with advanced physical and cognitive frailties, for periods of many months. The COVID-19 pandemic has highlighted the extreme vulnerability to infection of care home residents and the challenges that the current division between health and social care provision in the UK presents for optimal care at the 57 end of life. Recent UK legislation seeks to ensure that care home residents, as all other patients, will have access to NHS specialist palliative care services when needed (DHSC, 2022). Spilsbury et al. (2020) emphasised the compounded difficulties in accessing end-of-life medicines during the pandemic, underscoring the necessity for robust, adaptive systems capable of withstanding healthcare crises. 2.6.1 Patchwork prescribing The review highlights the heterogeneity in CD prescribing practices across care homes. For instance, Finucane et al. (2014) detailed a stark contrast between homes where all residents had anticipatory CDs versus homes where such provision was markedly low. This variation underscores the need for standardised guidelines which balance the flexibility required for individual patient needs with consistency in care quality. The evidence suggests a patchwork of approaches to prescribing, with certain homes demonstrating proactive strategies, as seen in the end-of-life CD stocking intervention by Morris & Hockley (2013). Conversely, the study by Kinley et al. (2016) reveals gaps where symptomatic residents lacked timely access to CDs, pointing to a critical need for integrated systems that ensure swift responsiveness to patient symptoms at the end of life. 2.6.2 Challenges of clinical uncertainty Difficulties in recognising the dying phase in frail older care home residents with non-cancer illness (Alcorn et al., 2020; Bowers et al., 2022; Murray et al., 2005), associated with concerns that CDs used for end-of-life symptom management might hasten death (Wilson et al., 2015; Bowers et al., 2019), are frequent causes for concern for care home staff. Prognostication in advanced frailty and extreme old age is very challenging and requires multi-disciplinary team assessment and decision- making and sensitive conversations with residents (where possible) and their families, with acknowledgement of the uncertainties involved (Alcorn et al., 2020). For older people there is limited discussion about the dose and frequency appropriate for symptom control using CDs, especially if they have not been taking such medicines previously (Wilson et al., 2016; Wilson & Seymour, 2017). This exacerbates known staff anxieties about using CDs at the end of life in social care settings where there is no on-site medical cover (Seymour et al., 2011; Wilson et al., 2016). On CD disposal practices, Breen et al. (2018) suggests a widespread lack of awareness and knowledge that contributes to clinical uncertainty and staff anxiety surrounding CD use in care homes, signalling a need for clearer communication, information dissemination, and education on disposal policies. Collectively, the studies underscore an urgent need for targeted training and support that 58 equips care home staff with the confidence to administer CDs effectively and keeps them abreast of regulatory changes and best practices. 2.6.3 Challenges in interprofessional working Close working relationships with GPs are highly valued by staff; when absent, or when GPs are reluctant to prescribe CDs ahead of need, symptoms are often poorly managed at the end of life. Wilson et al. (2016) and Seymour et al. (2011) reflect on the symbiosis between nurses' expertise in patient care and GPs' decision-making authority in prescribing: strong relationships with a supportive GP practice are of great importance to ensure that skilled clinical assessments are made with shared management decisions and that end-of-life CDs are available for administration in a timely manner. Where these are present, excellent symptom control and end-of-life care can be achieved in most cases: where absent, residents may suffer and care home staff struggle. The synthesis suggests that fostering strong collaborative networks could alleviate some of the challenges in securing and administering CDs. 2.6.4 Implications for research, policy, and practice The current literature concerning the use of end-of-life CDs in UK care homes is limited and of mixed quality. The studies included in the synthesis demonstrate that prescribing is primarily by GPs with substantial variation in prescribing practices between GPs, while administration is primarily by nurses. However, nurses frequently report feeling inadequately trained in the use of CDs. Thus, this review highlights the need for training and support for nurses in the administration of CDs, particularly end-of-life anticipatory medicines such as morphine. Additionally, this review highlights notable gaps in the evidence base concerning the storage, monitoring, disposal, and repurposing of end-of-life CDs. The absence of research concerning these areas in care homes, as well as the perspectives of residents and their family members on these issues, is a pressing research need. Additionally, several challenges related to symptom identification, prescribing, and administration of end-of-life CDs were reported, with limited inquiry and investigation into their underlying causes. Moreover, given the diversity of professionals involved in end-of-life CD processes, there is a pressing need to further investigate the perspectives of the wider, relevant range of health and social care professionals. Understanding their experiences and challenges can provide a more nuanced view of end-of-life CD processes. Thus, for the development of informed and effective policies, it is essential that future research not only addresses these gaps but integrates the feedback of both patients and providers. This will ensure that end-of-life care processes are shaped by the very experiences of those they aim to serve. 59 2.6.5 Limitations of the review The literature review has several limitations to consider. Firstly, the reliance on a single individual (MM) for the initial stage of title screening may have introduced bias into the title selection process, although measures were taken to mitigate this, such as consulting a medical librarian to inform the search strategy and following Cochrane guidelines to guide the inclusion of the second reviewer (BB). This reliance was a deliberate choice, given that the systematic review was conducted as part of my PhD. Secondly, the review's concentrated approach on UK care homes restricts broader generalisability, as global contexts differ in end-of-life care practices. Furthermore, while the review’s focus on the selected end-of-life CD processes is crucial, it may eclipse other important considerations of end-of-life medication systems, such as procuring the medicines, issues of access, and the emotional and psychological impacts of medicines management. Lastly, the adoption of Gough’s WoE Framework for quality assessment involves subjectivity on the part of the reviewers in determining the quality of studies included. It is paramount to weigh these limitations when interpreting the study's results and planning further research. Thus, while the systematic reviews are thorough and methodologically sound as a result of my comprehensive multi-level search strategy, these potential limitations suggest areas for refinement in future research and a need for readers to interpret the findings with these considerations in mind. 2.7 Conclusion There is a professional consensus that CDs are an essential part of end-of-life care for people dying in care homes. The prescribing, storage, monitoring, administration, and disposal of CDs, particularly AMs, are standard practice in care homes, though associated with critical challenges. The limited available evidence suggests that access to this important component of symptom management is variable and dependent on practitioner preference and whether staff have the necessary knowledge, support, and confidence to administer medicines. Established policy and practice lacks evidence concerning the crucial stages of storage, administration, and disposal of CDs, as well as the perspectives of residents and their families. Importantly, the COVID-19 pandemic highlighted challenges related to the 'timely administration' of end-of-life medicines (Spilsbury et al., 2020). Given the rapid deterioration associated with the virus, the prompt delivery of end-of-life CDs and AMs became even more critical where residents exhibited distressing end-of-life symptoms, as delays can profoundly impact a resident’s end-of-life experience and cause frustration and upset to care home staff (Antunes et al., 2022; Spilsbury et al., 2020). Local shortages or reliance on external health care professionals for drug administration (such as community nurses or palliative care specialists, as in residential care homes) can result in 60 challenges to timely end-of-life symptom relief for care home residents (Antunes et al., 2022; Spilsbury et al., 2020). The swift nature of COVID-19-related deterioration thus emphasises the need for timely access to and administration of end-of-life CDs, namely by way of appropriately prescribing in advance of clinical need (i.e. anticipatory prescribing). In summary, the review findings emphasise the challenges of anticipatory processes and the importance of multidisciplinary teamwork in this area of care home care; highlight the limited or absence of evidence concerning many aspects of the prescribing, storage, monitoring, administration, and disposal of end-of-life CDs, the experiences of various health and social care professionals involved in these processes, and the views of residents and family members; and reveal important challenges that health and social care professionals encounter working within the AP system. This dissertation seeks to explore those experiences and challenges to better understand the AP system in operation, and to build on the evidence base by addressing several identified knowledge gaps through a qualitative interview study with health and social care professionals working within the AP system in care home settings. 61 Part two: Methodology & Methods Chapter 3 Methodology & methods This chapter outlines my methodology, data collection and analysis methods, followed by research ethics, reflections on my role as the interviewer, and final remarks. 3.1 Introduction and research questions The literature review highlighted the current limited research, of mixed quality, on the prescribing, administration, storage, monitoring, and disposal of end-of-life CDs, particularly anticipatory medicines, in UK care homes. The experiences of residents, family members, and health and social care professionals concerning these issues are largely unreported to date. My research agenda planned to address several of these topics, building on the literature, with a focus on the experiences of health and social care professionals involved in the AP system and the challenges they encounter while navigating it. To understand and address these challenges, this study aims to explore the tensions which arise from complexities and contradictions in the AP system, experienced by health and social care professionals working in care home settings, with the following research questions: 1. What underlying systemic tensions and what wider interprofessional contexts underpin issues related to the wastage and shortage of anticipatory medicines in care homes? 2. What role do the algorithmic and discursive structures associated with care home-based AP play in shaping the tensions between individualization and standardisation in end-of-life care? 3. How do the structural logics underpinning AP paperwork contribute to the tensions between navigating bureaucratic regulations and providing good, patient-centred end-of-life care in care homes? In 3.2 to 3.5, I provide a comprehensive account of my methodology (3.2-3.3), methods (3.4), and analysis (3.5) used in this study. Firstly, I outline my ontological and epistemological positions (subjective reality and constructivism) in 3.2, followed by a discussion of the selected theoretical perspective (paradox) in 3.3. These sections provide a rationale for the philosophical and theoretical approaches taken as well as for the methods, and I compare them to other potential approaches to 62 justify their use in addressing my research questions. I go on to describe my research design (3.4) and methods of data collection (3.5). In 3.6, I detail my method of analysis (Reflexive Thematic Analysis supported by paradox theory). In the following sections, I provide information about ethical and personal issues relevant to the study (3.7) and reflections on my role as the interviewer (Box 3.1), concluding with final remarks (3.8). 3.2 Ontology & epistemology: Ontology of subjective reality and epistemology of constructivism Research paradigms are philosophical perspectives that shape a researcher's approach to understanding the world around them (Kuhn, 1962 & 1996). This involves two key beliefs: ontology, which is about what a researcher believes to be the nature of reality, and epistemology, which refers to how a researcher believes knowledge is created (Scotland, 2012). These perspectives guide each aspect of the research, from the methods used to how the results are analysed (Guba & Lincoln, 1994). Kuhn (1962 & 1996) emphasised that an individual’s observations and subsequent interpretations are influenced by their past experiences, entrenched beliefs, and the sociocultural milieu within which they are embedded. Central to this worldview is the ontological perspective that reality is subjectively created and the process of research is not impartial: the researcher and their participants influence the knowledge created and meanings made (MacKenzie & Knipe, 2006). Epistemologies including constructivism and interpretivism align with this particular ontological perspective. Contstructivism is rooted in the notion that knowledge is not a fixed construct; as observers interact with their environments, their understanding evolves, reflecting a continuous process of construction and reconstruction of knowledge (Bryman, 2016). Interpretivism posits that knowledge is created when observations are filtered through our personal and collective lenses, giving them meaning and significance (Ritchie & Ormston, 2003). The worldview that upholds the subjectivity of reality can be juxtaposed with the contention that reality is objective (Goldenberg, 2006); an objective conceptualisation of reality forms the bedrock of positivist epistemology. Positivist researchers may find that knowledge can be acquired and created independently of personal biases or subjective interpretations on the part of the researcher (Crossan, 2003). They maintain that knowledge is of unbiased, value-free, measurable, and replicable nature (Crossan, 2003; Mays & Pope, 1995). Drawing from my investigative journey, my philosophical position strongly resonates with the constructivist paradigm. The constructivist paradigm posits that reality and knowledge creation stem from individual perceptions and experiences, influenced by an interplay of personal, social, and cultural factors (Bryman, 2016). In this paradigm, reality is understood to be constructed rather than 63 discovered, emphasising the subjective interpretations and negotiations between participants and researchers, and encouraging meaningful examination of these complexities, contradictions, and diverse perspectives to co-create richer understandings of the subject under study (Bryman, 2016). Within the context of UK care homes, for instance, my research focuses on how the views and experiences of health and social care professionals affect the AP system. Here, the observed tensions in the system represent the constructed realities that these professionals navigate in their daily practice. Insofar as the AP system, as experienced by health and social care professionals, amounts to a labyrinth of constructed realities, it is important that research seeks to understand their experiences and challenges, and thus the system in operation. This co-construction of knowledge helps to expand our understanding of the system’s complexity and identify strategies to improve end-of-life care quality in care home settings. 3.3 Theoretical perspective: Paradox Paradox theory is aligned with the constructivist paradigm (Smith & Lewis, 2011); it offers a lens that is premised on understanding the tensions and forces operating in systems, and how they coexist in an interconnected web of understanding (Smith & Lewis, 2011). Developed by sociologists Wendy K. Smith and Marianne W. Lewis (2011) in the discipline of organisational studies, paradox theory draws from seminal works on paradoxes, notably those by Bateson et al. (1972) and Koestler (1964). Bateson et al. introduced the concept of ‘double binds’ to emphasise contradictions as core aspects of systems (Bateson et al., 1972), while Koestler's ‘bisociation’ highlighted the coexistence of seemingly incompatible perspectives (Koestler, 1964). These foundational insights are central to paradox theory, suggesting that tensions and contradictions are not merely problems awaiting resolution, but inherent aspects of systems, able to yield both positive and negative outcomes (Smith & Lewis, 2011). In paradox theory, Smith and Lewis define the concept of paradox as ‘contradictory but interrelated elements that exist simultaneously and persist over time’ (Smith & Lewis, 2011, p. 382). Such elements can include competing demands, objectives, interests, and viewpoints, which, when in conflict, generate tension (Miron-Spektor et al., 2018). Tensions can catalyse ‘virtuous cycles’, fostering creativity, resilience, and sustainability; however, they can also cause ‘vicious cycles’ that challenge individuals’ sense of stability and order, leading to defensive behaviours and decision paralysis (Vince and Broussine, 1996; Miron-Spektor et al., 2018). Smith and Lewis (2011) posit that gaining insight into these conflicting elements and their resulting tensions can furnish the conditions for more effective management of them - and a better understanding of the system at large. Paradoxes are not only theoretical constructs; they surface regularly, contravening established logics in everyday life (Papachroni & Heracleous, 2020). Determining how they do so is important 64 in paradox theory: while each contradictory element of a paradox in isolation may seem rational and logical, their co-existence can appear counter-intuitive. Health care contexts are marked by various paradoxes resulting from overt contradictions, ambiguities, conflicts of interests in and varying interpretations of complex systems (Farrant, 1991; Kennedy, 2015). Contradictions in health care emerge, for example, when one considers that on one hand, modern advancements lead to improved health and longevity, and on the other, the increased prevalence of chronic diseases associated with such longevity (such as dementia) presents a challenge which health care systems are often ill-equipped to manage (Oliver et al., 2014). Furthermore, ambiguities, differences in interpretation, and conflicts of interest abound in social care contexts such as care homes, where managers must often balance both clinical and administrative responsibilities (Warmington, 2014; Orellana, 2017). These dual imperatives can cause conflict between patient-centric values and the demands of funding bodies or governmental directives (Warmington, 2014). The resulting tension is evident in managers’ uncertainty concerning the identity of their primary client, with both patients (and their well-being) and regulators (and their need for compliance) vying for primacy (Warmington, 2014; Orellana, 2017). While tensions stemming from organisational paradoxes have been well-documented across various sectors at the macro-level (McHugh et al., 2023; Miron-Spektor et al., 2018; Smith & Tushman, 2005), the micro-level nuances – specifically, the human and contextual factors underlying individuals’ responses to these paradoxes – remain underexplored. In the domain of health care, few studies grounded in paradox theory seek to identify these systems-level paradoxes and better understand the human and contextual factors that determine individuals’ responses to them. The utility of paradox theory in the health care domain, while not focused on the AP system or care homes, has been demonstrated by Cleland et al. (2018) and Erthal et al. (2021). Cleland et al. revealed key tensions in the modern system of medical education, pinpointing the challenges of aligning a standardised curriculum with the changing needs of individual students in a rapidly evolving health care landscape (Cleland et al., 2018). Their findings underscored the importance of acknowledging and addressing paradoxes to facilitate a more adaptive and responsive system of education. Meanwhile, Erthal et al. (2021) identified paradoxes within a health care organisation during its transition to a new managerial modality. The authors highlighted how the nuances of organisational culture could play dual roles – as barriers or enablers – during such transitions. Through their research, paradox theory emerged as both a theoretical scaffold and a practical toolkit for managers attempting to understand and improve outcomes in their organisations by confronting and navigating key paradoxes. 65 Although paradox theory has been embraced in health care systems research, as in Cleland et al. (2018) and Erthal et al. (2021), it has not yet been applied to care home studies. Care homes, operating as distinct subunits within the health care landscape, exhibit specific challenges and paradoxes. These paradoxes often surface due to the interplay of conflicting elements, such as the pursuit of organisational efficiency and the need for individualized end-of-life care. Given their primary demographic of older individuals with complex care needs, care homes often face the challenge of providing personalised attention in the face of operational constraints (Kane et al., 2007). Investigating these paradoxes and grounding the exploration in scholarly research can yield insights that can inform and improve end-of-life care provision in these settings. Furthermore, while existing literature has not identified care homes or the AP system as paradox- laden domains, this thesis argues for this point of view. A nuanced understanding of the system’s paradoxes is crucial for a thorough grasp of how it operates ‘on the ground’. Paradox theory offers a relevant analytical lens, with its emphasis on co-existing tensions (Lewis, 2000). To distil these tensions, juxtaposing the perspectives of key health and social care professionals involved in the AP system is necessary – well-suited to my choice of a qualitative interview method, which I expand upon later in this section. Disparities between the real-world execution of AP by these professionals and its theoretical framing in policy and paperwork can breed such tensions. By anchoring the study in paradox theory, this may be further explored and provide insights into the nature of the system, and how professionals can navigate paradoxes within it. While paradox theory can offer valuable insights, it can also present challenges as it lacks a unified understanding and consensus among paradox theorists, as identified by Smith and Lewis (2011). This dissonance is clear in the varied language used to describe paradoxes. In this thesis, the term paradox is specifically applied to describe the interrelated and contradictory elements of the AP system in care homes. Another point of contention centres around the ontological nature of paradoxes and whether they are innate features of systems or social constructs born out of lived experience (Clegg et al., 2002; Smith & Lewis, 2011). Smith and Lewis (2011) advocate for a dual perspective, suggesting that paradoxes are both systemic features and experienced by individuals working in a system. My analysis in Chapters 5 to 7 supports and engages with this perspective. In considering alternative theoretical perspectives, paradox theory emerged as the most fitting choice—particularly for the analysis of my qualitative dataset and for writing up the discussion—in comparison to, for example, grounded theory and interpretive phenomenological analysis (IPA). While alternative approaches were explored earlier (in section 1.11), it is worth considering them further here, given the predominance of grounded theory and IPA as qualitative research approaches. Grounded theory, which involves understanding social relations in underexplored contexts and 66 generating new ‘explanatory theories’ in response to the data (Bryant and Charmaz, 2007), is misaligned with my epistemology, that meaning is co-created by the researcher and participant, as grounded theory does not recognise the key role that researchers play in meaning-making (Bryant and Charmaz, 2007). Furthermore, while grounded theory focuses on deriving new theories, the AP system contains existing structures and processes which require a lens that can accommodate and elucidate the inherent contradictions which permeate it. Similarly, while IPA explores the lived experiences of individuals to ‘explore, describe, interpret, and situate the participants’ sense making of their experiences’ (Tuffour, 2017), phenomenological methods do not centre the ‘hows’ or ‘whys’ underlying these experiences (Smith & Osborn, 2007). My research explores certain tensions in the AP system and why they arise and persist. Given this, the constructivist paradigm combined with paradox theory can be considered a more suitable framework for my study. Rather than cataloguing the lived experiences of my participants without exploring the underlying dynamics, as per Tuffour (2017), the layered nature of the AP system demands a theoretical approach like paradox theory guided by Smith and Lewis (2011) to distil its tensions. Paradox theory, in alignment with the constructivist paradigm, thus serves as the theoretical frame for the analysis of my empirical research. An inductive approach and qualitative interview strategy proves well-suited to this frame, as such a design facilitates an in-depth exploration of health and social care professionals' perspectives and experiences related to the AP system, and an expanded understanding of the various paradoxes inherent within it. An inductive, or ‘bottom up’ approach, aims to generate understanding – the data collected drive the development of explanations for patterns identified in the dataset (Bryman, 2016; Saunders, Lewis & Thornhill, 2019). A key strength of this study’s inductive approach and qualitative interview strategy is that it is well-positioned to collect and compare a wide range of accounts about the AP system, supporting the cross-examination of various professional roles and the identification of paradoxes arising from differences in perspective. While paradox theory can be applied to both qualitative and quantitative research, the decision to favour one over the other, or to integrate both in a mixed-methods framework, depends on the specificity of the research questions and the nature of insight sought. The mixed-methods design, upheld for its potential to offer a comprehensive perspective, presumes that the combination of qualitative and quantitative paradigms can yield richer insights than either method individually (Tariq & Woodman, 2013). One could argue for the deployment of mixed methods tools, such as questionnaires or surveys, to determine the prevalence and types of AMs in care homes prior to conducting interviews. However, findings from Finucane et al. (2014) and Kinley et al. (2016), as highlighted in my systematic review, already provide quantitative evidence of variation in the number of residents prescribed AMs across different care homes and in the types of AMs prescribed. 67 Notably, these quantitative assessments do not explore nor explain the 'hows' or 'whys' underlying these observed variations, which constitutes the essence of this thesis. This underscores the rationale for a qualitative interview-based design as the optimal approach to investigating the AP system and addressing my research questions. This approach complements existing studies that have identified variation in the numbers of residents prescribed AMs in UK care homes (Finucane et al., 2014; Kinley et al., 2016; Morris et al. 2013), the types of AMs prescribed (Finucane et al., 2014), and the doses at which they were administered (Morris et al., 2013; Tanghe et al., 2020). Thus, interviews, as opposed to surveys or questionnaires, provide the depth required to explore complex and often opposing elements, such as bureaucratic structures that simultaneously uphold individualization (through documented principles) and standardisation (through set protocols) of anticipatory prescriptions. Ultimately, anchoring the research in a qualitative interview design grounded in paradox theory and aligned with the constructivist paradigm supports a thoughtful, rigorous study of systemic tensions in AP, thereby supporting my aim of expanding understanding of the system in operation and providing actionable recommendations for AP policy and practice. 3.4 Methods: Research design 3.4.1 Introduction Data collection was qualitative in nature, comprised of one-to-one semi-structured interviews. I adopted an exploratory approach aligned with my constructivist epistemology, emphasising the subjective and context-bound nature of individuals’ experiences and reality, while allowing for adaptable and iterative data collection that facilitated the emergence of themes and identification of systemic tensions. By interviewing a range of health and social care professionals, I was able to gain a nuanced understanding of the paradoxes experienced within and inherent to the AP system in care home settings. Data analysis was conducted through a constructivist paradigm (Braun & Clarke, 2020), using Braun and Clarke’s (2006 & 2019 & 2021) six-phase approach to reflexive thematic analysis (RTA) underpinned by paradox theory (Smith & Lewis, 2011; Suter, Irvine & Howorth, 2022). My constructivist alignment and paradox perspective are consistent with RTA, which Braun & Clarke (2006) define as a method rather than a methodology, not inherently restricted to any single epistemological or theoretical perspective (Braun & Clarke, 2006). This approach is expanded on in subsequent sections of this chapter (3.5.3, 3.5.4, 3.5.5), which introduce RTA, provide a rationale for the method, and detail the RTA- and paradox-informed analysis. Importantly, RTA values the 68 subjectivity of researchers’ perspectives, allowing their insights to enrich data analysis (Braun & Clarke, 2006 & 2019 & 2021). Insofar as my epistemological position centres the subjectivity of knowledge creation and highlights individuals’ unique perceptions, contexts, and experiences (Bryman, 2016), researcher subjectivity proves to be a key component of my analytic process. This perspective is consistent with RTA, in which researchers undertake an active role in reflecting on their role in knowledge generation (Braun & Clarke, 2006 & 2019 & 2021). The outcomes of my analysis are represented by the paradox themes, which constitute the core findings of this study (Chapters 5-7). 3.4.2 Public and Patient Involvement The data collection method was informed by qualitative methods literature, namely Brinkmann & Kvale (2009), guidance from the supervisory team, and insights from the Palliative & End-of-life Care group in Cambridge (PELiCam) and a Patient and Public Involvement (PPI) group. PPI was important for establishing the relevance of this research, by integrating the viewpoints of those directly affected by AP. PPI is defined as a dynamic partnership between members of the public and researchers, often through meaningful engagements or public consultations (NIHR, 2019). Engaging with relevant stakeholders, particularly those directly impacted by the research topic, is important to ensure the relevance and impact of the study (Mockford, 2012). To guide PPI in this study, I sought guidance from established resources such as the NIHR (2019) and the Public Health Agency (2022), which offered information concerning PPI in care home research. In this study, my aim for PPI was to ensure that the research design was informed by the collaborative insights of those directly affected by the AP system in care home settings - namely patients and their family carers. Their perspectives were integrated into various phases of this study, including its design and the dissemination of results. Regarding the role of PPI in research development, various aspects of the qualitative study were influenced by a PPI meeting organised in December 2019, coordinated by my PhD supervisor (SB). Participants in the meeting included family carers of care home residents who provided feedback that influenced the development of the initial research objectives, participant recruitment strategy, study design, and the language of participant information sheets for CHMs, identification of ethical issues specific to care home research, and the development of the interview topic guide (Appendix 8). A notable addition, as a result of their input, was a dedicated sub-topic and series of questions aimed at understanding how health and social care professionals involve family carers and residents in decisions and discussions about AP. 69 As part of an ongoing engagement process, I plan to organise one further PPI event in September 2023 to discuss the key findings of my research and their implications and involve PPI collaborators in developing dissemination materials such as leaflets for care homes. This event will provide an opportunity for meaningful dialogue to help ensure that the outcomes of my study are effectively communicated and understood by the public. There is a continued need to identify and circulate best practices for PPI, especially in the context of care home research. Establishing meaningful collaborations and increasing opportunities for stakeholder participation ensures that research remains rooted in the experiences and needs of those directly impacted (NIHR, 2019). 3.4.3 Setting The study took place in care homes across the UK. While interviews were conducted remotely (given the pandemic context), the participants were based in care homes which varied in size, organisational structure, accreditation, and location, ranging from urban to rural environments. Recruitment extended beyond an initial geographic focus in Cambridgeshire to encompass health and social care professionals working in care homes across the UK by leveraging online resources and professional networks. This diversity in setting brought insights into the workings of the AP system in different care environments across the UK. 3.4.4 Population Participants eligible for this study were care home managers, care home nurses, and visiting health and social care professionals (including doctors, community nurses, pharmacists, and ambulance personnel) working in UK care homes. The inclusion criteria were clearly defined to ensure a focused and relevant sample. All participants were: adults aged 18 and over, proficient in English to provide informed consent and participate in interviews, and directly involved in the care of residents where anticipatory prescribing of end-of-life care medications was considered or had been implemented. The exclusion criteria excluded those not meeting those requirements, such as non- adults, those unable to consent or communicate effectively in English, and individuals not involved in care where AP is practiced. 3.4.5 Sampling Following principles of purposive sampling (Creswell, 2013), I identified care homes across the UK with good or outstanding Care Quality Commission inspection reports by searching online. I obtained the contact information of the respective managers from their organisational websites in the public domain. I sought to recruit approximately 10 to 15 CHMs, 10 to 15 CHNs, and 10 to 15 70 visiting health and social care professionals, as per my study protocol (Appendix 6). However, following recruitment challenges detailed in Chapter 4, I adapted my sampling method with reference to Patton’s (2015) ‘snowball’ or ‘chain’ strategy and enlisted the support of regional and national professional contacts through colleagues in the Queen’s Nursing Institute and Enabling Research in Care Homes (ENRICH) network, who circulated my Participant Information Sheet to their respective networks thus broadening the recruitment reach. This sampling strategy focused on achieving meaningful variation across stakeholders and care home organisations, based on factors such as professional roles and care home types, to ensure that the dataset could support robust interprofessional and interorganisational comparisons. Consistent with Braun and Clarke’s (2021) concept of ‘data adequacy’, the number of participants recruited was ultimately based on the substantial quality of data relevant to the study’s aims, rather than the quantity of data collected. Informed by the expertise within my supervisory team during the protocol development phase (Appendix 6), we projected that a sample comprising roughly 10 to 15 CHMs, 10 to 15 CHNs, and 10 to 15 health and social care professionals who routinely visited care homes would yield ‘data adequacy’ and thus facilitate a robust analysis. The final sample comprised 42 participants working in 29 care homes across the UK: 19 CHMs, 10 CHNs, and 13 visiting health care professionals. 3.4.6 Recruitment The recruitment for the study was carried out in two main phases. In Phase 1, CHMs were the initial point of contact. An introductory email was sent to these managers, providing an overview of the research 'Anticipatory Prescribing in UK Care Homes' with a Participant Information Sheet attached (Appendix 2). The managers who expressed interest in response to this initial email were sent a follow-up email to arrange interviews. In situations where managers were identified through chain referrals via contacts in professional networks, the email was sent after they reached out to participate in the study. Where CHMs were willing and available, I scheduled and conducted an online Zoom-based interview - to ensure their safety in light of the COVID-19 pandemic. To initiate Phase 2, care home managers who completed the interview were asked to share the study details with potential participants: care home nurses, visiting doctors, pharmacists, and other professionals. My contact information and Participant Information Sheet (Appendix 2) were provided to CHMs to distribute on my behalf. Interested parties then contacted me directly to participate. This direct approach was consistent with principles of criterion-based selection, wherein all participants fulfilled specific criteria relevant to the study’s research questions (Creswell, 2013). Prior to the interview, participants were provided with the Participant Information Sheet (Appendix 71 2) and consent form (Appendix 3), to establish informed consent and to address any ethical considerations in recruitment, including the avoidance of coercion in line with the GMC’s principles of process consent (GMC, 2021). To acknowledge the time given by participants, I offered a £30 Amazon voucher. The ethical issues associated with informed consent, the provision of vouchers, and other aspects of recruitment are discussed in section 3.7. 3.5 Methods: Data collection 3.5.1 Interviews The effectiveness of research is substantially influenced by its design (Gray, 2014) – particularly when investigating complex, understudied topics such as the AP system in care home settings. To achieve the core objective of expanding understanding of the AP system – and the challenges faced by health and social care professionals working within it – interviews emerged for me as an optimal research design for several reasons. They can provide an in-depth understanding of individual experiences, which is crucial when studying complex systems (Kvale & Brinkmann, 2009). They allow researchers to reflect on participants’ experiences as well as their own, and as an inherently qualitative method, an interview-based research design supports reflexive thematic analysis (Braun & Clarke, 2006 & 2019) and a paradox perspective. Unlike questionnaires or surveys, interviews – particularly semi-structured ones – offer flexibility to ask follow-up questions, enabling the researcher to elicit detailed accounts which might otherwise remain obscured and pursue emergent themes or ask for clarification, which is helpful when conducting exploratory research (Gray, 2014). Interviews, then, provide a platform for professionals to contextualise their experiences. Insofar as the AP system in care homes is intertwined with various individual, institutional, and socio-cultural factors, individual professionals’ accounts shed light on nature of the challenges they encounter and the causes underlying them (Denzin, 2001). Additionally, when a topic is understudied in the literature, professionals in the field may feel their experiences are invalidated or overlooked; thus, interviews serve as a platform for these voices to be heard - recognising and validating their experiences, and contributing to a more comprehensive understanding of the field in question (Seidman, 2013). The constraints of the COVID-19 pandemic further reinforced the choice of interviews. While observational methods such as ethnographies offer similar advantages and rich and grounded data, their feasibility during the pandemic was compromised since pandemic restrictions prevented the adoption of methods that required physical presence. Thus, interviews emerged as the most effective and viable method, ensuring the continuity of the research process while respecting health and safety guidelines. 72 Grounding the approach in Kvale and Brinkmann’s (2009) and Saunders, Lewis, and Thornhill’s (2019) conceptualisation, I understood interviews to be purposeful conversations where the researcher is strategically positioned to ask questions and actively listen to interviewees’ experiences on a given topic. Building on this and my own constructivist, relativist worldview, I approached my interviews as co-constructive processes, enriched by the reciprocal dynamics between researcher and interviewee (Kvale & Brinkmann, 2009; Saunders, Lewis & Thornhill, 2019). This view resonates with Denzin’s (2001) ideas about the co-production of knowledge during interviews, and the significance of the interviewer's role in shaping that knowledge. This approach was underscored by the use of open-ended questions, which were intended more to harness rich experiential accounts than to extract specific answers. Rabionet outlines the usefulness of such questions in semi-structured interviews to capture the authentic voices and associated sense- making processes of interviewees (Rabionet, 2011). While semi-structured interviews are guided by a set of questions, the flow of the dialogue can adapt organically based on the interviewees’ contributions (Gray, 2014; Rabionet, 2011). This fluidity was important for this research, as the AP system in care home settings is largely understudied in the literature and the underlying causes of the many challenges that professionals have been documented to experience while doing this work (Bowers et al., 2019; Majumder et al., 2022; Teggi, 2023; Wilson et al., 2017) have yet to be explored. These challenges identified by the findings of my literature review (Chapter 2) thus served as an anchor, spotlighting key topics that were then explored during the interviews. This adaptability is an advantage of semi-structured interviews, which allow for the interviews to be informed by existing literature (Gray, 2014; Rabionet, 2011). For example, insights from the literature review revealed that care home nurses are often challenged by decisions surrounding the administration of AMs (Chapter 2). To expand my understanding of the underlying causes of these challenges, the interview guide included questions prompting participants to describe cases where decisions regarding AM administration were difficult, followed by probing questions about why these situations were difficult. Beyond these outlined focus areas, however, I ensured participants had ample room to discuss issues that mattered most to them. The interview topic guide, then, while providing a framework, was not restrictive and supported freedom of discussion (Appendix 8). Interviews thus focused on understanding health and social care professionals’ views and experiences working with and within the AP system in care home settings, and exploring challenges related to the prescribing, dispensing, procuring, administration, storage, monitoring, and disposal of AMs. Interviews were conducted remotely at times based on participants’ availability and preferences. Care home staff were mainly interviewed during their breaks at work, while visiting 73 health care professionals were largely interviewed from their homes in the evenings (except for one doctor who was interviewed from his office). Interviews lasted between 45 and 75 minutes, with a mean duration of 60 minutes. Care home professionals often consulted patient notes during interviews, allowing them to describe experiences related to particular residents with a considerable level of specificity. Interviews of care home staff at times elicited strong emotional responses, particularly when participants recounted events around ‘bad deaths’ – in these situations, I took actions to ensure the safety and well-being of the participants, as detailed in section 3.7.6: Protecting research participants from harm. Conducting and analysing these interviews posed additional challenges in managing the emotions of participants and interpreting the data: I took appropriate measures to safeguard the well-being of both the participants and myself informed by best practice guidelines and the literature, further described in section 3.7: Research ethics. 3.5.2 Reflections on Zoom interviewing As previously mentioned, due to the COVID-19 pandemic, it was not possible to conduct in-person research such as face-to-face interviews, as had been my original plan. In collaboration with my supervisory team, then, I redesigned my methods and shifted to online Zoom interviews - as online modes of data collection were ideally suited to sensitive qualitative research during the pandemic (Braun et al., 2020). A growing body of literature highlights both the benefits and drawbacks of online interviewing. Thunberg et al. found that during the pandemic, digital interviews facilitated long-distance participation in research, yielded rich data regarding sensitive topics, and was cost- effective for researchers (Thunberg et al., 2022). The challenges that I faced in conducting virtual interviews resonated with the findings of Thunberg et al. (2022). There were logistical difficulties from the video-conferencing platform: two participants had unstable internet connections in their care homes, resulting in interviews interrupted by intermittent disconnections and muffled audio recording quality. These interviews were particularly taxing and led me to confront the difficulties and limitations associated with virtual interviews. I also encountered an initial sense of self-consciousness upon observing myself on the screen during early interviews, although as I conducted more interviews, I became more accustomed to the experience and was better able to concentrate more intently on the discussion. Another limitation of my remote interviews was the absence of physical presence, an aspect of online interviewing not often discussed in the literature. Pierre argues that ‘presence’ is the cornerstone of the interviewee-interviewer relationship in qualitative research (Pierre, 2008). Presence involves being in together-ness, or meeting the interviewee where they would like to be met, through the modalities of conversation, observation, and empathy. Pierre has stated: ‘We believe that our face- 74 to-face interactions with people make our work especially valid, we are present in our research, in the thick of things, talking with and observing our participants. Qualitative inquiry is not distant; it's live and in-person; it happens right now’ (Pierre, 2008). I was aware that the absence of immediate and tangible human presence on Zoom made it challenging to immediately connect with and build trust with my interviewees. Additionally, conducting Zoom interviews involved a degree of performative expressiveness on the part of myself and my participants, limiting display of body language, eye contact, and social cues. Pandemic restrictions had led to many participants describing feelings of isolation from their social networks, including family, friends, and colleagues. As a result, several participant appeared to regard me as a confidant with whom they could discuss their work-related challenges. The interview presented an opportunity for human connection with a researcher, one who was eager and willing to listen. Furthermore, as a newcomer to the country without a patient care role, it appeared that the one-off experience of the interview generated a ‘strangers on a train’ effect - a situation in which individuals feel more comfortable sharing personal thoughts in anonymous and brief encounters with strangers they do not expect to meet again (McKenna & Bargh, 2000). The combination of my newness and the pandemic-induced isolation of my participants may have made them more comfortable in sharing their experiences with me, despite the lack of tangible presence. 3.5.3 Recording Although Zoom has a built-in recording option, my prior experience with several practice recording sessions proved its unreliability. Therefore, I used a separate, disconnected phone recorder that was placed out of sight from the interviewees and near my computer speakers for audio recording interviews. 3.5.4 Transcription The audio-recorded interviews were transcribed verbatim by a professional transcriber. Upon receiving the transcripts, I listened back to all of the interviews and compared the transcripts against the audio recordings. Braun and Clarke (2006) caution against indulging the temptation to ‘skip over’ this phase, highlighting the importance of revisiting the recordings for a fuller contextual understanding of the dataset. During this process, I found the transcripts to be significantly disadvantaged in that they lacked some of the nuance and emotions conveyed during the interviews. As Brinkmann and Kvale (2007) have noted, the transcripts were ‘impoverished, decontextualised renderings of live interview conversations’ (p. 93). 75 I therefore annotated the transcripts orthographically, inserting the inflections, breaks, pauses, laughs, and tears, which were important for context-building for my thematic analysis (Braun & Clarke, 2013). I also revisited reflections in my field notes to mark the transcripts with information that could provide further context and ‘colour’. I removed filler words such as ‘erms’ and ‘uhs’ to optimise for coherence as opposed to linguistic accuracy (McLellan, 2003). 3.5.5 Summary My research enabled me to obtain health and social care professionals’ views and experiences of AP in care home settings, an area that has been little studied to date. I now move on from a description of the study methods to describe my data analysis process. 3.6 Methods: Data analysis 3.6.1 Reading At the outset of my PhD, I engaged in extensive reading on AP, end-of-life care, UK care homes, and sociological texts that related to the broader UK health and social care system. I also undertook the systematic literature review exploring processes related to end-of-life CDs in care homes presented in Chapter 2. The reading period also offered a better understanding of the environments in which these professionals worked; it helped to situate me within the complex milieu of the UK health and social care sector (given my limited prior exposure in it as a newly arrived graduate student from the US). With the onset of the COVID-19 pandemic, it was no longer possible for me to shadow health care professionals or volunteer in care homes as I had originally planned. The reading became an invaluable source of insight and empathy into the work and lives of my participants; it was especially helpful in identifying and recognising points of contradiction within and between accounts, particularly when professionals were recalling and reflecting on experiences of ‘bad deaths’. The reading proved particularly beneficial during the six-month period (March to August 2020) when international students were required by my college to leave campus premises and return home, in my case to New York in the height of the first wave of the COVID-19 pandemic. During this 6-month period of enforced absence from Cambridge, I kept in close contact with my supervisors and conducted the systematic literature review presented in Chapter 2. 3.6.2 Timeline Table 3.1 outlines the timeline of my PhD and the sequence of events and engagements that took place. Different but related tasks were performed concurrently, as the nature of my research could 76 not be neatly compartmentalised. For example, although I formally began writing this dissertation in January 2022, my preparatory work during reading and departmental and conference presentations significantly influenced my writing process. Moreover, the iterative and cyclical process of analysis was shaped by both my writing process and ongoing engagement with the literature. Table 3.1 PhD tasks by date Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sept Year 1 2019 2020 Reading Volunteering / Shadowing Initial ethnography study preparation (abandoned due to COVID-19 pandemic) New interview study development Return to New York (COVID-19 pandemic) Systematic literature review * * Year 2 2020 2021 Reading Interview study ethics Interview study recruitment Data collection and analysis Systematic literature review * * Year 3 2021 2022 Reading Data analysis * Writing * * Year 4 2022 2023 Data analysis Writing/Editing Writing/Editing * *=presentations to members of the Public Health and Primary Care department, PELiCam, PPI groups, and the public 3.6.3 Reflexive Thematic Analysis: An introduction & rationale 77 This research, situated within the constructivist paradigm and my relativist worldview, aims to expand understanding of the AP system in care homes. Given this methodological position, my decision to employ RTA as the method of analysis is well-founded. However, in the earliest stages of my study, I read widely across the methods literature to inform my approach. I considered constructivist grounded theory (Charmaz, 2006) for its potential suitability to the complexities of the AP system and alignment with my epistemological perspective (Wooley, Butler, and Wampler, 2000). Further reading and discussions with my supervisors led me to realise that my approach may inadvertently align more with what Braun and Clarke (2006) referred to as ‘grounded theory “lite”’ rather than the more robust, ‘“full-fat” grounded theory’, which they believe to be rarely employed in its intended form (p. 81). Moreover, grounded theory, with its emphasis on novel theory generation, did not fit with my study’s objective: to explore the AP system and the challenges faced by health and social care professionals working within it. IPA was another potential method I considered, given how it resonated with my philosophical position by centring the subjectivity of participants and their experiences in the analysis. I initially felt that phenomenological approaches would not meaningfully emphasise the researcher's impact in the analytic process, but further discussions with my supervisory team helped me differentiate between this more limited conceptualisation of phenomenology, labelled by Willig (2012) as ‘descriptive phenomenology’, and ‘interpretative phenomenology’ which accommodates researcher’s reflexivity and interpretation of participant experiences. During my reading, I found that that despite their distinctive labels, many qualitative methods are ‘variations of the same method’ (p. 10, Timulak & Elliott, 2019). The use of ‘brand-name’ methods such as IPA is occasionally believed to lend credibility to research; however, its use will not necessarily produce richer, more impactful insights in comparison to other qualitative approaches (Elliott & Timulak, 2005). Strict adherence to a specific qualitative brand may in fact restrict a researcher’s flexibility and reflexivity which could otherwise promote innovation in their analysis (Elliott & Timulak, 2005). This knowledge supported my decision to undertake thematic analysis within a constructivist framework. Thematic analysis provided a structured yet adaptable approach to guide my analysis within my epistemological frame and clear guidelines to provide a ‘rich and detailed, yet complex’ analysis of my data (p. 78, Braun & Clarke, 2006). It is important to point out that it is an analytic method rather than a methodology and is not restricted to any singular theory or epistemology: specifically, thematic analysis is defined as ‘a method for systematically identifying, organizing, and offering insight into patterns of meaning (themes) across a data set’ (p. 57, Braun & Clarke, 2012). To distinguish their method from other types of thematic analysis (as discussed below), Braun and Clarke termed it ‘reflexive thematic analysis’ in a seminal 2019 paper. Importantly, they acknowledge that RTA’s relative independence, while advantageous, can also 78 expose it to critiques for potentially lacking in philosophical commitment (Braun & Clarke, 2021). Thus, I was careful to align my thematic analysis with my paradox perspective, reflecting carefully on the theory that shaped my method to facilitate meaningful and rigorous engagement with the data and avoid the pitfalls of an ‘anything goes’ approach that occasionally characterises theoretically unmoored thematic analyses (Braun & Clarke, 2018). A defining aspect of RTA is its acknowledgment of the researcher’s position and contribution throughout the analysis. Rather than viewing subjectivity as a potential bias to be reduced, RTA considers the researcher's own positionality – their experiences, pre-existing knowledge, and social positions – as a crucial resource. By 'critically interrogating' these aspects, researchers harness their unique lens to gain deeper insights into qualitative data (p. 5, Braun & Clarke, 2021). Reflexivity in RTA demands the recognition of knowledge as situational – that is, derived from the interaction between the researcher and the data. This approach is geared towards recognising the influence of the researcher’s values and beliefs on data interpretation (Braun & Clarke, 2021). The understanding that researchers do not uncover themes in data passively, but actively engage in their identification, construction, and reporting resonated with me as a researcher, and aligned with the constructivist, relativist underpinnings of this research in its view towards the co-construction of knowledge and the iterative nature of qualitative inquiry (Braun & Clarke, 2006 & 2021; Bryman, 2016). RTA emerged, furthermore, as the best fit for my research objective and questions over other thematic analyses. Among the diverse approaches to thematic analysis, including RTA (Braun & Clarke, 2021, 2021b & 2022), coding reliability (Boyatzis, 1998), and codebook (King, 2012), each possesses distinct characteristics. RTA as a method recognises and even champions the role of the researcher's subjectivity and their role in data construction (Braun & Clarke, 2006 & 2019); this method emphasises the need for researchers to continually reflect and be aware of their influence on knowledge construction (Braun & Clarke, 2006 & 2019). Coding reliability thematic analysis uses a codebook to foster consistency across multiple analysts, striving for a sense of objectivity (Braun and Clarke, 2020; Boyatzis, 1998). In contrast to RTA, this approach does not emphasise the researcher's subjective involvement in the analysis. Codebook thematic analysis approaches data with pre-established areas of interest, often setting out to test specific hypotheses (King, 2012; Gale et al., 2013), and is therefore incongruent with this research’s core objective and methodology. Thus, RTA emerges as a fitting choice, given the congruence of its principles with the study’s methodology. While RTA offers numerous advantages in terms of depth and adaptability, it is important to also consider its limitations. The inherent flexibility of RTA, though advantageous in many contexts, may render analysis unclear (Boyatzis, 1998; Braun & Clarke, 2006 & 2022). Braun and Clarke 79 (2006) acknowledge this, noting potential criticisms of thematic analysis due to its perceived lack of clarity and epistemological infidelity. Further studies echo this sentiment, highlighting that the pronounced role of the researcher can sometimes lead to ambiguity in analysis and interpretation (Attride-Stirling, 2001; Finlay, 2021). Recognising these limitations, careful attention has been given to ensuring transparency and reflexivity throughout my analysis. Ultimately, RTA – specifically Braun and Clarke’s (2006 & 2019 & 2021) six-phase approach – offers a robust framework to understand and untangle the experiences and challenges of health and social care professionals in the AP system, given its alignment with the study's philosophical underpinnings and meaningful consideration of my influence as the researcher. 3.6.4 Doing RTA The RTA approach taken followed Braun and Clarke’s (2006 & 2019 & 2021) six-phase framework, guided by a paradox lens (Smith & Lewis, 2011; Suter, Irvine & Howorth, 2022). These phases included: familiarisation; coding; initial theme generation; theme development and review; defining, refining, and naming themes; and writing up the results (Braun & Clarke, 2006 & 2019 & 2021). Later in this section, I will describe this process in detail within a table (Table 3.2) for the purpose of clarity. These phases were iterative rather than linear. Additionally, coding was primarily done by hand and facilitated by the qualitative data management software NVivo 12 (QSR International Pty Ltd., 1999-2021). Figure 3.1 presents an example of my coding. Figure 3.1 Example of coding [Coding extract from the transcript of the interview with Dr. Kennedy, Consultant in Palliative Medicine, Participant 8] This analysis was inductive (data-driven), ensuring that my interpretations remained rooted in participants’ accounts. In the ‘Findings & Discussion’ chapters, the story of the analysis is written in such a way as to highlight their experiences, with direct quotations. Recognising the influence of 80 my own experiences, perceptions, and theoretical orientations (Johnson & Rasulova, 2017), I aimed for a reflexive stance, further detailed in section 3.6 of this chapter and a reflexivity statement in Chapter 8. To further strengthen the analysis, I locate my findings within the wider literature and the policy environment. Such triangulation, recommended by Braun & Clarke (2013) and Terry et al. (2017), ensures the findings resonate with both the dataset and broader academic conversations. Embracing the iterative nature of qualitative research, I kept a reflexive journal to support the development of my research and evolution of my themes. This practice, described by Braun and Clarke (2019) as a ‘continual bending back on oneself’, enriched the analysis by fostering deeper, more nuanced engagement with the data (p. 594). Excerpts from my reflexive journal at different stages of the research process including analysis can be found in Appendix 9. Importantly, data collection took place over four months and thus I began analysing the transcripts before data collection was complete. This allowed me to iterate the interview topic guide to explore insights identified in earlier interviews. For example, the interview guide largely focused on gathering experiences related to individual AP processes. During familiarisation, I noticed that participants described the processes with respect to the challenges they faced in carrying them out. Thus, as the interviews progressed, I sought to better understand the nature of the challenges described, as reflected in the core objective of this thesis. It is important to acknowledge, furthermore, that undertaking research that is largely exploratory in nature may requires reference to literature and theory that extends beyond the scope of the literature review, to provide a more comprehensive understanding of the topic (Stebbins, 2001). These connections serve as intellectual anchors for such research (Stebbins, 2001). Morgan et al. underscored this, noting a significant gap in theoretical insight within current end-of-life care research, suggesting that such perspectives may enrich the field (Morgan et al., 2020). In response to this, Part 3 of this thesis (Findings & Discussion) draws on and extends relevant sociological theory in this domain of end-of-life care research. 3.6.5 The paradox lens applied The challenges faced by professionals in the AP system, as highlighted in the literature review (Chapter 2), found resonance in the qualitative interview dataset with participants' accounts. Beyond corroborating the previously identified challenges, the data shed light on an important pattern: participants were not merely describing challenges but were actively grappling with systemic tensions. In this analysis, I demonstrate that these tensions are symptomatic of key paradoxes in the AP system. 81 The sociological and organisational literature further supports this understanding: complex systems, by their very nature, often incubate paradoxes (Miron-Spektor et al., 2018). Such paradoxes can, in turn, generate tensions (Smith & Lewis, 2011). These tensions manifest palpably in individuals as challenges, stressors, anxiety, and/or decisional dilemmas (Putnam et al., 2016). The recurring tensions in my dataset and their resonance with the wider literature led me to refine my analytical approach and adopt a paradox lens, inspired by Suter, Irvine and Howorth’s (2022) study which draws on Smith and Lewis’s (2011) paradox perspective. The authors explore the challenges of business managers addressing employee mental health issues using a tension-based lens (Suter, Irvine & Howorth, 2022). My analysis adapts this approach (see Table 3.2), highlighting the tension between conflicting elements of the AP system to surface the paradoxes within it. The paradox lens and a constructivist, relativist approach A constructivist epistemology holds that knowledge is a natural and ever-evolving precipitate of the process by which observers interact with their environments rather than a fixed construct (Bryman, 2016; Theys, 2017). This epistemology proved especially accommodating to a paradox lens, since an understanding of paradox begins at the point where apparently fixed and secure systems of knowledge end or break down, revealing their internal contradictions (Lewis, 2000; Smith & Lewis, 2011). Both the constructivist viewpoint and the paradox lens operate on the premise that knowledge is not static but fluid, always in a state of flux and informed by individuals’ lived experiences and interactions (Bryman, 2016; Theys, 2017; Lewis, 2000; Smith & Lewis, 2011). For example, while the AP system appeared to be founded upon a fixed epistemological structure made up of guidance and policy documents which aimed to consolidate all the knowledge (both specialist expertise and more practical knowhow) needed to navigate the care home space, participants revealed that this guidance was very challenging to implement without substantially reconstructing its meaning in line with their daily realities. My analysis is anchored in an understanding that their knowledge of the care home space was constructed as part of a continuous engagement with that space and with the guidance, but never limited to either one. This dynamic, where fixed guidelines demand continuous reconstruction in real-world contexts, underscores the paradoxes inherent in such systems. It becomes clear, then, that a paradox lens, focused on identifying and exploring these inherent contradictions, is well-suited to a constructivist approach. The paradox lens and RTA 82 Paradoxes are often an inevitable consequence of organisational systems (Lewis, 2000; Smith & Lewis, 2011; Gaim et al., 2022). A complex system such as AP, which exists at multiple levels (e.g. the levels of the regulator, prescriber, pharmacist, and administering nurse) can often breed paradoxes when two groups of professionals interpret the same idea differently (Lewis, 2000; Smith & Lewis, 2011). For instance, the concept of 'anticipation' was perceived differently by doctors and nurses in my study. While doctors equated it to rationing and deemed it a nurse's duty, nurses associated it with proactive prescribing and looked to doctors for direction. This chasm in understanding is revealed in a critical comparison of professional perspectives in the analysis (Chapter 5). In assessing the evidence provided of such paradoxes, which came through in my interviews, I was compelled to reflexively examine my own role in the process, understanding ‘the level of the researcher’ as a crucial aspect of my research process (p. 197, Thomas, 2004). In fact, it was only at this level that an understanding of these paradoxes was possible, since my interviews allowed me to cut across the different levels, analysing GPs, CHNs, and pharmacists alike, thus allowing me to cross-examine viewpoints and perspectives and note discrepancies between them. I also recognised the possibility of social desirability bias; that is, the potential that much of what participants reported during the interviews may have been influenced by a desire to make a certain impression on me as the researcher (Chung, 2003). RTA holds that the presence of the researcher should be viewed not as an impediment to objectivity but a crucial resource (Braun & Clarke, 2021), and I was able to appreciate that many of the paradoxes which emerged in my data may not have come into focus if not for my role as a researcher. Interviewing professionals at different levels of the AP system allowed me to relate what I had learned from one type of professional to another type of professional, thus stimulating valuable discussion concerning the differences between the two types of professionals, in which they reflected seriously on the complexities of their interrelationships. This sort of cross-examination of viewpoints may not occur organically in the absence of a researcher’s questioning and collating them together. RTA thus supports the paradox lens, which, in turn, enriches the analysis and offers a way of understanding the challenges that professionals face, the tensions they grapple with, and their underlying paradoxical dynamics. Smith & Lewis (2011) observed that such tensions are the very fulcrum on which systems innovate and evolve. By exploring the interplay of tensions in the AP system, we can better understand it and promote its continued development. Table 3.3 presents a list of themes which offers simple listing of the themes (paradoxes) and subthemes (tensions). 83 Table 3.2 Phases of data analysis, adapted from Braun & Clarke (2006 & 2019 & 2021), Smith & Lewis’s (2011) paradox perspective, and Suter, Irvine & Howorth’s (2022) tension- based lens Phase of RTA (adapted from Braun & Clarke, 2006 & 2019 & 2021) Methodological steps (adapted from Braun & Clarke, 2006 & 2019 & 2021) Doing RTA with a paradox perspective (adapted from Smith & Lewis (2011) and Suter, Irvine, & Howorth’s (2022)), with a worked example from this analysis Familiarising myself with the data • In this first phase of data analysis, I immersed myself within the data, ensuring I grasped the ‘depth and breadth of the content’ (p. 16, Braun & Clarke, 2006). This involved delving into the transcripts by repeatedly reading and reviewing them, cross-checking them with the audio files, anonymising transcripts, and listening again to audio recordings. • Additionally, I wrote notes on initial observations and reflections on emerging patterns helped to organise my analysis (Braun & Clarke, 2021). • Noticing contradictions: Immersion in the data revealed contradictions within and between participant accounts about AP processes that signalled deeper underlying paradoxes. • For example: during this phase, I noticed that when participants spoke of ‘challenges’, they often grappled with conflicting demands of the AP system. Such challenges resonate strongly with Smith and Lewis’s (2011) concept of paradox-induced tensions ‘pulling entities in seemingly opposite directions’ that manifest in challenging experiences for individual workers (Smith & Lewis, 2011; Miron-Spektor et al., 2018), underscoring the potential for a paradox theory perspective. • Through this deep engagement, the applicability of the paradox lens became clear, in guiding my understanding of the connections between participants’ experiences of challenges and tensions in the AP system, and the paradoxes underlying them. Coding • The coding process was inductive, with an emphasis on segmenting the data into meaningful expressions and describing them in either a single word or a short sequence of words to highlight participant meanings. Figure 3.1 presents an example of my coding. • Open coding was used: I did not use pre-set codes, but created and modified codes as I worked through my coding process. • I used both semantic (descriptive) and latent coding, with neither approach prioritised over the other. • A data-driven approach: A paradox lens during coding involved being sensitive to words, phrases, or paragraphs in the transcripts that suggested competing demands or contradictory elements in the data. These were the most compelling parts of the dataset, as the tensions participants wrestled with consistently animated them. These tensions thus became the foundation for the themes and subthemes represented in Chapter 5-7, and are reflected in the research questions in Chapter 1. • The importance of open coding: Open coding made it possible to accommodate a paradox lens because the codes could change and develop – if they were pre-set and I noticed a contradiction or tension as I coded the transcripts, this would not be reflected. • An example from my analysis: Some codes included: ‘Repetition in form-filling for AM disposal’, ‘Confusing medicines administration records’, ‘Policy advocating for individualized AP’, ‘Inadequate time for direct patient care provision’, and ‘Relationship with patient guiding end-of-life symptom assessment’, suggesting tension between distinct elements competing for participants’ time. Generating initial themes • Codes were grouped based on their deeper (latent) meanings, capturing what was important or interesting about the data. From this, initial themes were • Generation of earliest themes: As the earliest themes were developed from interconnected codes, navigating their many and loosely defined relationships made understanding challenging. The paradox lens brought them into focus. 84 generated, reflecting Braun and Clarke’s (2021) conceptualisation of a theme as ‘a pattern of shared meaning organised around a central concept’ (p. 76). • The following questions from Braun and Clarke (2021) guided initial theme generation: ‘Does this provisional theme capture something meaningful? Is it coherent, with a central idea that meshes the data and codes together? Does it have clear boundaries?’ (p. 84) • Exploring relationships: The paradox lens thus illuminated the tensions within and between the earliest themes, highlighting those in opposition to one another due to inherent contradictions. As such, the initial themes foreground tensions between contradictory elements. • An example from my analysis: As related codes were organised, the early theme 'excessive paperwork and policy' was developed from codes such as 'repetition in form-filling for AM disposal' and ‘time-consuming documentation’. Another group of related codes pointed to an early theme 'policy ideal of patient-centred care’. Coded separately, these aspects of my data appeared contradictory, since the time taken up excessive paperwork would detract from the time needed to provide patient-centred care. I realised that contradictions of this sort were not a result of a fault in my coding process but rather because certain contradictory states of affairs were present in the AP system at large. - In organising my codes into themes, then, I actively avoided choosing themes that would paper over these contradictions. ‘Excessive paperwork and policy’ and ‘policy ideal of patient-centred care’ could have been reconciled, for instance, under the theme ‘ideal of patient well-being’. Care regulators impose high volumes of paperwork on care homes in order to ensure patient safety by holding CHNs to account; they demand patient- centred care to improve patients’ quality of life – both of these phenomena, then, could have been subsumed under the theme ‘ideal of patient well-being’, thus resolving the contradiction. Yet this would have failed to account for the fact that these two demands were often felt, by CHNs operating ‘in the field’, as contradictory, in spite of the best intentions of the regulators who imposed them. - I therefore elected to combine codes into themes which embraced, rather than resolved, their contradictions, these took the form of oppositions. One such theme derived from this approach was ‘Paperwork vs. Principles of Patient-Centered Care’. While the excessive paperwork demands attention and time, its purported principles uphold the importance of patient- centred care, reflecting a key tension inherent in the work of AP, and a contradiction in the system at large. Another such theme was ‘Ambiguity vs. Over-specificity in AM administration guidance’. On one hand, ambiguity in guidance can create confusion and hesitation, but may allow for patient-specific adjustments to AM administration. On the other, over-specificity can enforce rigid protocols, impeding truly individualized care. - Therefore, the binding agent which connected codes into themes was precisely that which apparently set them at odds. This approach proved true to participants’ experiences, since contradictory policy goals, conceived somewhat independently, were often experienced simultaneously by them. My choice to treat the data in this way signified a wider methodological shift, a move from problem-solving to tension-navigating, which proved crucial thereafter. 85 Review themes • In this phase, I refined the themes by revisiting the coded data to ensure that the themes reflected the core of the dataset. This involved reviewing and modifying themes during supervisions, differentiating between overarching themes and subthemes, and understand the relationships between them. • The following questions adapted from Braun and Clarke (2021) guided theme review: Where are the boundaries of each of these themes? How rich and complex are they? How relevant is each theme, and what valuable insights can they provide? • Revisiting tensions: Here, I reviewed the initial tension themes and further interrogated their boundaries in light of Smith and Lewis’s (2011) understanding of tensions as manifestations of deeper paradoxes in complex systems. I revisited the data and consulted policy, documentation, and the literature, finding that several initial tension themes shared meaning and were rooted in distinct underlying paradoxes. • Paradox themes: From this understanding, I developed three overarching paradox themes: (1) the wastage- shortage paradox, (2) the individualiation-standardisation paradox, and (3) the paperwork-patients paradox. Within these paradox themes, I explored the tension subthemes. • An example from my analysis: I found that the initial tension themes of ‘Paperwork vs. Principles of Patient- Centred Care’ and ‘Ambiguity vs. Over-specificity in AM administration guidance’ were both rooted in the ‘paperwork-patients paradox’: while strongly endorsing personalised, patient-centric care, AP bureaucracy paradoxically imposes paperwork and policies that deviate from its core ethos of prioritising patients and personalising care. Refined and name themes • Themes were further refined and given descriptive names through a reflective process that resonated with the data's narrative, echoing Braun and Clarke's (2021) recommendation to identify names that are ‘informative, concise and catchy’ (p. 111) • I rearranged or merged subthemes in order to achieve a clearer and more coherent representation. I focused on understanding the story of the themes and subthemes, and the interactions between them. This approach was grounded in the aim to reveal the true 'essence' of each theme, as outlined by Braun and Clarke (p. 92, Braun & Clarke, 2006). • To ensure authenticity and depth, I frequently circled back to the transcripts (Braun & Clarke, 2021). This process further validated the themes and subthemes and enhanced the rigour of the analysis. • The following questions adapted from Braun and Clarke (2021) guided this phase: What is the theme telling you? How do • An example from my analysis: The overarching analytic theme described above was further refined and named the Paperwork Paradox. The theme encompasses the tensions that participants grapple with between AP bureaucracy and the direct patient care needed for good and caring AP, and also, importantly, the deeper systemic issue where the needs of bureaucracy coexist with, and sometimes overshadow, its ideals of patient-centred care, underscoring a core paradox in the system. This insight offers a profound understanding of the experiences and challenges professionals face working within it, potentially guiding actionable recommendations for system reforms that go beyond surface-level changes. • Continued iterative engagement with the data: As I refined the themes, I continued to revisit the transcripts to ensure that the paradox themes and tension subthemes authentically represented the experiences and challenges described by participants. • Bringing it back to the research aims: After identifying and refining paradoxical themes, I reflected on how they aligned with the core objective of the thesis and the research questions set out in Chapter 1. This ensured that the paradoxes identified were not only intriguing but relevant and foundational to achieving the aims of this research. 86 the subthemes interact and relate to the overarching themes? How do the overarching themes relate? Writing up • During this process, I explored the paradox themes and related tension sub-themes with reference to key quotes. Selecting interview extracts was challenging due to the richness of the data. In line with Braun and Clarke's (2021) recommendations, I chose extracts from the across the dataset for representation, a range of quotes to emphasise the depth and breadth of the central organising concept, and extracts that were clear, engaging, and concise to demonstrate my analytic claims. • To produce an analytical write- up, I meaningfully synthesised and contextualised the data as I presented it, within the wider literature (whereas a purely illustrative approach would link results to literature in a separate discussion section) (Braun & Clarke, 2013). • By locating my research in current academic conversations, I strived to contribute to broader understanding and the ‘rich tapestry of understanding that we and others are collectively working on, in different places, spaces and times’ (p. 120, Braun & Clarke, 2021). • In the 'Findings & Discussion' chapters, the paradox perspective focuses my writing on teasing out the contradictions in the AP system under each paradox theme and their tension subthemes. I build on ideas about paradox from the wider literature to do so. Central to this approach are the three paradox themes, aligned with my research questions. • Structured Analysis: My write-up addresses the research questions by contextualising my analytic claims about paradoxes and tensions in the AP system within participant accounts, drawing on relevant AP policy, care home documentation, the wider literature, and relevant theory. Such an approach builds on the emerging evidence base on AP in care home settings, contributes to the wider literature, and supports a holistic understanding of the tensions characterising the work of professionals in the AP system in care homes. • Practical Implications: In addition to expanding understanding of the AP system, my write-up underscores the practical implications of these paradoxes. By identifying root causes of professionals’ challenges, it paves the way for policy interventions that address underlying systemic problems, not just the symptoms. • An example from my analysis: My examination of the 'Paperwork Paradox' was moored within relevant AP policy (e.g. policy concerning controlled drugs cabinets), care home documentation (e.g. controlled drugs books), relevant theory (e.g. Habermas’s (1985) concept of the system and lifeworld), and wider literature (e.g. Froggatt et al.’s (2011) application of the ‘lifeworld’ concept to care homes). Synthesising these various elements, I explored the ways in which top-down regulations from the regulatory ‘system’ of care do not take into account the practicalities of navigating the ‘lifeworld’ of the care home; for instance, regulations which necessitate controlled drugs cabinets and controlled drugs books which care homes were reported to experience logistical challenges in implementing. Table 3.3 List of themes Theme Subtheme The Anticipatory Prescription Paradox Wastage vs. Shortage Nurse-boundary vs. Doctor-boundary Training vs. Support Anticipation vs. Reaction Uncertainty vs. Certainty Holding on vs. Pushing Away 87 The Protocol Paradox Individualization vs. Standardisation Micro-Decisions vs. Algorithmic Decisions Individualized Language vs. Jargon Oblivion Complex Cases vs. Bureaucratic Shortcuts The Paperwork Paradox Clock-based vs. Task-based Time Granular Certainty vs. Structural Stupidity Risk-Management vs. Error-Multiplication Coercion vs. Moral Principles Guidelines vs. Orders 3.6.6 Quality of the research Ensuring the quality of qualitative research is important for several reasons. At its core, the robustness and integrity of qualitative research determine its acceptance, relevance, and application within both the academic and professional disciplines (Birks, 2014; Cope, 2014). Specifically, high- quality research provides a foundation for the credibility and transferability of the findings. This study therefore seeks guidance from recognised frameworks to maintain its quality. My appraisal of the RTA is anchored in Braun and Clarke’s (2020) 'Twenty questions to evaluate the quality of thematic analysis.' This tool is instrumental in ensuring the trustworthiness and credibility of the analysis undertaken in this research. Trustworthiness in qualitative research offers a representation of participants' experiences that is comprehensive and authentic (Lincoln & Guba, 1985); credibility emphasises the alignment between the findings and the realities of the participants, ensuring that the interpretations of data genuinely depict the information shared by them (Morse et al., 2002). One noteworthy question from Braun and Clarke's (2020) tool asks: ‘Do the authors clearly specify and justify which type of thematic analysis they are using?’ Another question considers the clarity with which the theoretical foundations of the thematic analysis are laid out (Braun & Clarke, 2020). Addressing this, the research firmly grounds itself in a constructivist, relativist approach, enriched by a paradox perspective that resonates with both the theoretical foundations of this research and RTA. To guide the quality of this research overall, I adopt Tracy's comprehensive assessment criteria which draws together a range of best practices, outlining several universal criteria adaptable to various methods and paradigms (Tracy, 2010). These encompass a worthy topic (central, relevant subject matter), rich rigour (depth and breadth of the research process), sincerity (honesty and transparency in the research), credibility (the trustworthiness and believability of findings), resonance (the extent to which the research resonates with or influences its audience), ethics (the 88 integrity maintained throughout the research), and meaningful coherence (ensuring the research process and findings align and make sense within its context) (Tracy, 2010). Tracy (2010) outlines how each of these can be met through a various skills and techniques, inherently flexible and tailored to the specific aims of the study at hand. The use of such an encompassing framework is helpful for both guiding the parameters of quality in this research and in facilitating dialogue between the public and the scientific community, framing qualitative work in a structured and systematic way (LeGreco & Tracy, 2009). The conclusion chapter will offer a reflective overview of the degree to which this research aligns with Tracy's (2010) established quality criteria, carefully considering the strengths and limitations of the study in relation to them. 3.7 Research ethics This study was reviewed and approved by the Psychology Ethics Committee at the University of Cambridge (PRE.2021.017) (Appendix 7). Conducting research with care home staff and other health care professionals introduces specific ethical challenges, particularly concerning recruitment and during interviews (Wyld, 2017). Furthermore, researching and discussing end-of-life care experiences can cause emotional distress to participants due to the sensitive nature of the topic (Gysels et al., 2013). The context of the pandemic further exacerbated these ethical issues, as many health care and social care professionals, especially those working in care homes, were experiencing significant stress related to their circumstances (Birt et al., 2022; BMA, 2021). Thus, the vulnerability of the recruited professionals was an important concern, and understanding their experiences is important for expanding understanding of and improving the AP system to better serve care home residents and their families. The relevant ethical issues in this study are detailed below. 3.7.1 Recruitment in a pandemic context Ensuring the autonomy of potential participants and their understanding of the research is important for the conduct of ethical research (p. 98, Hammersley & Traianou, 2012). The study design and aims were thus communicated to participants either via email or over telephone, and any issues or questions were addressed prior to scheduling an interview. Kvale & Brinkmann (2009) suggest that a transparent recruitment process supports the development of trust and rapport with participants. Study participants were frequently reminded of their ability to abstain or withdraw at any point, with no explanation necessary. The interviews were scheduled for a time most convenient for the participants, thereby reducing any potential disruptions to their work and life. These considerations 89 were especially important given the stressful context of the COVID-19 pandemic (Hensen et al., 2021). Furthermore, issues of accessibility were considered, and it was important to ensure that virtual interviews did not marginalise participants without access to computers. To address this, options including phone interviews were made available alongside video calls. This decision balanced inclusivity with safety during the pandemic. Emerging literature emphasises the efficacy and safety of remote data collection during pandemics, noting the viability of phone interviews (Adom et al., 2020; Hall et al., 2021), email discussions (Amri et al., 2021), and video conversations (Dodds & Hess, 2020). Synchronous video interviewing facilitated by the Zoom platform mirrors face-to-face conversations, thereby promoting organic discussion (Nehls et al., 2015) and enhancing rapport (Deakin & Wakefield, 2013; Sy et al., 2020). All participants opted for Zoom interviews. In recognition of the time dedicated by participants in this study, 30-pound Amazon vouchers were offered as a gesture of gratitude. The inclusion of such incentives is supported by literature that states such gestures can encourage participation from a wider range of individuals (Thompson, 1996), and given the demanding nature of health and social care professionals’ roles during the pandemic, setting aside time for research may have been otherwise de-prioritised (Small, 2023). Thompson (1996) further stated that financial incentives could minimise biases that can occur if the perspectives of certain stakeholder groups are excluded because of non-inclusive recruitment strategies (Thompson, 1996). Such incentives thus helped encourage participation from groups where recruitment can be challenging, providing more holistic representation of stakeholder views. With the offering of financial incentives, it was important for me to clarify that these incentives were not meant to be coercive nor unduly influential in skewing a potential participant's judgment (Grady, 2005). The effectiveness of this approach was underscored by Giebel et al. (2020), who successfully used shopping voucher incentives to enlist care home staff for telephone interviews. The choice of Amazon vouchers in particular was informed by the ongoing pandemic. With physical shopping being discouraged or restricted, Amazon's expansive online marketplace presented a safe and convenient alternative, aligned to the pandemic context. 3.7.2 Informed consent Prior to the interviews, I set aside dedicated time for a phone consultation, typically days or sometimes weeks in advance. This initial briefing conversation, aligned with recommendations from Kvale and Brinkmann (2009), centred around the Participant Information Sheet (Appendix 5), where I explained the study in detail, including its purpose, the advantages and potential challenges of participation, research contacts and commitments, and protocols for data handling and 90 confidentiality. Thereafter, before starting the interview, I began the informed consent process. This procedure was focused on ensuring that participants felt valued, understood, and safeguarded (p. 70, Kvale & Brinkmann, 2009), keeping with the principles of the GMC Research Ethics Code and the British Psychological Society's (BPS) Code of Human Research Ethics (The British Psychological Society, 2014) and Code of Ethics and Conduct (The British Psychological Society, 2018). These codes emphasise the need to provide participants, especially those from vulnerable populations, sufficient time to understand the purpose and implications of the research, empowering them to provide informed consent. The process of informed consent specifically involved a three-fold approach, rooted in the principles and guidelines of the GMC Research Ethics Code of Conduct (GMC, 2021). The first step was a thorough review of the Participant Information Sheet, ensuring any concerns or uncertainties were promptly addressed prior to seeking verbal consent. Secondly, I employed process consent throughout, by continually seeking verification of consent and agreement to continue, and monitoring for any signs of discomfort or resistance. Participants were given the option to pause, opt out, or reschedule if they became distressed: in practice no participant did so, and each interview was completed in a single session. Thirdly, I ensured that participants were wholly aware of my role as the researcher. This three-fold strategy helped to ensure that participants were not only informed but also accorded the respect and protection they deserved throughout the research (GMC, 2021). 3.7.3 Interviews in a pandemic context In designing the interview topic guide, careful attention was given to ensuring that participants faced minimal risk or burden, especially considering the pandemic context. The topic of the interviews centred around end-of-life care – a sensitive subject that could potentially cause distress participants during discussions (Dickson-Swift, 2009). The interviews were thus conducted to accommodate participants’ needs. Prioritising their convenience and comfort, every virtual interview was scheduled based on their availability. If challenges arose, such as emotional distress or exhaustion, participants were offered the option of pausing or halting the interview, though none did so. In line with best practices put forth by the GMC (2021), participants were provided information about services including mental health counselling, resources, and helplines that could offer further assistance, described in detail below (section 3.7.6). 3.7.4 Socially desirable responses 91 One important ethical concern that surfaced during this research was the potential for participants to give socially desirable responses (SDR). Bergen and Labonté found that this inclination arises from a human tendency to present oneself favourably, especially in a research setting (Bergen & Labonté, 2020). They propose that participants may knowingly or unknowingly contour their responses to either match what they believe the researcher would want to hear or what they believe is socially accepted (Bergen & Labonté, 2020). In the context of this study, particularly given the sensitive nature of the topic, there was a risk that participants would share a more sanitised version of events, perhaps highlighting their good practices while glossing over or even omitting less commendable ones. This ethical concern is rooted in the pursuit of authentic information; SDR can skew findings and provide an unrealistic picture of the topic under study. While most qualitative research guides and textbooks emphasise the importance of developing rapport with interviewees and strategies for eliciting rich descriptions of experiences (Kvale & Brinkmann, 2009; Thomas & Pollio, 2002), the literature offered scant direction on managing issues of SDR. To mitigate this, I adopted a multi-pronged approach informed by recent recommendations outlined by Bergen et al. (2019). Creating an atmosphere that was informal, reassuring, and non- judgemental supported participants' confidence and encouraged them to share their experiences candidly and honestly (Bergen et al., 2019). Additionally, Josselson (2013) emphasised the importance of recognising one’s own positionality as the researcher, as accounts can differ depending on the context and the interviewer. Had I been a medical professional, participants may have felt an unspoken pressure to align their accounts with best practices. As a young and eager foreign research student with no clinical background, I was well-positioned to cultivate a relaxed and conversational environment, establishing a foundation for more transparent sharing. Thus, by providing reassurance of anonymity, posing interview questions in a balanced manner, and cultivating a conversational environment, I sought to diminish SDR in my dataset. However, given the sensitive nature of AP and the serious consequences of any errors related to the AP system in care homes, the possibility of SDR remained a concern. To work through this, I leaned on Mills’ theory of motivation. According to Mills, spoken accounts are shaped not only by the individual, but also by the social structures to which they are anchored (p. 909, Mills, 1940). This perspective became instrumental, as it underscored the idea that the motivations shaping accounts are inherently social and arise from the context in which participants are situated, just as the vocabularies they use to relate their accounts are socially constructed. When these vocabularies are recognised, the motivations are accepted and their accounts are validated. When the vocabularies are not shared, the motivations remain unfamiliar and are not accepted. Mills' argument that ‘what is reason for one man is rationalisation for another’ is particularly relevant (p. 910, Mills, 1940), given the diverse approaches to care provision and the differing vocabularies and contexts of health and social care 92 professionals. As a hypothetical example, a CHM who opposes the administration of morphine except for when a patient is actively dying for fear of hastening death is unlikely to welcome a DN’s intervention to administer morphine to a patient experiencing pain at an earlier illness stage. The CHM may instead label the DN's actions as hurried, lacking in empathy and care, or even dangerous. Mills contextualises accounts in participant vocabularies, which he argues are inextricably linked to participants’ social realities (p. 910, Mills, 1940). I therefore recognise that participants’ accounts are relative and affected by what participants were willing and able to share in light of their motivations, vocabularies, social worlds, and respective knowledges. While my research is limited in that I can only know what participants disclosed, their accounts provide valuable insights into the AP system in care home settings. By sharing their accounts, participants were able to give meaning to their own experiences, which in turn enriched the dataset (p. 636, Popay, 1998). 3.7.5 Confidentiality and anonymity In accordance with my study protocol (Appendix 6) and my informed consent process, I was committed to preserving the confidentiality and anonymity of participants. In compliance with the General Data Protection Regulation (GDPR) and Data Protection Act 2018 (EU, 2018; UKDHSC, 2010), I ensured the secure collection and management of data to maintain confidentiality and followed the Caldicott Principles when handling identifiable data of participants. James Brimicombe, Data Manager of the University of Cambridge Primary Care Unit, advised on the secure management and storage of study data. After each interview, audio recordings were removed from my phone and transferred to the University of Cambridge Secure Data Hosting Service. Prior to transcription, all participants were assigned numbers and pseudonyms matching their gender and ethnicity. The audio recordings were sent in encrypted form to the professional transcriber and returned to me in encrypted form. The transcriber signed a confidentiality agreement and destroyed any voice recordings they had on completion of the transcriptions. Subsequently, I examined the transcripts to remove any references to interviewees’ personally identifiable information, including details of their colleagues, care homes, or residents. I substituted all names in the interview transcripts with pseudonyms, approximating the familiarity of the original narrations. Any changes made to the transcript text were indicated by square brackets thus [ ]. 3.7.6 Protecting research participants from harm In line with established best practice guidelines for safeguarding participants in qualitative research from the British Sociological Association (BSA) (2017), I incorporated several measures to prioritise 93 participant well-being throughout the research process. Central to these guidelines is the need to ensure transparency and autonomy of participants, and to anticipate and guard participants from potential harm (BSA, 2017). Recognising that the COVID-19 pandemic brought a unique set of challenges to health and social care professionals, I emphasised flexibility in scheduling interviews to both minimise added stress or disruptions to their daily duties (Small, 2023) and bolster participants’ agency and autonomy. Arranging interviews to best suit their convenience also helped to mitigate the risk of added stress, taking into account the documented emotional and psychological strains on health and social care professionals during the pandemic (Greenberg et al., 2020). To further empower participants and promote inclusivity, I extended to them the choice of their preferred remote mode of interview (Skype, Zoom, or telephone), though all opted to use the Zoom platform. From a logistical standpoint, all interviews were conducted virtually with respect to NHS guidance during the pandemic (NHS, 2020), protecting both the participants and myself from risk of infection. Conducting the interviews, I drew from my own experiences working with health and social care professionals working in end-of-life care contexts. I developed the questions in the interview topic guide to be attuned to the emotional and psychological contours of the participants (Appendix 8). My academic training, work experience, and research equipped me with the confidence needed to navigate potentially sensitive topics during the interviews and prioritise participant well-being. As interviews can be both cathartic and potentially distressing for participants, particularly when discussing traumatic events (Dickson-Swift et al., 2007), I also incorporated a trauma-informed approach outlined by Carello and Butler (2014). This involved maintaining a non-judgmental disposition, recognising signs of emotional distress, offering breaks when necessary, and ensuring participants knew that they could decline to answer any questions and withdraw from the study without consequence, fostering a supportive environment throughout the interaction (Carello & Butler, 2014). Post-interview debriefing sessions were structured in accordance with recommendations from Kvale and Brinkmann (2009): I revisited and further discussed the Participant Information Sheet, addressing any new questions and concerns, and detailing support avenues including St. Christopher's hospice helpline, the Samaritans' confidential support line for NHS staff, and a spectrum of online resources for information for legal advice, advice on rights at work, and emotional health support. Re-emphasising participant rights, the process of data anonymisation, and providing channels for future communication was also covered in this debrief, ensuring participants left the study feeling informed, supported, and valued. 94 3.7.7 Minimising harm to the researcher Recognising the risk of potential emotional distress of this research on me, I employed self-care strategies outlined by Dickson-Swift et al. (2009), seeking regular debriefing sessions with my supervisors and colleagues from the PELiCam research group and scheduling rest breaks throughout my PhD training. Formal counselling support was also accessed during the data collection and analysis periods. These avenues helped me process my emotions, reflect on the experiences shared by participants, and ensure that I was mentally and emotionally equipped to continue this work (Dickson-Swift et al., 2009). For my own well-being, I consistently engaged in self-reflexive exercises throughout the research process. This not only allowed me to be self-aware of my emotional state, but also to understand how these emotions could potentially influence data interpretation (Finlay, 2002). As described by Finlay (2002), such reflexive engagements help maintain a researcher's well-being and support the integrity and authenticity of research outcomes. 3.7.8 Reporting procedures Adhering to established research ethics protocols, this research integrated reporting procedures consistent with the 2020 NIHR Safeguarding Guidance (NIHR, 2020). In addition, given the unique circumstances surrounding the COVID-19 pandemic, I also employed safeguarding guidance for research provided by UKCDR (UKCDR, 2020). These guidelines were developed to support prompt and appropriate action in the event of any serious clinical issues, safeguarding concerns, or circumstances that could compromise participant safety or that of another person. No such situations arose during this research that warranted the activation of these reporting procedures. Box 3.1 Reflections on my role as the interviewer _____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Throughout the interview process, participants expressed gratitude for my interest in their work and for being a young person investigating AP. Reflecting on my positionality, being a young woman from outside the UK appeared to be advantageous as experienced professionals felt it necessary to guide me through their work and experiences, recognising my newness to the UK and my position as a young researcher. My age and American accent appeared to be reassuring to some initially hesitant participants who shared more detailed experiences with AP to help explain the system to myself as someone unfamiliar with it. Care home professionals explained and occasionally conveyed opinions about other health or social care professions to better situate me as a researcher. As a newcomer to UK health and social care, their descriptions of the roles and responsibilities of these professionals provided valuable insight 95 that experienced researchers may have already known or taken for granted. However, my outsider status occasionally resulted in interviewees at times simply repeating the rules, regulations, and guidelines related to AP rather than sharing their personal perspectives and experiences. In this way, some interviewees avoided answering my questions directly, providing only idealised responses that lacked specific details about the AP system, such as how AMs were stored or disposed of in their care homes. While my age, gender, and newness to the UK were occasionally limiting, I believe that these qualities also helped me embody the role of the ‘naïve researcher’ with tact; a recommended approach in qualitative interviewing to effectively probe interviewees on familiar topics (Farooq & De Villiers, 2017). McDowell suggests that it is important to strike a balance between ‘expert and ignoramus’ as an interviewer, and for the most part, I was able to negotiate this balance (McDowell, 1998). My experience in qualitative research supported this balance, grounded in both academic training and practical exposure. During my undergraduate studies, I conducted semi-structured interviews with diverse stakeholders, including patients, family carers affected by neurodegenerative disorders, and their nurses and doctors, to inform the development of a medical device for nurses and generalists at non-neurology clinics intended to detect early signs of neurodegenerative disorders. This project provided first-hand experience in collecting, managing, and analysing nuanced and often sensitive data while ensuring participant comfort and ethical considerations. Furthermore, throughout my academic journey, I have actively sought courses and modules that build my understanding and skills in qualitative research. For instance, as an undergraduate, I enrolled in a course, 'Qualitative Research: Design, Implementation, and Methods' (UC Berkeley), which offered a foundational understanding of qualitative research methods. This initial exposure was supplemented by my work as a qualitative research assistant in the Bove Lab (UCSF), where I designed and conducted interviews with patients and family carers focused on exploring patient- provider relationships and communication issues where patients lacked capacity. My Masters program had a dedicated module on ‘Research Ethics’ (Harvard Medical School), where I explored the ethical dimensions of qualitative research, ensuring that my approaches always prioritise the well-being and autonomy of participants. Regular consultations with my project supervisor during this time provided personalised guidance and feedback on my interview strategies for my masters research project. My doctoral training further supported this foundation. Dr. Jenni Burt’s module ‘Qualitative and mixed methods approaches in primary care’ (University of Cambridge) developed my understanding of methodologies, ensuring that my qualitative interview strategy was consistent with my epistemology, ontology, and theoretical perspective. Regular meetings with my supervisory team also provided important insights that influenced my approach to data collection. Additionally, 96 during the pandemic, I participated in virtual qualitative research workshops: ‘Qualitative Research Methods Workshops 1 & 2’ and a conversational interviewing course (MIT online), where I had the opportunity to engage with qualitative researchers and learn from their experiences conducting interviews, which contributed to my learnings about sensitive interviewing in the pandemic context. Furthermore, I actively sought and accepted invitations to review qualitative studies in journals such as BMC Palliative Care, which contributed to my understanding of best practices in interviewing. All these experiences, taken together, equipped me with a robust toolkit for conducting interviews with sensitivity, depth, and integrity. However, my lack of a clinical background at times put me at a disadvantage during interviews, particularly when participants would mention a medical condition and expect me to have prior knowledge or understanding of it. In these instances, I would nod in agreement and encourage them to continue speaking, rather than disrupting the interviewee's train of thought: after the interview, I would research the condition to supplement my knowledge. This approach allowed me to avoid my being perceived as someone who could not understand or empathise, while also allowing me to engage meaningfully with the topics being discussed. My interviewing approach evolved during the course of the study. Initially, I sought to adopt a professional stance in order to remain as unbiased as possible during both interviews and analysis. However, upon listening to the first few interview recordings, I decided to adopt a more caring, conversational approach. Drawing from my academic training and research experience, I employed a range of communication skills during the interviews. Firstly, establishing trust, understanding, and mutual respect between me, the interviewer, and my participants was essential. This process of rapport-building took place over the course of each interview as it was important for participants to feel safe and in a non-judgmental environment where they could share their perspectives about the AP system openly - especially given the sensitive nature of my research topic and its position in end- of-life care (Elmir et al., 2011). Secondly, I engaged in active listening, which ensured that I was understanding and engaging with the meanings and emotions conveyed (Kvale & Brinkmann, 2008). Josselson stated that listening should be ‘attentive, empathic, non-judgmental, listening in order to invite, and engender talk’, and that silence, smiles, head-nodding, and utterances can also encourage further elaboration from participants (p. 66, Josselson, 2013), which resonated with my interview experience. Thirdly, the use of open questions supported a thorough exploration of the AP system and its paradoxes, as open questions did not limit participants to ‘yes’ or ‘no’ responses, but rather encouraged the sharing of in-depth accounts (Whiting, 2008). Lastly, I often repeated or paraphrased participants’ responses, demonstrating understanding and reassurance to participants that I was listening and that I cared (DeJonckheere & Vaughn, 2019). Each of these skills, supported by documented best practices in guidelines and the literature, provided me with the necessary 97 confidence to explore sensitive topics, gather meaningful insights, and ensure the comfort and trust of participants. As I conducted more interviews, I became more comfortable with the process and found it easier to transition between topics in a way that was more organic and aligned with the natural flow of conversation, at times going out of the order of the interview topic guide (Appendix 8), while remaining consistent with the overarching research goals. This flexibility was aided by the semi- structured nature of the interview method. Semi-structured interviews situate themselves as a middle ground between structured and unstructured interviews, and were thus well-suited to this research, given its exploratory nature. Semi-structured interviews allow researchers to balance consistency with flexibility, ensuring that core topics are covered, while also leaving room for participants to express themselves and highlight issues that might not have been anticipated (Gray, 2014; Kvale & Brinkmann, 2009). The topic guide was crucial in this: developed with the guidance of my supervisory team, input from PPI collaborators, and insights from the literature review, it served as a roadmap to help me navigate participants’ experiences and perspectives related to the AP system, covering important gaps highlighted in the literature review, such as the prescribing, procuring, administration, storage, monitoring, and disposal of AMs, while also allowing for spontaneity and uncharted discussions. To ensure that this guide was not used as a questionnaire, I took measures to ensure that it did not restrict nor define the flow of conversation (Gray, 2014). While a questionnaire might limit a participant's response, the open-ended questions on the topic guide were crafted to encourage elaboration and exploration (Kvale & Brinkmann, 2009; Gray, 2014). The iterative character of the interviews ensured that each conversation was dynamic and adaptable to the unique experiences and insights of each participant. Thus, rather than attempting to cover all of the topics and questions in the topic guide, I often took pauses to delve deeper into the topics that a participant was particularly interested in discussing, ask follow-up questions, or simply encourage the participant to reflect and add more. This approach, drawing on my training in active listening and responsive questioning, led to fluid, organic conversations, consistent with the principles of semi-structured interviewing (Kvale & Brinkmann, 2009; Gray, 2014). Regular reflections after each interview, involving the writing up of field notes in my reflexive journal, helped to refine my interview approach and frame the topic guide as a helpful tool and not a limiting one (Meyer & Willis, 2019). My reflexive journal comprised my personal thoughts and reflections that I had not expressed verbally during the interview, patterns related to interview content with respect to previous interviews, and notes on the locations and interviewees’ conduct. Extracts from my reflexive journal are available in Appendix 9. ___________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 98 3.8 Final remarks This research is the first study to focus on the AP system in UK care homes from the perspectives of health and social care professionals. Throughout the process, I grew increasingly aware of the weight of my responsibility to bring to the forefront the work of the participants, honouring the complexity of their experiences with the AP system in an environment as marginalised as care homes. My approach to analysis allowed me to centre this awareness, as well as flexibility, feedback loops, and my own role in knowledge production (Braun & Clarke, 2019 & 2021). This comprehensive process supported a nuanced exploration of the professionals’ experiences working with AP, locating their challenges in tensions between competing demands, underscoring deeper paradoxes in the system at large, described in the following chapters. In this chapter, I detailed the methodology, methods for data collection and analysis, and ethics for the empirical research. I will now move on to Part 3 of this thesis, the first chapter offering information about the outcomes of my recruitment strategy. 99 Part three: Findings & Discussion Chapter 4 Outcomes of recruitment This chapter is the first of four of the ‘Findings & Discussion’ chapters, describing the recruitment of participants and participant characteristics, including characteristics of their respective care homes. 4.1 Participant recruitment strategy Following Creswell’s (2013) principles of purposive sampling, I identified all adult care homes in Cambridgeshire through searching online, from where I obtained the contact information of their managers. I approached the 123 adult care homes in Cambridgeshire via email (Appendix 4) introducing the study and attaching the Participant Information Sheet (Appendix 5). The email invited care home staff involved in the AP system to participate in the study. This direct approach to all adult care homes is consistent with criterion-based selection, a form of purposeful sampling where all participants meet a specific criterion of importance for the study's research questions (Creswell, 2013). However, I encountered the often-cited challenge in qualitative research: gaining access (Bryman, 2016). Only 4% of Cambridgeshire care homes (5/123) responded to this initial outreach. I therefore reached out to regional and national professional contacts through colleagues in the Queen’s Nursing Institute and the Enabling Research in Care Homes (ENRICH) network, who generously circulated the Participant Information Sheet to their respective networks. This approach led to expressions of interest from 50 potential participants. Due to the broad nature of the requests through email newsletters and social media outreach, I am unable to calculate a response rate. This recruitment approach, leveraging existing networks or participants’ assistance in recruiting others, was conducted with reference to Patton’s (2015) ‘snowball’ or ‘chain’ strategy. While this approach may introduce biases as participants might suggest those with similar views, the strategy is particularly beneficial when targeting specific, niche populations such as care home staff (Jenkins, 2016), and can also lead to richer and more engaged participation (Patton, 2015). Expanding on methodological issues related to this approach, Spreen and Zwaagstra (1994) described that this ‘chain referral’ can overlook individuals who are not actively engaged with the broader community. While alternative approaches such as venue-based sampling have emerged to 100 address the limitations of snowballing, they demand that the researcher knows the population well and who exactly they want to recruit (Magnani et al., 2005), which I did not. Sudman and Kalton (1986) argued that the essence of effective snowballing lies in a profound understanding of the population or network access. My position as a foreign researcher new to the health and social care landscape of the UK meant that since I was lacking in a profound understanding of the landscape, I needed a more adaptive recruitment strategy. Given this context, I leveraged the extensive social capital from my supervisors (SB, CG, and KP) with well-established ties with institutions like ENRICH and QNI to gain access to relevant networks that could help with recruitment (Faugier & Sargeant, 1997). The determination of an appropriate sample size is a subject of ongoing debate in qualitative research. A review of the literature indicates that an 'accepted minimum' or 'ideal quantity' is elusive and influenced by various factors, including the research subject, objectives, methods, and differences in discipline standards and expectations (Braun & Clarke, 2021; Guest et al., 2006). In the context of applied health and health services research, there has been a tendency to rationalise smaller sample sizes often associated with qualitative research, with an emphasis on securing a diverse sample. In this study, discussions with my supervisory team centred on the point at which to ‘stop’ data collection, aiming to pre-empt the pitfalls of collecting too much data while still achieving ‘data adequacy’. Braun and Clarke (2021) highlight the importance of 'data adequacy' rather than ‘data saturation’ in the determination of a sample size in qualitative research, emphasising the quality of information collected rather than the quantity. This shift towards an understanding of data adequacy or depth appreciates the inherent subjectivity and uncertainties in the research journey, underscoring that qualitative research cannot always be rigidly planned (Braun & Clarke, 2021; Saunders et al., 2018). The concept of data saturation is thus not considered an appropriate measure for determining the adequacy of a sample size in RTA (Braun & Clarke, 2021). Specifically, Braun and Clarke (2021) suggest that ‘data adequacy’ is better evaluated based on the perceived quality of data relative to the study's aims, rather than the point at which no new codes or themes are found. Our strategy therefore focussed on achieving meaningful variation across stakeholders and care home organizations, based on factors such as professional roles and care home types, to ensure that the dataset could support paradox-informed analysis and robust interprofessional and interorganisational comparisons. Informed by the expertise within my supervisory team during the protocol development phase (Appendix 6), we projected that a sample comprising roughly 10 to 15 Care Home Managers, 10 to 15 Care Home Nurses, and 10 to 15 health and social care professionals who routinely visit care homes would yield ‘data adequacy’ and rich comparative insights. In Phase 1, I interviewed the CHMs recruited through the processes outlined above (n=19). In Phase 2, I interviewed CHNs (n=10) and visiting health and social care professionals (n=13), identified through a combination of referrals from the CHMs and the above professional networks. I provided 101 the CHMs and network leaders with my contact information and the Participant Information Sheet, which they distributed on my behalf. Interested participants who met the eligibility criteria (adults aged 18 years or over, able to provide informed consent, and involved in the care of residents for whom AP was considered) contacted me directly. 4.2 Recruitment outcomes Between April and July 2021, despite the limitations posed by the COVID-19 pandemic, I compiled an extensive qualitative dataset comprised of interviews with 42 health and social care professionals across 29 care homes in the UK. Combining my initial recruitment strategy focused on local Cambridgeshire care homes and a subsequent wider recruitment strategy through colleagues in professional networks, 50 potential participants were identified of whom 42 (84%) were interviewed, while 8 (16%) expressed interest in participating but did not respond to my follow-up requests for interviews. Table 4.1 presents the number of participants per professional group interviewed. Table 4.1 Participant numbers by professional role Participant role Count Care home manager (CHM) 19 Care home nurse (CHN) 10 District nurse (DN) 3 General Practitioner (GP) 3 Palliative Medicine Consultant 3 Pharmacist 3 Ambulance Paramedic 1 Total 42 Initially, I aimed to recruit 10-15 CHMs, 10-15 CHNs, and 10-15 HCPs as related case-units, forming 10-15 CHM-CHN-HCP teams working collaboratively in relation to individual care homes. Despite my requests that each CHM-interviewee connect me with at least one CHN and at least one visiting HCP, this did not prove possible in practice: pandemic restrictions and participants' workloads prevented me from recruiting them as case-units. Visiting professionals, such as district nurses, doctors, pharmacists, and ambulance staff often worked across several care homes; interviewees were asked to focus on experiences related to the care home they were primarily associated with as far as possible. 102 4.2.1 Reflecting on recruitment outcomes I recruited 42 health and social care professionals involved in AP, working in 29 care homes: 19 CHMs, 10 CHNs and 13 visiting health care professionals, thereby meeting my initial recruitment objectives. The participants provided information concerning their roles, regions (urban or rural), care home Gold Standards Framework (GSF) accreditation, and organisational structure (independently owned or part of a larger corporate chain). In alignment with Braun and Clarke’s (2021) concept of data adequacy, the number of participants recruited was ultimately based on the substantial quality of data relevant to the study's aims, rather than the quantity of data amassed. The final sample of participants showcased a diversity in roles which enriched the dataset, from 29 care home professionals to 13 visiting health care professionals (Table 4.1). However, such a composition could potentially overshadow the perspectives of relatively less represented visiting health professionals, including ambulance personnel (n=1). This imbalance was a limitation that required careful consideration during my analysis, and warrants future studies which centre the accounts of ambulance personnel to ensure more comprehensive representation in the study of AP. Managing and analysing a large dataset presented a unique set of challenges. Although advantageous for capturing a wide array of perspectives, the sample size involved negotiating the complexities of a substantial amount of qualitative data; this included preventing data redundancy and becoming overwhelmed by the sheer volume of my dataset (Easterby-Smith & Lyles, 2012). Thus, despite additional interest post-recruitment, my supervisory team and I decided to cap the number of interviews at 42, guided by the need to balance the depth and breadth of the dataset with the practicalities of conducting a rigorous analysis. While more data could potentially yield additional insights, it was essential to limit the scope to what could be thoroughly analysed within reasonable timeframes, thereby respecting the integrity of each participant’s contribution and supporting a high- quality analysis. The sizeable dataset and the extensive time I invested in collecting and analysing it—six months of interviewing participants (with each interview lasting between 45 and 75 minutes and a mean duration of 60 minutes), followed by one year of analysis—was important for both addressing the study aims and honouring the depth required for a paradox-informed RTA. The focus was not on accumulating information until no new themes emerged, but rather on gathering enough quality data to robustly support the study aims (Braun & Clarke, 2021). Given the sample size, I was acutely aware that an abundance of data does not inherently guarantee the quality of analysis. A substantial number of participants poses a risk of over-reliance on data extracts during analysis, potentially leading to what Braun et al. (2016) consider a descriptive rather than critical and conceptual analysis: ‘You need to determine a good balance between [data extracts and analytic commentary]…A 50:50 ratio works for fairly descriptive analyses; more critical/conceptual analyses often have a greater proportion of analytic narrative. Your narrative will also be proportionally greater if you combine the results and discussion’ (p. 13). In my analysis, I aimed for a balanced ratio of 1:2 of representative data extracts and their analysis to analytic commentary and 103 discussion. This was to ensure that the rich, detailed quotes did not eclipse the analytical rigour of my study. I offer further reflections on writing-up with a large qualitative dataset in Chapter 8. 4.3 Participant and care home characteristics Tables 4.2 and 4.3 show the data collected on participant roles, regions, and care home GSF accreditation and organisational structure in aggregate and per participant, respectively. For the purposes of confidentiality, participant names were anonymised. Location and regions data are not reported or are pseudonymised. The majority of participants (n=37) were located in England. Of the 29 care homes, 8 were GSF-accredited, 21 were non-accredited. 12 care homes were independently owned and operated; 17 were corporate chain care homes, owned and operated by the five largest care home providers in the UK (HC-One, Four Seasons Health Care, Barchester Healthcare, Care UK, or Bupa). The majority of independent care homes were GSF-accredited (n=8, 67%), while all of the corporate chain care homes were non-accredited (n=17, 100%). Where participants (e.g. GPs, DNs, and pharmacists) worked across multiple care homes with different characteristics, they are denoted as N/A in Table 4.2. While I ensured the accuracy of participant gender and ethnicity during the anonymisation process, these variables were not examined in the following chapters. The care home workforce is predominantly female, with white middle-aged women typically occupying nurturant positions, while women from ethnic minority backgrounds tend to hold non-nurturant positions (Skills for Care, 2020). Men in health and social care professions tend to occupy higher positions of power than their female counterparts. While the demographics of my dataset support these patterns, the present study did not focus on gender and ethnicity. Table 4.2 Aggregate table of participant characteristics Participants N = 42 Participant role Care home manager n=19 Care home nurse n=10 District nurse n=3 General Practitioner n=3 Palliative Medicine consultant n=3 Pharmacist n=3 Ambulance paramedic n=1 Setting of care home with which participant associated Rural n=20 Urban n=22 GSF accreditation of care home with which participant associated Non-accredited n=25 Accredited n=8 N/A=9 Care home type with which participant associated Independent n=12 Corporate chain n=20 N/A =10 104 Table 4.3 Individual participant and care home characteristics Participant Number* Participant Name Participant Role Urban/ Rural Care Home Region GSF Accreditation Independent/ Corporate Chain care home 1 Mr. Smith NHM Urban Accredited Independent 2 Mrs. Wright NHM Urban Non-accredited Corporate chain 3 Dr. Powers Consultant in Palliative Medicine Rural Non-accredited N/A 4 Mrs. Walker CHM Rural Non-accredited Corporate chain 5 Mr. Rowlands Pharmacist Rural N/A N/A 6 Dr. Cameron Consultant in Palliative Medicine Rural N/A N/A 7 Mrs. Fields DN Urban Non-accredited Corporate chain 8 Dr. Kennedy Consultant in Palliative Medicine Rural N/A N/A 9 Mr. Jeffries CHM Rural Non-accredited Corporate chain 10 Mrs. Kinley NHM Urban Non-accredited Corporate chain 11 Mrs. Bennett NHM Rural Non-accredited Independent 12 Ms. Wynne CHN Urban Non-accredited Corporate chain 13 Mrs. Haberman DN Urban N/A N/A 14 Mrs. Graham CHN Rural Accredited Independent 15 Mrs. Ainsley CHM Urban Non-accredited Independent 16 Mrs. Lancaster CHN Urban Non-accredited Corporate chain 17 Ms. Jeeves CHM Urban Accredited Independent 18 Mrs. Galway NHM Rural Accredited Independent 19 Mrs. Laney NHM Rural Non-accredited Corporate chain 20 Mrs. Larkin NHM Urban Accredited Independent 21 Mrs. Whitfield NHM Urban Non-accredited Corporate chain 22 Mrs. Gardner CHN Urban Non-accredited Corporate chain 23 Mrs. Barnes CHN Rural Non-accredited Corporate chain 24 Mrs. Jiro CHN Urban Accredited Independent 25 Mrs. Raymond CHM Rural Non-accredited Corporate chain 105 26 Mrs. Beverley NHM Urban Non-accredited Corporate chain 27 Ms. Jones CHN Rural Non-accredited Corporate chain 28 Mrs. Greensborough CHM Urban Non-accredited Corporate chain 29 Mrs. Perry CHN Urban Accredited Independent 30 Ms. Foster CHN Urban Non-accredited Corporate chain 31 Mrs. Whittaker NHM Rural Non-accredited Corporate chain 32 Mrs. Claremont CHN Rural Non-accredited Independent 33 Ms. Gonzalez DN Urban N/A N/A 34 Mrs. O’Connor CHM Urban Non-accredited Corporate chain 35 Mr. Anderson CHM Urban Non-accredited Independent 36 Mr. Reed CHM Urban Accredited Independent 37 Dr. Cairns GP Rural N/A N/A 38 Dr. Jameson GP Urban N/A N/A 39 Dr. Murray GP Rural N/A N/A 40 Mr. Thomas Paramedic Rural Non-accredited Corporate chain 41 Mr. Nanda Pharmacist Rural N/A N/A 42 Mrs. Sandy Pharmacist Rural Non-accredited Corporate chain *Participants are arranged in chronological order according to the sequence in which I interviewed them 4.4 Conclusion Forty-two interviews were conducted with a diverse range of health and social care participants concerning AP in UK care homes. The dataset provided detailed accounts of participants’ views and experiences with the AP system. The dataset also provided health and social care professionals’ perspectives related to caring for care home residents at the end of their lives more broadly. Despite the considerable challenges posed by the pandemic, the recruitment process yielded a substantial, rich, and diverse qualitative dataset, with a range of participants reflecting the diverse roles and responsibilities involved in the AP system in care homes. The following ‘Findings & Discussion’ chapters present the results of my analysis, exploring three key themes: three paradoxes in the AP system and its the resultant tensions. 106 Chapter 5 The Anticipatory Prescription Paradox This chapter is the first of three ‘Findings & Discussion’ chapters, presenting my analysis of the qualitative interview study. Here I focus on the Anticipatory Prescription Paradox theme and its tension-based subthemes. The chapter is organised thus: 5.1 Background, 5.2 Introduction, 5.3 Wastage vs. Shortage, 5.4 Nurse-boundary vs. Doctor-boundary, 5.5 Training vs. Support, 5.6 Anticipation vs. Reaction, 5.7 Uncertainty vs. Certainty, 5.8 Holding on vs. Pushing Away, 5.9 Chapter summary, and 5.10 Discussion, in which I outline my contributions to theory, to literature, and for policy, practice, and further study. 5.1 Background Anticipatory medicines shortage is one of the most pernicious of the many issues which blight care homes today. Williams et al. found that one third of participants (n=199) described problems with timely access to medication; for one quarter of participants (n=28), this resulted in a failure to administer much-needed medication to residents (Williams et al., 2018). The CovPall study found that, during the COVID-19 pandemic, medication shortages affected approximately 25% to 33% of the included study settings, including UK care homes (Oluyase et al., 2020). Rosoff found palliative care to be the only ‘effective tool we have available’ to help patients facing drug shortages (Rosoff, 2012). As the pandemic showed, such a recommendation becomes problematic when palliative care itself becomes overburdened with crisis-induced shortages. In order to address these pandemic- related shortages, Reuse of Medicines legislation was introduced in 2020; it has since been repealed (DHSC, 2020). Antunes et al. (2022) examined this pandemic context, and found that practitioners were moving towards embracing changes in AP practice in order to address stock shortages, including keeping a stock of medicines. Antunes et al’s (2022) study also found that ‘keeping stock’ was proposed by respondents as a way of addressing AM waste, another grave issue in palliative care. This issue is reflected in the EHCH’s key priority area (4.2.2): to reduce medication waste (NHS, 2020). Katz et al. found that the majority of the AMs prescribed by palliative care doctors in their study were never administered (Katz et al., 2018). ‘Keeping stock’ can mean that unadministered medicines are eventually used, rather than being wasted because the named resident they were assigned for has died (Antunes et al). A creative solution to waste was proposed by Donyai et al. (2021), who took inspiration from schemes in Greece and the US in attempting to ‘tackle negative perceptions about receiving what some might consider to be ‘second-hand medicines’, ultimately aiming to make ‘medicines reuse as acceptable as reusing 107 coffee cups and plastic bags’. A 2010 report estimated the national figure of pharmaceutical waste to be £300 million, revealing: ‘This sum represents approximately £1 in every £25 spent on primary care and community pharmaceutical and allied products use’ (NHS, 2015). On their own, however, statistics concerning wastage are less helpful than attempting to identify the causes of wastage. Statistics can produce the impression that wastage is a matter of more or less, something which need only be reduced to be fixed. In fact, measures which reduce wastage can inadvertently exacerbate the equally serious issue of shortage. To navigate this double bind, any analysis of wastage or shortage must address them in concert, rather than treating them as separate issues. Wastage and shortage can often be considered consequences of over-prescribing and under- prescribing, respectively. As Knights has shown, under-prescribing can develop directly from a fear of over-prescribing (Knights, 2013), so they must be considered in tandem, and wastage and shortage similarly must be considered in tandem. The under-analysis of the relationship between wastage and shortage may have to do with the fact that they exist in a paradoxical relationship with each other, thus defying intuition and discouraging deep contemplation. Studies centring on Reuse legislation, such as that of Antunes et al. (2022), simultaneously acknowledge issues of shortage and wastage without examining how they interrelate. A study by Morris et al. (2013) came to the conclusion that keeping stock in one care home did not materially reduce wastage, thus conflicting with Antunes et al. (2022), who found the opposite. Thus, the data on wastage appears at times to be conflicting, even paradoxical; as I will show later, resolving this contradiction involves understanding the relationship between shortage and wastage. 5.2 Introduction This chapter examines the anticipatory prescription paradox – the simultaneous occurrence of AM wastage and AM shortage in many UK care homes. While care home professionals and pharmacists in my dataset independently expressed frustration with both the significant wastage of AMs and the frequent AM shortages in care homes, it was when the two issues were contemplated in tandem that participants reported feelings not only of discontentment, but also of bewilderment: “It's like a double whammy, you know. On one hand, we're chucking away the anticipatories, and on the other we're scrambling to get our hands on them when we actually need them. If it wasn’t keeping me up at night, it’d be just downright silly. I think we just need a system that makes more sense”. (Ms. Jones, Care home nurse, Participant 27) Ms. Jones’s use of the word ‘system’ here points to the systemic nature of paradox. Systems, and especially bureaucratic systems, operate by dividing labour up into manageable portions and assigning them to separate, specialised branches or roles, thus resulting in the type of ‘organisational complexity’ which naturally predisposes itself towards paradox (Lewis, 2000). Such systems contain 108 many different moving parts and agents which operate with a degree of independence from each other, making it likely that two separate elements may inadvertently contradict each other. Disentangling a paradox therefore requires peeling back the layers of the system, and understanding the complex ways in which these parts and agents interact. In this chapter, the agents in question were typically prescribing GPs and administering care home professionals, whose specialised roles are sufficiently distinct from each other that they lacked awareness of each other’s expectations and capabilities, but not distinct from one another that they did not have to interact at one crucial professional intersection – the prescribing process. As its name suggests, the process of prescribing lies at the heart of the anticipatory prescription paradox. The cross-sectional view offered by my dataset allows me to examine the act of prescribing from multiple different perspectives – that of GPs, care home professionals, and pharmacists. Prescribing is always at least a bilateral process, and typically acts as an intersection between these various professions - a flashpoint where tensions and differences of perspective between different types of professionals are thrown into relief. Differences in perspective are a useful lens for the examination of paradoxes, since, once again, ‘the wide range of paradoxes in health care appears to result from different interpretations of the system’s complexity’ (Hoffman, 2001). In other words, paradoxical effects often ensue when different agents in a system view the same thing in conflicting ways. My data reveals that GPs and care home professionals involved in prescribing often had differences of perspective with regards to what they expected of each other during the process of prescribing. In order to analyse the differences in perspective between professionals which are illuminated by a close examination of the prescribing process, I will adopt a ‘negotiated order’ perspective to complement my paradox lens. ‘Negotiated order’ is a sociological lens tailor-made for the analysis of how professional boundaries shift and are tested during the process of negotiation that often attends anticipatory prescribing (Strauss et al., 1963) (Strauss, 1978). Nugus has provided important modern contributions to the theory, partly by applying it to the discipline of geriatric medicine (Nugus et al., 2010; Nugus, 2019). Negotiated order will support my analysis of the interprofessional relationships underlying the AP process. I will examine whether the strength or weakness of these relationships has an effect on AM wastage and shortage. Finally, I will explore how the issues of wastage and shortage penetrate deeply into the heart of a duality concerning AMs, which are simultaneously always seen as existing in excess and in deficit, as potentially doing good and potentially doing harm, as both predictors of death and reactions to death. This will in turn lead me to onto deeper questions about the nature of palliative and end-of-life medicine itself, and the paradoxes that ensue when an institution founded on the preservation of life is forced to accept death. To examine these issues in their wider context, I will situate my analysis within the wider history of medicine and death, provided by Shai Lavi (2004) and Michel Foucault. 109 Research question • What underlying systemic tensions and what wider interprofessional contexts underpin issues related to the wastage and shortage of anticipatory medicines in care homes? 5.3 Wastage vs. Shortage AM shortages typically occur when a care home resident deteriorates more quickly than expected, meaning that either a prescription cannot be put in place or that the necessary medicines cannot be procured quickly enough to manage that resident’s pain symptoms, often resulting in them experiencing a ‘bad death’: “I always try to make sure that before the weekend we’ve got stock so that we don’t get stuck with a controlled drug shortage. It’s better to get it from the pharmacy, but if they’re out of stock, we have to ring the out-of-hours doctors which is a long process. Sometimes we can’t get it locally. I remember there was a time when somebody had to go an hour to [city] and get this controlled drug for a resident who was in pain. They were his last hours”. (Mrs. Barnes, Care home nurse, Participant 23) “That’s the most frustrating thing, not being able to get things the same day, and some community pharmacies just don’t stock that sort of thing [anticipatory medications]. I had to go out to about five different pharmacies to get the prescription, and one could maybe give me one ampoule and one could give me like three ampoules… That is probably the most stressful thing: the actual getting of the medications”. (Mrs. Bennett, Nursing home manager, Participant 11) These shortages typically occurred during weekends, when far fewer pharmacies were open, and involved care home professionals undertaking the arduous work of locating pharmacies in the middle of the night in order to procure AMs for rapidly deteriorating residents. GSF participants reported dealing with fewer shortages in general, and when they did experience shortages, these shortages were caused by national supply chain issues. Non-GSF participants, on the other hand, experienced frequent AM shortages due to weak relational ties with community colleagues, especially prescribing GPs. While some of the shortages reported in my interviews can be specifically localised to the pandemic, the issue of shortage is a systemic issue woven into the fabric of the current care climate, and thus transcends the pandemic context. In my dataset, GSF-accredited care homes, since they were generally well-stocked, also more frequently reported wastage, since there was simply more medication to waste: 110 “We’re attached to Gold Standard Framework so we’re really good at getting anticipatory medication on site, but on the bad side it means we’ll waste a lot of medications because when that resident dies, then we can’t return the medications to pharmacy. It just needs disposal and the time and cost of that is horrendous…” (Mrs. Perry, Care home nurse, Participant 29) The fact that the more well-stocked a pharmacy is, the more well protected against shortages it is, and yet the more vulnerable to wastage it is, gives some measure of the interdependence of wastage and shortage. Wastage and shortage exist in a zero-sum relationship; they occupy two sides of the same coin. This can be seen in a piece of guidance issued by NICE during the COVID-19 pandemic: ‘Take into account potential waste, medicines shortages and lack of administration equipment by prescribing smaller quantities or by prescribing a different medicine, formulation or route of administration when appropriate’. (NICE, 2020, COVID-19 Palliative Care Scenario) The advice here appears somewhat paradoxical, as it attempts to address both wastage and shortage, and yet ultimately encourages prescribing smaller quantities of medicine – a measure that could minimise potential waste but would actually increase the likelihood of shortages. It is perhaps because medicine wastage and shortage are such inherently related issues that this guidance lumps them together, attempting to address both simultaneously. And yet since the exact nature of their interrelation takes the form of a zero-sum arrangement, any measure that mitigates one will actually exacerbate the other. This guidance is therefore the site of a paradox – that the same inherent interrelatedness between the issues of wastage and shortage that encourages one to attempt to address them simultaneously also makes it difficult to address them simultaneously. The futility of many waste-reducing measures within AP can be demonstrated not only by the fact that attempts to reduce waste often increase shortages, but by the fact that attempts to reduce waste can fail to reduce waste in the absence of shortages, again revealing the interdependence of the two issues. A study by Morris and Hockley (2013) came to the seemingly counter-intuitive conclusion that keeping a stock of AMs in a care home (thus allowing for medicines reuse) did not significantly reduce that care home’s overall wastage. Given that the care homes in this study were GSF- accredited, it is reasonable to assume that they did not suffer from medication shortages (Covington, 2013). It may therefore have been that care homes not suffering from shortages did not have enough incentive to reduce wastage. In a care home suffering from shortages, those disposing of medicines are condemned not only to the knowledge that the medicines in question could theoretically have been used by someone, but that they could be used for a specific resident who may die in pain without them. That such knowledge strongly incentivises waste reduction was suggested by one of my participants: 111 “…the nurse or the GP will prescribe another driver and you haven’t actually got enough for the next driver, so it’s swings and roundabouts and I think stock in care homes would take that pressure away. You know you might have some residents who don’t need it, you might have one resident who needs a lot and I think definitely stock is the way forward… Like I say it’s just, just because I think there’ll be less wastage”. (Mrs. Haberman, District Nurse, Participant 13) This account suggests that the desire to achieve ‘less wastage’ may stem from the knowledge that, though some residents ‘don’t need’ AMs, meaning these AMs are likely to be wasted, there may be another specific, known resident in the same care home who ‘needs a lot’, making the human cost of wasting that medication and failing to address that need far more immediately obvious. The concept of waste is therefore arguably dependent on the concept of need; the disposal of a medicine is felt as waste only in proportion to how in demand that medicine is, and this kind of immediate need, shaped by supply and demand, is often the result of shortages. So interdependent are the concepts of wastage and shortage that wastage may in fact require shortage to already be occurring for it to even be felt as wastage in the first place. Care home managers who made use of the March 2020 Reuse of Medicines legislation (NHS, 2020) reported in my interviews that keeping stock did help them to reduce both wastage and shortage: “What happened during Covid, which was brilliant, was that we were able to use other people’s medications that had passed away… we’ve been trying to do this for a long time, but the NICE guidelines wouldn’t let us, and the CQC wouldn’t let us. Because the waste was appalling, you know… We’ve always got the stock in the home, because there’s nothing worse than flying around looking for a chemist when our resident is dying in pain”. (Mrs. Galway, Nursing home manager, Participant 18) This is a rare and fortunate case of an intervention simultaneously addressing wastage and shortage. Notably, this participant’s care home had been experiencing AM shortages, meaning that there was an incentive to ‘reuse’ medicines and thus reduce waste. This participant was not alone in associating keeping stock with waste reduction: “In the same building you could have twenty people with twenty packs of medicines where half of them would probably not get used. They might be used in a dose of one, and then the rest thrown out because it's got someone's name on it. It makes much more sense to keep care home stock…” (Mr. Rowlands, Community pharmacist, Participant 5) 112 Aside from wastage, another participant highlighted the medicines reuse scheme’s potential to address shortages: “I absolutely think that care homes should have just a small stock. Most of the time it’s fine. People don’t just deteriorate in a day. Generally you do see that deterioration over a matter of, you know, a week or something before they need that medication. But it does happen that people do deteriorate very quickly every now and then, so I think if you did have even just 5 ampoules of each potentially, just so you don’t have that stress, like at weekends when we’re always at risk of running out of something or not having a drug we need”. (Mrs. Bennett, Nursing home manager, Participant 11) Keeping stock prevents wastage by allowing medicines to be reused rather than thrown away, and these reused medicines can also prevent shortages when directed to residents experiencing sudden deteriorations. The interdependence of wastage and shortage can thus sometimes be put to good use. Just as a measure taken to alleviate wastage can simultaneously exacerbate shortage and vice versa, so too can a measure taken to alleviate wastage also alleviate shortage and vice versa. Medicines waste differs from other everyday forms of waste (e.g. food waste) in one key way. Whereas food waste is generally an unintended consequence, a passive side effect, medicines waste is a process which must be actively performed, usually via highly involved processes such as CD disposal and CD destruction. As the following account demonstrates, the very CHNs who suffer most from waste (in that it takes away medications which could address their shortage problems) are also legally required to put the most effort into carrying out the waste, into becoming complicit in the waste: “We take off the labels of any name, details, address, destroy them and then put the drugs into a returns box, which is locked and stored away… It’s a lot of paperwork to make a lot of waste so the whole thing is just painful”. (Ms. Foster, Care home nurse, Participant 30) CHNs most actively and directly involve themselves in waste when they ‘destroy’ medicines, a process which often involves chemically denaturing medicines using a DOOM (Disposal of Old Medicines) kit after a resident has died. One participant reported her frustration at having to destroy medications which had not yet been opened: “We know that the law states… that you have to destroy. But in my opinion we should look at the possibility to call hospice or pharmacy. You know, the box wasn’t even open. But the answer is always: ‘was prescribed for this individual, you have to destroy’”. (Mrs. Barnes, Care home nurse, Participant 23) 113 This participant also reported an even more counter-intuitive scenario in which medicines were ordered destroyed before they had even left the pharmacy: “The resident had passed away on a Sunday and we were receiving a new batch of medication on Monday morning and I rang the pharmacy to say, please don’t bother sending this resident’s medication because she’s passed away. They said, ‘Well the medication has already been packed and we can’t go back into the bags and remove the medication, and anyway they’ve been labelled so they can’t be used for anybody else.’ I thought the pharmacy could just remove the label and put it for somebody else. And we just received those mediations, honestly and the next thing they were destroyed. This could have been sorted out with the pharmacy, they shouldn’t have sent them knowing that the resident has died”. (Mrs. Barnes, Care home nurse, Participant 23) The contradiction with these CHNs’ struggle is that the acts of name-labelling or packing a medicines bag change only the bureaucratic status of the medicine; they do not alter the medicine’s actual effectiveness, its chemical composition. In spite of their knowledge that the medicines would still be perfectly effective for residents who desperately need them, these purely bureaucratic changes force the CHNs to treat the medicines as if they were worthless and subsequently actually change their chemical make-up by denaturing them. The earlier-cited account describing medicines being deemed unusable before they had been taken out of the box, as well as another in which a participant (Mrs. Lancaster, CHN) remarked, ‘We’re having to throw this medication away even though it’s not been touched.... And for me I just think if it’s not been touched and it’s totally sealed in, you know, the vials and the packaging…’, both point to the sentiment that a medicine cannot reasonably be deemed compromised if it has not even been touched. This sensitivity to touch may be related to fears and measures put into place during the COVID-19 pandemic, when touching medicines introduced a risk of contamination: “You know, if we had a syringe driver during the outbreak, we had to be addressing it as if it’s significantly likely to be a contaminated piece of equipment. The medicines themselves, we have to dispose of them in a very particular way with very clear guidance, and we do our very best to do that”. (Mrs. Bennett, Nursing home manager, Participant 11) Donyai et al.’s (2021) study on medicines reuse and waste suggests that any act which involves handling or touching medicines is treated as if it in some sense ‘contaminated’ the medicine with the risk of error: 114 ‘[U]sing multi-compartment compliance aids (MCCAs) as a practice is thought to be wasteful; in MCCAs, individual doses, having been removed from their packaging, are placed with other medicines within distinct compartments a month or more in advance of actual consumption, rendering the medicines ‘expended’ as soon as the MCCA is prepared. Then if a patient’s medication regimen is changed (e.g., a medication is discontinued), the entire content of their MCCA becomes unusable for that patient—this is because it is too risky to remove individual discontinued tablets from the compartments (risk of error) so the entire contents have to be discarded once the regimen changes’. (p. 2, Donyai et al., 2021) This situation, in which the mere act of touching something renders it ‘expended’ and, eventually, compromised, brings to mind cultural and religious rituals discussed in anthropologist Mary Douglas’s Purity and Danger (1966), where the act of touching can either ritualistically purify or instantly contaminate. Douglas points out that pharmacology, and medicine generally, often overlaps with rituals around cleanliness, noting that ‘the resemblance between some of their symbolic rites and our hygiene is sometimes uncannily close’ (p. 33, Douglas, 1966). Discussing an ancient medical text, Douglas cites a pharmacologist’s interpretation that some ‘instructions’ contained within it – namely, instructions ‘to “purify” the simples before pulverisation’ – suggested ‘a step which must have required several chemical operations’ (p. 33). Douglass points out that ‘purifying here’ could instead ‘mean sprinkling with holy water or reciting a spell’ (p. 33). Bureaucratic rituals such as the name-labelling of medicines, the packing of a medicines bag at the pharmacy, the removal of medicines from packaging, and the preparing of MCCAs, can all be considered in some senses similar to ‘symbolic rites’ (p. 33). These meticulous routines are charged with bureaucratic meaning to the point where merely touching a medicine (something which occurs at a level of abstraction similar to that of ‘sprinkling holy water’) can compromise it and cause it to need to be destroyed – a process which requires complex chemical operations. A major cause of medicines waste is AM standardisation. The GSF opts for standardised ‘anticipatory boxes’, ‘kits’, or ‘packs’, to be prescribed by GPs in packs of four. This can cause wastage when a resident does not need all four AMs. Taking a more personalised approach to a resident’s individual symptoms, and accordingly prescribing only the medicines out of the full box that are specifically needed by that resident could therefore reduce this wastage. Furthermore, once GPs become aware that the standardised packs they are prescribing can cause wastage, this can lead them to overcorrect and become more conservative with their prescriptions, leading to shortage. This was seen in the following account: “Getting a prescription in place has always been hard… I wish I could have an anticipatory box for everybody that’s on an end-of-life schedule. But they don’t do that because then it’s overprescribing and it’s waste, all of which I get. Nobody wants to have 115 medicine waste, but what I also don’t want to be doing is rattling around at 3am on a Saturday morning trying desperately to find a 24-hour pharmacy...” (Mrs. Raymond, Care home manager, Participant 25) This account chimes with data that suggests that under-prescribing can develop directly from fear of over-prescribing (Knights, 2013). Since AMs are prescribed four at a time, in a ‘box’, GPs must constantly choose between prescribing too much and too little. Standardised packs of four both increase wastage, by putting more medicines in circulation, and paradoxically also increase shortage, by causing GPs to under-prescribe in response to fears of causing this very wastage. This process may have its counterpart in a tendency towards under-requesting on the part of the CHNs on the other side of the prescribing process. In one study, a NHM reported that access to the stock ‘transformed’ her nursing staff’s practice, as they became comfortable requesting prescriptions for dying residents knowing that the medications would not go to waste (Griggs, 2010). This implies that, before this, this nurse manager may have been actively under-requesting prescriptions due to fears of waste. 5.4 Nurse-boundary vs. Doctor-boundary In general, GPs’ desire to prescribe conservatively in order to avoid wastage (as per Mrs. Raymond’s account above) and care home professionals’ conflicting desire to obtain more generous prescriptions in order to avoid AM shortages were described by my participants as texturing the prescribing process with interprofessional tension. Mrs. Raymond recounted having found it ‘difficult to get 111 to prescribe end-of-life medication’, showing that she encountered difficulty in convincing an out-of- hours GP to prescribe AMs to resolve her shortage issue. Other participants reported a very similar experience: “We did have to contact an out-of-hours team, who I did have a lot of trouble with. I had a big argument on my hands about why I considered that person to be end-of-life and why I was asking for end-of-life medications… In the end the doctor did listen to me and he did prescribe the medications”. (Ms. Jeeves, Care home manager, Participant 17) “The number of times I’ve had to invoke my position as a clinical nurse to get them [GPs] to give us the anticipatory medications for our residents who were clearly at that point…” (Mrs. Whitfield, Nursing home manager, Participant 21) Such accounts point to the importance in the AP process of what Strauss et al. (1963) term negotiated order (Strauss et al., 1963; Strauss, 1978). Strauss’s concept of negotiation, which he described as ‘the continual working out, together, of who [is] to do what, how and with whom’ (p. 107, Strauss, 1978), was intended to sensitise sociologists to the ways in which social order in complex 116 organisations is reconstituted or ‘worked at’ continually through negotiation and bargaining. The fact that Ms. Jeeves’s had to make a ‘big argument’ in order to convince a GP to ‘listen to’ her and prescribe necessary AMs indicates that the effectiveness with which care home professionals make their case can materially impact whether or not a certain prescription is granted. Negotiated order makes it so that prescribing decisions are always, to some extent, up in the air, since whether medicines are prescribed generously (leading to ‘over-prescribing’ and ‘waste’) or conservatively (leading to shortages and out-of-hours emergencies) becomes dependent not on any reliable constant found within AP guidance or legislation, but on a variable – the negotiating capacities of individual nurses. In health care, prescribing decisions act as a flashpoint in which the boundaries between different professions are tested. Negotiations concerning prescriptions often highlight differentials in power between prescribing professionals and CHNs, with the latter, in the accounts just cited, appearing often to be ‘at the mercy’ of the former. The power to prescribe is an example of what is known as ‘medical dominance’, the hierarchical professional advantage supposedly enjoyed by physicians over other clinical groups (Freidson, 1988; Willis, 1989). In my data, prescribing decisions were reported in some cases as partially redrawing the lines of these hierarchical boundaries, rendering them porous. During the COVID-19 pandemic, when the Reuse guidance was temporarily put in place, one CHN participant reported being invested with power traditionally allotted to prescribers: “… Our GP sent us the reuse guidance. Management said, ‘Right, just write a policy.’ I mean, I’m only a nurse. I don’t have any extra qualifications, no degrees, no prescribing, nothing. She said, ‘Just write a policy and get going. Do what you need to do.’” (Mrs. Graham, Care home nurse, Participant 14) That Mrs. Graham not only brings up the differences between professional roles, but goes on to consider the ‘qualifications’ and ‘degrees’ underlying those roles, suggests that the issue of prescribing touches deeply on the question of how power and responsibility are distributed within health care. Allen’s (1997) study on negotiated order argues that ‘there is a sense in which some blurring of the nurse-medical boundary is unavoidable’ (p. 511), and the above quotation shows that this certainly applies to GPs and CHNs. A perhaps surprising finding of Allen’s study, however, was that boundaries between doctors and nurses remained porous even without any direct, that is face-to- face, negotiation occurring at all: ‘According to Strauss et al. (1963), negotiations are the key to understanding how order and change fit together. The findings of this study, however, reveal that formal organisational structures can be modified in the absence of face-to-face negotiations.’ (p. 514, Allen, 1997, ‘The nursing‐medical boundary: a negotiated order?’) 117 Doctors, handling overwhelming workloads, were reported by Allen to respond negatively to being bleeped by nurses for issues they deemed unworthy of medical attention. Thus, ‘[t]he difficulties nurses experienced in getting medical staff to come to the ward… led nursing staff routinely to violate organisational policies by undertaking medical work’ (p. 511). This constitutes a shift, however minor, in the social order of the hospital, but one which Allen deems ‘non-negotiated informal boundary-blurring’ (my italics) (p. 516), insofar as nurses and doctors did not directly confer over this transfer of responsibilities, the latter being too busy to partake in this negotiation. During the COVID-19 pandemic, GPs were reluctant to visit care homes, leading to a situation similar to what Allen’s study describes, in which face-to-face interactions between doctors and nurses halted. The pandemic has changed the landscape of health delivery, substantially altering structures of negotiated order between prescribers and CHNs by forcing them to fall back on telehealth and remote consultation options such as telephone, FaceTime, or Zoom. Such remote consultations may remain routine post-pandemic (Richardson et al., 2020) My participants reported that telehealth introduced difficulties to symptom identification: “Through COVID, GPs weren’t coming out. They wouldn’t come and see anyone. They tried video calling but it was just very jumpy… We’d been doing a lot of conversations over the phone. It’s hard to explain how somebody’s presenting without actually seeing them…” (Mrs. Laney, Nursing home manager, Participant 19) Such accounts accord with various studies on telehealth during COVID-19, including one study examining communication styles between patients and providers during virtual visits. This study reported an increased ratio of physician to patient conversation, indicating physician verbal dominance, and increased requests for repetitions, suggesting perceptual challenges during telemedicine visits (Agha et al., 2009). In another study, health care providers expressed that video consultations failed to capture changes in facial expressions quickly enough and attributed most of their challenges to technology (Maslen, 2017). In my data, one GP was reported to have responded to these communication difficulties by granting care homes greater freedom regarding prescribing decisions: “After Covid, GPs don’t come in. They just rely on our information. If we ask them for Just in Case medicines, they just give them. They don’t want to come in and see anybody”. (Mrs. Wright, Care home manager, Participant 2) Thus, as in Allen’s (1997) study, a lack of face-to-face interaction resulted in nurses picking up the responsibilities of GPs who could no longer fulfil them. These responsibilities included, firstly, routine in-person symptom identification (‘our information’), and secondly, the ultimate power to trigger AM prescriptions. 118 For some participants, GPs’ unwillingness to visit care home during the pandemic also extended to an unwillingness to provide those care homes with anticipatory medicines: “We did struggle with the GPs… We didn’t really have much contact with them because they weren’t willing to come out… nobody was there to put the Just in Case medicines in place. (Mrs. Larkin, Care home manager, Participant 20) My data thus displays two varying responses from GPs to their newfound inability to carry out face- to-face interactions during the COVID-19 pandemic. These were prescribing freely (‘they just give them’), a form of over-prescribing which could have led to wastage, and under-prescribing (‘nobody was there to put [AMs] in place’), which caused shortages. An earlier-cited participant, Mrs. Raymond, similarly described a double bind wherein being prescribed an ‘anticipatory box’ would have constituted ‘overprescribing’ and thus ‘waste’, whereas not being prescribed one led to the equally severe issue of shortage. My participants’ experiences during COVID-19 thus once again illuminate the simultaneous issues of wastage and shortage that constitute the anticipatory prescription paradox. Tellingly, Mrs. Raymond remarked that the issues she encountered while navigating and negotiating this difficult prescribing process would have been alleviated ‘If they [GPs] visited more often’, that is, if face-to-face interactions had been more frequent. Allen’s (1997) study also reported nurses having to navigate double binds while dealing with the difficulties that ensue in the absence of face-to-face negotiations. Here, she explains: ‘Rule-oriented boundary-blurring occurred when nurses worked in the spirit of one rule even if this meant breaking another. Nurses routinely gave saline flushes after the administration of intravenous antibiotics even if it was not prescribed. Here, nurses were honouring the hospital policy that recommended saline flushes to follow intravenous antibiotics, although they were prohibited from prescribing medications’. (p. 511, Allen, 1997) This can be compared to the examples just discussed, in which GPs were unable to meet with residents face-to-face, leading to some CHNs in my dataset assuming some GP responsibilities and thus engaging in ‘boundary-blurring’. Just as the nurses in Allen’s study had to contend with a paradoxical double bind in which nurses could not honour one rule without breaking another, so too did the CHNs in my dataset (such as Mrs. Raymond, just mentioned) have to navigate the anticipatory prescription paradox, wherein accomplishing one goal (preventing shortages) meant failing at another (preventing wastage). Allen describes the ‘saline flush’ double bind she observed as a consequence of the fact that the professional boundaries of the hospital hierarchy were not fixed in place and were instead subject to the sort of improvised boundary-blurring associated with negotiated order. Allen’s nurses could either keep to their circumscribed professional roles but in doing so allow 119 a patient to go without a hospital-sanctioned saline flush, or they could administer the unprescribed saline flushes but in doing so overstep their professional bounds. Similarly, in my data, professional boundaries were flexible due to negotiation. CHNs who lacked negotiating power enjoyed no influence over prescribing decisions, leading to shortages; on the other hand, CHNs who did have negotiating power were able to essentially take over aspects of prescribing responsibility normally reserved for GPs, thus transcending their professional bounds, but potentially leading to wastage. AP can be understood as a system wherein the power attached to the nurse role is in flux and its place in the social order is, at any given moment, subject to the vagaries of negotiation. It is possible to see how such a system could also produce a state of affairs in which both shortage and wastage paradoxically occur at the same time. The anticipatory prescription paradox – that is, the simultaneous existence of shortage and wastage – follows directly from the status of the nurse as an agent who, during the prescribing process, enjoys either little negotiating power (leading to shortages) or much negotiating power (leading to wastage), depending on the nurse in question. 5.5 Training vs. Support My data suggests that whether or not a nurse successfully or unsuccessfully negotiated, thus causing either wastage or shortage, depended above all else on perceptions of that nurse’s level of expertise: “I think having stocks of controlled drugs at nursing homes without direction is potentially risky… Nurses might feel tempted to administer outside of their level of comfort or level of expertise...” (Dr. Murray, General Practitioner, Participant 39) This participant, Dr. Murray, also explained that part of the reason that wider ranges, rather than smaller fixed doses, have lately been discouraged in palliative care as a whole has to do with nurses being perceived as ‘inexperienced’: “In the community it’s also probably the least safe place to have ranges because you’ve often got very inexperienced district nurses going out to see these patients. We are supposed to prescribe to the least experienced person…” (Dr. Murray, General Practitioner, Participant 39) The convention of ‘prescribing to the least experienced person’ may bear the unintended side effect of causing prescribing professionals to default to the assumption that CHNs are inexperienced and thus under-prescribe, often leading to the latter being forced to negotiate for a prescription. This was described by the following participant: 120 “The GP couldn’t come out for a week at least. He just wrote me off. Then I came out and said that I’m a nurse manager, and I’m asking for medication because our resident needs it imminently. That he would die over weekend, and he did. Everything was sorted in hours. The GP came out to see us and prescribed the anticipatories with no worries. As a residential care home, we’re not seen as professionals who know what they’re talking about so I’ve had to invoke my credentials to get things moving”. (Mrs. Whitfield, Nursing home manager, Participant 21) The account also offers a valuable insight into how GPs’ concerns about care home professionals’ ‘levels of expertise’ – such as those just cited from GP Dr. Murray – are experienced from the point of view of the care home professionals themselves. A care home manager remarked: “There’s not really any prior…training? You know what, that’s not the word as such. It’s support, really. We need some support with the medications, in terms of what to expect and maybe the reasons why certain end-of-life medications are prescribed as opposed to others…” (Mrs. Larkin, Care home manager, Participant 20) Thus, from the point of view of GPs, prescribing decisions are accompanied by an implicit assessment of the expertise of the care home professionals being issued the prescription (leading to CHNs having to ‘invoke their credentials’); from the point of view of care home professionals, however, these same prescribing decisions may be accompanied by an assessment of how supportive that GP is. The fact that Mrs. Larkin, in the above account, brings up ‘training’ – an issue of expertise – and then revises her choice of word, substituting it with ‘support’, suggests that she believes, to some extent, that a deficit of training can be compensated for by a surplus of support. GPs, however, are far more likely to see support and training as mutually exclusive, and to believe that CHNs are not entitled to their support (in the form of generous prescribing) until they have received adequate training. The fact that prescribing is a negotiated process introduces considerable variability to prescribing outcomes. One nurse may negotiate a more generous prescription, while another may experience under-prescription because they are unaware that the onus is on them to negotiate – they may instead expect that a GP will approach prescribing with an eye towards filling their gaps in knowledge (i.e. supporting them), rather than effectively punishing those gaps with under-prescription. Thus, one nurse may secure more medicines, eventually leading to wastage, while another may secure very little medicine, leading to shortage. Both wastage and shortage can thus, paradoxically, occur simultaneously because the role of a nurse simultaneously commands so much and so little power within prescribing negotiations. 121 Paradoxes are often the result of two parties within a decentralised system interpreting one thing in two different ways (Smith & Lewis, 2011; Hoffman, 2001), leaving each party unable to reconcile the contradiction between their expectations and reality. A CHN who views providing support as a GP’s role will view the decision to deny or under-prescribe anticipatory medications in spite of clear clinical need as paradoxical, since this GP will be acting in opposition to their supposed purpose. Cross-examining GPs’ and CHNs’ expectations of each other, as revealed in interviews, allows one to understand the contradiction between the two sets of expectations – that one party views prescribing as a possible opportunity to educate, while the other views prescribing as a privilege which can only be provided to those who are already sufficiently educated. This contradiction ripples through the prescribing system as a whole, eventually leading to the paradox of simultaneous shortage and wastage. Another mismatch between GPs’ and CHNs’ expectations of one other concerns what ‘support’ actually means to both parties. Mrs. Larkin, cited above, expected support from GPs not only in the form of prescribing but also in the form of information. In particular, she suggested that GPs might support her staff by instructing them on ‘what to expect and maybe the reasons why certain end-of- life medications are prescribed as opposed to others’. While GP participants largely based prescribing decisions on their perception of a CHN’s ‘level of expertise’, CHNs viewed their expertise not as something fixed, but as something which GPs had the power to augment through the provision of prescription-related information that would allow them to administer more effectively. ‘Expertise’ is arguably not solely a matter of specialised training; it also encompasses knowhow. This knowhow is often picked up here and there, ‘on the job’. A hospital-based study by Nugus (2010), for instance, describes ‘care delivery’ being ‘negotiated informally in corridors, offices and specific treatment areas throughout the unit’. Allen (1997) revealed the importance in hospitals of the informal process of asking questions to colleagues. One of my participants attributed high standards of care to the fact that he was positioned within a close-knit island community which fostered the conditions for such informal conversations: “I’ve got a working relationship with the hospice consultant, who on the mainland I probably wouldn’t even know, let alone have actual conversations with them about patients and life. It might only work if you’ve got that personal relationship”. (Mr. Thomas, Senior paramedic, Participant 40) The advantage of informal questioning is that it covers the minutiae of clinical experience, in a way standardised, official training may miss. Mrs. Larkin earlier suggested that ‘support’ might be a more important consideration than ‘training’. Indeed, if prescribing professionals began to perceive CHNs’ ‘inexperience’ not as the result of a lack of specialised training but also a lack of knowledge¸ this might help them to realise that this lack of knowledge is something they themselves can help to redress, via informal information-giving processes which occur outside the bounds of official 122 training. Unfortunately, one participant suggested that this kind of informal information-giving was lacking in care homes as opposed to hospitals: “In a hospital setting, they’re supported by a bigger team… there are other people on the ward that they can go and talk to, or there are other people within their setting that they can pass a decision up to or sideways to”. (Mrs. Raymond, Care home manager, Participant 25) Another participant gave an insight into the importance of being able to ask colleagues for information, and also acknowledged that this kind of support was more available for nurses working in hospitals: “I think there probably could be a bit more training on End of Life, specifically just recognising when they might need medication… When I first started in the care home, I was a bit worried to give anticipatory medications, because I didn't really know how to do anything… So, it did feel impossible. But there were senior nurses and nurses that have been here for a long time, so I could always ask them for help and I did that for a while. And the hospice are really great too. They are there if we need them, but there’s more pressure, more risk in making those decisions, like when to start someone on morphine. Whereas obviously in the hospital, you know, we'd have palliative care nurses that would come to the ward every day, see how they were, and they’d set up the syringe drivers. They’d do sort of thing… In a way it's a little bit easier in hospital, just because you've got that team around you”. (Mrs. Bennet, Nursing home manager, Participant 11) This participant pointed to the need for more training in order to fill gaps in her knowledge, but also indicated that, in practice, what improved her ‘knowledge’ of AM administration was not official training, but the ability to ask her colleagues for help. Once again, this points to the need to acknowledge that CHNs often gain expertise through the interprofessional transfer of information – a process which often occurs informally – rather than, as GPs often assume, through formal training. Mrs. Larkin’s earlier-cited account offers a specific example of a topic about which GPs could profitably educate CHNs, thereby facilitating interprofessional information transfer – namely, ‘what to expect and maybe the reasons why certain end-of-life medications are prescribed as opposed to others’. My dataset provides some answers to this question. In interviews, participants often cited their reasons for opting for one drug as opposed to another as being related to supply shortages or cost: “We've moved now nationally to using morphine instead of diamorphine, in part because of the diamorphine shortages”. (Mr. Rowlands, Community pharmacist, Participant 5) 123 “Our first line opiate is morphine because it’s the cheapest. We only prescribe oxycodone if there is a reason to prescribe oxycodone… and it’s not that easy to get out-of-hours either”. (Dr. Cameron, General Practitioner, Participant 6) If not explicitly given information regarding why one drug is being prescribed over another, it would be possible for CHNs to develop misapprehensions about certain drugs. They could, for example, make the assumption that morphine, on evidence of its frequent provision, is an inherently more effective form of pain relief than oxycodone, thus making clinical correlations where only practical or logistical considerations regarding stock issues are relevant. Dr. Cameron described his experience explaining to nurses that his prescribing decisions were partly motivated by stock availability of drugs, and not always simply by ‘which is the appropriate drug for that symptom’: “It does make it difficult for medics, as well as the patients themselves, if you can’t get hold of a drug at the weekend. Because then you’re worrying about what you can change it to so they can get hold of it, and is it safe? … And that’s outside of which is the appropriate drug for that symptom. So, if someone’s got neuropathic pain or bone pain or visceral pain, then you might sort of opt for different things with regards to that. There’s a lot of factors that go into the decision about prescribing that you don’t think about until someone asks you the question. I’ve been taking nurses through their prescribing over the last few years and when I’m out with them, I’m having to justify what I’m doing when I’m doing it”. (Dr. Cameron, General Practitioner, Participant 6) By explaining his prescribing decisions to nurses, Dr. Cameron – rather uniquely within my dataset – answers to the need voiced by Mrs. Larkin for nurses to be educated by GPs as to why certain medicines are prescribed as opposed to others. This can be considered a positive instance of information concerning medicines being freely given where it is sorely needed. In stark contrast, another participant expressed frustration about the scant information she sometimes received from pharmacists regarding stock issues: “The pharmacy we’re allied to were a bit vague. Pharmacies were going out of stock of medications we needed like morphine and GPs were unaware of stock issues… unaware of how that might impact on us or our residents”. (Mrs. Raymond, Care home manager, Participant 25) A much more positive experience was reported by the GP Dr. Jameson, who sought out information regarding stock shortages in advance of prescribing: 124 “The prescriptions are initiated by us at the surgery. We then liaise with the care home and the chemist to make sure that any drugs that are being given on a PRN basis are available…” (Dr. Jameson, General Practitioner, Participant 38) This measure was designed to prevent prescriptions being written for drugs which ultimately prove to not in fact be available – a scenario reported by many participants, including one (Mrs. Haberman) who remarked: ‘Then the nurse-prescriber or GP will prescribe another driver and you haven’t actually got enough for the next driver’. By actively seeking out information about stock availability from the chemist, Dr. Jameson appeared to be conceptualising the transfer of information as an inherent part of the prescription process. He thus arguably conforms to the ideal of GP ‘support’ held by care home professionals such as Mrs. Larkin, who suggested that prescriptions be provided to her nurses packaged with supplementary information about the prescription. Another participant aligned even more closely with this ideal: “I’ve always been quite interested in therapeutics, but also interested in medicines management, in how you use the tools that we have to try and address some of those problems... I write the anticipatory prescribing chapter for our formulary because I’ve got the most interest in the community side of things”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) By taking the information and expertise picked up from his experience in palliative care and circulating it to the wider care sector, Dr. Powers exemplifies the ideal of a medical professional who freely volunteers information and thus improves care all around. In situations where professionals take the initiative to transfer important information to each other, my data showed that interprofessional relationships are strengthened, which in turn expedites the prescribing process, often preventing shortages: “The care home that I work at has one practice, not different GPs, and because of that, we get to know that GP well, they get to know us well and respect our judgement”. (Mrs. Kinley, Nursing home manager, Participant 10) “We have a fabulous relationship with our GP surgery. We have a good relationship with our community matrons, with all our district nursing team. We have a really fabulous pharmacy that provides all our medication quick, quick, quick. We’ve got good communication with everybody. The thing is, none of us are afraid to challenge or chase things up. We’re all here to look after the residents, they’re our main concern…” (Mrs. Ainsley, Care home manager, Participant 15) 125 It is notable that these relationships were characterised by ‘good communication’ and professionals ‘getting to know’ each other well – both examples of healthy information transfer. My data shows that when interprofessional relationships were poor, care homes were far more liable to out-of-hours shortages: “Running around town for prescriptions and medicines and dealing with [pharmacist] colleagues who are disrespectful and don’t care anymore, it can’t be sustained… We need better connections with the community teams but there’s so few people who stick around long enough to keep them up”. (Mrs. Claremont, Care home nurse, Participant 32) Mrs. Raymond, cited earlier, reported her experience of a pharmacist not adequately informing either care homes or GPs about stock shortages. Yet information is a ‘two-way street’, and should be requested as freely as it is provided (Pham & Williamson, 2018). The GPs in question could also have prevented shortage-related out-of-hours crises by seeking information about potential shortages in advance. For instance, by keeping pharmacists ‘in the loop’ while initiating his prescription, Dr. Jameson avoided prescribing a drug that would not, in practice, be available. When professionals freely provide each other with information, each party acts as a system of checks and balances for all other parties, thereby addressing gaps in knowledge. The 2010 Mid Staffordshire Inquiry, which unearthed ‘conditions of appalling care’ in a Staffordshire hospital, in part attributed this lapse in standards to a culture in which ‘[o]rganisations relied on others to keep them informed rather than actively seeking and sharing intelligence’ (Francis, 2010). My dataset showed that end-of-life care improves across the board when information is freely volunteered – more information, that is, than is strictly needed by law. Dr. Jameson, for instance, was under no obligation to seek information from pharmacists regarding whether the medicines he was prescribing were available. His approach contrasted with that of other GPs, for instance one described by Mrs. Raymond: “Often, medical professionals are more concerned with, ‘Do they want this medication or that medication? Do we wait for the rattle on the chest before we give this? Or do we make sure there’s an anticipatory box in place in the care home?’ It really only extends as far as ‘Do we need anticipatory meds or not? Is it a weekend? Do we have to do it now, just in case? Is it going to be a bit difficult? Do you need a prescription?’ That’s as much input as they want to give. But there’s more to it than that, if you want to support somebody with dignity”. (Mrs. Raymond, Care home manager, Participant 25) Thus, prescribing professionals were identified as both seeking and providing only the minimum amount of information needed to be able to carry out their prescriptions. Mrs. Raymond suggests that ‘there’s more to it than that’, and what these professionals might have done that would have exceeded the bare minimum information required to prescribe would have either seeking contextual and supplementary information regarding resident conditions or providing supplementary 126 information regarding why they prescribed the medicine that they prescribed as opposed to any other medicine. In practice, such information was seen as surplus to requirements. An example of poor information transfer in my dataset can be seen in the following account: “Sometimes the GPs won't tell us directly when our residents are end of life. They'll just prescribe the Just in Case pack when they think so and move on”. (Mr. Anderson, Care home manager, Participant 35) When professional groups do not adequately inform each other of issues (e.g. stock shortages), it becomes possible for a prescription to pass from GP, to pharmacist, to care home professionals, and finally to the intended resident, with only as much information passing from each agent to the next as is needed to allow that process to continue. For all intents and purposes, this amounts to the AP process inadvertently placing crucial clinical information on a need-to-know basis. UK care homes are quite reasonably required to follow Caldicott Principles, meaning that only the minimum amount of ‘patient identifiable information' concerning residents is exchanged, and only when absolutely necessary. Yet for this same paradigm – calculated to ensure personal information stays private – to unintentionally extend to clinical information suggests a system working at minimum capacity. GPs, pharmacists, and care home professionals all find themselves pressed for time and, quite understandably, have an incentive to streamline all interactions rather than to ask questions. However, when care home professionals seek out information about, for instance, medicine shortages in advance, this can prevent them from wasting more time later than they otherwise would have, should they be forced to attempt to locate a pharmacy in the middle of the night. The best way to combat an inadvertent ‘need-to-know’ approach to prescribing, then, is to reason not the need – to volunteer information even when it is not strictly needed. My data shows that, when relationships between GPs, pharmacists, and CHNs were sufficiently strong, information did not even necessarily need to be consciously transferred any more, since most of the relevant information was already intuited, simply by virtue of familiarity between individuals. In one case, a pharmacy became familiar with the prescribing footprint of a GP: “I tend to prescribe the same stuff again and again and again for the people, so the pharmacy that deals with it has got used to my prescribing”. (Dr. Murray, General Practitioner, Participant 39) In a similar case, a pharmacist became familiar with the routine of CD destruction, streamlining a complicated process that could otherwise have caused frequent delays: 127 “Part of the service to care homes that we provide involves giving a CD destruction kit to them. When they need it, they just need to give us a ring and we put in the delivery for them… With my care homes, I know the staff who come in. We have rapport and they know what to do. They’ve got a book where they record the medications; administration, death, expiry, everything. And they will bring back all the medications to us with the book, and I will sign it”. (Mr. Nanda, Pharmacist, Participant 41) Another participant, Dr. Jameson, similarly attributed positive outcomes to interprofessional familiarity, to what Mr. Nanda above calls a ‘rapport’: “I think actually that the biggest thing is the consistency of the relationships. I’ve been going into the care home now for 15 years so I know all the nurses really well”. (Dr. Jameson, General Practitioner, Participant 38) Interprofessional relationships are strengthened by repetition; after enough of the same type of professional interaction, a professional can begin to anticipate what a colleague will do. Dr. Jameson’s account demonstrates that when interprofessional lines of communication were maintained, resulting in strong relationships, problems automatically began to be spotted before they arose: “We… liaise with the care home and the chemist to make sure that any drugs that are being given on a PRN basis are available, and that the drug charting arrangements are robust. We have specific charts for things like syringe drivers and anticipatory drugs that we can use as well. In fact, our local drugs charts are really good because they’ve been redesigned relatively recently, so they are very, very safe to fill out. They’re quite prescriptive and so it’s easier for the out-of-hours service. But as a rule, I think we’re quite good at spotting problems before they arise and for most of those patients, it’s unusual for there to be a surprise that the out-of-hours service are going to need to pick up on. It does happen rarely, and in that case they prescribe, but most of the time we’ve been expecting something”. (Dr. Jameson, General Practitioner, Participant 38) Here, Dr. Jameson lists multiple examples of taking measures to improve interprofessional relationships, even ones not strictly needed: liaising with pharmacists, double-checking charting arrangements, redesigning local drugs charts, etc. None of these measures, individually, go a particularly long way towards improving the relevant professionals’ ability to actually predict when a given resident may reach the end-of-life stage and require medication, but together, they do seem to do so. Counter-intuitively, it may be that the best way to anticipate problems is not to isolate a specific problem and attempt to address it in advance, but rather simply to account for a spread of possible variables and contingencies in order to get a sense of all the things that can go wrong – 128 something which can be achieved by becoming familiar with the roles of all professional groups involved in the AP process. Anticipating a specific need to some extent implies already knowing what is needed, and if this is the case, then what is being done is not really anticipation at all, it is, rather, reacting to a need which already exists. Counter-intuitively, it may be that true anticipation cannot be actively accomplished; it emerges bottom up out of a general spread of actions that can be taken to maintain contact with other professional groups. True anticipation can only be a side-effect of this practice. 5.6 Anticipation vs. Reaction I will now explore the concept of anticipation – nominally, the cornerstone of the anticipatory prescription paradox – in greater depth. Differences in understanding between professional groups regarding what anticipation entails likely contribute to the paradoxical nature of anticipatory prescribing. Mrs. Raymond’s interview showed that CHMs believe that anticipation is partly the responsibility of GPs: “The decision to start anticipatories is usually a bit rushed and it’s not always thought about as far in advance as I’d like it… GPs tend to be reactive, not proactive… and certainly clinicians in the community tend to be reactive, not proactive, and only think about anticipatories when we’re approaching the last few days of life, which doesn’t really give anybody much time to have any real significant input”. (Mrs. Raymond, Care home manager, Participant 25) “The GPs aren’t forward-thinking about what’s going to be needed, so getting a prescription in place has always been hard”. (Mrs. Raymond, Care home manager, Participant 25) These accounts suggest, paradoxically, that arrangements for medicines called ‘anticipatories’ are typically poorly anticipated – made on a last-minute basis, prescribed, dispensed, and administered in haste. In fact, my data shows that the process of anticipatory prescribing was usually only set in motion once residents were already in their last days, making that process ‘anticipatory’ in name only. One participant, Dr. Kennedy, reported explaining to family members of residents ‘that we want to have anticipatory medicines in place before they’re required’. Intuitively, it seems that something called anticipatory medicine should by its very nature be prescribed before it is required; that Dr. Kennedy would need to specify this indicates the extent to which anticipatory medicines fail on their own terms. One participant account reveals an interesting discrepancy between how GPs and CHNs conceptualise the process of ‘anticipating’ AM shortages: 129 “If I’m prescribing an end-of-life patient Oxycodone during the week, we have to consider what’s going to happen at a weekend if they run out of medications. And I’ll be begging the manager or the district nurse, ‘Please don’t run out of medication,’ [laughs] ‘do not let this run out, make sure that you anticipate before the weekend how much you’re going to need”. (Dr. Cameron, General Practitioner, Participant 6) Here, Dr. Cameron uses the word ‘anticipate’ to mean not allowing medicines to run out, which may amount to an encouragement to ration medicines. Dr. Cameron’s advice echoes that of NICE (2020), whose aforementioned guidance also advocates rationing (‘prescribing smaller quantities’) in order to head off shortages, and the CQC, which states: ‘Staff must make sure ordering processes are robust enough so that people do not run out of these medicines’ (CQC, 2022). Doctor participants such as Dr. Cameron similarly implicitly place the onus of anticipation onto care home professionals. The idea behind anticipatory prescribing is that of taking advance action, of anticipating a care home resident’s trajectory so that said resident can be furnished with enough medication to meet their clinical need. Yet in his account, Dr. Cameron uses the word ‘anticipate’, referring to CHNs, in an effectively opposite sense; for a CHN to ‘anticipate’, in his view, means administrating as little medication as possible. Paradoxes often set in when individuals operating at two different levels of the same system interpret the same thing in different ways. A cross-examination of the CHN’s and GP’s accounts just cited reveals that, whereas CHNs view ‘anticipation’ as a GP’s responsibility, which they can fulfil by prescribing early, GPs may view ‘anticipation’ as CHNs’ responsibility, something which they can fulfil by administering conservatively. Within the internal logic of the term ‘anticipatory prescribing’, the implied thing being anticipated is the end of life, but the end-of-life period is often conflated with the last days of life – a time scale so narrow that nothing resembling genuine anticipation is truly possible. In other words, the implied object of the anticipation mentioned in the term ‘anticipatory prescribing’ itself precludes the possibility of that very anticipation ever occurring in the first place, making the term self- contradictory, even paradoxical. As one participant (Mrs. Raymond) put it, ‘clinicians in the community tend to be reactive, not proactive, and only think about anticipatories when we’re approaching the last few days of life’. The patient’s last days thus trigger a process which, nominally at least, purports to be ‘anticipating’ those very last days – thus putting the cart before the horse, essentially. While the end-of-life stage can in fact take days, weeks, or months, NICE’s 2015 guidance conflates the end-of-life period with only the last days of life throughout (NICE, 2015). One typically only consults this guidance after a patient has already reached their last days. Yet when one reads this document up to article 1.6.3, which advises to ‘Ensure that suitable anticipatory medicines and routes are prescribed as early as possible…’, this implies a timescale (that of weeks or months) which in fact outstrips the timescale to which the document as a whole confines itself (that of days). To really anticipate, to prescribe as early as possible in a true sense, one would have 130 needed to have already initiated the process; the instruction is thus not to do something but to have done something already. It is as if the implied ‘anticipation’ is being constituted retroactively. This oxymoron, retroactive anticipation, perhaps points to a deeper paradox within the concept of anticipation itself, one best understood via Aristotle’s paradox of future contingents, a problem of formal logic. Following Papachroni & Heracleous (2020), who integrate an understanding of Ancient Greek paradoxes into Smith and Lewis’ (2011) paradox theory, I take this paradox as a reference point. Aristotle’s formal logic is concerned with codifying the rules for true statements. Statements about the present, such as ‘England is a country’, or about the past, such as ‘The Battle of Hastings occurred in 1066’, can both be considered true statements. The problem of future contingents sets in once statements are made concerning future, or anticipated, events. Aristotle’s own example is the statement ‘There will be a sea battle tomorrow’. The statement cannot be proved true or false until ‘tomorrow’ comes, and yet, since the statement continues to exist today, and since a founding assumption of formal logic is that a statement cannot exist that is not either true or false (the so- called ‘Law of the Excluded Middle’), the statement still poses a logical problem. What the paradox demonstrates is that an anticipatory statement ultimately requires that the thing it is anticipating have already happened in order to hold any objective truth-value or conceptual stability, thereby defeating the very point of making an anticipatory statement in the first place. This paradox is felt in the world of anticipatory prescribing insofar as the act of anticipation is fraught with uncertainty, as it deals with ‘contingent’ events that have not happened yet (specifically, a resident’s death, which can happen at any time), while the act of prescribing implies, even prerequires, a degree of certainty (one comparable to the degree of certainty prerequired by statements of formal logic, which must be entirely and unobjectionably true to be made at all). That the act of prescribing prerequires a certain degree of certainty and finality is suggested by one of my participants: “I’m sure you’ll be familiar with the anticipatory prescribing processes around the Just in Case medicines. But those generally don’t get used until someone is really deemed very definitively end-of-life by a clinician”. (Mr. Smith, Nursing home manager, Participant 1) Indeed, prescribing produces the sense of a decision being locked in (in a manner analogous, perhaps, to a stock bet, another kind of ritualised prediction). The sense of finality attached to prescribing is also reinforced by the fact that the prescription is immortalised by the bureaucratic record kept of it, to which a prescriber can be held accountable forever after. The word ‘trigger’ – often used by participants, such as one (Mr. Jeffries) who spoke of ‘trigger[ing] the anticipatory medicines process’ – also carries a sense of a definite, irreversible action being taken. That prescribing is often reduced to a fixed, conclusive, binary choice was also suggested by Mrs. Raymond, who remarked: 131 “Often, medical professionals are more concerned with… Do we need anticipatory meds or not?... Do you need a prescription?… That’s as much input as they want to give”. (Mrs. Raymond, Care home manager, Participant 25) Ultimately, the need for certainty as a prerequisite for prescribing can conflict with the realities of end-of-life care, since death is to some extent inherently uncertain. One participant usefully framed end-of-life medicine as a field which should be able to accommodate this uncertainty, stating: “I think ultimately, the aim should be… that people can become ill at the most inconvenient of times and we are still able to respond to that”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) To this end, this participant brought up the Patient Group Direction (PGD), a legal document which allows paramedics to be remotely authorised to administer AMs, thus conferring the adaptability required to negotiate this uncertainty: “…it’s enabled us to respond to people deteriorating who, for a number of reasons, haven’t got Just in Case drugs in-situ. It might be that somebody should’ve done it and hasn’t… Or somebody’s changed their mind about wanting a very active treatment and then right at the eleventh hour when death is very close, they say ‘Actually, I don’t want to go to hospital. I want to stay here. I know I’m dying’, so a very last-minute change. That’s one group of people who I think have benefitted from the PGD, who were never going to look into the future and say, ‘Yeah, a palliative care plan, that’s for me, that’s great, I’d like some of that’. They were never going to do that, not until it’s right in their face. And the other group are people where things have just unravelled really quickly and so there’s just not been time”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) Rather than requiring certainty that a patient is ‘definitively end-of-life’, or requiring something as fixed as a palliative care plan, in order to become confident enough to provide patients with much- needed AMs, Dr. Powers acknowledges there that some aspects of death cannot be ‘definitively’ predicted, and instead devises a solution – the PGD – flexible enough to account for the possible range of variables and contingencies which can play out during death (e.g. a mistake regarding AM procurement, a last-minute change of location, a rapid deterioration or ‘unravelling’). As I argued earlier, accounting for a range of contingencies in this way can be considered a higher form of anticipation. Death – especially one’s own impending death – is a difficult thing to come to terms with, and it therefore stands to reason that a significant proportion of patients are ‘never going to look into the future’ and actively plan their own death. Rapid deteriorations, which cause out-of- hours emergencies, were reported frequently across my dataset, suggesting the need to view 132 unexpected outcomes like this not as anomalies, but as part and parcel of end-of-life care. One participant remarked the following: “Very occasionally someone can go and can kind of deteriorate without giving prior warning, so in that case, you sometimes end up trying to do everything quickly. There's been occasions when sometimes drugs aren't in stock at pharmacies and trying to find the medication has delayed things...” (Dr. Cairns, General Practitioner, Participant 37) It might be observed that this GP is implicitly shifting the burden of responsibility on the patient, to give ‘prior warning’ of their deterioration, once again calling into question the very concept of anticipatory prescribing – the provision of medicines in advance of clinical need. It is perhaps more appropriate to conceptualise ‘anticipate’ as an intransitive rather than a transitive verb; one anticipates, rather than anticipating some specific thing. As I suggested earlier, anticipation can be understood as a general state of heightened awareness, an increased sensitivity to the possible range of contingencies which can arise, a by-product of one’s familiarity with the full scope of a system, rather than a prediction, which by nature commits itself to one thing. In line with the findings of Etkind and Koffman, true anticipation requires tolerating uncertainty, remaining alert, adaptable, and open to the full range of possible contingencies that can arise (Etkind & Koffman, 2016). Death is inherently uncertain, and one way to combat this uncertainty may be to expect the unexpected. Yet prescribing requires certainty – a firm commitment to a specific course of action. In order reach the level of certainty that prescribing seems to demand, prescribers often simply wait until the thing their prescription is meant to anticipate (a resident’s last days) has essentially already begun, until residents are ‘very definitively end-of-life’, so that it is already known and no longer a surprise, to prescribe. Since, nominally, ‘anticipatory medicines’ are what they are prescribing, it can be said that this reaction, after the fact, to oncoming death is being retroactively constituted as anticipation (at least, within the internal logic of the term ‘anticipatory medicines’). In summary, the paradox that anticipatory prescriptions tend in practice to be one of the least anticipated forms of prescription has to do with an inherent tension between prescription, which requires certainty, and anticipation, which requires tolerating uncertainty. The fact that the two components of anticipatory prescribing – anticipation and prescription – contradict each other manifests in a paradoxical situation where AP is reactive (beginning after a resident enters their last days) yet retroactively describes itself as ‘anticipation’. The fact that AMs are also known as Just in Case medicines attests to the fact that successfully prescribing in anticipation means prescribing flexibly, prescribing before there is any absolutely concrete sign that the end of life is approaching. Some tolerance for the uncertainty inherent in anticipating something as unpredictable as death is thus baked into the concept of AMs. On the other hand, prescribing is an act which implies a certain level of fixedness and certainty (i.e. certainty that the medicine will be used). It is no accident that 133 the word ‘prescribe’, outside of the context of medicine, means ‘to lay down a fixed rule or law’. This incompatibility can ultimately be called upon to clarify the paradoxical relationship between wastage and shortage in AP. Wastage can be considered an inherent part of end-of-life care; AMs are prescribed ‘just in case’ or ‘as needed’, implying some expectation that they may not be needed, and that they may thus be wasted. Furthermore, it is typically the death of a resident that triggers the process of disposing of, and thus wasting, that resident’s medication. In practice, however, prescribers are heavily predisposed to avoid the wastage that the concept of AMs, by nature, seems to be designed to tolerate. As Mrs. Raymond put it: ‘I wish I could have an anticipatory box for everybody… But [GPs] don’t do that because then it’s overprescribing and it’s waste’. Thus, prescribers forgo anticipation and instead wait until a resident has already deteriorated, at which point they can be certain that what they are prescribing will not be wasted. However, waiting until the eleventh hour like this can causes shortages, as Mrs. Raymond’s account went on to show. Thus, the inherent tension between anticipation (which tolerates uncertainty) and prescription (which demands certainty) feeds into the wider anticipation paradox, the simultaneous occurrence of wastage and shortage. 5.7 Uncertainty vs. Certainty To summarise, when prescribers wait until the last minute – until a resident is only days away from death – to prescribe a form of medication named anticipatory medication, this creates, at the level of that name, a paradox, since a fundamentally reactive, belated measure is being retroactively imagined as something proactive, as anticipation. That death acts as the site of a paradoxical retroactive anticipation is a phenomenon which has been observed by Michel Foucault (1973), who contextualises the phenomenon within the historical emergence of autopsy as a method of understanding disease. As he explains: ‘Death, which, in the anatomical gaze, spoke retroactively the truth of disease, makes possible its real form by anticipation’ (p. 195). Foucault suggests that the end-of-life period which directly precedes death is a time when symptoms are most in flux, most uncertain (Foucault, 1973). However, the end result of this period of flux – death – fixes the body as a static text in time and space, thereby giving physicians a stable base from which to reverse engineer or reconstruct those symptoms and that flux via autopsy, even to the point of restoring the time axis of the disease by working out in what order the symptoms progressed. Thus, the symptomatic flux of the time of dying makes the time of death entirely uncertain. Yet this death itself, when it finally occurs, retroactively renders those symptoms certain, by allowing them to be fixed into a chronology via autopsy. The symptomatic flux of the end-of-life scrambles and defies all attempts to predict with certainty the time dimension of death – at what rate symptoms will develop, when death will come – yet the outcome to which it inevitably leads counter-intuitively restores that certainty. Foucault (1973) describes this reversal in paradoxical terms: ‘How could a simultaneous set of spatial phenomena 134 establish the coherence of a temporal series that is, by definition, entirely anterior to it?’ (p. 133) An analogy may be said to exist with anticipatory prescribing. Prescribers, tasked with navigating inherently uncertain end-of-life symptoms which take place during a dying period that can last months, can forgo this responsibility in favour of waiting for the exact moment (or at least, the exact days) of a resident’s death, at which point those symptoms are retroactively rendered certain. Throughout the patient’s period of symptomatic flux, it is uncertain whether their symptoms really do point to the approach of death; the patient’s condition could at any point change course and improve. Only when the patient is very definitively dying does the prescriber feel emboldened to prescribe, at which point the preceding period of flux is retroactively rendered certain. Looking back from the vantage point of the death, now reliably set in stone, it appears as if those symptoms were always pointing towards death, as if the anticipatory prescription had always accounted for that. One participant, describing a series of sudden deteriorations, suggested that some of these dynamics were at work during the COVID-19 pandemic: “…So the drugs were essential in that case because at that point, no one had predicted these patients were going to die. They didn’t fall under the pathway of an expected death, although the GP had the right to decide whether it was expected or not. That’s their call when they’re writing the death certificate”. (Mr. Thomas, Senior paramedic, Participant 40) Here the reality of how unexpected and uncertain death can be clashes with the certainty required by the bureaucratic architecture of end-of-life care (exemplified by a document known as a ‘pathway’ colloquially or ‘protocol’ formally which segments the dying process into a set of neat, expected steps). Though residents deteriorated and died in an unexpected fashion, the GP was able to retroactively declare these deaths as expected or anticipated (thereby preventing the legal need for a coroner’s investigation (Teggi, 2022)). According to Foucault (1973), autopsies, which allowed doctors to reconstruct living symptoms using a dead body, led to a change in the perception of death, which involved ‘see[ing] death in life… the beginning of a reversed time swarming with innumerable lives’ (p. 210). Death represents a fixed point in time; it places what Foucault calls ‘a black border’ around a life. From the outlook of a patient entering the indeterminate zone in which death approaches, the certainty of a specific time of death may appear elusive. Once that death has occurred, though, it provides a position from which to look back on the days or weeks before it with far more clarity in hindsight. The moment of death acts a fulcrum which, paradoxically, both promotes looking backwards and looking forwards. On the one hand, it allows for a hindsight which reverses the direction in which one contemplates time, and on the other hand, it cements once and for all the teleological, one-way flow of the life that had led up to the death, rendering it ‘finished’. Traditionally, prescribing was only advised within guidance once ‘irreversible dying symptoms’ had presented. Yet this approach was criticised for being too teleological and assuming that dying trajectories only operate in one direction. It was 135 found, for instance, that ‘insufficient attention [was] paid to signs of reversibility in the patient’s condition’ by practitioners using the Liverpool Care Pathway (Neuberger et al., 2013). One participant (Mrs. Larkin) demonstrated the non-teleological character of the end-of-life by recounting an experience in which ‘the anticipatory medications that [a resident] was prescribed were never needed… within about 24, 48 hours, he’d massively improved and they were like ‘Oh yeah, he isn’t end-of-life’. Anticipatory prescribing guidance has recently attempted to account for the reversibility of death, yet it has not reconciled this with its pre-existing emphasis on administering AMs only in the event of ‘irreversible’ dying symptoms, instead allowing the two contradictory instructions to co-exist in the same protocol (NHS, 2015; NICE, 2017). In essence, guidance and protocols advise that the process of anticipatory prescribing can only be triggered once irreversible dying symptoms are established. While the process is underway, however, CHNs and GPs are also instructed to monitor for reversible dying symptoms, whose existence appears to be precluded by the very premise of the AP process, so defined. Death is thus paradoxically imagined by the AP process to always be both reversible and irreversible, something which ties into the wider paradox between the uncertainty of a death which can always change course and prove reversible, and the certainty required by the prescribing process (and by rigid ‘pathways’ like the LCP), which requires that death to already be established as very definitively irreversible to be able to commit to a prescribing decision. That death will occur is certain, and this means a certain degree of AM wastage is inevitable. However, as Barclay et al. have shown in the context of care home patients, when death will occur is deeply uncertain (Barclay et al., 2014). The uncertainty of when death will occur makes shortages far more likely, as it prevents CHNs and GPs from being able to predict when the end-of-life period will start well enough in advance to put AMs in place in a timely fashion. The anticipatory prescription paradox – the co-existence of shortage and wastage – is thus attributable in some part to the paradoxical nature of death itself, whose co-existing yet contradictory elements of certainty and uncertainty cause wastage and shortage, respectively, to sit together similarly uncomfortably. The uncertainty of death presents a prescriber with two equally undesirable options – a paradoxical ‘double bind’. A prescriber can either hold back from prescribing in order to eliminate one consequence of uncertainty, the possibility of medicines being wasted if a patient proves not to really be end-of-life; or they can prescribe generously so that CHNs have enough medicines in stock to deal with the other consequence of uncertainty: shortage, caused by a resident suddenly deteriorating and requiring more medicine than previously expected. The duality of AMs, a feature which appears to pull prescribers in two different directions at once, is reflected even in the competing names used for them. They are known both as ‘Just In Case’ medicines – a name which, in acknowledgement of the uncertainty of death, implies they may well not be needed – and anticipatory medicines – which implies that they are anticipating some definite 136 event, and thus will be needed. My data shows that the sense that AMs are likely to go unused and may thus be wasted influences prescribers to prescribe conservatively in order to forestall wastage, thus potentially contributing to shortages. Conversely, a sense that AMs will be needed may influence a prescriber to prescribe more generously in order to prevent those AMs not being unavailable when they are needed (i.e. to prevent shortages); this generosity may contribute to wastage. These co-existing yet contradictory demands, concentrated within one and the same type of medicine, were reported by participants to result in seemingly paradoxical prescribing behaviours: “Or you might get a GP that says, ‘Well, best practice has always been to prescribe all four of the last medications.’ Or the GP that says, ‘Well it’s end-stage dementia, we’ll just give them some Midazolam and some Morphine Sulphate. They won’t need the other two,’ even when they do, I think. Sometimes it’s the same GP that says both”. (Mrs. Fields, District Nurse and Community End-of-Life Care Facilitator, Participant 7) “Sometimes I feel that the doctors prescribe too much, because sometimes they will prescribe 5 ampules for morphine and then, arbitrarily, it seems, they will prescribe 10 for the other resident. But they won’t listen if I say it’s too much or too little for the residents. Mostly the ones who are prescribed less run out but that one who has got 10 may not even need to take a single dose of them and they will have to be destroyed”. (Mrs. Barnes, Care home nurse, Participant 23) The subtraction of medicines other than Midazolam and Morphine Sulphate from the standard anticipatory pack of four AMs is likely motivated by a desire to reduce waste. If the very same GP elects at other times to prescribe the full pack, this may be to reduce risk of shortage in another care home. The simultaneous threats of wastage and shortage, concentrated in one and the same class of medicine, seem to oblige GPs to simultaneous straddle paradoxical opposites, to prescribe both ‘too much’ and ‘too little’ at the same time. 5.8 Holding on vs. Pushing away Another aspect of anticipatory medicines which was felt by my participants as a duality was their simultaneous ability to do good and, if over-administered, to cause harm, as demonstrated by the following account: “Anticipatories can be contradictions because you’ve never been told how to use them for end-of-life but you have been taught to use them at other times. So you have to know and be comfortable with the fact that a lot of our medicines do have a double edged sword, and that, actually, if you use them correctly, they can be very beneficial for our patients. It’s when we don’t use them correctly that it becomes a problem, like Gosport. But it’s, you know, it is 137 difficult for a lot of people to kind of get their head round that…” (Dr. Cameron, General Practitioner, Participant 6) This GP goes so far as to directly describe AMs as contradictions, thus providing some of the clearest evidence of the anticipatory prescription paradox discussed so far. The phrases ‘contradiction’ and ‘double-edged sword’ can be considered examples of the language of paradox, capturing as they do the phenomenon of two ‘contradictory yet interrelated elements that exist simultaneously’ (Smith and Lewis, 2011). The phrase ‘difficult for a lot of people to kind of get their head round’ also corresponds to the way in which paradoxes often defy intuition, since their constituent elements ‘seem logical when considered in isolation but irrational, inconsistent, and even absurd when juxtaposed’ (Lewis, 2000). Dr. Cameron’s account references a scandal at the Gosport War Memorial Hospital, wherein 450 patients had their lives shortened by being given large doses of opioids that were not medically indicated (Gosport Independent Panel, 2018). This event appears to have lived on in the minds of care professionals, prompting a shift towards more conservative prescribing, as indicated by Dr. Cameron’s observation that ‘after the Gosport issues and enquiries, as a profession, palliative care decided that ranges weren’t a good idea any more’. This culture of under-prescribing was frequently reported in my study to cause shortages. My data showed that CHNs often associated AMs with the fear of hastening death, as occurred at Gosport: “For most of these drugs, we never get a chance to use them, so they usually go out of date… We don't want to cause any complications or kill them. Those medications can be very dangerous to the person and us too, so most of the time we don’t use them”. (Mr. Smith, Nursing home manager, Participant 1) “I think once the GPs know that they’re at end-of-life, they’re very easily put in place because they’ll say straightaway, ‘Oh, let’s put anticipatory meds in place.’ But we don’t look at it like that… We use them very, very cautiously because we know what the outcome is and we know that once we start on that slippery slope, it is very unlikely that you ever get them back”. (Mrs. Lancaster, Senior care home nurse, Participant 16) “I think a lot of people here think it's safer for them for the resident to die in pain than make a mistake administering controlled drugs. They're dangerous drugs, to the resident and to us but in different ways”. (Mrs. Walker, Care home manager, Participant 4) This last participant suggests that AMs are dangerous to patients clinically (by hastening death) and to CHNs professionally (since this kind of clinical mistake can cost a CHN their job). The fact that 138 AMs appear in different ways to the two parties on either side of the administration process once again attests to their ‘double-edged’ or contradictory nature. This is also suggested by the following: “So sometimes, we feel like, you know, you’re sort of pushing them towards the edge when you’re supposed to be holding on to them. But then you can see this person is in pain and they really do need the Morphine. It’s a fine line”. (Ms. Jones, Care home nurse, Participant 27) Again, this participant grapples with the manner in which AMs paradoxically can simultaneously do good (by easing pain) and to do harm (by potentially hastening death, when used inappropriately). At the end-of-life, the duty to ‘hold on to’ patients and the side-effect of ‘pushing them towards the edge’ (by treating their pain) often paradoxically co-exist, pulling CHNs in two different directions at once. This duality is also reflected in NICE guidance (NG31), which contains the following paradoxical pair of guidelines: ‘When deciding which anticipatory medicines to offer take into account: • the benefits and harms of prescribing or administering medicines • the benefits and harms of not prescribing or administering medicines’ – NG31: Care of dying adults in the last days of life (NICE, 2015) Without AMs, residents can die in pain, meaning it would be harmful to not administer them; yet AMs are also imagined to hasten death, suggesting it would be harmful to administer them. AMs are therefore paradoxically both necessary and undesirable, leading to a situation where guidance paradoxically instructs care home professionals both to use them and not to use them at the same time. One participant’s account suggests that the issue of euthanasia, the deliberate hastening of death, may be a concern amongst CHNs as they negotiate the dangers associated with AMs: “Sometimes nurses are under the illusion that anticipatory medications are given so that a patient can die basically, like euthanasia, but that’s just not what it is. We’re just trying to keep them settled and comfortable”. (Mrs. Beverley, Nursing home manager, Participant 26) One ethical principle frequently called upon when discussing the topic of euthanasia, especially as it applies to the scenario of hastening death using pain relief, is the Doctrine of Double Effect (DDE). The DDE holds that an act can be considered acceptable if its primary intention (e.g. alleviating pain) is good, even if its secondary effect (e.g. the hastening of death) is not and/or if the secondary effect is not known about (McIntyre, 2004). The ‘double effect’ in question once again recalls the ‘double- edged sword’ of AP described in the account cited earlier. The DDE exhibits one of the main 139 hallmarks of a paradox, the phenomenon of one and the same action being simultaneously viewed in two different ways. The act of administering AMs, if rationalised using the DDE, is simultaneously viewed as both inherently good, and as the direct cause of a self-evidently bad effect. Shai Lavi contextualises this tension between medicines as easers of death on the one hand and hasteners of death on the other hand within a historical shift in end-of-life care (Lavi, 2004). As Lavi explains, death was traditionally handled exclusively by the clergy, whose job was to administer hope related to the consolations of the afterlife. In the early 19th century, however, dying instead became a medical issue, handled by medical doctors. Lavi ponders: ‘how could the medical physician become the minister of hope, when it was medicine itself that announced that the dying patient was incurable?’ (p. 16) As he goes onto explain, this paradox led to the emergence of a troubling 19th century practice – that of deliberately hastening death: ‘It is precisely this paradox—that physicians did not have the means to cure dying patients, but nevertheless became the new governors of the process of dying—that characterises the rise of dying as a medical problem. And it is precisely out of this paradox that the final shift in euthanasia takes place: from the benign duty of easing death to the troubling practice of hastening death’. (p. 14, Lavi, 2004, Autopoiesis, Nihilism and Technique: On Death and the Origins of Legal Paradoxes) Whereas the clergy were content to ‘wait passively for death’, this practice was ‘no longer fitting for the modern science of dying, which demanded action’. Medicine demanded that ‘something must be done’, Lavi argues, and the ‘something’ ultimately chosen during this historical period was the hastening of death: ‘The overambitious desire to profess hope at the deathbed, despite the incurable condition of the dying patient, was the origin of the medical hastening of death as a last resort to the problem of dying. It is simultaneously the moment when all hope is lost, but also the moment in which a final effort is made to overcome the helplessness of the deathbed by hastening death’. (p. 19, Lavi, 2004) Lavi’s explanation shows that the tension between easing and hastening death has been a feature of end-of-life medicine for the last century. This tension was felt strongly by CHNs in my data, for instance, one already quoted (Ms. Jones) who felt as if she were ‘pushing [the residents] towards the edge when you’re supposed to be holding on to them’. Lavi’s account of the contradictions that arose when doctors sworn to preserve life were forced to accept death points to the wider paradox that afflicts end-of-life care as a field of health care. Modern medicine is founded on the imperative to cure disease, to preserve life at all costs, and as such mobilises all of its resources and infrastructure towards achieving that goal. In end-of-life care – the 140 only field in healthcare where a final cure is by definition impossible – this infrastructure is forced to struggle against its own nature. A GP cited earlier was quoted remarking that ‘[a]nticipatories can be contradictions because you’ve never been told how to use them for end-of-life but you have been taught to use them at other times’ (Dr. Cameron). Spending possibly the majority of his professional life prescribing medicines for the purpose of curing or at least relieving conditions, it stands to reason that a sense of contradiction would ensue from having to use these same medicines in a situation where no cure is possible. Another participant, a paramedic, revealed that the instinct for preserving life was so ingrained that it led paramedics to interfere in the process of a natural death: “Paramedics are very, very strongly drilled towards the preservation of life. It’s absolutely written into us. If you cut a paramedic in half, that’s what it would say inside. And one of the biggest problems I’ve faced as a paramedic in End-of-Life medicine is stopping my colleagues from trying to actively treat dying patients…” (Mr. Thomas, Senior paramedic, Participant 40) Once the importance of preserving life in medicine as a whole is acknowledged, one of the core concepts of end-of-life care – that of a ‘good death’ – is thrown into question: “As a nurse, you need to recognise that End of Life stage and make sure you have got the anticipatory medication in, just so that you can ensure a good death, if there is such a thing”. (Mrs. Kinley, Nursing home manager, Participant 10) A ‘good death’ is generally imagined to be a pain-free and comfortable one, yet there is an ingrained bias within the care professions towards viewing death as a failure or at least a worst-case scenario, something which could never be considered ‘good’. In this way, the phrase ‘good death’ is often felt to be paradoxical, as suggested by this participant qualifying her statement (‘if there is such a thing’). Nor is this the only contradiction created by healthcare’s implicit insistence on the preservation of life. Another contradiction, for instance, is that, on the one hand, this insistence results in the improved longevity and health of the population, and on the other hand, it increases the prevalence of chronic, age-related conditions like dementia, which can only occur if a patient’s lifespan extends long enough to accommodate them (Rechel, 2009). Thus, the problems faced by the patient population of end-of-life care can paradoxically be considered a negative effect of something entirely positive – the preservation of their lives. Foucault’s (1973) history of the discovery of the autopsy (by French anatomist Xavier Bichat) contains another useful insight into the ambivalent relationship between medicine and death, specifically that making use of dead bodies in order to increase medical knowledge prerequired that 141 doctors – whose entire field was predicated on preventing death – must go against their nature and accept death: ‘It was no doubt a very difficult and paradoxical task for the medical gaze to operate such a conversion. An immemorial slope as old as men’s fear turned the eyes of doctors towards the elimination of disease, towards cure, towards life: it could only be a matter of restoring it. Behind the doctor’s back, death remained the great dark threat in which his knowledge and skill were abolished… With Bichat, the medical gaze pivots on itself and demands of death an account of life and disease… Was it not necessary that medicine should circumvent its oldest care in order to read, in what provided evidence of its failure, that which must found its truth?’ (p. 146, Foucault, 1973, The Birth of the Clinic) (My italics) Within the ‘paradoxical task’ described here is contained much of the mystery and logic of the anticipatory prescription paradox. Anticipatory prescription entails taking an institutional infrastructure (that of prescribing) that has always been geared towards the preservation of life, and instead mobilising it towards the acceptance of death. As I have shown, this inherent contradiction in motivations produces paradoxical prescribing behaviour, with AMs being prescribed both too much and too little at the same time, a practice which also explains the co-existence of both AM wastage and AM shortage. Ultimately, however, my decision to entitle the theme of this chapter the ’anticipatory prescription paradox’, rather than the more direct ‘wastage-shortage paradox’, is that – as I have hoped to show – the paradoxical co-occurrence of wastage and shortage of AMs points to deeper paradoxes and internal contradictions within the entire idea of end-of-life medication, and beyond that, the field of end-of-life care itself. 5.9 Chapter summary In this chapter, I aimed to explore the simultaneous occurrence of AM shortage and AM wastage, which I termed the anticipatory prescription paradox. I broke apart and independently examined every component of the paradox: wastage, shortage, anticipation, prescription, and, finally, death, the thing which the term anticipatory prescribing implies is being anticipated. I covered these topics in the following order: shortage (and the difficulties in procurement that result from it), wastage (and the extent to which it can be considered an inherent part of AP), the independence of wastage and shortage, the prescribing process (and how it fosters negotiated orders which breed paradoxes), the concept of anticipation (which, paradoxically, often appears to foster less anticipatory or proactive practices than other forms of prescribing), and finally, death (which medicine typically strives to prevent at all costs but must, in the case of palliative medicine, paradoxically accept). 5.10 Discussion 142 o Contributions to theory In my analysis, my decision to apply a paradox lens to negotiated order yielded fruitful insights into the ways in which paradoxical tension emerged from within the loose and shifting boundaries of the interprofessional structures of the AP. In particular, the indeterminate negotiating power of CHN was shown to result in a corresponding indeterminacy concerning whether a given GP would prescribe generously or conservatively; at the level of the AP system as a whole, this resulted in AMs, paradoxically, being likely to result in both wastage and shortage at once. Of course, claiming that the indeterminate or negotiated interprofessional boundary between CHNs and GPs shown in my data was anything entirely out of the ordinary would be naïve; an important development in the modern debate around negotiated order is the realisation that processes of negotiation should not be considered the ‘other’ of institutional structures – that is, the point at which the rules that bind those structures break down – but rather as part of the essence of structure itself. Nugus (2019) contextualises negotiated order alongside symbolic interactionism, a sociological theory which explores the subjective meanings individuals attach to their environment, and defends the latter against charges of ‘astructuralism’, arguing that ‘interactionist researchers’ such as Dennis and Martin’s empirical focus on situated activity is due to the ‘manifestation of social structures in this very activity’ (Dennis & Martin 2007). Similarly, Allen (1997) notes that ‘a frequent charge’ levelled against negotiated order ‘is that the approach assumes everything is indefinitely negotiable’ – essentially, another accusation of astructuralism . Allen’s finding that nurses and doctors in a hospital were ‘negotiating’ the professional boundary between them without directly interacting at all ‘raises the question whether the concept of negotiation is simply a convenient shorthand for diverse processes of social interaction or whether its meaning is more restricted’. Again, there is a sense in which negotiated order is too loose, since practically anything can be considered negotiable, and this negotiation itself can be considered to entail practically anything – any form of interaction, or even non-interaction. The debate around negotiated order thus appears to be stuck in a stalemate between the binary of negotiation as subverting structure by operating outside of the system and negotiation as constitutive of structure, maintaining it from within the system. I have aimed for this chapter to contribute the notion of paradox towards the theory of negotiated order precisely so that this binary can be circumvented. Only paradox, which acknowledges the ability for contradictory elements of a system to co-exist, can account for the ways in which negotiation both subverts structure and subsequently constitutes it. Only paradox can accommodate the way in which elements internal to a system – internal contradictions - can nevertheless go against the grain of that system’s very essence, without being expelled from it. Since negotiated order emphasises the actual act of verbally negotiating so strongly, the fact that the nurse-doctor boundary could have been blurred in the absence of this kind of explicit, face-to-face 143 dialogue proved puzzling to Allen (1997). Paradox theory can be called upon to address this confusion, since paradoxes find their natural habitat in situations where direct communication is lacking. Paradoxes often result when different parts of or agents within a decentralised system are separate but communicating, such that enough information passes between them to remain in contact, while not being enough information to prevent differences in perspective (Hoffman, 2001; Lewis, 2000; Smith & Lewis, 2011). These differences in perspective are the direct result of the extent to which professionals who must collaborate do not communicate. My analysis showed that when GPs and CHNs were unable to communicate face-to-face, the status of the latter became indeterminate, such that the role of CHN simultaneously and paradoxically held both much and little power within the AP system. This paradox rippled out into the wider fabric of the care system, resulting in the simultaneous build-up of waste and shortage. While ‘negotiation’ as a concept tends to be difficult to constitute any other way than positively, as explicit verbal exchange, paradox is a concept with enough latitude to accommodate unintuitive situations in which something comes from nothing, in which social order is negotiated through the absence of communication. This is because paradoxes are by nature conceptually unstable, always both there and not there, something and nothing, since they consist of elements which mutually deny each other, rendering their existence logically impossible yet nevertheless undeniable. Cross-examining interviews to excavate paradox-inducing differences in perspective was a vital aspect of my methodology in this chapter. The importance of cross-examining accounts to unearth differences in perspective and thus understand systemic tensions is suggested by Hoffman (2001). It is in the interstitial spaces between professionals that zones of non-communication – sites of non- negotiated order, as Allen (1997) describes – emerge, leaving asymmetries in knowledge and understanding. Allen’s study on nurse-doctor boundaries in negotiated order, furthermore, notes a discrepancy between how nurses described their work (in interviews) and the work itself (described in observational work). To rationalise the discrepancy, Allen cites Silverman’s (1993) idea that interview data can be treated as ‘displays of perspectives or moral forms’ rather than ‘literal representations of reality' (p. 107, Silverman, 1993). Allen concludes: ‘we can understand the interviewees’ accounts in this study as reflecting real conflicts of interest and perspective, but this should not necessarily lead us to expect these tensions to be manifested in their daily interactions’ (p. 511). Indeed, it may be reasonable to go further and to argue that discrepancies between interview data and action can be considered rich data in themselves, as rich in their own way as are the paradoxical discrepancies that emerge when different participants’ interviews are cross-examined. Such discrepancies between what one says and what one does can be the result of paradoxical double binds, in which professionals must act in a certain way in spite of the fact that it conflicts with their own moral principles (Bateson et al., 1972). In Chapter 7, which covers the ‘paperwork paradox’ theme, I explore a scenario of this sort: namely the way in which CHNs, who subjectively feel that 144 patient-centred care is of utmost importance must act in a way that belies that impulse entirely and spend time that could be spent with residents completing bureaucratic paperwork (which, paradoxically, also purports to uphold the importance of patient-centred care), due to external regulatory pressure. Only by treating the contradictions between word and deed as natural consequences of the paradoxes of the AP system, rather than flaws in qualitative data, is it possible to understand why the issues of wastage and shortage can paradoxically coincide with a universal belief, seen throughout my data, that these issues should be avoided at all cost. At some point between people’s deeply held beliefs and the practical constraints of the system that binds them, paradox intervenes. o Contributions to literature Within paradox theory, paradoxes are described as ‘contradictory but interrelated elements that exist simultaneously and persist over time’ (p. 382, Smith & Lewis, 2011) (my emphasis). It follows that most particularly persistent paradoxes will span long enough to develop histories. I have attempted to show the historical background of two paradoxes which appear to be foundational to end-of-life care – the tension between medicine’s duty to preserve life and its helplessness in situations where it is forced to accept death; as well as the tension between easing and hastening death, one concentrated within AMs themselves. These two paradoxes were contextualised within two successive historical instances of modern medicine attempting to accommodate death: firstly, by making use of autopsies (in the 18th century), and secondly, by placing the dying phase under the control of doctors rather than clergymen (in the early 20th century). While palliative care as a specific discipline has its roots in the hospice movement of the 1960s (Clark, 2007), the latter of these two historical shifts can be considered a crucial chapter in its prehistory. To this extent, it can be said that the paradox constituted by the competing demands to ward off and to accept death is part of the historical foundation of palliative care. The idea that systems or disciplines can be built on paradox has slowly been gaining acceptance within the wider conversation between paradox theorists. Smith (2001) has argued that paradox is not a defect of a complex, organised system, but an inherent feature of it; Hoffman (2001) has gone further and suggested that paradoxes do not only relate to the systemic organisation of health care, but rather ‘appear to be integrated in medical thinking’ and ‘clinical’ reasoning, supporting this claim with references to the work of Jacob and Moore (1984) and Hunter (1996). Hunter suggests that ‘because uncertainty in all its variety is the central if never quite stated theme of clinical education’, this uncertainty can only manifest in the form of paradoxical maxims or epigrams which enshrine ‘practical wisdom’ (p. 225). Furthermore, she describes the fact that ‘medicine is a science; medicine is an art’ as the ‘overarching paradox of medical knowing’ - defining an ‘art’ as ‘an uncertain field of knowledge’, in which ‘no matter how sound the science may seem to be, exceptions inevitably arise’. I have attempted to add to Hunter’s work on this paradox of uncertainty by applying it end- 145 of-life care, which can be considered an uncertain clinical situation par excellence insofar as it can never be known with certainty when, how, and even if an end-of-life patient will die. This uncertainty often clashes with a medical infrastructure built on certainty – for instance, the prescribing system, which requires a bare minimum level certainty to make a diagnosis and thus justify prescribing at all. This state of affairs breeds the paradox in which ‘anticipatory prescribing’ (a form of prescribing designed to accommodate greater uncertainty and foster forward-thinking practice) paradoxically renounces all claim to anticipating anything and starts to be used more reactively than other forms of prescribing, only kicking in once a patient is already certainly going to die. o Recommendations for policy and practice 1. Reinstatement of Reuse of Medicines guidance In this chapter, I examined multiple accounts provided by care homes who made use of the Reuse of Medicines guidance during the pandemic (DHSC, 2020). The picture provided was positive overall, with issues of wastage and shortage sometimes appearing to be resolved in situations where a stock of anticipatory medicines was kept. A study by Antunes et al. (2022) similarly posits keeping stock as a potential solution to these issues. The Reuse of Medicines legislation was repealed in 2022 (DHSC, 2022). Based on my participants’ positive experiences with it, its reinstatement would appear to be worthwhile, although I attach to this recommendation the caveat that an expectation that wastage will be significantly reduced should be avoided should a care home not be suffering from shortages that provide an actual incentive to reduce wastage, as indicated by my cross- examination of Morris et al. (2013) and my own data. 2. Fostering strong interprofessional relationships My analysis highlighted the importance of building relationships between care homes and community colleagues. Through stronger relationships, care home participants were able to centralise AP bureaucracy to align regulations with local priorities and care home cultures. This aligns with a 2020 national report that emphasises the role of relationships, culture, and support between health and social care professionals (DHSC, 2020b). These relationships have been conceptualised as a form of ‘social capital’ that can bridge, accelerate, and enhance collaborative networking (Wallis, 1998). My data showed that strong interprofessional relationships are supported by robust lines of interprofessional communication, through which information is freely and organically volunteered, e.g. between pharmacists and GPs. There is arguably no better way of improving interprofessional relationships than placing different professional groups in the same space. My analysis suggested that the type of informal interactions between professionals – which occur, for instance, in corridors, as Nugus et al. (2010) highlights – can facilitate the free spread of information and thus improve health care outcomes across the board. The Medicines Optimisation in Care Homes (MOCH) program aimed to place two professional groups – pharmacists and care home professionals – in the same space by integrating pharmacists into care homes. The MOCH 146 program involved structured medication reviews, end-of-life care support, and frailty reviews, and reduced polypharmacy from an average of nine medications per resident to seven, with 17.4% of medications ceased (NHS, 2018). One CHM participant cited in this chapter reported not being properly kept abreast of information relating to stock shortages by pharmacists. Recent professionalisation projects have recognised the potential of pharmacists as key sources of information about medicines and have subsequently focused on improving knowledge transfer, cooperation and communication between pharmacists, GPs, and nurses (Breen et al., 2018; Goodman et al., 2016). 5.11 Conclusion Having interrogated the paradoxical relationship between wastage and shortage which forms an inherent part of the paradoxical dynamics of anticipatory prescribing, I will now move onto an analysis of the protocol paradox: the fact that, though dying is by nature an individualized process, end-of-life care is a field beset by standardised models of health care provision. 147 Chapter 6 The Protocol Paradox This chapter focuses on the Protocol Paradox theme and its tension-based subthemes. The chapter is organised thus: 6.1 Background, 6.2 Introduction, 6.3 Individualization vs. Standardisation, 6.4 Micro-Decisions vs. Algorithmic Decisions; 6.5 Individualized Language vs. Jargon Oblivion; 6.6 Complex Cases vs. Bureaucratic Shortcuts, 6.7 Chapter summary, and 6.8 Discussion, in which I summarise my findings and outline its contributions to theory, to literature, and for practice and policy. 6.1 Background The tension between individualization and standardisation in end-of-life care has been subject to vigorous debate in recent times. Since the Liverpool Care Pathway Review (2013) found evidence of a standardised, ‘pro forma’ approach to care and recommended its discontinuation, the true parameters of what constitutes standardisation in health care have been thrown into question. Knights defends the LCP, suggesting: ‘could ‘individualized care’ simply be a pleasant-sounding phrase…?’ (Knights, 2016) Negative responses to the demise of the LCP can be explained by the ‘clinical guidance vacuum’ it left (Jones, 2016), raising the troubling possibility that the alternative to standardised care may, practically speaking, tend to end up being, not individualized care, but an absence of care. The debate around standardisation goes back further than the LCP, however, and often hovers around the issue of protocols. A protocol, also known as a pathway or an algorithm, is a form of documentation which contains fixed instructions for carrying out certain practices, including prescribing and administration. Protocols often contain diagrams such as flowcharts, which connect symptoms with medicines using boxes and arrows, and which organise instructions into step-by-step guides. Rycroft-Malone et al. (2008) noted that while some professionals expressed concern that protocols could result in a ‘tick-box mentality’, protocolisation also allowed nurses to take on new roles which streamlined care. Pinder et al. (2005), on the other hand, argued that standardised ‘pathways’ may be inherently incompatible with personalised, patient-centred care. Timmermans and Berg (1997) noted no disparity between protocols and personalisation, arguing that the process by which a protocol is ‘tinkered with to make [it] workable in practice’ is ‘a sine qua non for the functioning of the protocol in the first place’. It is also important to acknowledge the importance of the literature on standardisation outside of the LCP context. A study by Nugus (2019) found evidence of standardisation within the bureaucratic patient-processing mechanism of a hospital, noting that complex conditions, largely afflicting 148 elderly patients, were being deflected by overworked doctors into generalised ‘catch-all’ departments. Such findings point to the need for a greater collective understanding of how standardised practices effect elderly patients, including patients in palliative care. There is evidence to suggest that end-of-life patients, who suffer from a high rate of dementia, have more individualized needs than that of other types of patient. Dementia at the end-of-life can result in an inability for a resident to articulate their pain, rendering it difficult for nurses to determine which AMs are appropriate for administration (Gilmore-Bykovskyi et al., 2013; Lundin & Godskesen, 2021) – a challenge which amplifies the need for an individualized care approach attuned to recognising the subtlest of symptoms. A study by Sheaff et al. (2017) found that patients were far more likely to take their pain more seriously than doctors were, describing their inability to articulate pain as a ‘discursive gap’ between doctors and patients. This discursive gap has been covered extensively by a body of literature centring on the issues of medical jargon and euphemism, standardised forms of language often used in order to shirk the responsibilities of personalisation. Barlet et al. (2022) have classified four types of euphemism for death used in medicine. Pitt et al. (2020) have similarly codified euphemisms for death, building on their earlier work in which they propose a classification system for seven types of medical jargon in order to protect against ‘jargon oblivion’. More recently, Batten et al. (2019) have observed the effects of jargon at work in the term ‘treatability’, which they found being understood differently by patients and by doctors. 6.2 Introduction This chapter will approach questions of euphemism, jargon, and, generally, the language of AP with recourse to the discursive theoretical lens offered by Michel Foucault’s The Birth of The Clinic (1973). The current literature on the discourse between doctors and patients, while substantial, would benefit from being contextualised with greater historical depth. Foucault’s approach, which situates developments in clinical language amidst shifts in clinical and cultural attitudes to death, proves tailor-made for a rich analysis of this area. His focus on how power structures are organised along linguistic lines also facilitates a fruitful exploration of the way in which non-verbal dementia patients at the end of life are placed at a disadvantage by their inability to communicate their pain, while health care professionals are placed in a position of power over them by their mastery of a well- gatekept and mystifying clinical jargon. My findings indicate that the prevalence of dementia and non-verbality at the end of life render dying patients especially deserving of individualized care that can accommodate their more involved needs. They also indicate, however, that end-of-life care is disproportionately dependent on protocols which render care overly standardised, as well as on nonspecific clinical jargon and euphemism. This leads me to formulate the protocol paradox: the 149 fact that, though dying is an inherently individualized process, end-of-life care is – as my findings will show - one of the more standardised areas of health and social care. This points to a wider systemic paradox within health care which results from the attempt to organise the fundamentally personalised, person-specific business of patient care according to a bureaucratic system which operates by means of standardised protocols and guidelines. This chapter suggests that the tensions which emerge from the protocol paradox can be explained by tracking how they manifest in language, particularly the forms of standardised language that collect around standardised care conventions such as protocols and guidelines. Research question • What role do the algorithmic and discursive structures associated with care home-based AP play in shaping the tensions between individualization and standardisation in end-of-life care? 6.3 Individualization vs. Standardisation One of the areas of end-of-life care where the protocol paradox is most immediately felt is within guidance documents governing best practices concerning AP. Paradoxically, these documents often deal in standardised rules, protocols, and prescribing algorithms, while simultaneously advocating for a personalised, individualized approach to AP for dying patients. One participant noted the discrepancy: “…we do a session for GPs on anticipatory prescribing. For ease of teaching we have to have some tools to be able to say, ‘This is what you do’. But, actually, you know, NICE have said that we’re supposed to individually prescribe. So how can you anticipate on an individual basis? It’s really difficult to do. So there are our standard four drugs, that we say, these will help these symptoms and these are most likely symptoms in the End of Life. But, actually, if your patient is stable on Metoclopramide, why not use that for End-of-Life, why do you have to use our standard drugs?” (Dr. Cameron, General Practitioner, Participant 6) One of the clearest examples of standardisation within the AP system is the standardised pack of four AMs which is typically prescribed at the end of life. These vary slightly by region, but often include the following four: morphine, for pain, midazolam, for agitation, hyoscine butylbromide, for excess respiratory secretions, and levomepromazine, for nausea (NHS, 2015, 2017, & 2019). Dr. Cameron suggests that other medicines may fall beneath the attention of prescribers simply due to the fact that they are not part of the familiar and expected pack. Thus, he expresses difficulty in 150 reconciling two aspects of AP which conflict with each other and yet paradoxically co-exist – regulatory bodies’ insistence on individualized prescribing, and the realities of standardised prescribing. Dr. Cameron’s confusion stems from the fact that he is being expected to ‘individually prescribe’ AMs, a type of medicine which is almost always seen as already standardised, as they are often prescribed as kits comprised of four medications at their standard doses. Dr. Cameron also helpfully localises the source of this paradox to NICE (2015) guidance, whose section on ‘Anticipatory Prescribing’ contains the following instructions: ‘1.6.1 Use an individualised approach to prescribing anticipatory medicines for people who are likely to need symptom control in the last days of life. Specify the indications for use and the dosage of any medicines prescribed. 1.6.2 Assess what medicines the person might need to manage symptoms likely to occur during their last days of life (such as agitation, anxiety, breathlessness, nausea and vomiting, noisy respiratory secretions and pain)’. – NG31: Care for dying adults in the last days of life (NICE, 2015) This guidance advocates for an ‘individualized approach’ to prescribing, but the array of symptoms mentioned as examples are roughly the same ones typically targeted by the familiar pack of four AMs. In the Gold Standards Framework Good Practice documentation, this standardised pack is similarly placed alongside a vague appeal to the importance of individualization: ‘Patient’s GP will prospectively prescribe appropriate medications on form FP10, which are likely to include diamorphine for pain; cyclizine, haloperidol or levomepromazine for nausea and vomiting; midazolam for agitation; glycopyrrolate/glycopyrronium or hyoscine hydrobromide for respiratory secretions; oral lorazepam tablets. Prescriptions and the medicines supplied will reflect the individual needs of each patient’. – Gold Standards Framework Good Practice Guidance (GSF, 2006 & 2019) The standardised pack of four AMs enjoys pride of place in this guidance, with the reference to individualization relegated to a single line, as though an afterthought. This document greatly clarifies the work of prescribing, clearly and transparently enumerating a navigable set of medicines and symptoms, which map cleanly onto one another. The reference to individualization, meanwhile, is murky, underdeveloped, and has the air of a platitude. It largely pays lip service to the idea of individualized prescribing, rather than providing any sort of instructions concerning how to achieve it, let alone instructions approaching the level of immediately actionable clarity exhibited by the symptom-medicines list directly above it. While a prescribing professional could easily begin 151 following a standardised prescribing route upon receiving the symptom-medicines list, they would have to actively go out of their way to interpret how exactly to follow the gesture towards individualization tacked on below it. The mismatch between these two articles of guidance may have to do with an internal tension, common within guidance documents, between two duties which they must fulfil simultaneously: that of fixing specific rules in place, and of representing a general body of values. Patient-led and personalised care can be generally considered a value or ethos, an abstract ideal towards which to aim, whereas an instruction to prescribe ‘midazolam for agitation’ takes the form of a specific guideline. It is common for guidance documents to open with a list of overall aims, or an overall ethos. The NHS’s Enhanced Health in Care Homes model (EHCH), for instance, states: ‘The EHCH model has three principal aims: (1) delivering high-quality personalised care within care homes …”, etc. (3.4, NHS, 2020) These, however, are vague, where others, including prescribing guidance which outlines the four anticipatory medicines within their standard doses, is specific. Intuitively, it makes sense to be specific about specificity and general about generalities. Within these guidance documents, however, individualization, a concept which entails catering to specificities, is maintained, paradoxically, as a general principle. Pinder et al. (2005) describe the way that personalised patient care is insistently yet vaguely upheld as the overall principle at the heart of standardised protocols – which are in fact fundamentally incompatible with that principle – as a kind of ‘tokenism’. Of all the general principles espoused by guidance documents, furthermore, perhaps the most vague and general is the principle of ‘good practice’. As one participant put it: “I think everybody loves to talk about good practice, but when it comes to actually giving the meds, some of my colleagues seem really unsure. I think some of them actually end up under-dosing residents who really need the pain relief. It's like playing it good and safe backfires and the ones we're supposed to help suffer”. (Mrs. Barnes, Care home nurse, Participant 23) By deploying a term as ambiguous as ‘good’ to direct or define an individual's practice, guidance documents create situations in which healthcare professionals have so little instruction regarding how to conduct themselves that they tend to default to doing as little as possible. Where there are no clear instructions, one is more likely to be penalised for doing the wrong thing than for doing nothing at all. It can thus be said that the term ‘good practice’, in being so ambiguous that it promotes inaction which often harms residents, paradoxically tends to result in bad practice. Like individualization, ‘good practice’ is a principle circulated within guidance documents that proves so general and nonspecific that it actually promotes its own opposite. As one participant observed, furthermore, the problem of nonspecific language of AP extends to other terms, such as ‘as directed’: 152 “…if I ever get a controlled drug that says as directed, nine times out of ten, I can’t administer it because ‘as directed’ means nothing. But getting the community pharmacist to understand that…” (Mrs. Raymond, Care home manager, Participant 25) Here, the jargon term ‘as directed’, like ‘PRN’, leaves decisions such as how often and how much of a medication to administer up to CHNs, thus theoretically giving them the freedom to better individualize their administration practices by reacting to resident issues in real time. In practice, however, since CHNs are typically punished severely for administering inappropriately, they do not feel confident in using this freedom, and thus tend to default to under-administering or not administering at all. Ultimately, this term proves so broad, ambiguous, and empty (it ‘means nothing’), that – like the vague appeals to ‘good practice’ and ‘individualization’ which appear in guidance documents – it undermines the very individualization it is intended to support. The basic unit of the guidance documents I have been discussing is the guideline. One participant (Dr. Cameron) remarked: ‘This is a guideline and it’s supposed to guide you, yes. But your patient should guide you more’. My data often showcased the conflict between individualized or patient- centred care – being ‘guided by the patient’, as Dr. Cameron put it – and guidelines, which are inherently standardised. The reason that guidelines can be considered inherently standardised is that they enshrine a particular suggestion in the form of a singular, authoritative, referenceable rule or instruction which applies widely across an organisation. A standardised rule has the useful quality of being replicable, meaning various actors across a given jurisdiction can adopt it, sparing each one from needing to individually come to their own conclusions about best practice. Guidelines can often be too rigid to be applicable to diverse situations; however, they are useful when this drawback is outweighed by the benefits of providing professionals with a useful reference point regarding what to do in any given situation, thus eliminating confusion by placing as many people as possible on the same page. One participant gave an example of a move towards standardisation which, unfortunately, did a disservice to the many individual care homes which it lumped together: “…we’ve just adopted Pan-London MAAR authorisations, which everybody hates. They are just diabolical because they’re different from what we usually do under the GSF… Normally the prescriber would say where to start and then give increments. But with a syringe driver, we don’t have that now. It’s up to the nurse that’s starting the syringe driver to make the clinical decision. So the new regulations and the Pan-London MAAR charts distort nurses’ confidence levels and now there’s fewer syringe drivers being put in place… There’s a PowerPoint that they did locally which explains why we went from one to the other… it means that you can nurse anywhere in London and you’ll have to do the same paperwork… 153 But everybody’s individual areas were tailored before… We don’t want to adopt and take on other areas’ at all”. (Mrs. Fields, District Nurse and Community End-of-Life Care Facilitator, Participant 7) Bureaucratic organisations (such as Pan-London, NICE, the NHS) work by consolidating separate institutions into one interconnected system. This account reveals one of the inherent flaws of bureaucracy: the fact that, the more varied are the institutions it places all under the same umbrella, the more general a bureaucracy’s standardised guidelines must become, often spreading themselves thin to account for all possible situations that might arise in any one of these very different institutions. Here ‘individual areas’ had developed individualized practices which were tailored to these areas. This tailoring was undone by the onset of guideline standardisation, whose purpose was ostensibly to make paperwork less confusing and put everybody on the same page by ensuring that all care homes throughout London were using the same forms. In fact, this standardisation made matters more confusing by introducing new rules which conflicted with old, well-known ones, causing CHNs to avoid taking action altogether in order to avoid being penalised. As one participant (Mrs. Raymond) put it, referring to CQC regulations and guidelines, ‘these one-size-fits-all policies never work’. She went on to speak directly to the issue of gaps in guidance documents: “I think the policy was written with the best of intentions, but like most of these things, it was written by somebody at the top of the company who’s never stepped foot in a care home… we get a piece of guidance or we have a hard and fast rule thrown at us, and the conversation will be like, ‘What?... Who said this? They want that of us? That’s not possible’”. (Mrs. Raymond, Care home manager, Participant 25) Indeed, guidelines traffic only in ‘hard and fast’ rules, in certainties. This allows them to standardise practice in order to reduce uncertainty. This can mean, however, that guidelines only focus on the few aspects of fundamentally uncertain situations that one can be certain about. An example can be found in a regional guidance document offering protocol for PRN medications: ‘‘PRN’ or ‘when required’ medication is medication which is… administered according to need of the service user. Service users who have capacity will indicate need for themselves...’ – When Required’ Medication (PRN) Good Practice Guidance for Care Homes (NHS, 2022) PRN medication, by its very nature, only needs to be used in uncertain situations; in end-of-life care, these situations typically relate to the inherent uncertainty around when exactly a resident will die. This guideline deals only with situations in which care home residents ‘have capacity’ and can 154 ‘indicate need for themselves’. However, given that it has been estimated that 80% of UK care home residents have dementia (Alzheimer’s Association, 2022) and may lack decision-making capacity, this can be considered an idealised scenario. Faced with the uncertainty inherent to end-of-life care, guidance documents often have no choice but to find a foothold somewhere within the small proportion of things they can be certain about, things which can be standardised. This, however, has the effect of disproportionately amplifying considerations of the least important parts of care work by formalising them in an authoritative document. A guideline about residents who do not have capacity would be far more useful to nurses, since it is a more uncertain situation and thus requires guidance, but this uncertainty is incompatible with the sort of definite, certain, standardisable rules in which guidance documents traffic. As CHM Mrs. Raymond points out, furthermore, these guidelines may often be produced by individuals who ‘never stepped foot in a care home’. Other participants echoed this sentiment: “We get the weekly updates to controlled drugs guidance over email now… I think it could be less quantity and more quality? If whoever was writing it all up came up and talked to us, the landscape would really feel different”. (Mrs. Jiro, Care home nurse, Participant 24) “I think there are individuals who’ve got a fixed belief that named patient supply is the gold standard, and they’re not ‘in the trenches’ so they’re not seeing the problems with that… it’s really frustrating”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) Generally speaking, if one can produce a rule or piece of guidance (e.g. the rules around ‘named patient supply’) which instructs care home professionals on how to handle some aspect of care work without having directly experienced the day-to-day complexities of care work, that guideline may not be particularly necessary in the first place. Care home professionals can typically handle more straightforward matters themselves, simply via a mixture of common sense, training, and experience. 6.4 Micro-decisions vs. Algorithmic decisions Another inherent tension within the AP system I located was that of the standardised nature of AM prescribing often conflicting with the individualized nature of AM administration. The ‘trial and error’ approach described by various CHNs throughout my dataset paints a picture of individualized care as involving iterative dosing and complex, intuitive, and adaptive administration: “It’s all about treating the pain, adjusting the doses. In a way I suppose it’s trial and error…You give a certain dose and you monitor how long the effects of the medication lasts, 155 then that guides what we go with for the next dose, so we’re always learning”. (Mrs. Kinley, Nursing home manager, Participant 10) “I would always start on the smallest dose prescribed to see how they react to that. Some people can be quite opiate sensitive, so within half an hour of even a tiny dose of morphine they’ll have pinpoint pupils and be utterly confused. That’s possibly not helping them, so that’s the kind of thing that would guide my decision making. It’s not just one decision. It’s a lot of decisions and they can change”. (Ms. Wynne, Care home nurse, Participant 12) By taking the time to ascertain important details about the individual nature of the resident to whom she was administering – that is, whether the resident was or was not ‘opiate-naïve’, for example – Ms. Wynne was able to avoid simply treating her residents in a standardised fashion. Ms. Wynne was also one of many CHNs who reported that smaller doses proved useful in ‘testing the waters’ during AM administration. Their accounts suggest that administering using a personalised ‘trial and error’ approach appears to involve finely modulating doses – frequently switching between small amounts of many medicines, rather than giving fixed amounts of one or two medicines. This individualized approach, however, often proves to be incompatible with the prescribing of standardised ‘stock doses’: “…it may be appropriate to give them the Levomepromazine if the Midazolam has not been effective. But also, I do find that they do actually need Glycopyrronium… which is why I’d rather the GPs prescribed all of them, the whole anticipatory suite, and just a small amount of ampules of each rather than ten ampules of two of them”. (Mrs. Fields, District Nurse and Community End-of-Life Care Facilitator, Participant 7) Stock doses were associated in my data with not only over-prescribing but also under-prescribing, as evidenced by one CHN’s account of an incident in which her care home lacked sufficient medication to effectively control a patient’s distressing end-of-life symptoms during a ‘bad death’: “…we just had to, sort of, maintain him on the 2.5 to 5 because a lot of GPs, if you can get one to come out, will give stock doses of things, but they’re not very helpful really”. (Mrs. Graham, Senior nurse, Participant 14) It may be, then, that the prescribing of standardised of stock doses acts as a bottleneck within the AP process that denies CHNs the flexibility needed for individualized, finely-modulated, ‘trial and error’ administration. The fixing of dose ranges also appears to be inimical to individualization, as suggested by one GP participant: 156 “In my area, we’ve got protocols that guide us through end-of-life prescribing, but the ranges can be narrow and there’s only so much I can do in the way of personalising…” (Dr. Cairns, General Practitioner, Participant 37). A final aspect of prescribing worth mentioning is the practice of prescribing to individual, named residents (i.e. ‘named patient supply’), something which the following pharmacist participant criticised: “Sure, we can prescribe anticipatory medicines to individual residents, but really it doesn't make sense. In the same building you could have twenty people with twenty packs of medicines where half of them would probably not get used. They might be used in a dose of one, and then the rest thrown out because it's got someone's name on it. It makes much more sense to keep care home stock, but still need each individual patient to have a prescription for the medicine. It is just common sense to do it”. (Mr. Rowlands, Community pharmacist, Participant 5) Intuitively, it may seem as if prescribing to ‘individual residents’ would promote an individualized approach to care. In practice, however, this prevents CHNs from using the various medicines which they actually do have in their possession but are not allowed to use. CHNs are less able to tailor their care towards an individual – to match different specific medicines to an individual’s specific needs and specific symptoms as they appear – if they have fewer medicines to select from. Paradoxically, then, the practice of prescribing to named individuals appears to hurt a CHN’s chances of being able to provide individualized care. The problems that the various prescribing practices just discussed (prescribing stock doses, prescribing dose ranges, and prescribing to named individuals) caused for administering CHNs speak to a deeper tension between AM prescribing and administering. As CHN Ms. Wynne puts it, administering involves not ‘one decision’ but ‘a lot of decisions’ – small, finely-modulated, micro- decisions ‘which change’, and which take as many different forms as there are different possible contingencies which can arise during the inherently unpredictable process of dying. Within the anticipatory prescribing process, however, the range of choices often tends to be restricted to a small set of possible doses and dose ranges, and generally to four standard medications. In contrast with ‘micro-decisions’, prescribing decisions are by their very nature fixed, definite, and all-or-nothing; one either prescribes or does not. Indeed, Mrs. Raymond remarked of prescribing professionals: ‘“Do we need anticipatory meds or not?” ... “Do you need a prescription?” That’s as much input as they want to give’. Thus, the question to ask is how such a fixed and singular prescribing decision 157 can be made to bear the weight of all the potential administering micro-decisions which will have to be made in the shadow of it. This tension is exemplified by the following account: “[AMs] generally don’t get used until someone is really deemed very definitively end-of-life by a clinician. We have our own sort of criteria about what that looks like. For example, if someone is no longer able to respond, if their swallowing reflexes slow or become compromised, if their skin’s breaking down… they’re requiring a lot of bed rest…” (Mr. Smith, Care home manager, Participant 1) A CHN must constantly be able to detect even the most minor fluctuations in a resident’s symptoms in order to decide whether to administer more, or less, of a given medicine. For every ‘micro- decision’ a CHN must make, then, there tends to be a corresponding micro-symptom which inspires that decision once detected (e.g. ‘pinpoint pupils’). These micro-symptoms are so inherently subtle and minute that they may often fail to meet the minimum threshold of what, in a prescribing algorithm or protocol, constitutes a symptom. Subtle fluctuations in skin tone and sleeping patterns occur at a level of magnification beyond what most protocols can control for; a protocol-approved ‘symptom’ will typically be something more on the order of ‘excessive respiratory secretions’ (NICE, 2014). Such symptoms belong to a binary, since they will either be there or not there, rather than to a modulated spectrum, which requires a more discerning eye to appreciate. The fact that prescribing professionals’ need to deem a patient ‘very definitively end-of-life’ before they can prescribe AMs may be a consequence of their overreliance on standardised prescribing algorithms which, insofar as they traffic in certainties and absolutes, may fail to appropriately account for the actual contingencies of the dying process, which are unique to each individual. When the abstract, idealised criteria put forth by prescribing algorithms collide with the realities of a situated death, what Mr. Smith calls ‘our own sort of criteria’ becomes the only criteria specific enough to be useful. It is in the nature of the dying process, which always varies from one individual to the next (Lawton, 2002) to generate new and unique ‘sets of criteria’ depending on the individual care home, and even the individual resident. The protocol paradox points to the fact that, in spite of this, end-of-life care patients receive especially high levels of standardised care, as exemplified by prescribing protocols or algorithms. An account provided by one nursing home manager showcased that the ‘micro-decisions’ made during AM administration often stem from considerations of the complex interrelations that can occur between medicines: “Our lead nurse is skilled at knowing how much analgesic they currently are on from a patch or orally. When we see that a patch or oral administration isn’t cutting it, we start 158 with subcut [subcutaneous injections], leaving the patch on. Then, when we see how frequently that’s being needed, we work out how many subcuts we’ve had and what we need in our syringe driver. Only then when that’s working do we take the patch off. So by assessing and readjusting every few hours, we know that we’re not leaving them with a risk of breakthrough pain. And we continue to tweak it up and down”. (Mrs. Whittaker, Nursing home manager, Participant 31) This is another example of finely-modulated, adaptive, intuitive administration. The lead nurse described makes the decision to administer a subcutaneous injection only after ascertaining whether the fentanyl patch already administered has been effective. With every new medicine introduced, this nurse re-assesses whether or not the fentanyl patch should remain on or off. At each stage of this process, the decision of whether or not to administer a certain medicine is dependent on the status of other medicines already in use. Each of these ‘micro-decisions’ illuminate how the interrelations and interdependencies between different medicines shape the administration process. In contrast, prescribing protocols often presume one-to-one correspondences between medicines and symptoms (e.g. ‘pain’ and ‘morphine’), thus failing to account for the relationships between medicines and other medicines. The administration of one medicine, furthermore, may imply the non-administration of another. In Mrs. Whittaker’s account, the complex interplay of syringe drivers, fentanyl patches, and so on, indicate that the presence of one medicine always depends on the presence or absence, and the effectiveness or non-effectiveness, of others. Other CHMs in my dataset reported a desire to acquire precisely the intimate knowledge of medicines and administrations demonstrated in this account. One CHM participant (Mrs. Larkin), for instance, reported that she might have benefitted from some guidance from GPs on ‘why certain end-of-life medications are prescribed as opposed to others’. Yet GPs themselves may actually suffer from the same knowledge gap, as shown by Dr. Cameron’s account, cited earlier: “It’s really hard, when people are not doing the job day in, day out are being told that, ‘Oh, we’ve got Levo for sickness. Oh, we’ll prescribe sickness.’ Well, actually, they’re on Metoclopramide, so why don’t you just keep them on that”. (Dr. Cameron, General Practitioner, Participant 6) Dr. Cameron made this remark in the course of criticising the practice of prescribing standardised packs of four AMs (including Levomepromazine). However, as indicated by Figure 6.1 below, it is ‘not advised clinically’ to administer Metoclopramide alongside Hyoscine Butylbromide, which is part of the standardised pack in most areas across the UK, including Dr. Cameron’s: Figure 6.1 Compatibility chart for two drugs in water for injections 159 Source: The Syringe Driver: Continuous subcutaneous infusions in palliative care (p. 120, Dickman & Schenider, 2016) If a resident were to be administered AMs in a standardised pack which includes Hyoscine butylbromide but which substitutes Levomepromazine with Metoclopramide – as Dr. Cameron appears to be suggesting – this could cause an adverse synergistic reaction. This is another example of an interrelation between medicines: the administration of Metoclopramide implies – or ought to imply – the non-administration of Hyoscine butylbromide. Figure 6.2 GSF-provisioned symptom control guidelines 160 Source: Palliative Care and the GMS Contract Quality and Outcomes Framework (p. 11, GSF, 2006 & 2014 & 2019) Several participants described using GSF-provisioned protocols or algorithms to guide AM prescribing and administration (p. 11, GSF, 2006 & 2014 & 2019). These algorithms take the form of flowcharts linking symptoms to medicines. One route through this GSF flowchart leads to Levomepromazine and Haloperidol – two drugs which are incompatible, as the above diagram indicates – being prescribed in tandem. Similarly, a regional guidance document for AMs reported to be used across several CCGs provides information about prescribing and administering both cyclizine and oxycodone, another combination treated with caution by the above diagram (NHS, 2017). That synergistic reactions and interrelations between medicines are often unaddressed by these prescribing algorithms may be a by-product of the fact that algorithms tend to consider medicines and symptoms in isolation. Within one and the same guidance document, for instance, there might be a prescribing algorithm drawing an arrow from the symptom ‘pain’ to the medicine ‘morphine’, and another, separate algorithm linking the symptom ‘excessive respiratory secretions’ to the medicine ‘Hyoscine butylbromide’. Yet confining the two medicines to their separate compartments obscures the overlap between them, namely that Hyoscine butylbromide can also be used to treat a kind of pain – abdominal pain (Tytgat, 2007) – and thus encroaches to some extent on the territory marked out for morphine. If a care home resident is suffering from abdominal pain, a medicine other than morphine in the standardised AM pack – hyoscine butylbromide – may instead be effective pain relief. However, in a situation where prescribing algorithms earmark hyoscine butylbromide for respiratory secretions alone, and consign all pain symptoms to morphine in a 161 separate protocol, an administering nurse may interpret abdominal pain simply as pain and administer only morphine. One account provided by a CHN revealed the tendency within prescribing algorithms towards viewing medicines in isolation: “GPs still love patches, which can be great, but the GPs don’t recognise that they are slow release. If somebody’s on a patch, you actually need to say, ‘Okay, that’s fine, we’ll give them a patch, but they need something else on top and in addition to.’ But the GP is part of prescribing a clear pathway because I think they get nervous that we’re going to overdose people…” (Mrs. Perry, Care home nurse, Participant 29) This account shows how a GP’s prescribing ‘pathway’ – a type of prescribing algorithm – failed to account for the interrelations between fentanyl patches and other medicines. Mrs. Perry’s knowledge that the fentanyl patch may need to be administered in combination with other medicines to provide effective pain relief – rather than being prescribed in isolation, as prescribing algorithms often encourage by visualising drugs in isolation – is the natural result of her experience as an administering nurse. The process of administration, as we have seen, involves picking up great deal of knowledge regarding the idiosyncrasies of drugs, and the interrelations between drugs. The highly complex ‘trial and error’ of the administration process, being vulnerable to contingencies and unpredictable elements, is difficult to represent in any standardised way, and thus poorly served by a simple algorithmic diagram. The downside of what Mrs. Perry calls a ‘clear pathway’ is that said clarity may be achieved only at the cost of oversimplifying the complexity of administration. The simplistic 1:1 ratio between symptoms and drugs purported by prescribing algorithms does not acknowledge that one medicine may treat many different symptoms, just as one symptom may require many different medicines, that medicines and symptoms can overlap in complex, even idiopathic ways. This simplistic 1:1 ratio between symptom and medicine runs the risk of treating morphine as a panacea and standardising the idea of pain. In reality, ‘pain’ is a deeply individualized, subjective, and diverse sensation. No two sets of ‘pain symptoms’ are truly identical, and the reason one care home resident may need morphine will never be exactly the same as the reason why another may need it. Different individuals may also react to morphine in different ways – for example, one may be ‘opiate-naïve’. One participant observed that CHNs were often reticent to acknowledge that pain is an individualized phenomenon, one which varies from person to person: “…if you tell people that neuropathic pain doesn’t respond to opiates, then it’s like, ‘Oh, no, we can’t prescribe opiates for that.’ Well, why can’t you? Because, actually, some neuropathic pains do respond to opiates. Why not give it a go? But, if the nurse is like, ‘Oh, 162 no, no, we’ve got to prescribe only this for neuropathic pain and that for…’ [CHNS] are so stuck to guidelines. A lot of their work is kind of guideline-driven, isn’t it?” (Dr. Cameron, General Practitioner, Participant 6) This is a prime example of the simplistic 1:1 ratio between symptom and medicine often found in prescribing protocols. This participant’s sentence could be filled in with something such as: ‘we’ve got to prescribe this medicine for neuropathic pain and that medicine for respiratory secretions…’ These two drug-symptom pairings, linked by an arrow, would be found on separate pages of a protocol. Crucially, this would obscure the fact that the medicine earmarked for one symptom may be able to treat the other symptom. Dr. Cameron attributes this oversight to CHNs’ excessive adherence to rigid and standardised ‘guidelines’, a word often used for the instructions given within prescribing protocols. Another participant reported a similar scenario in which a nurse prioritised a rigid, schematised idea of how care should be delivered over the specific needs of the resident being treated: “I’ve had one palliative care nurse helping to look after a lady with dementia, and she wanted to give a syringe driver. I said, ‘I don’t think she’s at that stage. She’s not really having those anticipatory medications at the moment to warrant a syringe driver. She’s still very alert and she’s got dementia.’ But the nurse was adamant. She said that I was withholding medication for this lady. I said, ‘Right okay, we’ll go with the syringe driver,’ and straight away, my resident broke the syringe driver. She threw it and broke it, but I knew that would happen, because she was too agitated. She wasn’t at that stage where she was calm and peaceful. But this palliative care nurse was having none of it. And I think she was just looking at it as a stepwise picture: this is what we do now, this is the stage she should be at. But it’s all about knowing your resident well enough. And there is a place for syringe drivers in dementia, but they’ve got to be at that calm, very much End-of-Life stage”. (Mrs. Kinley, Nursing home manager, Participant 10) The ‘stepwise picture’ described here is typical of prescribing protocols, which often reduce care to an algorithm or flowchart comprised of steps (or stages) to be followed one-by-one. This account reveals what occurs when an individualized model of care (centred on ‘knowing your resident well enough’) collides with a highly standardised, protocol-driven model of care, which subjects each resident to the exact same treatment, the same ‘stepwise’ trajectory. A similarly stepwise view of syringe driver use was found by Neuberger (2013) in the LCP Review, which found that ‘too often it appears that a syringe driver is put in place as the ‘next step’ on the LCP, overlooking the needs and wishes of the patient’ (p. 30). Another instance reported in my data in which a standardised 163 ‘pathway’ was prioritised over the wishes and needs of families of dying residents (which were reported in ‘ReSPECT forms’) was the following: “The problem with the ReSPECT form in its current format and also the DNAR is that ultimately, it’s a clinical decision. So a family member can be consulted but they’re not allowed the ultimate decision. I understand the reasoning behind that, because, you know, it’s a doctor, it’s a medical professional who gets to make that call. However, families are often presented with it as a fait accompli I suppose… And that might be the case, that pathway may be the right route, but there’s not an awful lot of room in that discussion for questions or to take into account family members who don’t really know that that’s the right thing for them. They may have questions that don’t get answered…and they walk away feeling bad that what they wanted didn’t matter anymore”. (Mrs. Raymond, Care home manager, Participant 25) ReSPECT forms record patient preferences and clinical recommendations developed through conversations between the patient, their families, and their care teams (Eli et al., 2021). Since families of residents tend to know them very well, taking the families’ views into account can be a useful way of individualizing that resident’s care. In this case, this individualized approach was overridden in favour of the inherently standardised approach represented by the ‘pathway’. As one participant (Mrs. Perry) put it, the practice of ‘prescribing a clear pathway’ is often prioritised by GPs. Yet, the apparent ‘clarity’ of a pathway or protocol – typically made up of well- organised, neat boxes and arrows – may in fact oversimplify the complexity of a resident’s trajectory. Pinder et al.’s (2005) study found that ‘As soon as the contingencies of patients’ lives were overlaid onto the hard rectangles and straight lines of the pathway, the illusion of authoritativeness began to crumble’ (p. 768). Because prescribing algorithms are created in the interests of clarity and replicability, they possess a powerful internal logic, meaning they run the risk of appearing self- evident and self-sufficient, when in reality they should only be considered meaningful alongside a specific example of a resident trajectory. In his Birth of the Clinic, Michel Foucault argues that once a disease has been broken up into a set of ‘symptoms’ – isolated snapshots abstracted from the organic flux of the body – these symptoms tend to crystallise into ‘well-defined structures’, which exist only in the abstract but which tend to be more readily comprehensible, and thus potentially more attractive to clinicians than the organic processes they are supposed to map: ‘A man coughs; he spits blood; he has difficulty in breathing; his pulse is rapid and hard; his temperature is rising; these are all so many immediate impressions… Together, they form a disease, pleurisy: ‘But what, then, is pleurisy?… ‘Pleurisy’ has no more being than 164 the word itself; it ‘expresses an abstraction of the mind’; but, like the word, it is a well- defined structure...’ (pp. 118-119, Foucault, 1973, The Birth of the Clinic) This process can be seen at work in what Pinder et al. call ‘the rationalising and atomising practices of the care pathway’ (p. 765). Prescribing algorithms carve up the organic processes of disease into isolated symptoms, and afford to symptoms the appearance of self-contained reality by confining them to their own ‘boxes’, matching each one up with a corresponding medicine. They are, above all else, a well-defined structure. When symptoms are formalised within the simplified and logical geometries of a prescribing algorithm, they take on an objective and unobjectionable aspect. The symptom ‘pain’, for instance, may appear to be a stable and self-consistent category simply because it has been neatly organised into a box connected with an arrow to another box recommending morphine. Pinder et al. note that ‘pathways… abstract the patient and reify the condition… they all define their situation by their impairment, simplifying the complexity of their lives’ (p. 765). To underscore this point, they quote a Director of Services at a Primary Care Trust explaining the rationale behind the construction of ‘pathways’ for cataracts: ‘It's an easily diagnosable condition, there's lots of them around and it's quite straightforward. There's a large cohort of people all needing the same thing, they need to be funnelled through a process… a cataract is a cataract is a cataract. You can quite nicely develop a pathway for that’. (p. 765, Pinder et al., 2005) As this quotation demonstrates, a condition is considered to be a good match for a ‘pathway’ only if it can present across different patients yet appear in each one more or less the same, thus taking the condition as the constant and the patient as the variable. Yet, this leads to a ‘paradoxical position’ described by Michel Foucault (1973), wherein ‘[i]f one wishes to know the illness from which he is suffering, one must subtract the individual, with his particular qualities’ (p. 14, Foucault, 1973). In essence, the patient is considered to be ‘only the negative of the disease’: ‘Paradoxically, in relation to that which he is suffering from, the patient is only an external fact; the medical reading must take him into account only to place him in parentheses’. (p. 8, Foucault, 1973, The Birth of the Clinic) The paradoxical situation Foucault identifies relates to the protocol paradox, which encompasses the difficulties that arise when trying to standardise techniques with which to deal with individuals. To understand a disease which can only ever really be experienced directly by the patient, something 165 absolutely unique to that patient, doctors must focus on all the ways that the disease is like other diseases – since it is only when a specific type of pathology afflicts a certain number of people that it is classed as a disease at all - and thus the ways that the patient is ultimately just like any other patient. The ‘pathways’ described by Pinder et al, then, resemble an old paradigm of medicine described by Foucault (1973), ‘in which the individual in question was not so much a sick person as the endlessly reproducible pathological fact to be found in all patients suffering in a similar way’ (p. 97). Elsewhere in The Birth of the Clinic, Foucault also documents a more positive, person-centred paradigm of medicine. As he recounts, a paradigm shift in the perception of death that resulted from the discovery of autopsy at one point led to a new kind of medicine centred on specific patients rather than general disease categories. This marked a shift from an older conception which viewed death as a universalising force, a ‘leveller’ that brought all to the same pass. Modern death was, rather, individualized, ‘constitutive of singularity’: ‘…it is in that perception of death that the individual finds himself, escaping from a monotonous, average life; in the… approach of death, the dull, common life becomes an individuality at last…’ – (p. 211, Foucault, 1973) This idea that every death leaves its unmistakeable, individualized signature chimes with the experiences of CHNs in my dataset; for instance, the following: “I always explain to people, people pass away in their own time. Just as we all live our lives in a very individual way, people pass away in a very individual way… I think the one thing I have learnt from working in nursing homes is that people have a lot more choice about dying than I thought they did… I’ve seen people who’ve just decided, ‘I’ve had enough, I want to die’ and they’ve died very quickly”. (Ms. Wynne, Care home nurse, Participant 12) This participant went on to describe the effects that AMs can have on the control or ‘choice’ that residents have over how their own deaths play out: “I think it can whittle that control away if you sort of over-sedated someone. I try to titrate doses very carefully, and I’ve had so many of these pre-discussions with them. Some will say, ‘I want to be awake as possible’, in which case we’d be keeping to the very minimum doses when they’re dying, whereas if somebody’s been able to say, ‘I really don’t want to be in pain, I don’t want to know what’s going on,’ then you might feel happier giving a slightly larger dose”. (Ms. Wynne, Care home nurse, Participant 12) 166 This account describes a highly individualized approach to the administration – characterised, for instance, by a sensitivity to the patient’s own wishes (‘I really don’t want to be in pain’), and by the careful, patient-specific, finely-modulated iterative titration of doses which I discussed earlier. This individualized approach to end-of-life care fits in well with her belief that ‘people pass away in a very individual way’. Interestingly, another CHN participant similarly remarked – ‘dying takes a lot of patience, a lot of understanding’ as part of an anecdote about family involvement: “One time, we had the two daughters of a poorly resident visiting from Canada. They were looking more at the dates on their return tickets more than mum’s comfort and safety. They were trying to get me to ‘rush things along.’ And I understand the frustration, but dying takes a lot of patience, a lot of understanding”. (Mrs. Gardner, Care home nurse, Participant 22) Indeed, where finely-modulated, individualized administration is concerned, the dimension of time was often reported to be paramount. As the following participant’s account reveals, iterative dosing requires ample time and a great deal of patience: “You might have someone whose symptoms are not controlled at all, so you start them on a driver. Say you’ve ordered ten ampules, they’ve got 30mg of Morphine in the driver. That driver takes four hours to see any sort of benefit to the patient. And then they might have a couple of PRNs. The staff might have a review after six hours, and then before you know it, you’ve used six or seven of the vials”. (Mrs. Haberman, District Nurse, Participant 13) This account also reflects another cited earlier, in which recognising the ‘slow release’ nature of fentanyl patch formed an important part of a personalised administration trajectory. In general, it is important to understand that behind the intuitive and adaptive administration practices which I have been discussing in this section, there often lies a concrete philosophy of death – the belief that every individual dies in their own individual way. 6.5 Individualized language vs. Jargon oblivion It is important to consider the extent to which various conditions associated with the end of life compromise a resident’s communication faculties, thereby increasing the need for individualized care. Non-verbality in care home residents is often the result of dementia (Corbett et al., 2016), a condition which greatly problematises pain identification. As Lundin and Godskesen note: 167 ‘…pain management becomes very complex. Pain is a subjective sensation and not being able to communicate with persons with advanced dementia makes nurses uncertain of whether there is pain and, if so, whether it is being adequately managed’. (p. 10, Lundin & Godskesen, 2021) A similar phenomenon was observed in one participant’s account of her experience with a resident: “She just keeps saying ‘I want to go home’ and I think she’s got a level of dementia where, ‘I want to go home’ means, ‘I’m in pain, I don’t like what’s happening with me, I want to get away from it all.’ So it’s trying to identify what her non-verbal signs or even what some verbal signs of pain actually mean and that can be quite difficult”. (Ms. Wynne, Care home nurse, Participant 12) Pain relief is among the most important concerns within end-of-life care. Unlike other clinical conditions, pain can rarely be assessed using the usual techniques of scientific observation; there are no tell-tale biomarkers to work with (e.g. lesions or swellings). Pain thus largely depends upon language in order to be identified and relieved (Corbett et al., 2016). Fundamentally, pain is a signal sent by the body, and pain relief largely depends on a doctor or patient’s ability to translate that internal biological signal into an external linguistic signal by properly articulating it. As Mehta et al. (2008) put it: ‘Pain is a subjective perception. Pain is what the patient says it is’. Foucault (1973) argues that pain is so subjective that medicine’s relationship with it can be considered a ‘fantasy link between knowledge and pain’ (p. 10). End-of-life care thus presents a dilemma: the dying phase, during which pain identification and relief is most necessary, is also the most likely time for a resident to lose the ability to communicate, thereby making that pain – which by its nature relies on its articulation to be identified – inarticulable. This in turn sets off the protocol paradox: that end- of-life care patients – especially prone to pain whose inherent subjectivity makes the individualization of care that much more important – are likely to receive more standardised care precisely because that pain has become inarticulable at the end of life, leaving many clinicians or carers at a loss regarding the appropriate clinical action to take, often resulting in them falling back on the most generic one. Mehta et al. frame nurses’ experiences with pain identification using the concept of ‘total pain’: ‘Dame Cecily Saunders… coined the term “total pain” and suggested that pain [can] be understood as having physical, psychological, social, emotional, and spiritual components. The combination of these elements is believed to result in a “total pain” experience that is 168 individualized and specific to each patient’s particular situation’. (Mehta et al., 2008) (My italics) The idea that the blind spot opened up in the identification of pain by non-verbality at the end of life can be compensated for to some degree by individualized knowledge of residents was supported by my data; for instance the following account: “We use an initial pain assessment, and then we use the Abbey Pain Scale for daily assessments of pain. But also we’re in a very fortunate position of knowing our residents very, very well, so even though they can’t communicate verbally and a lot of them have quite advanced dementia, we’re looking for signs all the time that they might be in pain or discomfort. We watch their face for frown lines and sometimes it’s as straightforward as somebody actually saying ‘ow’ or moaning”. (Mrs. Graham, Care home nurse, Participant 14) Since pain is inherently subjective and thus unique to each individual, standardised pain scales can arguably only go so far in identifying symptoms. The shortfall in communication which dementia introduces between resident and CHN can much more effectively be offset by seeking out more individualized, person-specific knowledge regarding the resident in question – by ‘knowing our residents very, very well’. Non-verbality in dying residents is an issue of language, and is therefore well-served by the theory of the relationship between language, death, and medicine outlined in Foucault’s Birth of the Clinic (1973). As discussed earlier, Foucault observes that, after the discovery of autopsy, death was reconceptualised as a more individualized phenomenon. This shift also had an effect on medical language. ‘Language and death have operated at every level of this experience’, writes Foucault, ‘to offer at last to scientific perception what, for it, had remained for so long the visible invisible—the forbidden, imminent secret: the knowledge of the individual’ (p. 170). Foucault argues that the reconceptualisation of death led to a ‘new turn given to medical language’ (p. 169) – specifically, it moved medical language towards individualization. He explains that, when the parameters of death shifted, this had the effect of ‘opening words to a certain qualitative, ever more concrete, more individualized, more modelled refinement; the importance of colour, consistency, texture… the value of intersensorial qualities (smooth, greasy, bumpy)’, and so on (p. 169). Foucault argues that words like ‘smooth’, ‘greasy’, ‘bumpy’, though less technical and formal than the clinical discourse often used by doctors, are better able to capture the nuances and idiosyncrasies – the individuality – of a patient’s condition. Though these words referred to the qualities of internal organs exposed to autopsy, they can reasonably be extrapolated to other forms of symptom identification. Indeed, 169 nurses in my dataset often used this kind of informal language – words like ‘bubbly’, ‘chesty’, and ‘rattly’: “What we don’t want is a surprise, so that if the pain increases or they become bubbly or chesty, nobody’s like, “Oh, God, what do we do?” (Mrs. Galway, Nursing home manager, Participant 18) “Our resident then developed a bit of a rattly chest and the GP immediately went in with Hyoscine…” (Mrs. Raymond, Care home manager, Participant 25) The strength of such individualized language, Foucault suggests, is that it can better tune into the minutiae of patient conditions associated with death – considerations such as ‘colour, consistency, texture’. My data showed that identifying imminent death does indeed often rely on the recognition of changes in skin texture: “There are some things that you can notice, like the mottling of the lower extremities, the cyanosis of the lower limbs, the nails…” (Mrs. Barnes, Care home nurse, Participant 24). “And the stages are quite common and for someone in their old age, we recognise those signs… You know, we’ll start to see extremities become… much more cyanosed…” (Mr. Smith, Care home manager, Participant 1) The importance of texture and colour was also mentioned by a participant who described their difficulties with remote consultations during the COVID-19 pandemic: “It’s hard to explain how somebody’s presenting without actually seeing them, like the skin colour, the tone, the texture of their hands and things like that. So they can’t have that conversation with the GP and say ‘I’m feeling this, I’m feeling that, It hurts here, it hurts there.’ You need to sort of know the individual to do it remotely, especially if somebody non- verbal”. (Mrs. Laney, Nursing home manager, Participant 19) As just discussed, Foucault (1973) argues that a reconceptualisation of death had the effect of shifting language towards individualization. This account demonstrates a similar but slightly modified relationship between death, language, and individuality. Essentially, my data showed that the approach of death often compromised language in residents, thus forcing CHNs to fall back on their individualized knowledge of residents, i.e. ‘know[ing] the individual’. In this account, the added strain of COVID-19 preventing GPs from seeing residents directly added a second obstacle to 170 language and communication generally. This account resonates with others in my dataset which demonstrate the attention that must be paid to micro-symptoms such as ‘the skin colour, the tone, the texture of their hands’ in situations where residents are non-verbal and unable to describe their own condition: “We look for non-verbal expressions, moaning, grimacing, the death rattle, secretions, getting restless”. (Mrs. Jiro, Care home nurse, Participant 23) “We have our own sort of criteria about what [the end of life] looks like… if someone is no longer able to respond, if their swallowing reflexes slow or become compromised, if their skin’s breaking down, if they’re beginning to demonstrate that their lives are beginning to sort of come to a close, they’re much more sleepy...” (Mr. Smith, Care home manager, Participant 1) In both these accounts, since patients were rendered non-verbal, care home professionals were forced to individualize their care – that is, to adapt a radar far more finely attuned to subtle modulations of condition (including, once again, changes in skin texture). To summarise, Foucault (1973) observes that a new appreciation of death in medicine at one point led to medical language becoming less technical or jargonistic and more individualized, informal, and intuitive (e.g. words like ‘smooth’, ‘greasy’, ‘bumpy’). CHNs in my dataset often used similar words (‘bubbly’, ‘chesty’, ‘rattly’). One care home manager participant explicitly stated that he uses such language (e.g. ‘sleepy’) because he believes that it is more ‘person-centred’ (i.e. individualized) than jargonistic or technical medical language: “I personally am not a big fan of the medical language that’s usually around anticipatory prescribing. The language that I prefer is much softer. I talk about people slowing down, closing down. I talk about people becoming more sleepy, less able, less responsive. I talk about the need for us to be anticipating their needs as they start to die. I talk about the fact that we recognise that they may be approaching the end of their life. And we use sensitive language that is very person-centred”. (Mr. Smith, Nursing home manager, Participant 1) Indeed, the medical jargon around AP was often revealed in my data to be inimical to the individualization of care, as seen in the following account: 171 “…when families ask them why mum's been started on morphine, I’ve seen some of my staff saying it’s 'just in case.' Well 'just in case' of what exactly? I'm not really sure how many of them know why they're giving it”. (Mr. Anderson, Care home manager, Participant 35) Here, instead of providing the resident with individualized care – i.e. care which is carefully planned and amply informed by specific knowledge of the resident in question – AMs were prescribed and administered automatically and almost without thinking, simply because they are a step along a standardised end-of-life protocol. This account shows that the ambiguity of the jargon term ‘just in case’ is part of what encourages this behaviour in CHNs. The accounts just cited suggest the strong influence that language – which ranges from non-specific, standardised jargon to the hyper-individualized language of symptom-identification – holds over the dynamics of individualization and standardisation within care homes. One participant granted an important insight into the role language plays at the end of life by revealing his preferred jargon terms and phrases: “Oh, I guess I do have a stockpile of phrases... Among professionals, I'll often use the term 'unmet palliative needs' or 'unmet end-of-life needs' simply so we can figure out the symptoms to treat. Sometimes trying to look at the overall package of end-of-life care misses the problems that dying patients frequently have: pain, secretions, nausea, bowels and bladder, all of that lot. So focusing on unmet needs is often a phrase I'll use among professionals. Among relatives or indeed patients themselves, I'll use the term 'TLC' or ‘tender love and care' a lot... I'll often use phrases like that, but frequently my consultation will adjust according to who I'm speaking to, how they reply to me, their level of understanding, their level of communication and so on”. (Dr. Murray, General Practitioner, Participant 39) Here, Dr. Murray resists the standardising effects of ‘the overall package of end-of-life care’ (i.e. the standardised pack of four anticipatory medicines), which often treats all end-of-life patients as if they suffered from the same four symptoms, ignoring other symptoms like ‘bowels and bladder’, etc. Furthermore, by making a clear distinction between the term he uses with professionals (‘unmet palliative needs’) and the term he uses with care home residents and families (‘tender love and care’), Dr. Murray individualizes his care by tailoring his language to the individual. My data shows that issues often arose when care was not taken to use different language for medical professionals and for families in this way – for instance, when terms more well-suited to medical colleagues, typically jargon terms such as ‘controlled drugs’, were instead used with families: 172 “I have a real issue with the term ‘controlled drugs’ personally. They’re all locked away and I think that it gives them a sinister feeling… we’ll say a resident needs his controlled drugs, and then a family member will go, ‘Controlled drugs? What are they?’ And then we explain to them that actually they’re locked away for people’s safety… and right there the fear starts”. (Mrs. Raymond, Care home manager, Participant 25) As Dr. Murray appeared to intuit by separating the terms he uses for other doctors and family members, the same linguistic term can often be received differently by different parties within the care home space. It is in the nature of jargon terms, for instance, to present differently to medical professionals and non-medical professionals. To those in the know (i.e. medical doctors), jargon terms hold a specialist meaning, whereas to laymen, they represent an unknown, a specialist knowledge base inaccessible to themselves. A paradox is the result of two contradictory elements existing simultaneously within one and the same thing, and because a jargon term holds two conflicting meanings within itself at the same time – one for medical professionals and another for laymen – jargon terms can often be experienced as paradoxes. For instance, the meaning of the term ‘controlled drugs’, which indicates how a drug should be stored, will be different for CHNs (to whom the responsibility of drug storage falls) and families (for whom the storage conditions of the drug are irrelevant). The use of the term ‘controlled drugs’ was thus experienced in a paradoxical way by the family members who were forced to simultaneously view the same drugs as both potentially beneficial to a loved one and potentially dangerous. In this instance, individualized language, tailored to the family in question, could have prevented issues. This situation could have benefitted from some consideration of the family’s point of view, and of what aspects of the medicines in question were actually relevant to them, individually; instead, the general, standardised term for all controlled drugs was used. It may have been better to simply name the specific drugs being used. A more positive example of a participant dealing with the issue of family members attaching sinister connotations to AMs was described in the following account: “There’s always that stigma from relatives attached to syringe drivers. They think that you set them up and then people die. Midazolam has that awful association cloud around it, doesn’t it? People think once you start, that’s it. But since we started using Gold Standard Framework, we involve families a lot more and we tend to have really good relationships with them and with the GPs. It’s rarely a struggle to get them onboard with anticipatories anymore”. (Mrs. Graham, Care home nurse, Participant 14) 173 Many participants noted the GSF’s usefulness in guiding them towards having early conversations with family members which encouraged them to ‘increase the certainty of the language’ around AMs, as one participant (Mr. Smith, CHM) put it. The LCP, on the other hand, was found by the LCP Review to be far less successful at dispelling negative connotations around its language. The jargon term ‘integrated care pathway’ in particular was revealed to have a negative impact on families. Like the term ‘controlled drugs’, it was associated with what the participant just cited called an ‘awful association cloud’: ‘The Review panel has reluctantly concluded that the term ‘Liverpool Care Pathway’ is most unhelpful: anxious and upset relatives cannot be expected to understand what an ‘integrated care pathway is... A ‘pathway’ suggests to most people a road that leading somewhere. When someone is ‘put on’ a pathway, it sounds like, as one carer put it, they are being placed on “a conveyor belt to death”’. (Neuberger et al., 2013) My own data suggests that the association between the Liverpool Care Pathway and the hastening of death via anticipatory medicines has lived on in the minds of residents’ families: “There was this one particular family who felt that if we’d brought in the end-of-life medications, it would mean that their mum would pass away quicker. Obviously, the Liverpool Care Pathway has not done anyone any favours. But we had to involve the doctor and the doctor did speak to them and they finally agreed to the medications, but it was very hard for them. I think it was guilt more than anything, but we appreciate that and we understand it. It’s not personal”. (Mr. Smith, Care home manager, Participant 1) Other tense exchanges of this sort reported in my dataset suggest that ‘Liverpool Care Pathway’ itself continues to be an extremely loaded term, linguistically speaking: “We spent quite a bit of time sort of defending the Liverpool Care Pathway and trying to explain it… you actually had some patients where we’d say to the family effectively they’re not on the pathway but we were doing everything the pathway said, but we’re just not calling it the Liverpool Care Pathway. And I think the media tension around that was really unhelpful and that did create some tensions with relatives on occasion”. (Dr. Jameson, General Practitioner, Participant 38) Since the LCP Review concluded that ‘the term ‘Liverpool Care Pathway’ is most unhelpful’, the term ‘pathway’ has fallen out of use within the modern care landscape, and has been largely replaced with the word ‘protocol’. However, it may be that merely changing the name used for protocols does 174 not address the underlying issue – the protocols themselves, which are inherently standardised and thus so unconducive to individualized, patient-centred care that their adoption (as in the case of the LCP) can often lead to patients’ individual needs being neglected at a mass scale. The above participant’s admission to essentially continuing to use the LCP while merely calling it something else corroborates reports that the LCP has been ‘rebranded, not axed’ following its fall from grace (Donnelly, 2013). The tension and debate around names and words during the LCP affair attests to the importance of remaining attuned to the dynamics of language within the care space. The above discussion of jargon terms like ‘controlled drugs’ and ‘pathway’ introduces a new dimension of the protocol paradox: the fact that end-of-life care, in which patients’ needs are inherently individualized, also tends to be an area of medicine where these patients and their wider family are subjected to a great deal of standardised, nonspecific, jargonistic, or euphemistic language. Pitt et al. (2022) have drawn attention to the pernicious effects of medical jargon, coining the term ‘jargon oblivion’ to describe ‘a clinician’s failure to recognize and mitigate their use of medical jargon despite having a broad awareness that patients may find it confusing’ (p. 1861). Jargon is by nature standardised: it is powered by a consensus which ensures that everyone is on the same page about which words to use, and to which medical phenomena these words refer. Forgoing terms whose meaning is standardised between professionals allows for the use of more patient- centred, individualized, tailored language in its stead. Jargon terms are often calculated to produce an aura of neutrality, and are so frequently and unthinkingly used by professionals that the sense of the words’ meaning is largely dulled (Pitt et al., 2022). Yet what appears neutral to professionals may conceal hidden metaphors when encountered outside of professional circles. The scrutiny of emotionally affected family members may touch off latent connotations in terms (e.g. ‘a conveyor belt to death’) which professionals have begun to take for granted. In the case of the term ‘pathway’ in the LCP, these connotations had a notably sinister character. One form of jargon included within Pitt et al.’s (2022) widely cited framework, which codifies seven categories of medical jargon, is euphemism. They cite a study from Barlet et al. (2022), which found that references to death in medicine are very often couched in euphemism, a finding supported by my own dataset. This is partly because the discursive space around death is such a charged one. One participant (Mrs. Graham), for instance, remarked: ‘I’m sorry if this is offensive terminology, but we had what I would describe as one bad death during Covid’. If the phrase ‘bad death’ is considered so offensive that it is typically avoided, this may make it harder to frankly discuss ‘bad deaths’ and thus take measures to prevent them. There appears to be a strong bias against mentioning death directly within many care homes. As one participant revealed, the tendency to use death euphemisms (e.g. ‘pass away’) is so widespread that care home professionals must often go out of their way to ensure they are being direct and clear: 175 “We don’t really shy away from language that is the right language. Some people talk about passing away. But we’ve got to be really clear”. (Mr. Smith, Care home manager, Participant 1) Care home professionals often reported that they used language which could be considered euphemistic to cushion the blow of death in conversations with families: “We normally use ‘passing away’ instead of ‘dying’, to make it as gentle as possible for the family. We don’t say anything like ‘They’re going to die now.’ Maybe that ‘The time is getting near’ or ‘We don’t know how long now.’” (Mrs. Wright, Care home manager, Participant 2) Some care home professionals, furthermore, informed family members that their loved ones were on end-of-life medications (e.g. morphine) as a euphemistic way of suggesting that they were on the verge of death: “We’re really very busy so if we have time to update family, then we will, but mostly there just isn’t the time… Many of our nurses and carers don’t tell the family when their mum or dad is dying because it’s a long, difficult conversation to have. And this isn’t something that I think is right, but some of them will tell family that Mum is on morphine now, you might want to say goodbye soon”. (Ms. Foster, Care home nurse, Participant 30) This account shows that care home professionals’ reluctance to engage in direct conversations about a resident’s death often stems from the desire to avoid the emotionally difficulty task of communicating a loved one’s death to family. Another participant also flagged this as an issue, one which could be counteracted by holding open, honest, and proactive conversations with families: “I think everybody’s scared to talk about dying. Or there’s the perception that people are scared and we find a lot in care homes that actually we don’t have the conversation and it’s because we say, oh people don’t want to discuss it. Well actually, people do, but we’re still afraid to have the discussion. So I think a lot of it is down to communication, involving family members right when they come in, having an open and honest conversation about what dying entails, exploring their care preferences and end of life meds, and not being scared to talk about it". (Mrs. Perry, Senior care home nurse, Participant 29) 176 Mrs. Perry’s reference to ‘exploring [resident’s] care preferences’ points to the important role that conversations with families play in making care more individualized. Since families often know residents better than anyone else, they can communicate that resident’s person-specific ‘care preferences’ to care home staff, which are then written down within a document known as a ReSPECT form. Therefore, when care home staff are reluctant to openly and honestly discuss death, this can lead both to euphemistic language, as discussed earlier, and to a failure to make proper use of familial knowledge when recording care preferences – both of which result in less individualized care. Indeed, euphemism, which is an inherently non-specific, ambiguous, and imprecise form of language, can be considered diametrically opposed to individualization, which involves precisely tailoring care to specific residents. Barlet et al. (2022) codify four different types of death euphemism: survival framing (e.g. ‘not live’), colloquialisms (e.g. ‘let her go’), medical jargon (e.g. ‘code event’), and pronouns without a clear antecedent (e.g. ‘it may happen soon’). My dataset raises the possibility of another class of euphemisms for death: pharmacological euphemisms, in which the reality of a resident’s death is laundered through the language of medicines (specifically AMs). My data shows that, during the dying phase, care home professionals communicate with family members in two crucial ways: firstly, by breaking the news that a loved one is dying, and secondly, by providing an explanation of the AM process and the terminology around AMs. My data revealed that the latter was often used as a way of cushioning or even obscuring the former. In other words, families were often informed that their relatives were being put on end-of-life medicines as a way of euphemistically informing them that their deaths were imminent: “I would usually explain that there might come a point when they’re unable to take oral medications…” (Dr. Kennedy, Consultant in Palliative Medicine, Participant 8) “…some of them will tell family that Mum is on morphine now, you might want to say goodbye soon”. (Ms. Foster, Care home nurse, Participant 30) End-of-life medicines are referred to by many competing terms – anticipatory medicines, Just in Case medicines, controlled drugs, etc. – all of which can be considered euphemistic in the end-of- life context insofar as they do not explicitly mention death. It is partly because there is so much jargon around AMs that they are so conducive to euphemism. Since jargon terms are so naturally ambiguous and thus conducive to the obscuring of meaning, it is easy to clandestinely smuggle potentially unwelcome news through them, such as news concerning death. To some extent, pharmacological euphemisms are a natural consequence of discussions about medications being at the centre of end-of-life care, something suggested by the following account: 177 “While their loved one is dying, we have to spend generally quite a lot of time explaining why the changes to medicines are taking place… explaining the terminology is very much at the heart of our end-of-life caregiving”. (Mr. Smith, Care home manager, Participant 1) Since medicines, and ‘explaining the terminology’ around them, are already central to end of life discussions with family, in seeking a euphemism for death, the language around medication naturally puts itself forward as the path of least resistance. 6.6 Complex cases vs. Bureaucratic shortcuts Pharmacological euphemisms are attempts to substitute emotionally difficult conversations about the realities of death with generic language relating to prescribing and medication administration. It may be interesting, however, to consider to what extent these prescribing and administration processes themselves can also perform this substitutive function: “For families, us giving the medication conveys the message that their loved ones are being kept comfortable”. (Mrs. Perry, Care home nurse, Participant 29) The idea that giving medication can be said to convey a message at all is intriguing. Discussions with families which involve the explanation of pharmacological terminology – a major locus of death euphemism, according to my dataset – are both about language and occur in language. Yet the discussions themselves are not strictly necessary to the euphemism, the simple act of prescribing or administering a certain drug can also be a way of tacitly conveying the message to a family member that their loved one is dying. This can be seen in the following account, cited in part earlier: “Sometimes the GPs won't tell us directly when our residents are end of life. They'll just prescribe the Just in Case pack when they think so and move on... So when families ask them why mum's been started on morphine, I’ve seen some of my staff saying it’s 'just in case.' Well 'just in case' of what exactly? I'm not really sure how many of them know why they're giving it”. (Mr. Anderson, Care home manager, Participant 35) Here, the phrase ‘just in case’ is a linguistic pharmacological euphemism which removes the need to directly mention death. Yet an even deeper euphemism is located at the level of the prescriber prescribing the ‘Just in Case’ pack, thus also removing the need to directly mention death on his part. My data shows, therefore, that busy GPs can use AP as a kind of professional shortcut; by letting their prescriptions speak for themselves, they are able to save time and essentially skip the 178 task of informing CHNs of important resident information. Another account revealed a similar phenomenon: “The hospital will send them back to us with anticipatory medications for end-of-life because it’s easier, especially when the residents don’t communicate. They don’t have a voice for themselves. It’s really difficult”. (Ms. Foster, Care home nurse, Participant 30) In the account cited earlier, the act of prescribing AMs fulfilled a role normally fulfilled by language – that of euphemism – taking the place of what would have been conveyed in a conversation. In this account, AMs substitute the function of language in an even more pointed way. Since the residents in question could not communicate verbally, and were thus unable to advocate for their own needs and preferences, hospital-based prescribing professionals let the medicines speak for these residents, branding them as end-of-life out of convenience. Conditions related to aging, such as dementia, opened up a vacuum in the language capacities of these residents, a vacuum into which AMs, the physical embodiment of a fatalistic prognosis, were subsequently enabled to flow. The fact that hospitals ‘send [patients] back…with anticipatory medications for end-of-life because it’s easier’, furthermore, points to an example of hospital professionals using AP as a power structure. In particular, they used their ability to prescribe – a form of ‘medical dominance’ harnessed by doctors but not nurses – to wield power over CHNs, reducing their own patient rosters and thus their crowded workloads by deflecting patients into care homes and increasing the workloads of CHNs. The reason for which anticipatory prescribing offers this type of bureaucratic shortcut to some professionals has to do with the fact that practically any patient of a certain age and condition can be classed as end-of-life, allowing them to be discharged from the hospital. The findings of Nugus et al.’s (2019) hospital-based study resonate with this observation: ‘The generalised character of the conditions of frail older patients, or patients with chronic or otherwise complex illnesses, made these patients more vulnerable than other patients to the formal, bureaucratic influence of the hospital on ED [Emergency Department] work’. (p. 387, Nugus et al., 2019) ‘The status of geriatric medicine and the aged care department as a ‘catch-all service’ was also reflected in the perspective of ED doctors. During observations, an intern said to the observing researcher: “It’s interesting, you know, that after a certain age, no matter what the condition, they come in under Aged Care”’. (p. 387, Nugus et al., 2019) 179 Ms. Foster’s account, just cited, gave the example of a hospital deflecting an elderly patient onto her care home in a manner similar to what Nugus’ study shows. These revelations penetrate to the heart of the issue of the balance between individualization and standardisation in end-of-life care. As we have seen, conditions related to old age such as dementia often greatly complexify end-of-life caregiving by rendering residents non-verbal, forcing CHNs to undergo the difficult and inherently individualized work of interpreting non-verbal signs and symptoms, which vary from patient to patient. ‘Complex’ conditions of this sort, which Nugus et al. identify most commonly with aged patients, thus typically require more individualized care than more simple ones. Yet the authors report that the more complex a patient’s condition, the more likely they were to be pawned off to a less individualized, ‘catch-all’ department. This was also seen in my data: “Unfortunately here, there’s not very many care homes that do syringe drivers. It’s risky and expensive. They’ve got to pay for individual training which they won’t, so they just don’t do them. At a maximum, maybe 6 out of the 18 nursing homes would do syringe drivers. The rest not only don’t, they choose the less complex patients to admit to their homes”. (Mrs. Fields, District Nurse and Community End-of-Life Care Facilitator, Participant 7) This is perhaps the clearest example yet of the protocol paradox: the fact that, even though end-of- life care might be considered the area of medicine most in need of individualized care, as it deals with elderly people who are especially to likely to be ‘complex patients’, it proves instead to be one of the more standardised areas in medicine, with nursing homes both foregoing ‘individual’ (i.e. individualized) training for CHNs and opting to take in a more standardised, simpler, less unique patient population. Nugus et al. reveal the tendency within care to view older patients as a monolith, a ‘catch-all’. This may have to do with the variety of possible age-related conditions (for instance, dementia) in care homes being overshadowed by the presence of cancer at the end of life (Higginson, 2016). The LCP Review states that ‘diagnosing imminent death is a far more imprecise science than people realise. An accurate prediction in non-cancer patients is particularly difficult’ (Neuberger et al., 2013). This statement both acknowledges that end-of-life care diagnoses can be significantly more complex than diagnoses in other medical specialties, and demonstrates that this fact is generally obscured by expectations that end-of-life patients are always cancer patients. My data shows something similar: “I think it’s very easy to overlook terminal care needs of patients with dementia because by definition they’re often poorly communicating. A lot of their symptoms can be quite internal and they don’t have the language. I think it’s easy to forget that palliative care is not just cancer and so I think the need for care is important for everybody, but it’s important to 180 emphasise it for groups who sometimes get overlooked”. (Dr. Jameson, General Practitioner, Participant 38) As Dr. Jameson suggests, these patients’ ‘lack of language’ – precisely the thing that should entitle them to more individualized care, the only measure which could excavate their inarticulable needs and wishes by adapting to their specific non-verbal cues – is what in actuality renders them vulnerable to being treated as part of a monolith (in this case, cancer patients) and thus treated in a more standardised manner. Another participant account showed that providing individualized care to patients who lack language can be the difference between life and (premature) death: “When Covid happened and an individual was in the hospital, we did find that they were put on end-of-life care without an explanation. He went in just not feeling very well and then he’s end-of-life all of a sudden. The difficulty is that we know the individuals that we support very, very well, especially this resident, who is non-verbal and can’t communicate his needs to somebody who doesn’t know him. We explained to the Consultant that the reason that he’s deteriorating is because he’s not having his needs met in hospital, and we asked for him to come back to his home environment. When he came back, he was absolutely fine. We were able to support him with eating and drinking, and the anticipatory medications that he was prescribed were never needed. He came out with a syringe driver and all the end-of- life medications but then within about 24, 48 hours, he’d massively improved and they were like ‘Oh yeah, he isn’t end of life.’” (Mrs. Larkin, Care home manager, Participant 20) This participant’s experience mirrors that of another participant cited earlier, who revealed that ‘hospital will send [patients] back to us with anticipatory medications for end-of-life because it’s easier, especially when the residents don’t communicate’ (with the important difference that the care home requested that the resident be sent back in this instance). In both cases, because the patients lacked language and were unable to articulate the nuances of their conditions – a problem which can only be solved by providing individualized care, by ‘know[ing] the individuals that we support very, very well’ – the hospital appeared to give up on understanding these conditions altogether, instead consigning them to the catch-all category of ‘end-of-life’ and prescribing anticipatory medicines. Multiple accounts already cited reveal instances of anticipatory medicines being prescribed far too readily, apparently because the existing bureaucratic structures of AP had simply made certain prescribing routes convenient. For instance, Mr. Anderson’s earlier account describes busy GPs prescribing AMs as a convenient, time-saving shortcut, using this practice as a means of avoiding challenging, time-costly conversations with care home staff and family about how their resident had 181 reached the end-of-life stage. Ms. Foster’s earlier account describes a hospital deflecting patients to care homes, ‘because it’s easier’. Further accounts fit into this pattern: “Whenever I’m with our residents with DNACPR [Do Not Attempt Cardiopulmonary Resuscitation] some GPs automatically think that it’s end-of-life and start prescribing away. So I have to say, ‘No it’s not,’ and that’s a challenge I’m dealing with every day”. (Mrs. Jiro, Care home nurse, Participant 24) “Our resident then developed a bit of a rattly chest and the GP immediately went in with Hyoscine and said, ‘Let’s just treat him as End of Life.’ I wasn’t necessarily sure that that was the right decision, but that was the clinician’s decision in that situation”. (Mrs. Raymond, Care home manager, Participant 25) One reason that the knee-jerk, ‘catch-all’ prescribing described in these accounts might be self- perpetuating is that it lacks a meaningful feedback element in the law – particularly the law around coroner’s investigations – that could assist in retroactively detecting instances in which it was carried out inappropriately. Teggi (2023b) established that, under current law, if a resident who dies within 14 days of the last doctor’s visit and had the standardised pack of prescribed end-of-life AMs in place, no Coroner’s investigation will take place and the GP can release a Medical Certificate of Cause of Death (MCCD), removing the chance of a post-mortem investigation. Teggi concludes that ‘care home managers and senior staff sought to pre-empt the possibility of a Coroner’s investigation by having a GP regularly visit a resident with EOL medication prescribed’ (pp. 245-247). Instead of anticipatory prescribing being used as a bureaucratic shortcut, then, this constitutes an example of anticipatory prescribing being used as a legal loophole which ensures that no coroner’s investigation could possibly determine any fault in that very prescribing. This can be considered perhaps the furthest extreme of the protocol paradox, where not only the dying process with its inherently individualized needs but death itself is treated as a vehicle for standardisation. The use of the standardised pack obscures the truly individualized nature of the death by precluding the possibility of a post-mortem investigation excavating the peculiarities and anatomical idiosyncrasies that are common to all deaths. The precedence taken by AP over post-mortems recalls Foucault’s (1973) history of medicine, which suggests that autopsies opened up the idiosyncrasies of death to the clinical gaze, leading to what he describes as the ‘discovery of the individual’ via death. It appears that in modern times this process is being reversed, with the element of individuality in death appearing ever less important within the wider bureaucratic AP system, despite remaining as deeply and fundamentally personalised as it ever was. 6.7 Chapter summary 182 In this chapter, I examined the protocol paradox, a contradiction which emerges from the tensions between individualization and standardisation within the AP system. Initially, I explored how this manifested in tensions inherent in AP guidance documents, which tout individualization as a general value yet fix specific rules in place to ensure standardisation. I then moved on to consider medicines administration, examining the paradox that the finely-modulated, iterative administration practices (‘micro-decisions’) needed for personalised care must depend for their supply of medicines on a standardised, algorithm-dependent prescribing system geared towards deceptively clear-cut models of symptom-medicine relationships and the prescribing of stock doses. My data showed that finely- modulated administration depends on a method of symptom identification which remains hyper- attuned to the minutiae of resident symptoms, especially in cases where dementia renders the resident non-verbal. I explored how the words that CHNs use to describe symptoms often predispose symptom identification towards individualization. On the flipside of this, I explored accounts in my data which described the use of inherently standardised and nonspecific forms of language – jargon and euphemism. I went onto show how the jargon of anticipatory prescribing was specifically conducive to euphemism, partly because the systemic structure of the AP process was conducive more generally to the gaming of professional vested interests (e.g. patient deflection). Finally, I connected examples of professionals using AP to game the institutional system of medicine to the generalised, ‘catch-all’ manner in which elderly patients are often treated throughout medicine. 6.8 Discussion o Contributions to literature and theory Perhaps the most crucial finding in this chapter was the simple notion, reported by a participant, that ‘people pass away in a very individual way’. The fact that the incidence of dementia during the end- of-life period is so consistently high, with one report estimating that 80% of care home residents suffer from it in some form (Alzheimer’s Association, 2022), points to a greater need for individualized care. In spite of this, it appears that at the moment when patients are most likely to need personalised care, they are instead often funnelled, at the first sign of an end-of-life symptom, into a standardised prescribing protocol which tends to treat all end-of-life patients as if they only suffered from the four same symptoms. This paradox – the protocol paradox - was contextualised alongside Nugus’ (2019) finding that complex cases within hospitals are often ‘deflected’ onto geriatrics, which is treated as a ‘catch-all’ department. The protocol paradox deals with the problem of creating a standardised set of rules with which to deal with dying individuals each one of whom, due to the fact that every death is an individualized process, will always introduce some minor 183 variation to the system in which they are embedded, thereby nudging the ideal parameters of those rules in one direction or another. The protocol paradox also points to a deeper paradox located within the concept of individuality itself – specifically, within the problem of reconciling one’s own individuality with its place in the collective. As Giddens (1984) puts it: ‘while the continued existence of large collectivities or societies evidently does not depend upon the activities of any particular individual, such collectivities or societies manifestly would cease to be if all the agents involved disappeared’ (p. 24). The problematic relationship between individuality and collectivity was most famously formalised by the Ancient Greek philosopher Zeno in his paradox of millet grain, which observes that a single grain of millet does not make a sound when it falls, whereas a thousand grains do make a sound. In drawing parallels with Zeno, I join Papachroni and Heracleous (2020) in attempting to integrate Ancient Greek paradoxes with Smith and Lewis’ (2011) paradox theory. Papachroni and Heracleous (2020) have themselves done this with Zeno’s paradoxes of motion. In a previous chapter, I argued that weak lines of communication and differences in perspective between agents in a system or institution led to individuals viewing the same thing in different ways, resulting in paradoxes. In this chapter, I have explored how jargon and euphemistic language operate at the intersection of these different perspectives. Jargon terms and euphemisms inherently tend towards paradox insofar as they often consist of two competing meanings within the same term, one for laymen and another for those in the know. By uniting the lenses of paradox and a Foucauldian appreciation of language and discourse, this chapter found an apt framework through which to understand this phenomenon. Furthermore, by using interviews to excavate significance from linguistic terms used within the AP space (e.g. ‘controlled drugs’) – terms often taken for granted and used without thinking - this chapter undertook what my method, Reflexive Thematic Analysis, calls ‘latent meaning’, i.e. meanings which are underlying or not immediately visible within interview data. The emphasis on discourse and the history of medicine exhibited by Michel Foucault in his Birth of the Clinic (1973) has proved extremely fruitful as a theoretical lens in this chapter. Most valuable have been his insights into the problematic relation between the body and language, which set up my discussion of the individualized language of symptoms and the standardised language of medical jargon. Foucault’s unique contributions to the history of medicalised death also render his work uniquely suited to the topic of end-of-life care specifically. Foucault demonstrates a remarkable sensitivity to the way that death, insofar as it constitutes the limit or boundary to medical knowledge, has the ability to reconstitute those boundaries when public perceptions toward it change, or when a 184 new discovery related to death is made, thus causing major paradigm shifts to ripple through the whole institution of medicine. Foucault (1973) even goes so far as to posit his triad of death, language, and the individual as the original historical foundation for modern positivistic medicine, arguing that that modern medicine came into being when it was able to ‘rest upon the stable, visible, legible basis of death’ (p. 196). This chapter contends with what might be considered positivistic medicine’s furthest extreme – the assumption that human pathology is so fundamentally reducible to the ‘stable’, ‘visible’, and ‘legible’, that it can be whittled down to prescribing algorithms. Tracing this paradigm to its current instantiation in end-of-life care, as this chapter has done, closes the loop, examining how a paradigm which originally emerged out of a certain conception of death perfectly serves, perpetuates, and intensifies modern medical conceptions of death as a standardised condition in which symptoms are easily predictable. By drawing on Foucault’s (1973) conceptualisations of death and language, I was able to unearth useful insights concerning the point at which they converge – euphemism. Medical euphemisms surrounding death are well studied in the literature (Barlet et al., 2022; Pitt et al., 2015; Xafis et al, 2016). Indeed, such euphemisms are so common that general surveys of jargon and euphemism often end up inevitably touching on the topic of death euphemisms. Batten et al. (2019), for instance, note that the term ‘treatable’ can euphemistically obscure the reality of a prognosis; “your cancer is treatable” might mean ‘your cancer is curable’ to a patient, while instead meaning ‘although your cancer is likely incurable, we can still treat you with chemotherapy’ to a doctor. This chapter provided an original contribution to Barlet et al.’s (2022) four classes of euphemisms for death in medicine – a new category which I have termed pharmacological euphemism. Pharmacological euphemisms involve using the language of anticipatory prescribing to euphemistically break the news that a resident is dying. An interesting further finding of my analysis was that, unlike most death euphemisms codified in literature – e.g. ‘your heart rate may drop and not recover’ or the simple term ‘pass away’ (Pitt et al., 2020) – the use of the language of anticipatory prescribing as a euphemism could be substituted by the actual act of prescribing, and still accomplish the same goal (usually, the goal of making things easier for the professional using the euphemism: prescribing AMs, for instance, can eliminate the need to inform CHNs and family about a resident’s end-of-life status). o Recommendations for practice and policy 1. Integration of pharmaceutical knowledge regarding synergistic reactions into protocols 185 A finding of this chapter was that, in order to meet the demands of individualized care, CHNs utilise finely-modulated, time-sensitive, adaptive, intuitive, and iterative methods of AM administration which involve remaining cognizant of the synergistic interrelations between the different medicines which they introduce into residents’ systems at any one time. However, it also appears that some CHNs are less aware of these interrelations, with some highlighting the need for education regarding when to prescribe one AM as opposed to another. This chapter also found that the prescribing protocols participants described using during my interviews often do not adequately warn of the possible synergistic reactions between medicines. The diagram (Figure 6.1) I provided earlier from Dickman and Scheider’s (2016) The Syringe Driver (p. 120), which maps out synergistic reactions between commonly-prescribed AMs, is an example of the type of information resource which prescribing protocols could benefit from integrating. An example of such integration can be found in the Medical Handbook from the Isle of Wight NHS Trust (Lynch, 2017), in which The Syringe Driver is listed as further reading. Dickman and Schneider are Consultant Pharmacists, and The Syringe Driver can thus be considered an excellent example of pharmacists acting as information- givers within the AP space, which multiple studies suggest could improve health care outcomes for all parties involved (Breen, 20118; Ibrahim et al., 2023; Latter et al., 2022). A drive to integrate pharmacists’ specialist knowledge into protocols could work in tandem with the integration of pharmacists into the care home space, a recommendation which I made in Chapter 5, citing the MOCH study, which accomplished precisely this (NHS, 2018). 2. Adoption of ReSPECT documentation for families of residents Pinder et al. (2005) emphasise the importance of families in assisting with individualized care, summarising their study of ‘pathways’ (i.e. protocols) thus: ‘What had started out as a one-size-fits- all model of professionally-delivered care turned out to be dependent for its success upon a complex (and unmapped) network of relatives, friends and informal carers in the community’ (p. 762). In this chapter, my data showed that GPs were occasionally reported to override the wishes of families, contained in ReSPECT forms, in favour of using standardised protocols. ReSPECT documentation records patient preferences gleaned through conversations between the patient or their families (Eli et al., 2021), and proves particularly valuable if a patient is unable to make or express choices about their care. ReSPECT forms therefore capitalise on family members’ potential to provide individualized knowledge concerning residents, which enables individualized care. In light of these factors, this chapter recommends the adoption of documentation of this kind wherever possible. 3. Combatting Jargon and Improving Communication The data concerning family members covered in this chapter also pointed to numerous instances of families experiencing confusion or fear in the presence of medical jargon and euphemism relating to AP, including terms like ‘controlled drugs’, or ‘just in case’. This finding links to wider research 186 on how families interpret the use of AMs at the end of life, their symbolic significance, and the importance of tailored communication. Payne et al. (2015) found that inadequate and inaccessible communication, information, and support regarding end-of-life medications can result in additional burden and distress for family carers. They emphasised the need for effective communication between primary care teams and families about end-of-life medications and their respective roles and responsibilities to alleviate anxieties related to the administration of end-of-life opioids, which have symbolic significance associated with deterioration and approaching death (Payne et al., 2015). Furthermore, Spacey et al. found that when there are fewer discussions about care planning, and patients and families are unable to express their preferences regarding end-of-life care, there is an increase in unnecessary hospital admissions, as evidenced by the many care homes during the COVID-19 pandemic (Spacey et al., 2021). Tailored communication could perhaps be improved by circulating Pitt et al.’s (2020) classification system of seven types of medical jargon, intended to support clinicians in recognising and eliminating jargon-oblivion. 6.9 Conclusion Having interrogated the paradoxical relationship between the individualized nature of the end of life and the standardised practices which dominate end-of-life care, I will now move onto an exploration of the paperwork paradox: the fact that the excessive bureaucratic documentation imposed on care homes detracts from the time needed for CHNs to provide good and patient-centred care, while simultaneously demanding that care. 187 Chapter 7 The Paperwork Paradox This chapter focuses on the Paperwork Paradox theme and its tension-based subthemes. The chapter is organised thus: 7.1 Background, 7.2 Introduction, 7.3 Clock-based Time vs. Task-based time, 7.4 Granular Certainty vs. Structural Stupidity, 7.5 Risk Management vs. Error Multiplication, 7.6 Coercion vs. Moral Principles, 7.7 Guidelines vs. Orders, and 7.8 Discussion, in which I summarise my findings and outline its contributions to theory, to literature, and for practice and policy. 7.1 Background The ever-increasing role of paperwork and bureaucracy in UK care homes points to a wider landscape of health care regulation by now so established that it may be difficult to remember a time when things were ever different. As early as the 1980s, medical anthropologist Arthur Kleinman noted: ‘The immense profusion of bureaucratic paperwork is meant to control the practitioner’s actions through a regulatory approach taken over from various governmental sectors. It ends up numbing them and usurping huge blocks of time better spent with patients’ (p. 264, Kleinman, 1988). John Burton, in a blistering critique of the over-regulation and bureaucratisation of care, dates the issue to around 2004, contrasting it unfavourably with the more person-centred care culture in which he participated in the 1970s (Burton, 2016). This turn towards over-regulation can be contextualised within the wider flourishing of a national culture of risk management which has been dated to the mid-1990s (Power, 2004) and to the publication of the Department of Health’s 2003 National Minimum Standards for Care Homes (Popham et al., 2011). One of the major precipitates of bureaucracy in the care sector is paperwork. Paperwork issues from three regulatory levels: the national government (including NHS, NICE, and the CQC); local authorities (social services, health service Trusts and CCGs); and the organisations operating care homes. Poldrugovac et al. found that that regulatory bodies tend to focus on assessing compliance with bureaucratic requirements, which may not always prioritise resident care (Poldrugovac et al., 2023). The organisational, bureaucratic culture which has overtaken care has been described as directly interfering with the one-on-one contact and communication which, according to Kitwood, is needed to maintain dementia patients’ sense of identity (Burton, 2016; Kitwood, 1997; Killett et al., 2014). Especially inimical to person-centred care is the ‘checklist’ or ‘tickbox’ approach typically encouraged by many regulators and much paperwork (Popham et al., 2011). The 2020 report ‘Busting Bureaucracy’ identified six areas where bureaucracy was found to be particularly burdensome: duplicative data requests, complex regulations, lengthy appraisals, mandatory training, time-consuming staff processes and information management, out-of-date and prescriptive 188 legislation, and time-consuming procurement processes (DHSCb, 2020). Studies have found that documentation and administrative responsibilities limit managers’ capacity to meaningfully engage with residents (Dewar et al., 2019; McGilton et al., 2016; Penney et al., 2018). A 2019 King's Fund report revealed an alarming disconnect where frontline practitioners were frequently uninformed about updates to guidelines due to the sheer amount and complexity of regulations (The King’s Fund, 2019). 7.2 Introduction Many studies, including those quoted above, diagnose the problem with the current regulatory climate of care to be excessive bureaucracy, rather than bureaucracy in itself. The 2020 report ‘Busting Bureaucracy’, for instance, considers bureaucracy to be an issue of more or less, ‘minimum’ or ‘excess’: ‘There is an irreducible minimum level of bureaucracy in any complex system. Rules and processes are essential means to manage risk… But excess bureaucracy reduces the time that staff have for care… This negatively impacts staff well-being, morale and retention while hindering the very outcomes the processes aim to support’. – Busting Bureaucracy, Department of Health and Social Care (p. 1, 2020b) This chapter argues that paradoxical situations – such as the way bureaucracy ‘hinder[s] the very outcomes [its] processes aim to support’ – are not a side-effect of excessive bureaucracy, but rather fundamental to all bureaucracy. It has been shown that paradoxes are a natural and inescapable consequence of organisational complexity (Lewis, 2000). Bureaucracies rest on the logic of ‘divide and conquer’ – i.e. that tasks can be made easier by dividing them into smaller components and assigning them to separate individuals. In general, bureaucracies are complex structures which consist of many parts which are separate yet interconnected (e.g. professionals, institutions, buildings). When several elements interact, it is likely that at least two of them enter into a contradictory relationship, particularly when these elements are interconnected enough to be able to communicate in the first place but also distinct enough for that communication to be imperfect. Different professionals – e.g. CHNS and GPs – sometimes develop different ideas about the same issue (e.g. the concepts of ‘training’ or ‘anticipation’), as I have shown in previous chapters. A paradox can be defined as ‘the simultaneous presence of contradictory, even mutually exclusive elements’ (p. 2, Cameron & Quinn, 1988). The paperwork paradox - the way in which excessive paperwork detracts from the time needed for the patient-centred care demanded by that very paperwork - stems partly from the inherently contradictory nature of bureaucracy; as a consequence of the decomposition of the care home bureaucracy as a whole into separate parts which do not always communicate (e.g. NICE, the CQC, local authorities), these different institutions often 189 produce separate documents all requesting the same information, forcing care home professionals to honour these ‘duplicative’ data-requests at the expense of time with their residents. Following Morgan et al.’s (2020) observation that end-of-life care research lacks sufficient engagement with theory, this chapter engages with paradoxes of bureaucracy and paperwork formulated by sociologists and anthropologists. Paradoxes often spring up at the point where intuition breaks down (Lewis, 2000); while purely scientific studies which do not use theory are typically oriented towards yielding lucid, certain, and conceptually stable ‘findings’, the wider theoretical work done on paradoxes can be useful in navigating the often opaque and counter- intuitive territory of care home bureaucracy. The literature on the topic of bureaucracy largely discusses the abundance of paperwork rather than engaging with paperwork in itself (i.e. specific files, forms, and documents). Drawing on the work of media ecologists Marshall McLuhan and Walter Ong, as well as more recent media historians Craig Robertson and Ben Kafka – all of whom are devoted to foregrounding aspects of our information storage media we instinctively deem ‘neutral’ vessels and take for granted – this chapter homes in on something deeper, the structural specifics of paperwork: its design, its technological medium, and the way in which it carves up and controls time. It also explores the bodies which disseminate the paperwork – regulatory bodies such as the CQC – and the legal dimension of this regulation. This exploration is grounded in legal theory, including interpretivism. Craig Robertson, in The Filing Cabinet: A Vertical History of Information, makes the claim that ‘storage is never neutral practice.’ (p. 27, Robertson, 2021), focusing specifically on information storage, specifically the medium of paperwork or files. He argues against the tendency to view ‘an object like the filing cabinet as inert or frozen’ (pp. 15-16), as a neutral or empty vessel for the storage of information: ‘To think about the filing cabinet from the perspective of filing is to argue that “ways of doing” are important because they “engender systems of knowing and modes of social organization.” [Shannon] Mattern uses the concept to connect the histories of technologies such as bookshelves, desks, card catalogs, and server racks’. (p. 14, Robertson, 2021, The Filing Cabinet) In focusing on paperwork in this chapter, I also argue that the forms, files, sheets, logs, and charts circulated throughout care homes, which act as media for the storage of a complex admixture of rules, records, and reminders, can be made to yield interesting revelations about the ‘social values’ of the system of care as a whole, and especially that system’s emphasis on regulation, compliance, law, and punishment. Research question 190 • How do the structural logics underpinning AP paperwork contribute to the tensions between navigating bureaucratic regulations and providing good, patient-centred end-of-life care in care homes? 7.3 Clock-based Time vs. Task-based time AP paperwork acknowledges, even centres, the importance of individualized, patient-centred care. CQC guidelines, for instance, emphasise that ‘People and communities should always be at the centre of how care is planned and delivered’ (CQC, 2023). What these guidelines appear to overlook, however, is the dimension of time. The basic constraint that time is zero-sum means that time spent doing paperwork detracts from the time needed to provide the individualized, patient-centred care mandated by that very paperwork. This phenomenon, which I term the paperwork paradox, was described thus: “I’m always buried in paperwork. It just takes up so much of my time… it’s supposed to help us give good and personal end-of-life care to our residents. We're supposed to do that and be present and be compassionate, but here I am shackled to forms that keep me away from the important stuff”. (Mrs. Laney, Care home manager, Participant 19) Another participant noted the time-cost of implementing another form of paperwork, Reuse of Medicines guidance, during the pandemic: “In each individual case you’ve got to speak to the practitioner and the family members and get everything in order and log it compulsively and so on and so forth… We have to spend more time working our way round it… than we spend actually implementing it, if it’s implemented at all”. (Mrs. Raymond, Care home manager, Participant 25) Yet another participant, speaking of a specific form of paperwork known as a Medication Authorisation and Administration Record (MAAR chart), remarked: “I do feel the amount of documentation has led to the GPs prescribing less anticipatory medicines” (Mrs. Fields, District Nurse, Participant 7). This suggests that the excessive demands of paperwork can often result in residents receiving fewer AMs, potentially compromising the quality of care. Madeleine Bunting’s (2020) Labours of Love describes a common situation in which care home staff, forced to perform tasks within narrow time slots, find themselves ‘caught between clock-based time and task-based time’. She notes that ‘Half an hour may not be long enough to wash and dress a client with dementia’ (p. 42). In Dr. Oliver Sacks’ (1991) Awakenings, he similarly observes the radically different perception of time held by patients suffering from another neurodegenerative 191 disease typically associated with aging populations – Parkinson’s Disease. ‘Parkinsonian Space and Time’, he suggests, is contradictory: it ‘gets bigger and smaller, it bounces back on itself, and it loops itself round’ (p. 340). Care staff attempting to wash or dress a resident with a neurodegenerative condition are tasked with negotiating between a resident’s lived and situated experience of time and space and the more abstract, standardised form of time built into MAAR sheets – and, indeed, timesheets. Emma Dowling (2022) observes, in the wider landscape of community-based care work, the widespread ‘practice of not paying for travel time’ (p. 119) - that is, travel between the houses of elderly clients - again indicating a model of clock-based time which expects tasks to be completed at certain times without regard for the logistics of doing so. The paperwork paradox arises, in a simple sense, out of this tendency within care home organisations to underestimate the time needed to provide patient-centred care. The model of clock-based time used within AP paperwork, specifically the MAAR chart, is best described by one of my participants: “The way the MAAR chart works, each item’s got the quantity on there. It’s also got the timings, and it’s done in a 28-day kind of circle. So, for instance, if a patient takes Paracetamol for two tablets four times a day, it would have the times, 8 o’clock, 12 o’clock… 3, 6, basically. And then each day, it would have the little 8 boxes basically, along those lines”. (Mrs. Greensborough, CHM, Participant 28) This model of time largely corresponds to the description of ‘printed calendar time’ given by media theorist Walter Ong (1982). According to Ong, printed calendars are the best example of the standardisation and abstraction which paperwork inflicts upon time, disrupting time’s fundamental essence – dynamism, ephemerality – by anchoring it to the static space of the printed page: ‘Time is seemingly tamed if we treat it spatially on a calendar or the face of a clock, where we can make it appear as divided into separate units next to each other. But this also falsifies time. Real time has no divisions at all, but is uninterruptedly continuous: at midnight yesterday did not click over into today’. (p. 76, Ong, 1982, Orality and Literacy) Ong’s conceptualisation of time accords with that of Giddens, who points out that, within the ‘time- space’ in which human agents are situated, time is perceived as a durée, a continuous flow (p. 35, Giddens, 1984); it is lived. The clock-based timeslots of a MAAR chart carve up the durée of time into manageable portions, viewing care home staff and residents only in a series of snapshots or ‘timed assessments’ (Bunting, 2020). The paperwork paradox is partly attributable to the tension between clock-based time, the abstract time of regularly spaced snapshots, and task-based time – the time which it actually takes to complete tasks (including the task of completing paperwork itself), a time which is felt not as an average rate (e.g. ‘two tablets four times a day’), but as a duration. 192 Paperwork is the site of this paradox because it straddles both conflicting forms of time at once – on the one hand, clock-based time, which it disseminates, and on the other, task-based time, the time in which it is filled out. The act of completing paperwork in order to meet the demands of clock-based time is something which can only occur in the very task-based time to whose complexities that paperwork appears constitutionally blind. The tension between clock-based and task-based time also plays out in prescribing. One of my participants, cited earlier, reported: “GPs still love patches, which can be great, but the GPs don’t recognise that they are slow release… But the GP is part of prescribing a clear pathway because I think they get nervous that we’re going to overdose people…” (Mrs. Perry, Care home nurse, Participant 29) This participant suggests that the GP’s reluctance to prescribe the additional medication needed to manage patients’ pain while awaiting the effects of an opioid patch may potentially stem from a desire to keep to a ‘clear pathway’. ‘Pathways’, or prescribing protocols, are documents comprised mostly of boxes and arrows, which thus exhibit the same properties of atomisation and rigid internal subdivision as do MAAR charts. They typically either overlook time or view it in abstract, discontinuous increments. The above account can therefore be considered an example of the abstract clock-time of AP paperwork taking precedence over the realities and complexities of the task-based time of care. In particular, the time dimension of AM administration (their ‘slow release’ nature) is overlooked. Foucault (1973) attributes to clinical medicine the assumption that ‘There is no process of evolution in which duration introduces new events of itself and at its own insistence; time is integrated as a nosological constant, not as an organic variable. The time of the body does not affect, and still less determines, the time of the disease’ (p. 12). Indeed, while the clock-based time of protocols deal with neat average rates (e.g. timed assessments which occur every three days), the task-based time of AM administration (i.e. time as it is actually experienced) proceeds in less predictable ways, involving as it does the complex process of simultaneously negotiating between the various different irregular timescales represented by various medicines, as well as negotiating the time of the body, an ‘organic variable’, which may react to one medicine at a different rate to another medicine. Giddens’ (1984) model of time-space is interesting in how it feeds into his conception of information storage: ‘”Storage” is a medium of “binding” time-space involving, on the level of action, the knowledgeable management of a projected future and recall of an elapsed past… Storage here already presumes modes of time-space control, as well as a phenomenal experience of “lived time”…’ (pp. 261-262, Giddens, 1984, The Constitution of Society) 193 Thus, Giddens argues that forms of information storage can be used to organise and indeed control time. One example he provides of this is the way bureaucratic institutions use paperwork to control employee time: ‘Control of time is characteristic of bureaucracy in general… Modern bureaucracies, Weber asserts, could not exist without the collation of documents which are both records of the past and prescriptions for the future – the “files”’. (p. 152, Giddens, 1984) This idea of control perhaps sheds some light on the paperwork paradox – the fact that paperwork both mandates good care and detracts from the time needed for care, a contradiction that arguably stems from an inability to conceptualise paperwork as something which takes time to complete. Perhaps it is precisely because bureaucratic paperwork organises, polices, and records time that it is unable to view itself as being subject to the conditions of time. If time is being conceptualised as a form of control to be imposed on employees via paperwork, it stands to reason that this paperwork would not appear to be itself bound by the very tool it uses to bind others. Duration is a condition of dynamic lived time-space. That paperwork considers itself exempt from the effects of duration is evident from the way in which it slices up lived time-space into abstract but homogenous snapshots, from which any sense of duration has been drained away. For Giddens, furthermore, focusing on how paperwork, an information medium, controls and shapes the experience of time is a way of applying Marshall McLuhan’s well-known maxim that ‘the medium is the message’, that ‘the character of [an] information medium… directly influence[s] the nature of the social relations which it helps to organize’ (p. 261-262, Giddens, 1984). Marshall McLuhan’s maxim aims to correct for the same tendency to view information storage as neutral, self-effacing, ‘inert and frozen’, also identified by Robertson (2021) – essentially, our tendency to grow so familiar with the media we use that we become blind to its influence and effects. It is likely that this tendency contributes to the paperwork paradox. The contradiction that the time it takes to complete paperwork detracts from the care that paperwork seems to demand can be considered a by- product of the fact that paperwork in general is considered a ‘neutral’ form of information storage or record-keeping. What Robertson calls ‘information labour’ – the labour associated with sorting, managing, processing information – is not intuitively considered to be labour at all, in the sense of being a form of action that occurs in time and can thus take time in the same way that, for instance, physical activity can. Bureaucratic records are always records of something or another; there exists a strong bias towards considering that something to be the real event, and to consider the record to be a mere side-effect of it, one which does not therefore possess the chief characteristic of a real event – duration. This accounts for why paperwork carves out timeslots in which physical tasks are performed in care, yet makes no provision within these timeslots for the time needed to fill out the paperwork itself, implicitly dismissing the (bureaucratic rather than physical) information labour expended on recording the event as a sort of atemporal non-event. 194 The NMC’s (2009) guidance states: ‘Good record keeping is an integral part of nursing and midwifery practice… It is not an optional extra to be fitted in if circumstances allow’. And yet, insofar as paperwork allocates specific tasks to specific intervals of time but does not carve out any specific room within those intervals for the meta-task of recording the task itself, perhaps this meta- task can only ever be felt as an ‘optional extra’. Though the NMC’s own phrase ‘in real time’ seems to assume that the record kept of a task and the task itself can be exactly coincident, there will always be some small lag separating one from the other. The very fact that a task must already be completed before it can be recorded makes the act of recording appear as a coda, supplementary to the task, even in excess of it. Clock-based time conceives of time only as an average rate (‘two tablets four times a day’), and as such, can only conceptually accommodate ‘neat’ and predictable forms of time which occur at regular and repetitive intervals (Bunting, 2020). By definition, then, the few odd minutes it takes to record tasks will fall beneath the radar of this average rate – an unassimilable, uneven fragment or chunk of time which will tend to be perceived as the overhang of a task rather than an actual task in itself. 7.4 Granular Certainty vs. Structural Stupidity An account which details the CD disposal process sheds light on the sheer volume of paperwork produced by any given task related to anticipatory medicines: “They’ve [CHNs] got like a book where they record the medications; administration, death, expiry, everything. And they will bring back all the medications to us with the book, and I will sign it. There is also a separate sheet just for the CDs for me to sign as well. What I will do is check it, make sure the amounts of CD administered checks out with how much they’ve brought back. Then it’s destroyed. I initial and date it, make two copies, one for the care home and one for me. I just keep it with the rest of my CD invoices, just to be on the safe side. It sounds meticulous because we have to be”. (Mr. Nanda, Pharmacist, Participant 41) The practice of ‘making two copies’ here exemplifies the duplicative logic of paperwork. This logic is also exemplified by duplicative data requests, which were reported by my participants to be a major problem within AP bureaucracy: “NHS and Pan-London want the care homes to use the lovely new Pan-London MAAR authorisations, but they already have their MAAR charts. So they’re double-counting their stock. If they’re administering something off the MAAR sheet on their own records, they then forget that they’ve got to administer it off of the, so there’s a lot of duplication”. (Mrs. Fields, District Nurse and Community End-of-Life Care Facilitator, Participant 7) 195 The 2020 DHSC report ‘Busting Bureaucracy’ cited duplication as not just one of the most, but the most taxing problem created by care home bureaucracy (DHSC, 2020b). It is important to contextualise this problem geographically; one Welsh participant, for instance, identified duplication as a uniquely English problem: “We try to do as much as we can at the national level, reduce any duplication and do it once rather than seven times… I guess it is easier for us to do it rather than England because England is a big old complicated machine. We can get seven people around the table represented the health boards pretty easy… So we can try to work on one document rather than a number of them… And there's always subtle differences between them all”. (Mr. Rowlands, Community pharmacist, Participant 5) Duplication is rare in Wales due to the centralisation of the institutions constituting the Welsh care sector. When said institutions are not on the same page, they can often independently request the same information. This is demonstrated in Figure 7.1, which provides a side-by-side comparison of two CD logs that are required to be completed within a single care home: Figure 7.1 A comparison of two CD logs, that are both required to be filled out Source: (NHS, 2021b) Evidently, the two CD logs above require nearly identical information, yet both must be completed to satisfy different regulatory bodies, one internal (the care home organisation) and another external (the CCG and CQC). Duplication can be considered something fundamental to the very design of paperwork. This design, the basic anatomical skeleton of any form or file, is made up of ‘boxes’ or ‘cells’. These cells act as standardised, easily repeatable modules that can be effortlessly copied, duplicated, or multiplied to make up the form. Cells can be cloned and transplanted across a range of contexts. In Figure 7.1 196 above, for instance, cells for ‘Date’, ‘Time’, ‘Received From’, ‘Quantity/Amount Given’, and ‘Quantity/Amount Received’ are standard across both documents. It is perhaps because cells are so easily repeatable that they can start to proliferate in a way which puts them out of touch with reality. Cells can be duplicated at a much lower energy cost than is required by the majority of the activities to which those cells refer, activities carried out in the real world. Thus, the single act of administering a drug may spawn ten boxes relating to different aspects of this one act (‘Time’, ‘Quantity Given’, etc). As one participant put it: “In my head, every decision I make with those drugs corresponds to filling in about ten boxes”. (Mrs. Graham, Care home nurse, Participant 14) In the long run, this 1:10 ratio between ‘real’ activities their ‘paper’ equivalents can build up, such that CHNs find themselves spending more time creating records of their care than actually performing that care. The duplicability and replicability of cells can also be traced in part to the fact that paperwork belongs to the medium of print, a technology which specialises in producing identical or ‘exactly repeatable’ copies of standardised documents (p. 129, Ong, 1982), as well as to the electronic medium, which can circulate copies of documents, saved as files, with speed and ease. This can help to explain the paperwork paradox: it is perhaps because technologies make circulating paperwork quick and easy for the regulators who issue this paperwork that these regulators seem unaware of the time required to complete it. For regulators, circulating paperwork involves pressing a button and producing a copy of an existing document, whereas the CHNs filling in the paperwork must constantly supply fresh information to each individual cell within that document, manually, longhand. One participant even reported having to produce paperwork from scratch: “Yeah, I mean, it’s hard because there’s only a certain amount of pharmacies that you can use. Not only that, because we’re a nursing home, we have to have MAAR charts attached to that for us to actually give it out. And if a care home doesn’t have a MAAR chart and a pharmacy can’t provide a MAAR chart then we have to handwrite all the MAAR charts, and it has to be two nursing professionals, it has to be two secondary signatures, there’s so many loops to jump through”. (Mr. Jeffries, Care home manager, Participant 9) That paperwork can be so easily duplicated and circulated via technology perhaps produces the false impression of it being a lightweight affair. Situations where CHNs were forced to write out an entire MAAR chart longhand shed some light on the sheer amount of information they were actually receiving in every time they received a printed copy in the past. David Graeber alludes to the duplication – or, as he puts it, ‘accretion’ – of cells in his essay on the ‘structural stupidity’ of paperwork: 197 ‘Paperwork… is designed to be maximally simple and self-contained. Even when forms are complex, even bafflingly complex, it’s by an endless accretion of very simple but apparently contradictory elements, like a maze composed entirely of the endless juxtaposition of two or three very simple geometrical motifs’. Utopia of Rules, Graeber (p. 52, 2015) The cells of MAAR charts and other documents exist so that CHNs ‘can tick that box to say, yes, that’s been given at that time on that day’, as one participant put it (Ms. Jones, CHN). Boxes or cells exemplify what Robertson (2021) terms the ‘granular certainty’ central to the logic of paperwork, the idea ‘that breaking things down into small parts to produce a high degree of detail or specificity will result in efficiency’ (p. 132). Cells represent a ‘conception of information as a discrete unit’, which applies to many different types paperwork (p. 132). According to Robertson, ‘[Granular] Certainty indicates the conviction that greater specificity will reduce individual discretion and increase the likelihood that a task will be completed efficiently’ (p. 132-133). Paperwork breaks information down to small, easily digestible units, each one of which one can theoretically be more ‘certain’ about since they are so manageable. However, as one participant’s description of a form of paperwork known as a CD register demonstrates, this fosters an appetite for accounting for every single example of even the most minute of things: “Compared to hospital settings, where you would count the stock daily and sign a drug check book, here we have to write individual lines in the CD register for every individual medicine. I mean, you’re only writing stock checks, but if you’re writing stock checks on twenty, thirty or more medicines, it’s time consuming”. (Mrs. Greensborough, CHM, Participant 28) Here the way in which granular certainty is ensured by breaking down information into its absolute smallest units – individual medicines that are normally viewed in aggregate – is likely a result of the fact that the design of the paperwork is broken down into small units (i.e. lines) which support this approach. As Graeber observes, precisely because cells, lines, and other such granular units are so simple and manageable, they tend to multiply or undergo ‘accretion’, causing them to build up and, paradoxically, become far less manageable at the macro-level. This participant observes that, though the information being recorded here is fundamentally simple and manageable (‘you’re only writing stock checks’) because the act of recording an individual medicine is so minute, these acts build up, eventually reaching ‘twenty’ or ‘thirty’ medicines, which naturally proves ‘time consuming’. Several participants thus reported remaining highly alert to the ever-present threat of making errors while completing paperwork: “I check them and I date them and I sign them and I make sure that all the anticipatory medications are all there, because anybody can make a mistake. The controlled drugs book 198 gets double checked by both medication-trained staff because they’re important and they’ve got to be right”. (Mrs. Ainsley, Care home manager, Participant 15) “You have to sign and check the CDs after every shift, but in some homes that’s done weekly… The former can burn you out, but checking only weekly means you’ll catch more errors and have less time to fix them”. (Mrs. Gardner, Care home nurse, Participant 22) In Graeber’s ‘Dead Zones of the Imagination: An Essay on Structural Stupidity’, he recounts the negative experience of assiduously filling out a form for a health care insurance provider, only to have the entire form rejected on the basis of his having made a single, small error (Graeber, 2015). Graeber puzzles over why seemingly intelligent people make ‘silly mistakes’ when completing paperwork, concluding that the ‘structural stupidity’ (i.e. the poor design) inherent in much paperwork is experienced by individuals as their own ‘stupidity’. Ben Kafka, in his introduction to The Demon of Writing (2020) explores paperwork errors through the lens of contradiction or paradox: ‘This work is about paperwork and its contradictions. It begins with the observation that notwithstanding its reputation for tedium, paperwork is full of surprises… Behind each [document] are hundreds or thousands or hundreds of thousands more opportunities to misspell a word, miscalculate a number, misread a blank, misunderstand an instruction, misaddress an envelope’. (p. 9, Kafka, 2020) Indeed, paperwork demands attention to detail, but since this detail manifests in such a tedious, pedantic way, with each cell giving the impression of merely splitting hairs, one must disengage one’s higher reasoning and sink to the level of a rule-following automaton in order to honour it. Yet disengaging one’s higher reasoning leaves one less alert and thus far more likely to make errors, precisely the type of errors so rigorously prohibited by the very paperwork which dulls one into this dronelike state in the first place. As one NHM participant (Mrs. Bennett) put it: ‘To do as much paperwork as we do, I think you’d need to be on autopilot’. Essentially, paperwork dulls the mind at the very moment it demands the greatest exertions from it (in terms of raw concentration). This problem ties into the dynamics of ‘granular certainty’. Precisely because the cells, the basic units, of paperwork, are so simple, there is no longer any excuse for making errors, and paperwork thus demands a high degree of ‘certainty’ and sustained focus from those who fill it out, thereby paradoxically rendering the act of filling in these cells far less simple. ‘Granular certainty’ differs from ordinary certainty in that it takes the form of the sort of hyper-attuned, eagle-eyed, attention to detail needed to follow the specific rules set out by each individual cell or guideline: “You have to pay very close attention to details. You have to check everything and if you are administering controlled drugs you have to fill so many forms… If even one tablet goes 199 missing, one patch, the company can take you down, and the authorities can take the care home down”. (Mrs. Walker, Care home manager, Participant 4) This participant came to associate the administering of CDs so heavily with paperwork that she came to see care in terms of its smallest units – one tablet, one patch – each of which must be individually accounted for in individual cells of a MAAR chart. 7.5 Risk Management vs. Error Multiplication Earlier, I pointed out how the paperwork paradox is in part a consequence of the fact that paperwork rigidly organises and controls time while failing to take into account the time that is needed for it to be completed. In the case of error, it similarly appears that paperwork neglects a crucial part of the equation – itself. By isolating every aspect of care work which can possibly yield error and containing it within a cell, paperwork aims to prevent that error, ensuring that CHNs will be reminded of it. Yet this cell also creates the possibility for error, namely paperwork error. For each cell (e.g. a tick-box) there is a corresponding error that can be made by failing to deal with it in the appropriate way (e.g. ticking the tick-box). One participant remarked, in reference to the procedure around CD disposal: “It’s a lot of paperwork. Some of it is on paper, like the CD logs, and some of it is electronic. So there is a lot of room for error, on the CD logs and the MAAR sheets”. (Ms. Foster, Care home nurse, Participant 30) As this account shows, the paperwork around CD disposal, intended to prevent CD disposal errors, itself creates the possibility of various kinds of errors native to the paperwork itself – errors relating to CD logs and MAAR sheets. Ultimately, a CHN can either administer a medicine correctly or incorrectly; the cells of a MAAR chart, however, break the act of administration down into many sub-components (‘date’, ‘time’, ‘quantity given’, etc.), such that marking each of these incorrectly becomes its own error. An act experienced as one continuous motion is thus separated out into a great many different components, the result being that ten possible errors are bred out of one act. This is the consequence of any one decision involving AMs fanning out into ‘about 10 boxes’, as Mrs. Graham put it earlier. One unified act is sliced up into its various different aspects (e.g. the time at which it occurred, the amount of medicine that was given), as if the surface area of the act – in terms of disciplinary or legal exposure – were being ‘stretched out’, as Mrs. Graham went on to state: “The disposing logs are really important, yes, but doing it over and over in the day, well that's the challenge. We have to do it, but it's like they've taken one simple task and it gets stretched out endlessly”. (Mrs. Graham, CHN, Participant 14) 200 One CHM participant (Mrs. Walker) reported, for instance, that ‘Many of our staff get fired over the smallest errors with the controlled drugs… if a care assistant [doesn’t] sign the MAAR chart even if they still administered the medication, they’ll get sacked’. Thus, for every possible error in care work that paperwork protects against, then, it creates the possibility of at least one new error merely by virtue of existing. Paperwork is both a vessel for institutional rules and a system with its own set of rules, what might be considered a meta-level of rules. These rules may be as simple as ‘place a tick or cross in this box’ or ‘sign on this dotted line’. For every possible risk or error, an action can be taken to protect against it; at the meta-level, however, this very action can itself be the occasion for possible risk or error, and another action must thus be taken to manage its risk. The following accounts demonstrate this: “…if you are administering controlled drugs you have to fill so many forms, and so many signatures and then you have to get the countersign from somebody…” (Mrs. Walker, Care home manager, Participant 4) “We’ve got a controlled-drug cabinet, but it is locked and it’s inside another cabinet. And then there is a door that is locked to get access to that controlled-drug cabinet”. (Ms. Jones, Care home nurse, Participant 27) The act of administering CDs requires signatures; these signatures in turn require counter-signatures. Similarly, CD storage requires a cabinet, which itself requires a cabinet, which itself requires a locked door. An earlier-cited account of CD disposal showed that this process requires that pharmacists sign both a general drugs book and a separate sheet for CDs specifically, which must be initialled, dated, and duplicated, i.e. made into two copies, for the pharmacy and the care home. The duplicative logic of this paperwork thus dovetails with the duplication of error-categories; the more risks are managed, the more risk – and the more documentation associated with risk – there is to manage. Graeber (2015) describes as a basic ‘law’ of bureaucracy (specifically, in economics) the idea that: ‘any government initiative intended to reduce red tape… will have the ultimate effect of increasing the total number of regulations, the total amount of paperwork’ (p. 9). Theoretically, reducing risk should render affairs safer and more certain, thus reducing the need for regulatory paperwork. In practice, however, reducing risk paradoxically requires paperwork (e.g. risk assessment forms). This conforms with Rothstein’s observation that risk-management tends to beget further risk-management, as more and more error-categories are recognised and codified in paperwork (Rothstein, 2006). The following CHM’s account offers useful insights into the meta-level of risk-management: 201 “Some of the company guidance is just really stupid. If you look at our controlled drugs and then you look at guidance, where the drugs are expected to be stored, I don’t know how big they think my controlled drugs cabinet is. I’ve had to put another controlled drugs cabinet just to meet storage requirements which is fine, but now that cabinet needs to be behind a locked door, which also has to be behind a locked door. How many of those do I have? How many keyholders do they think I have and how many controlled drugs books do I have? And the level of competence with the staff as to what gets stored where and when and for how long and then how should it all be disposed of and what should we be doing if somebody was End of Life and now isn’t, and do the anticipatories go back or don’t they? How can we sensitively explain to somebody whose partner is dying that actually now he needs his pain relief, morphine, when there’s so much to be done”. (Mrs. Raymond, Care home manager, Participant 25) In Mrs. Raymond’s account, the risk of CDs is managed via a locked cabinet. Yet the potential risk posed by the cabinet itself must be managed via a locked door, which must in turn be managed by another locked door. These locked doors in turn require keys; the risk associated with making an error with these keys must be managed by appointing multiple designated key-holders. For every possible error, one can institute a check, but then one must institute a further check at the meta-level to check that check, and so on in an infinite regress. This issue can be usefully analysed alongside the ‘Two Generals Paradox’ (Gray, 2005). The paradox imagines that two generals each command an army stationed on either side of an enemy camp. In order to coordinate an attack on the camp, they must both reach a consensus on a time of attack. However, if General A sends General B the message ‘Attack at 6AM’, he cannot be sure this message has been received, since it may have been intercepted at the enemy camp between them. He cannot attack without knowing the message has been received, otherwise he risks attacking alone and losing the battle. To tackle this issue, General B can send a receipt of confirmation back to General A, but there is no reason why this cannot also be intercepted, meaning the receipt of confirmation also requires a receipt of confirmation, which itself requires one, and so on, leading to an infinite regress. The Two Generals paradox was first formulated by computer scientist Jim Gray, who applied it to the real-life example of a computer operating a bank account communicating remotely with a bank machine. In any situation where bilateral certainty is required due to the imperfect nature of communication posing major risks, all information storage systems, not just computers, fall prey to this basic weakness. Paperwork – a form of information storage, often stored, in fact, on computers – requires that all aspects of care be communicated to upper levels of management, but perfect certainty and perfect communication is ultimately impossible. The Two Generals Paradox can be interpreted as the paradox that information systems designed to allow for greater inter-organisational certainty and communication via the recording of information in fact create the conditions for information overload when the amount of information recorded (in paperwork) grows too large. This 202 information overload can counter-intuitively reduce certainty and hinder the transparent communication of information. To the extent that this build-up of paperwork detracts from arguably the more important form of communication – face-to-face time with residents – the Two Generals Paradox can be considered an important counterpart to the paperwork paradox. Analysing care home paperwork through the lens of information science, furthermore, throws into question the role of CHNs as not only caregivers, but, increasingly, information-givers, forced to consolidate their every action into a stable, retrievable form of information. As discussed earlier, CHNs’ information labour typically goes unacknowledged as it is not associated with the same sort of exertion as, for instance, physical labour. The following account, in which a CHM details the painstaking measures his care home adopted in order to store AMs as securely as possible, underscores how much of the work performed in the AP system is oriented towards risk-management: “It’s just to cut out that we have not got no mistakes, our stock is right, anybody can walk in at any time and they can check our stock and we’re fine, and that’s one of my biggest things that I push for, especially if you’ve needed to use agency nurses… There’s such a massive scope of something going wrong, so I’ve tried to eliminate it in every which way possible you can think of. I mean we’ve even had to put bars on the outside of the med room window to protect the morphine and stuff”. (Mr. Jeffries, Care home manager, Participant 9) Mr. Jeffries cites the practice of placing bars on a window to protect morphine as the most extreme possible example of protecting against uncertainty. In Mrs. Raymond’s recently cited account, on the other hand, the very similar practice of using a locked cabinet to protect controlled drugs is described as the quintessential example of a measure which, attempting to reduce uncertainty, paradoxically increases it. Once the uncertainty associated with CDs is managed via the measure of a controlled drugs cabinet, this cabinet produces its own uncertainties – namely, what ‘level of competence’ staff need to store medicines within it, and the question of ‘what we should be doing if somebody was End-of-Life and now isn’t, and do the anticipatories go back or don’t they?’ These questions are only as old as the CD policy itself, thus showing how new risk-management measures like the policy bring in tow a whole host of contingencies unaccounted for by the guidance which enforces them. As Mr. Jeffries’s account shows, risk-management involves trying to eliminate every possible cause of uncertainty, ‘every which way possible you can think of’. As Rothstein (2006) observes, however, this can lead to the ‘massive scope of something going wrong’ growing even larger, as new error-categories breed more error categories. Rothstein argues that, in cultures dominated by risk management, an appetite for catching errors may have the effect of continually broadening the scope for what constitutes an error: 203 ‘That increasing emphasis on scrutiny and accountability has amplified and routinized the management of institutional risks, as failures have to be recorded, potential failures have to be anticipated and new categories of failure are defined. From that perspective, ‘good governance’ gives rise to risk itself’. (p. 216, Rothstein, 2006) Risk management is the main function of paperwork within the care home space (Warmington, 2014). Yet since every risk-reducing measure (e.g. that CDs must be locked in a cupboard with a key) tends to breed a further risk-reducing measure that regulates the first measure (e.g. that this key must be assigned a designated key-holder), reducing risk can have the effect of continually increasing the appetite for more risk-reduction. This leads to what Rothstein describes as ‘the spiralling regulatory logic of societal and institutional risk’ (Power, 2004). This issue feeds into the paperwork paradox, wherein the time spent heading off one category of failure, paperwork errors, leads to CHNs being implicated in failure at the most basic possible level – the failure associated with neglecting residents. 7.6 Tick-Box Compliance vs. Duty of Candour A useful case study of end-of-life care paperwork can be found in the now discontinued Liverpool Care Pathway’s generic document. Having just analysed how paperwork treats risk, error, and uncertainty, I will now investigate how this document deals with uncertainty – something which it terms ‘variance’ and addresses in a section known as the ‘variance analysis sheet’. The review of the LCP, ‘More Care, Less Pathway’, noted the following: ‘…the design of the form allows clinicians to record plans and findings, and there is a separate section at the back for variations to the plans. Because it exists as a separate record on top of the continuation notes written by doctors and nurses, it was easy for staff to believe that both sets had to be completed. The Review panel was told that a lot of energy was initially put into completing the form, but after a while some clinicians tended to ignore it, and reverted to standard contemporaneous note keeping’. (p. 18, Neuberger et al., 2013) The variance document in question can be seen in Figure 7.2 below: Figure 7.2 Liverpool Care Pathway variance document 204 The neglect of this document can be considered a natural consequence of the design of the form. The physical separation of the variance document from the rest of the document gives it the appearance of a sort of coda, shoehorned in, clearly demonstrating how often uncertainty is treated as an afterthought in paperwork. As with the MAAR chart, explored earlier in this chapter, this document slices up and pigeonholes time – and indeed, uncertainty itself – into manageable portions, each to be time-stamped precisely. The document also attempts to force uncertainty into the teleological structure of certainty, supposing that complex situations can be broken up into a simple three-stage process of ‘variance’, ‘action taken’, and ‘outcome’, the expectation being that the issue is solved. Though the document attempts to make room in its design for uncertainty, the cells it offers in which to report uncertainty appear too small for any genuine nuance to be expressed. The LCP’s generic document as a whole can be considered an example of the failures of paperwork. The LCP review notes that ‘filling this form in could be delegated to the most junior doctor who was sometimes tasked with “completing it at 3 o’clock in the morning’, and that ‘It is a complicated form; the Review panel was shown examples only partially completed’ (p. 19, Neuberger et al., 2013). The LCP review lays blame for this behaviour at the feet of clinicians rather than the document itself, remarking that ‘the duty remains with clinicians, so far as they are able, to overcome the challenge’ presented by it. However, it is clear that the document was neglected due to its unwieldy length (21 pages in total), its complicated nature, and the severe time constraints faced by clinicians. As with many other examples of the paperwork paradox, a document was imposed upon clinicians without any regard for whether there would be any opening within their demanding schedules to complete it. The LCP Review also found that the LCP’s official ‘form had been filled in wrongly – recording discussions with relatives or carers which they denied had taken place’ (p. 19). Such behaviour can 205 be explained by the fact that failing to complete paperwork can incur sanctions whereas failing to produce accurate accounts of reality within paperwork is something less likely to be discovered. Paperwork, in a way, demands that it be told what it ‘wants to hear’. One participant remarked: “In our company, there’s judgement attached if we ask out about these policies, this avalanche of guidance. If we moan about it, then we get seen as the upsetters, the trouble- causers, and actually, we should all just pretend that we know what we’re doing and everything’s working really well all of the time”. (Mrs. Raymond, Care home manager, Participant 25) Similarly, Burton (2016) pointed out that ‘evidence of care is not found in records or action plans or, indeed, of 'statements of purpose' (all of which may be used to disguise its absence); it is found in lived experience’ (p. 12). Paperwork sets out, in neatly organised compartments, a set of conditions for the fulfilment of certain specific practices. However, in opening up a dedicated space for the reporting of a certain practice, this paperwork tacitly presumes, even takes for granted, that this practice is already being carried out. From there, it is only a matter of not going against the grain of the form, of not disabusing the form of the preconceived notions towards which it is already heavily predisposed. Forms make clear which actions are expected and which will, on the contrary, be met with professional sanctions; fear of the latter may thus naturally inspire fraudulent claims to the former. Paperwork has the effect of giving whatever version of events is reported within its clearly defined fields, boxes, and dotted lines the appearance of objective fact. This is because paperwork not only acts as the vessel for rules governing care, but comes with its own set of rules for how it is to be completed – some explicit (e.g. the rule that MAAR charts must be signed after each drug is administered), and some more implicit (such as the internal procedure for correctly filling out the form, e.g. ‘please sign here’, ‘fill these boxes with a tick or a cross’). Thus, honouring the internal logic of a form by filling it out correctly (carrying out what the ‘Cutting Red Tape’ report calls ‘Tick Box Compliance’) can have the side-effect of making it appear as if one has also honoured the actual real-world rules contained within the form. As one participant put it: “It doesn’t matter how good we are to our residents. It matters if we tick all the boxes”. (Mr. Anderson, Care home manager, Participant 35) Paperwork’s tendency to elevate its contents to the appearance of objective fact has to do with its basis in print. As Ong (1982) argues: ‘Print encourages a sense of closure, a sense that what is found in a text has been finalized, has reached a state of completion’ (p. 132). Furthermore, it ‘presents utterance and thought as uninvolved with all else, somehow self-contained, complete’ (p. 132). 206 The LCP review’s response to its findings regarding the mishandling of paperwork provides insights into the wider regulatory structure surrounding care homes: ‘The reports of incomplete and wrongly completed forms that the Review received are of grave concern, reminding the Review panel of a key finding of the Mid Staffordshire Public Inquiry – insufficient openness, transparency and candour. The Review panel strongly supports Robert Francis and the CQC’s call for a duty of candour, and recommends that clinicians be reminded by their registration bodies that the deliberate falsification of any document or clinical record, in order to deflect future criticism of a failure of care, is contrary to GMC and NMC guidelines and therefore a disciplinary matter’. – More care, less pathway: a review of the Liverpool Care Pathway (p. 19, Neuberger et al., 2013) Here the LCP Review invokes the Mid Staffordshire Public Inquiry’s ‘call for a duty of candour’, yet misses a key piece of context for it. The Inquiry found evidence of ‘a culture of fear in which staff did not feel able to report concerns’. Their ‘duty of candour’ is intended as a corrective for this ‘culture of fear’, as indicated by the juxtaposition of the two: ‘1.180 The common culture of caring requires a displacement of a culture of fear with a culture of openness, honesty and transparency, where the only fear is the failure to uphold the fundamental standards and the caring culture. 1.181 A statutory obligation should be imposed: On healthcare providers, registered medical and nursing practitioners to observe the duty of candour…’ – Report of the Mid Staffordshire NHS Foundation Trust Public Inquiry (Francis, 2013) Recommendation 1.180 above refers to the fact that the fear associated with breaking the strict rules of paperwork – which, when broken, can spell professional disaster – often seems to take focus away from actually caring for residents. The LCP Review emphasises the importance of the ‘duty of candour’ by lacing it with the coercive threat of disciplinary action, a subtle move towards the sort of rule by fear typical of the many regulatory bodies who impose paperwork on care homes. Yet in doing so, they miss the entire original point of this ‘duty of candour’, which was to combat a culture of fear. The culture of fear described by the authors of the report - in particular, fear of disciplinary action imposed by the regulatory bodies who issue paperwork - has a clear parallel in my own dataset. One of the most significant findings of my study was that the threat of liability paralysed CHNs to the point where they deliberately under-administered AMs, even in cases of clear clinical need: 207 “I think a lot of people here think it's safer for them for the resident to die in pain than make a mistake administering controlled drugs. They're dangerous drugs, to the resident and to us but in different ways”. (Mrs. Walker, Care home manager, Participant 4) “….in a care home, all the risk is on your head… if I were to make a decision and it was the wrong one, it would be on me… Which is why I don’t think we give things unnecessarily in care homes. If anything nurses probably hold back a bit...” (Mrs. Bennet, Nursing home manager, Participant 11) “Whereas there’s no blame culture in the NHS, it definitely exists in care homes”. (Mrs. Raymond, Care home manager, Participant 25) The fear of being punished for making even minute errors was associated not only with administration itself, but the paperwork governing administration, as per an account cited earlier: “Many of our staff get fired over the smallest errors with the controlled drugs. I have to raise the safeguarding flag, it’s part of my job. The lower the person who makes the mistake, the harsher their penalty. Like if a care assistant didn’t sign the MAAR chart even if they still administered the medication, they’ll get sacked”. (Mrs. Walker, Care home manager, Participant 4) One participant reported a rather dramatic incident in which a paperwork error led to the involvement of the police: “They [two staff nurses] couldn’t account for some of the controlled drugs, and they told me two days later. I said, ‘Okay, well have you called the police? ... It is the law, you have to do this, there’s no argument about it…’ Actually it turned out to be a recording error that went back so many months… People weren’t paying attention, somebody in the stock check clearly hadn’t checked the stock and had miscalculated it”. (Mrs. Perry, Care home nurse, Participant 29) This account demonstrates that part of what promotes a culture of fear within the care home space is the extent to which paperwork is bound up with law. The presence of the law is deeply felt within care homes; one participant (Dr. Cameron) revealed that, ‘for our controlled drugs we have an ex- policeman who comes in here and does it with us’. CHNs are held accountable to CQC regulations, and the CQC in turn is held accountable to certain laws, for example the Health and Social Care Act 2008 and the Care Quality Commission (Registration) Regulations 2009. The content of these laws filters down to care homes through guidance documentation and other such paperwork. Care home professionals in my dataset often reported bearing the brunt of legal liability: 208 “If even one tablet goes missing, one patch, the company can take you down, and the authorities can take the care home down. So it’s a lot of responsibility that gets pawned on us”. (Mrs. Walker, Care home manager, Participant 4) “…if I were to make a decision and it was the wrong one, it would be on me. The company would say they support you, but in realistic terms, they wouldn’t. They would say it was our decision to administer that medication. So we don't have a lot of backing. Maybe none, not really”. (Mrs. Bennet, Nursing home manager, Participant 11) By shifting liability for errors onto front line staff such as administering nurses, care home organisations can avoid facing legal repercussions themselves. As one participant (Mrs. Raymond) explained, ‘As an organisation we have to prove that we know what we’re doing in order to stay registered, and the CQC has to regulate us appropriately in order to be seen as a legitimate commissioning body’. This participant attributed the high rate of staff attrition in the modern care landscape with the culture of fear created by disproportionate legal liability: “Fear has eclipsed everything else, fear of medication errors and losing jobs, and we pander to those fears more than we pander to our people. I think that’s why swathes of people are leaving the industry. I think they’re finding it difficult to reconcile the two, you know, the fear underlying everything we do and their real desire to do good by our residents. We’re burnt out and alone”. (Mrs. Raymond, Care home manager, Participant 25) Mrs. Raymond’s phrase ‘reconcile the two’ exhibits the hallmarks of a paradox, i.e. a situation in which ‘contradictory yet interrelated elements that exist simultaneously’ (Smith and Lewis, 2011). The two competing impulses which Mrs. Raymond detects in care work are, firstly, the fear of legal discipline and punishment, and, secondly, the desire to ‘do good’, i.e. the moral dimension of care. The sociologist Jurgen Habermas has identified a similar tension within law itself, a duality between coercion (the threat of punishment) and moral principles. This duality creates a paradox – that the reason for which law can justify enforcing rules by coercion is, counter-intuitively, because those rules are so reasonable that they seem to ‘deserve their addressees’ uncoerced… recognition from the moral point of view’ (my italics) (p. 29, Habermas, 1992). It is only when society as a whole believes strongly enough in a moral principle that it becomes deserving of universal recognition and thus becomes formalised in law; however, since the way in which the law promotes that moral principle is by enforcing it (i.e. by means of coercion, by threatening rule-breaking with punishment), rather than by convincing individuals to believe in it, this can bring about a state of affairs in which the principle is more often honoured out of fear of punishment than out of authentic 209 moral conviction. Paradoxically, then, what appears to be the summum bonum for any given principle – its formalisation in law – can actually damage its overall credibility within hearts and minds. For instance, it is only because it seems so paramount to uphold a moral principle (the principle that the safety of care home residents should be protected at all costs) that the laws and regulations around controlled drugs administration are so stringent, and yet my data shows that this very stringency paralyses nurses into under-administration and non-administration out of fear of breaking these regulations and being punished, thus paradoxically rendering residents less safe and defeating the point of the original principle. Ultimately, my data shows that the very CHNs who most wholeheartedly believe in resident safety as a moral principle are often made to associate it more with the pedantic and draconian rules which enforce it than its own moral content. As Mrs. Raymond suggests, CHNs often ‘pander’ to ‘fear of medication errors and losing jobs’ more than they ‘pander to our people’. This recalls the ‘carrot and stick’ approach taken by the LCP Review when discouraging the mishandling of paperwork. The LCP Review first appealed to a general moral principle – the ‘duty of candour’ – but then, finding this appeal insufficient on its own, cited GMC and NMC guidelines in order to back up the principle with the coercive threat of disciplinary action. Here, the coercive threat of punishment paradoxically conflicted with the very underlying principle which supposedly justified it – the ‘duty of candour’, which was specifically designed to combat a culture of fear of punishment. This points to the paradoxical relationship between coercion and moral principle in law and regulatory documentation in general. Moral principles are arguably fundamental to nursing. In discussing the concept of vocation in nursing, White refers to Blum’s (1993) ideas of morality, where individuals feel a ‘moral-pull’ towards the profession (p. 282, White, 2002). Studies by Eley et al. (2012) and Genders and Brown (2014) further support the notion that nurses are drawn to their profession due to a moral imperative for caring. One CHN cited low pay as another indicator that joining the nursing profession tends to follow from an authentic drive towards caring: “…you can earn the same amount of money working in a supermarket… But, at the same time, I think if it paid a lot more you might get people just doing it because it’s good money… When you consider how much you get paid as a nurse for the amount of responsibility that you have, it ensures that only people who really do want to be a nurse and really do want to care become nurses”. (Mrs. Graham, Care home nurse, Participant 14) Blum (1993) differentiates between industrial vocation, where individuals perform their work dispassionately and without personal connection, and social vocation, where professionals are primarily driven by a moral imperative to provide compassion. Whereas caring directly for a resident can be felt as a moral act, the process of filling out paperwork has no moral character, and can even 210 take away from the time needed for one-on-one care work. It is arguably because the bureaucratisation of care work via paperwork puts care work at war with its own nature, its own moral core, that paperwork fundamentally denatures care work. Paperwork thus not only encumbers but poses a fundamental existential threat to care work, causing nurses to resign en masse. The result of this is the remarkably high staff attrition rates which have afflicted care in recent times, with approximately 30% of CHCs leaving their positions annually (CQC, 2020). Wilson (2017) refers to this issue as the ‘elephant in the room’. Indeed, in an earlier-cited quotation, Mrs. Raymond cited the inability to reconcile ‘fear’ with moral principles, i.e. the ‘real desire to do good by our residents’ as the main reason ‘why swathes of people are leaving the industry’. There were several reports of resignation throughout my dataset, particularly in relation to the pandemic context: “Over the course of this year, I’ve lost thirty percent of my staff… When the first five nurses put in their resignation, that left the rest of us with double the workload because we can’t afford agency nurses. So that burned out the rest of the team and another six resigned”. (Mrs. Gardner, Care home nurse, Participant 22) Another account offered further insight into this issue, presenting a snapshot of a CHN suspended on the verge of professional ‘burnout’: “Sometimes I feel like, I can’t keep doing this, jumping pharmacy to pharmacy for medications like a headless chicken while the resident is dying in so much pain. But in the middle of that, I just remember, I am here for a reason and this is it. I’m a spiritual person and I take it as a privilege to be able to be in that position and give them a peaceful death, even if it’s at an expense to me”. (Mrs. Jiro, Care home nurse, Participant 24) This account suggests that moral obligation may act as a kind of failsafe within care: no matter how much emotional burden a CHN endures, it can be rationalised as a necessary moral sacrifice. When, on the other hand, participants reported spending their time completing paperwork, which has no moral character and poses no material benefit to residents, it appeared more difficult to rationalise: “For disposal, we have to use a returns book to log everything. We take off the labels of any name, details, address, destroy them… We write a list of what’s in there, sign it, then the driver collects it, signs it, and then takes it back to the pharmacy. They sign it... It’s a lot of paperwork to make a lot of waste so the whole thing is just painful”. (Ms. Foster, Care home nurse, Participant 30) 211 This is another example of the paperwork paradox: here the burden of regulatory bureaucracy (‘a lot of paperwork’), whose purpose is supposedly to keep residents safe, both likely took away from time spent with those residents, and was only carried out in the first place as part of the CD disposal process, another purely regulatory process which is itself often felt as contrary to residents’ interests insofar as it involves ‘wasting’ CDs which could have been repurposed by other residents, simply out of compliance to laws and regulations. The perfunctory nature of menial bureaucratic and administrative chores of this sort likely distracts and detracts from the emotional resonance of the event which triggers it – the death of a resident. Another participant described the CD disposal that follows a resident’s death in notably emotive terms: “If you really love the job, and I love my job, honestly it’s kind of a very sad moment because it’s like another spirit has gone, and you kind of just tick in the book and, you know, after 7 days, destroy”. (Mrs. Whitfield, Nursing home manager, Participant 21) This participant’s use of the word ‘spirit’ echoes the earlier-cited participant who attributed her desire to work in end-of-life care to being ‘a spiritual person’. Both participants appeared to view being involved in a human being’s death as a way of being part of something bigger than themselves, an experience bordering on metaphysical. The stark contrast between the emotionally and metaphysically charged phrase, ‘another spirit has gone’ and the mundanity of another ‘tick in the book’ – a bureaucratic chore which cannot possibly prove equal to the gravity of the event to which it responds – gives some insight into the extent to which bureaucratic paperwork proves incompatible with the moral, spiritual core of care home work. Ultimately, CHNs must grapple with two opposites – the moral and spiritual weight of working with death and the workaday tedium of complying with the regulations around death – which paradoxically co-exist in the same profession. 7.7 Guidelines vs. Orders Aside from forms and logs (e.g. disposal logs), another form of paperwork reported to be time- consuming across my dataset was guidance documentation: “We get the weekly updates to controlled drugs guidance over email now, but actually who has the time to read those? … I think it could be less quantity and more quality?” (Mrs. Jiro, Care home nurse, Participant 24) The overwhelming quantity of guidance was pointed out by another participant (cited earlier), who described it as an ‘avalanche’: 212 “In our company, there’s judgement attached if we ask out about these policies, this avalanche of guidance. If we moan about it, then we get seen as the upsetters, the trouble- causers… (Mrs. Raymond, Care home manager, Participant 25) As this account shows, guidance often proves highly ambiguous, and therefore requires clarification. Mrs. Raymond was not the only participant in my dataset who decried unclear guidance, or who reported having unsuccessfully attempted to seek clarification on unclear guidance from the higher levels of her company: “There’s quite a few policies that don’t really make any sense. You just live with those and pick your battles… I have questioned them in the past but you don’t really get an answer, and being part of a very large, global company, everything has to go up so many layers to the top for decisions to be made”. (Mrs. Bennet, Nursing home manager, Participant 11) “I’ve never quite understood that policy. The reuse one and all the ones about controlled drugs disposal. Especially the syringe driver disposal. They make no sense whatsoever. It’s just one of those policies you have to follow but it doesn’t make any sense”. (Ms. Wynne, Care home nurse, Participant 12) Mrs. Raymond recounted an experience in which, faced with ambiguity around a guideline (concerning GPs visiting care home), she felt unable to clarify the issue not only with regulators, but also with other managers experiencing the same issue in a ‘manager’s forum’ held on Zoom: “Until people admit to those failings in the systems and the processes and the policies and people can have open and honest conversations without being penalised, I don’t think the support will really work. Because, you know, you can only talk to a counsellor or a support person about so much before again it becomes a, ‘ooh, if that happened in care homes, that’s really bad…’ I was talking to them [other managers] about the unwillingness of the GP to come in and do a ward round… like is this some sort of guideline or is this just my GP being difficult? And I expected other managers to say either, ‘Yes, it happened to us and this is how we got over it’, or ‘No, that doesn’t happen, she’s coming in for us and for me’ – to have that kind of open conversation. In the round table format on Zoom, nobody spoke up at all, but individually after that Zoom call, three people contacted me and said ‘Yeah it’s happening to me too, but we’re afraid to speak up on that’”. (Mrs. Raymond, Care home manager, Participant 25) In practice, because CHNs are unable to clarify guidelines with regulators or colleagues, they tend to be forced to undertake the process of interpreting and clarifying guidance alone: 213 “We get a piece of guidance or we have a hard and fast rule thrown at us, and the conversation will be like, ‘What? Who said this? They want that of us? That’s not possible.’ We have to spend more time working our way round it, interpreting it, clarifying it and explaining it. More time than we spend actually implementing it, if it’s implemented at all. Because before you can get to the implementation, you’ve got to figure out how the hell you implement something that’s unworkable, and our medicines reuse policy was definitely one of those”. (Mrs. Raymond, Care home manager, Participant 25) Guidance documentation is somewhat distinct from the kind of paperwork I have been analysing so far – i.e. forms and charts (such as the MAAR chart). Whereas forms are actively filled in, guidance documentation contains rules to be passively followed. In practice, however, guidance documentation appears to require a great deal of input from CHNs to be made workable – input, that is, in the sense of interpretation. One participant indeed described the process of interpreting guidance (specifically, the reuse of medicines guidance) as a form of time-consuming paperwork: “Actually, it’s going to take companies such a long time to take in, digest and then bring that into their own policy, that, actually, it’s not really easy. Especially at a time when things are going wrong and things are busy, people unwell. You’ve not got time for extra paperwork, extra pressure, looking at things that you’ve never done before”. (Mrs. Greensborough, Care home manager, Participant 28) Guidance documentation can indeed be considered a form of paperwork, and one which contributes to the paperwork paradox. As Mrs. Raymond reveals, guidance documentation similarly requires ‘more time’ to interpret, i.e. more time spent engaging with guidance paperwork, than CHNs ‘spend actually implementing it’, i.e. spend enacting the policy one-on-one with the resident. Some participants reported interpreting laws, such as the COVID-19 Reuse of Medicines legislation, by means of ‘internal policy’ documents. One participant, for instance, remarked the following: “The Government said you could [reuse medicines], but didn’t actually publish any guidance to go along with it till a few weeks or a few months later, and the guidance that the Government did publish was very cloudy. As most things with care homes and Covid are, it was very fuzzy. We made an internal policy, and some guidance and some documentation to go with it because the problem with it was keeping track of it all. Because the law says each drug has to be prescribed for that individual, we’re not allowed to hold stock of controlled drugs unless it’s prescribed for somebody. Transferring ownership actually gets really messy. So it wasn’t just policy, there was documentation… a lot of learning through it”. (Mrs. Perry, Care home nurse, Participant 29) 214 This care home had to deal with three kinds of paperwork: firstly, the frequently changing guidance documentation around the reuse of medicines legislation; secondly, an internal policy document they created to interpret this legislation; and, thirdly, documentation which was used to officially transfer ownership of medicines. All of this goes to show how law and guidance contributes to the paperwork burden. This participant also reveals that, in care homes, ‘the law’ is often ‘messy’ or ‘fuzzy’ and must be worked out or ‘learned through’ via interpretation. A school of thought in philosophy of law known as interpretivism argues that interpretation is inherent to law. Ronald Dworkin, for instance, suggests that ‘law is an interpretive concept’ (p. 87), that law is inherently indeterminate (i.e. ‘fuzzy’) and that one cannot therefore not apply a law at all without interpreting it in some way (Dworkin, 1987). Many laws were found in my data to be extremely indeterminate. Aside from the ‘messy’ reuse guidance mentioned in the above account, the worst offender consistently appeared to be syringe driver policy: “There was an issue where I was involved in removing syringe drivers from people and then there was a lot of confusing, mixed messages about the disposal of the contents of syringe drivers… there were people saying, ‘Oh, well, you should really have disposed of that by yourself.’ I’d have been in breach of the professional code of conduct if I’d have decided, independently, to dispose of a significant amount of Morphine without it being witnessed and without it being subject to due process. I think there is big anxiety about that because you could lose everything making one mistake with the anticipatories”. (Mr. Smith, Nursing home manager, previously a care home nurse, Participant 1) “My company’s policy says we shouldn’t put more than three meds in a single syringe in a syringe driver, whereas the local policy has always been that you can put up to four but you should discuss it with them first. So you’ve got policies that don’t agree with one another which confuses everybody, and nobody wants to get into trouble, so some of them just don’t deal with it”. (Ms. Wynne, Care home nurse, Participant 12) In a situation where one rule suggests a maximum of four medications in a syringe driver and another suggests a maximum of three, two conflicting options are both presented as equally desirable, placing the CHN navigating this situation in a paradoxical double bind. Within his theory of law, Habermas (1992) attempts to normalise these kinds of ‘collisions between rules’ arguing that they ‘are quite probable—and yet do not betray a deeper-lying incoherence in the legal system itself’ (p. 217). He acknowledges that laws are inherently indeterminate and ‘fuzzy’ and thus by nature cannot be enacted without being first interpreted; he draws on the interpretivist school’s idea that interpretation is a normal and healthy part of law. Because laws are indeterminate, it is common for contradictions or ‘collisions between rules’ to occur, yet this is acceptable if, as Klaus Günther proposes, ‘the indeterminacy of legal norms’ is ‘reflected in the degrees of freedom enjoyed by a mobile set of principles’ (p. 219). Indeed, it would be easier to resolve the contradictions between 215 rules encountered by the CHNs in the above accounts if they were given assurance that they were legally and professionally justified in exercising their professional judgment, their own interpretation of the law, and choosing between them. Unfortunately, the opposite appears to be the case: CHNs are, as the accounts above suggest, always at risk of ‘getting in trouble’ or ‘los[ing] everything making one mistake with the anticipatories’. As another participant made clear, guidelines are often associated with the threat of punishment: “It’s very difficult for people to see guidelines as guidelines, rather than what most people see them as: orders. Like, ‘This is what you have to do and if you don’t, you’ll be in trouble’”. (Dr. Cameron, General Practitioner, Participant 6) Guidelines therefore paradoxically straddle two contradictory positions at once – they are seen simultaneously as pieces of friendly advice and as draconian edicts. Guidelines encourage their addressees to take them as guidelines – as mere helpful suggestions – thereby encouraging freedom of interpretation, while at the same time harshly punishing any deviation from them. When guidelines contradict each other, furthermore, this creates an even more paradoxical situation. Following one rule will break another rule, making it impossible to avoid making some kind of error. Yet, in spite of this, CHNs are still likely to be punished for errors, meaning that they are exposing themselves to liability simply by attempting to interpret or implement the guidelines at all. One CHN participant (Mrs. Barnes) located the contradiction inherent in guidelines which are ambiguously written yet also strictly enforced: “A lot of it is not clear enough … When it comes to controlled drugs, I do think that the law is strict and necessary. But how to do it and why it needs doing has to be explained and nobody here will do that for you”. (Mrs. Barnes, Care home nurse, Participant 23) Another participant, Dr. Powers, appears to have devised a solution to this problem. Dr. Powers’ proposal involves making use of internal policy documents to deal with confusing or paradoxical clinical situations in which two contradictory choices seem equally desirable, without exposing CHNs to liability, the risk of ‘getting in trouble’, in doing so: “Vicarious liability means that the liability transfers from the individual to the organisation… if I ask you to mix this drug and this drug, you know it’s all in there [the book] so that’s covered legally. But when you’re confronted with an ampule that might say ‘for IV use only’ and I’m saying, ‘No, no it’s fine, you can give that sub-cut’, and you look down on the sub-cut page and see, oh yeah, I can give that sub-cut. As well as giving you some legal cover, it’s also trying to make you feel a bit more confident…” (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) 216 Since CHNs are not afforded freedom of interpretation in navigating contradictory situations, Dr. Powers reasoned that an internal policy document can carry out the work of interpretation for them, and thus absorb the liability associated with that interpretation. In general, Dr. Powers provided particularly rich material for my analysis due to his engagement with the legal dimension of care. A notable example is his experience with the Patient Group Direction (PGD), a legal document which provides written instructions for the sale, supply and administration of medicines to groups of patients who are not required to have a prescription in place. Dr. Powers developed a PGD which licensed end-of-life mediation to be administered by experienced licensed paramedics, explaining his rationale thus: “I don’t think there’s a legal difference, I think it’s more that there’s a precedent to paramedics carrying drugs… They carry a whole load of drugs anyway, so you know, we were only asking if they can carry a couple more. For our nurses it was slightly different… In essence what we said was, if they’re looking after somebody at home, we’re considering that patient to be in our virtual ward, so we can authorise that nurse to administer our own drugs to that patient… It just happens to be that the nurse is driving to the patient instead of walking down the corridor. So, and that’s perfectly legal, there’s a clear legal precedent for that. In March last year, we just said well, actually that applies to controlled drugs too, it’s just we’ve never been brave enough to do it”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) Dr. Powers embodied what might be characterised as a broadly interpretivist approach to the law – engaging his freedom of interpretation in order to sound out the reasonable scope of pre-existing laws, while taking care never to veer far from the spirit of those laws, the moral principle underlying them. He demonstrates sensitivity to the importance of ‘legal precedent’ – a concept central to case law, a form of justice in which past court decisions (that is, previous instances of judges interpreting or applying the law in a particular way) inform present court decisions. Dr. Powers moves piecemeal by a process of extrapolation, arguing that because paramedics are already licensed to carry drugs, it is ‘logical’ that they should be able to expand the range of drugs carried; that, if a patient’s home could be considered a ‘virtual ward’ in light of the COVID-19 pandemic, then controlled drugs, usually confined to the care home space, could be acceptable to administer. Dr. Powers also at one point appealed to an actual legal precedent: “Have you come across the JRCALC Schedule 17 Exemptions? It’s really a murky area of absurdity but it’s a curious set of law, I think. That schedule came into being after… a doctor gave a mountain rescue team Morphine, years ago, and he broke the law, but what he was doing was clearly sensible, though he was prosecuted. I think that’s when people said, this is crazy, we need to enable people in certain situations to use these drugs. And so they 217 therefore exempted mountain rescue teams and a whole set of other people, and that’s how they dealt with paramedics using these medicines in the early days”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) Dr. Powers justified authorising paramedics to administer morphine in care homes by referring to the ‘precedent’ that they are already authorised to do the same; yet, paramedics in turn appear to have acquired these powers only as a result of the precedent that mountain rescue teams were invested with those powers, something which itself was illegal at one time. Dr. Powers points to the fact that interpretation – and specifically, the process by which one’s freedom of interpretation is used to stretch the parameters of a law to accommodate a slightly different situation than it immediately appears to account for – is in fact fundamental to law. Laws gradually evolve by evoking precedents, judges’ interpretations of the law, to cover new ground. A JCALC guidance document which interprets the ‘Schedule 17 Exemptions’ mentioned by Dr. Powers above, states that ‘[a]s pre-hospital models of care evolve, the legal system will need to adapt, and will always follow practice. While the law is well-defined, it will continue to require interpretation and amendment to enable good patient access to medicines.’ The paradox in law Dr. Powers discovered can be stated in the following way: one must break the law (in some reasonable way) in order to pave the way for actions which become the law. A participant based in the same area as Dr. Powers reported an advantage of the PGD being that it removed the need for time-consuming paperwork to be completed during out-of-hours shortages: “Nobody wants to have medicine waste, but what I also don’t want to be doing is rattling around at 3am on a Saturday morning trying desperately to find a 24-hour pharmacy that’s in the area that isn’t the hospital pharmacy where I’m going to have to sit and wait for two hours to get all the relevant paperwork in place before I can give somebody something that’s going to make their departure from this life a bit easier. I think the PGD is a good start to a solution. It certainly gives me hope and I think it’s a better option than what we were doing”. (Mrs. Raymond, Care home manager, Participant 25) Whereas the law around controlled drugs often adds to the paperwork burden by necessitating time- consuming interpretation and internal policy creation, the PGD appeared to offer a refreshing contrast to my participants. It appears to be a form of law designed to account for the possibility of bureaucratic delays, including paperwork. Dr. Powers was careful, when crafting his internal procedure for evaluating and reflecting on the usefulness of the PGD, to remain sensitive to the constraints placed by paperwork on time: “We literally give the paramedics and nurses just a single sheet of paper: ‘What went well, what went badly, and have you got any suggestions?’ We wanted it as brief as possible, 218 because obviously when they’re using it, they’re really in the middle of it and they can’t be doing a great big document. And then I go back and do a note trawl”. (Dr. Powers, Senior Consultant in Palliative Care, Participant 3) Though this ‘single sheet of paper’ amounted, as Dr. Powers admitted, to a ‘crude’ and underdeveloped procedure, it seems likely to have been designed to correct for the mistakes of documents like the earlier-discussed LCP generic document, which, at twenty-one pages, proved very difficult to accommodate within doctors’ schedules. 7.8 Chapter summary In this chapter, I attempted to ground my analysis of the paperwork paradox within the material realities of the files and forms that constitute said paperwork. To grapple with the temporal dimension of the paradox – the fact that the time taken to do paperwork takes away from the time needed for care – I interrogated some underlying assumptions commonly held about paperwork, namely that it is a ‘neutral’ form of information storage which does not enter into the felt dynamics of duration. Exploring the morphological structure of files and forms, I broke them down to their smallest unit – the cell. I discussed both the duplicative properties of cells and the duplicative properties of the categories of error which paperwork inaugurates. I then moved onto an analysis of the LCP’s generic document. Finally, I discussed guidance and internal policy documents as forms of paperwork which often necessitate the time-consuming process of interpreting laws around controlled drugs, and explored how the liability associated with these laws promotes a culture of fear within care homes. 7.9 Discussion o Contributions to literature and theory In this chapter, the emphasis on regulating error, uncertainty, and risk seen in much AP paperwork led me to view the paperwork paradox partly through the lens of risk management. Ulricht Beck’s (1992) idea of a ‘risk society’ has been problematised by Rothstein (2006), who argues that a great deal of what counts as risk in modern society is the effect, not of real-world risk, but risks produced by a ‘feedback loop, in which attempts to manage institutional risks changes the way that societal risks are managed, and which, in turn, leads to the identification of new institutional risks’ (p. 105). This ties into meta-risk, i.e. ‘the risks of risk management’ (Power, 2004), and the extent to which ‘regulation is itself subject to “regulation”’ (p. 106, Rothstein, 2006). I detected this process - ‘the spiralling regulatory logic of societal and institutional risk’ (Power, 2004) – in the multiplication of error-categories within AP paperwork. 219 By applying Robertson’s (2021) concept of ‘granular certainty’, furthermore, I was able to better capture the morphological structure of risk regulation by isolating its smallest unit – the boxes or cells of an average form. I focused on the paradoxical way that certainty at the ‘granular’ level of the cell allows for uncertainty on a large scale at the level of the entire form. My analysis of this issue formed part of my ambition to contribute a greater understanding of issues of information storage to the current literature on bureaucracy in care homes. Concepts like granular certainty and Ong’s (1982) analysis of print media allowed me to clarify not just that there is too much paperwork in care homes, but what precise mechanisms of repetition allow this accumulation to build up. As I showed, these mechanisms are both technological (print, which is duplicative by nature) and structural (granular ‘cells’, which are also by nature repeatable). In paradox theory, Smith and Lewis (2011) define the concept of paradox as ‘contradictory but interrelated elements that exist simultaneously and persist over time’ (p. 382). I have attempted to contribute to this idea by considering to what extent time itself can be implicated in paradoxes. I explored how paperwork regulates time while also itself taking time to complete, time which it paradoxically does not account for. Another dimension of paradox studied in this chapter overlaps with a study entitled ‘Speak! Paradoxical Effects of a Managerial Culture of “Speaking Up”’, which focused on a multinational health care company that encouraged its employees to ‘speak’, yet paradoxically treated them with hostility when they chose to take them up on this offer in order to offer constructive criticism (Cunha et al., 2018). My data similarly showed that CHMs who questioned the regulations set and enacted by paperwork were treated with hostility by top-level company regulators. Both of these examples indicate that some regulators are more concerned with creating the appearance of managing risk, possibly to avoid liability, than meaningfully managing it. It may be because the mere act of imposing paperwork on a care home can produce the appearance of managing resident risk and ensuring resident safety that regulators impose more and more paperwork on care homes, even though this very abundance of paperwork paradoxically takes away the time needed to ensure resident safety. Finally, this chapter engaged in an ongoing discussion about the LCP. Bowers (2022b) has observed that the LCP review did not recommend changes in AP specifically and that ‘Anticipatory prescribing effectively became decoupled from the ‘toxic brand’ of the Liverpool Care Pathway in practice and subsequent policy’ (p. 51). This may contribute to a tendency to treat the ‘brand’ of the LCP as the problem rather than the overly standardised, ‘tick box’ approach to care it was demonstrated by the Review to represent (Neuberger et al., 2013). I went directly to the source of the problem, analysing the LCP’s generic document itself. I questioned the way that the LCP Review coupled the coercive threat of disciplinary action with a ‘duty of candour’ which was itself rooted in an attempt to correct a culture of fear caused by the threat of disciplinary action. This dichotomy between coercion and (moral) duty was analysed alongside a paradox of law observed 220 by Habermas (1992), namely that laws are legitimised by the soundness of the moral principles underlying them, yet laws are also precisely what make it possible for individuals to simply outwardly conform to those principles in order to avoid being punished by the law, rather out of moral conviction. Habermas terms this phenomenon ‘the paradox of rules of action that, without regard for their moral worthiness, only require a behavior that objectively corresponds to the norm’ (p. 29), and it does indeed appear to exhibit the structure of a paradox, a structure made up of two irreconcilable yet co-existing elements. Habermas uses the language of paradox when he states that, in a given law, ‘two things are guaranteed at the same time: both the legality of behavior, in the sense of an average norm compliance that, if necessary, is enforced by sanctions; and the legitimacy of the rule itself’ (p. 31). In tracing back the phenomenon of AP paperwork to its source – the law, and the regulatory bodies who enforce it – this chapter found that this paradox was felt deeply by many CHN participants, who believed deeply in the principle of resident wellbeing, and yet were effectively steered away from this principle by the very regulations which attempted to promote them via law. o Recommendations for policy and practice 1. Participatory Governance Multiple participants cited in this chapter suggested that those who produce guidelines are out of touch with the lived realities on the ‘front lines’ of care. The current prevailing 'evidence-based practice' policy paradigm primarily relies on research conducted by ‘experts’ (Melnyk & Overholt, 2022). In contrast, involving care home professionals directly in policymaking may prove a less roundabout approach than extracting clinical research from the care home system and hoping that it eventually finds its way back there. A 2018 WHO report emphasises the importance of practitioner involvement in policy development and implementation (WHO, 2018). A 2022 WHO report recommended the adoption of a participatory governance model within health and social care systems, stating that ‘a resolute emphasis must be placed on consulting not only those with a more evident stake in health system operations, but also those who will bear the brunt of any operational decisions’ (WHO, 2022). Hammersley (2005) and Smith (2013) similarly emphasise the importance of linking policy and policymaking to the knowledge and experience of those who do the work of policy and practice. 2. Centralised Guideline Interpretation Resource This chapter showed that CHNs exert a great deal of effort in interpreting (often unclear) laws and guidelines across the care sector, and furthermore risk exposing themselves to liability in doing so. Dr. Powers’ interview data suggested that CHNs are empowered to administer medicines more confidently when they can fall back on a centralised resource (an internal policy document) that absorbs the liability associated with interpreting external guidance for them. It may be beneficial, therefore, to devise some way of collectively interpreting guidance. This could be done by means of 221 a centralised online resource which would act as a public forum on which CHNs and CHMs could share their internal policies (which are, by their very nature, interpretations of external policies). Initially, this online resource could fulfil the simple and much needed function of placing all relevant guidance in one place; moreover, by adding comments sections to each document, it would be possible to discuss how, in specific situations, they interpreted and implemented specific guidelines. An unconfident CHN struggling to interpret confusing, unclear, or conflicting guidelines would be able to navigate such a challenge with more confidence if they could refer to testimony provided by other CHNs of how they interpreted the same guidelines. Mrs. Raymond observed in this chapter that her manager counterparts in other care homes generally felt too fearful of ‘speaking up’, thus risking liability, to share their experiences with ambiguous guidance. Yet conversations between CHNs covering how a rule was actually interpreted could prove extremely valuable. Dr. Powers described his internal policy as a form of vicarious liability, a mechanism which serves to spread liability throughout his organisation so that it does not disproportionately impact any one individual. A centralised online resource, if used widely enough and perhaps partially anonymised, could serve to evenly distribute liability throughout the entire care sector. Using the proposed resource, CHNs could share accounts which detail how a certain rule was interpreted in a specific situation. These accounts would be analogous with the model answers for exam papers often used in schools. There is precedent for the use of accounts in the ‘Case Studies’ section of the GSF Care Homes Training Programme Good Practice Guide (pp. 24-25, 2014). These case studies deal with the nuances of specific patients’ conditions and show how GSF guidelines are implemented in practice. The rich accounts provided by my own dataset – especially those which detail the finely-modulated administration techniques explored in my last chapter – are a testament for the potential of accounts provided by CHNs. Thinking reflexively about the nature of my own data – about the process of collecting interview accounts – is a cornerstone of my RTA method. Upon such reflection, I have concluded that it may be beneficial for CHNs to inform each other in the same detail that they have informed me. Unlike guidelines, which inspire fear of error when taken as orders, CHNs using the proposed resource could either take or leave advice contained in accounts provided by other CHNs detailing how they interpreted a certain rule without feeling held to those accounts. These accounts could play the role that legal precedent and case law play in the legal system. The fact that legal cases frequently call on interpretations of law made in past legal rulings constitute the foundation of legal interpretivism’s claim that interpretation is fundamental to law. CHN accounts could therefore function as the interpretive arm of the care sector, an area of healthcare in which laws and regulations play a highly prominent role. A good example of an interpretation-friendly legal framework is the United States’ government’s Federal Register, which provides the public the opportunity to leave online comments on proposed rules and regulations, thereby giving the public a chance to provide their interpretation of a rule and, in doing so, contribute towards shaping it. As 222 discussed earlier, the centralised resource being proposed could contain a comment section which would allow care home professionals to comment on and thus share their interpretation of guidelines with other care home professionals. 7.10 Conclusion Having interrogated the paradoxical relationship between the bureaucratic and regulatory logics of paperwork and the realities of care, I will move onto my conclusion, beginning with the story of the thesis as a whole. 223 Part four: Reflections & Conclusion Chapter 8: Conclusion In this chapter, I present the story of the thesis, recommendations for policy and practice, my reflexivity statement, a quality appraisal (encompassing meaningful consideration of strengths and limitations of the empirical research), future research directions, and some final remarks on this body of work. 8.1 Story of the thesis This thesis set out to expand understanding of the challenges and experiences of health and social care professionals working within the anticipatory prescribing system in UK care homes. This research is important and timely because, with an aging population and a concurrent rise in care home placements, there is an incentive and even an imperative to improve end-of-life AP in care home settings (Bone, 2018; Teggi, 2023; Wilson, 2016). Identifying and addressing the infrastructural faults in the AP system can improve the quality of end-of-life care for residents. Moreover, in the backdrop of the COVID-19 pandemic and the rapidly shifting tectonics of guidance and regulation related to care home-based AP, understanding the experiences and challenges of these professionals can guide future policymaking, supporting a holistic approach to end-of-life care that values both patients and providers. The story of this thesis begins with a systematic review which evaluated the published evidence concerning the prescribing, storage, use and disposal of end-of-life care CDs for care home residents in the UK—the first study to explore these drugs inclusive of their various processes and thus collate an evidence base on end-of-life care controlled drugs in the care home context, specifically. The review findings showed that health and social care professionals working with AP experienced a wide range of challenges, particularly concerning end-of-life symptom identification (Bowers, 2022; Finucane, 2014; Kinley, 2010; Kinley, 2016; Morris, 2013), medicines supply management (Breen, 2018; Seymour, 2011; Watson, 2006; Wilson, 2016), and interprofessional relationships (Kinley, 2016; Wilson, 2015; Wilson, 2017; Seymour, 2011); however, the literature did not explore the reasons underlying them. The review also revealed a lack of evidence related to health and social care professionals’ experiences with prescribing, administration, storage, monitoring, and disposal of end-of-life care controlled drugs in UK care homes, particularly anticipatory medicines. 224 Building on these findings – and responding to recent NICE (2015) recommendations for further research into AP – this thesis focused on understanding of the challenges of health and social care professionals’ face working within the AP system in UK care home settings. To do so, I conducted semi-structured interviews health and social care professionals working with end-of-life care AP in care home settings across the UK. Despite the challenges posed by the pandemic in participant recruitment and data collection (Kim et al., 2021), I was able to cultivate a rich dataset consisting of numerous interviews with a diverse group of health and social care professionals working in care homes. Within the dataset, participants emphasised the challenges they encountered in this work; my reflexive, paradox-informed analysis made clear the tensions between contradictory demands underlying those challenges, as well as the paradoxes in which those tensions are sourced. More specifically, a paradox lens, adapted from Smith and Lewis (2011) and Suter et al. (2022), proved well-suited to the analysis I undertook on my empirical dataset. Paradox theory is especially sensitive to the dynamics of complex systems, which are considered the native ground of paradox (Miron-Spektor et al., 2018). My analysis undertook to investigate the AP system at all levels – prescribing, administration, supply, procuring, storage, monitoring, disposal, regulatory apparatus, and bureaucracy. Paradox theory allowed me to incorporate the inconsistencies in viewpoints which arose at the intersection between these levels within a framework of understanding that encouraged me to embrace them rather than shying away from them. The misunderstandings and differences of interpretation which occurred at the boundaries of interprofessional levels proved especially valuable for my understanding of paradox in Chapter 5. In general, the paradox lens forced me to clarify my understanding of key concepts that emerged as themes. When two phenomena enter into contradiction with each other, attempting to think through this conceptual difficulty can allow a researcher to develop an unusually precise and nuanced understanding of them. This was the case, for instance, with the paradox of simultaneous wastage and shortage (the anticipatory prescription paradox of Chapter 5), wherein my attempts to think through the paradoxical manner in which wastage seemed to be the inverse of shortage while still paradoxically being dependent, as a concept, on shortage, had the effect of sophisticating my understanding of the issue of wastage, which would otherwise have appeared simply as an abstraction dispersed across wastage statistics. From my findings, I located three core paradoxes: 1. The Anticipatory Prescription Paradox: that anticipatory medicines wastage and shortage often occur simultaneously – in the same care homes and at the same time (Chapter 5) 2. The Protocol Paradox: that end-of-life care home residents, whose needs are most individualised, are among the most likely to receive the kind of standardised care exemplified by catch-all protocols and nonspecific clinical jargon (Chapter 6) 225 3. The Paperwork Paradox: that the demands posed by excessive paperwork in care homes detract from the time needed for the patient-centred care demanded by that very paperwork (Chapter 7) Chapters 5 to 7 explored the nature of these paradoxes, their related tensions, and how health and social care professionals experience them. Each chapter addressed one of my research questions, which covered the following topics: (1) the underlying systemic tensions and wider interprofessional contexts which underpin issues related to the wastage and shortage of anticipatory medicines (Chapter 5), (2) the roles of algorithmic and discursive structures in shaping the tensions between individualisation and standardisation in end-of-life care (Chapter 6), and (3) the structural logics underpinning AP paperwork and how they contribute to the tension between navigating bureaucratic regulations and providing good, patient-centred end-of-life care in care homes (Chapter 7). I centred participants’ accounts and drew meaningfully from policy, documentation, theory, and the wider literature to support my analytic claims. Furthermore, I proposed recommendations to help professionals navigate these paradoxes. These learnings may be valuable for policymakers in identifying and addressing the underlying paradoxical factors in the AP system which contribute to health and social care professionals’ challenges. Importantly, Braun and Clarke (2021) encourage RTA researchers to think about not only what lies within themes, but also what lies between themes. The overarching three themes of this thesis – the Anticipatory Prescription Paradox, Protocol Paradox, and Paperwork Paradox – interrelate in various ways. The anticipatory prescription paradox and the protocol paradox, for instance, intersect at the topic of standardised ‘packs’ of AMs. The clearest formulation of the protocol paradox came from the account of a participant who struggled, specifically, to reconcile guidance to individualise AP with the fixed and near-universally accepted nature of standardised packs. Furthermore, these ‘packs’, which prescribe AMs four at a time even when sometimes fewer than this are needed, cause wastage; foreknowledge of this phenomenon on the part of prescribing professionals, however, have led some to over-correct and under-prescribe, causing shortage in care homes. These packs thus also contribute to the co-existence of AM wastage and shortage – that is, the anticipatory prescription paradox. Another point at which these two paradoxes meet is prescribing. The protocol paradox encompasses the fact that care home residents most in need of individualised, finely-modulated administration practices are dependent for their supply of medicines on an inherently standardised, protocolised prescribing process which is not structured to accommodate nuance. Similarly, the anticipatory prescription paradox deals with an internal tension between anticipation, which requires some tolerance for the uncertainty associated with the end of life, and prescription, which typically implies and prerequires certainty from a prescriber. The paperwork paradox similarly treats of the topic of uncertainty. The granular certainty which attends paperwork on its most basic level can in fact set 226 off a chain-reaction of error categories which paradoxically increases uncertainty. The paperwork and protocol paradox are also linked by the similarity between the duplicative logic of standardised guidelines (where standardised means replicable across wider contexts, more or less) and the cells and wider documents that make up AP paperwork. Having considered the relationships between the themes and the story of this thesis, I now move on to consider its implications and recommendations for practice and policy in Box 8.1. 8.2 Recommendations for practice and policy Box 8.1 Recommendations for practice and policy ___________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Chapter 5: Recommendations for tackling the anticipatory prescription paradox o Recommendation 1: Reinstatement of Reuse of Medicines guidance In Chapter 5, the counter-intuitive co-occurrence of AM wastage and shortages was explored as the anticipatory prescription paradox. My data brought to light various care homes’ experiences with the Reuse of Medicines guidance, implemented during the COVID-19 pandemic in order to address issues related to potential AM shortages and wastage in care home settings (DHSC, 2020a). CHNs who utilised this guidance credited their ability to do so to supportive community colleagues, particularly pharmacists: in this minority of care homes, the anticipatory prescription paradox appeared to be temporarily resolved. These findings are consistent with the literature that during the pandemic, many care homes introduced temporary changes in their standard operating procedures (SOPs) or developed new SOPs on medicines reuse, facilitated by the active involvement of local health care professionals and teams, including pharmacists and hospice services (Oluyase et al., 2021). A study by Antunes et al. (2022) explores how growing acceptance towards the Reuse legislation stemmed in part from a desire to reduce wastage. In light of the apparent substantial benefits conferred by the Reuse of Medicines guidance according to my own participants, I recommended its reinstatement. While the guidance was specifically tailored to the pandemic, the issues of shortage and wastage which it helped to mitigate are not pandemic-specific, but rather perennial within the care home landscape. This literature on Reuse legislation points to an important growing evidence base for the effects of the pandemic on the health and social care landscape, which will eventually prove critical in extracting lessons from the crisis. My own primary contribution to this evidence base took the form of a set of accounts, explored in Chapter 5, which catalogued the effects of GPs’ inability to visit care homes in person, and their subsequent adoption of telehealth – specifically, the way this introduced indeterminacy into the prescribing process which appeared to exacerbate the anticipatory prescription paradox. I also highlighted a lack of understanding concerning the issue of telehealth and how it altered the patient-provider relationship within the wider literature on this topic, 227 recommending that more research be done in order to support care homes and equip them with the tools required to be able to realise the potential of video consultations, echoing the recommendations of Lim et al. (2016). o Recommendation 2: Fostering strong interprofessional relationships I recommended positive, supportive relationships between the different professional groups that make up the AP system (Bowers et al., 2023). A useful resource on this topic is a 2020 national report which emphasises the importance of culture, support, and relationships between health and social care professionals (DHSC, 2020). My analysis focused on how moments of tension, friction and indeterminacy which occurred at interprofessional boundaries within the AP system perpetuated the anticipatory prescription paradox, whereas instances of collaboration tended to facilitate AM shortages being anticipated far enough in advance to be forestalled. My data showed that, in some instances, pharmacists’ ability to recognise GPs’ prescribing footprints and keep the necessary medications in stock served to prevent AM shortages, and to assist CHNs in the work of procurement. Positive relationships between GPs and pharmacists, and between pharmacists and CHNs, were indispensable for CHNs as it spared them from having to visit multiple pharmacies to procure AMs for their residents. The CovPall Study found that during the pandemic, when care teams were integrated and prescribers had knowledge of local shortages, clinicians could prescribe appropriate alternative medications (Oluyase et al., 2021). The ActMed study found that when GPs are aware of community pharmacy stock, mismatches between prescriptions and available stock are less likely to result, thus resulting in fewer dispensing delays (Latter et al., 2022). Given the interdependency inherent to the AP system, relationships between care homes, GPs, and pharmacists are crucial for effective procurement and dispensing of AMs. A key finding of my analysis was that these relationships were particularly robust where information was perceived as something that should be freely volunteered, rather than sought out only when needed. Chapter 6: Recommendations for tackling the protocol paradox o Recommendation 3: Integration of knowledge regarding synergistic reactions into protocols I defined the protocol paradox as a mismatch between the fundamentally personalised, person- specific business of patient care and a system which operates by means of standardised protocols and guidelines. This system-level contradiction can be understood in terms of the equal pressures it places on professionals to individualise and standardise anticipatory prescriptions and doses. Prescribing protocols and administration algorithms have limitations, and while they can offer valuable guidance to care home staff and health care professionals, there are dangers in using them without accounting for a resident's unique situation and preferences. One possible blind spot of prescribing protocols which emerged from my analysis was the synergistic reactions which can occur when certain incompatible medicines are prescribed in combination. A helpful resource on 228 this topic, Dickman and Scheider’s (2016) The Syringe Driver (p. 120), contains a diagram which visualises the many different possible synergisms that can occur between AMs. This resource, written by pharmacists, can be considered a good example of the potential of pharmacists to act as what Dingwall and Wilson (1995) call ‘information-givers’. In the interests of rendering anticipatory prescribing and administration safer and more informed, I recommended the integration of this type of information into current protocols to render them safer for patients. o Recommendation 4: Adoption of ReSPECT documentation for families of residents During my analysis of the protocol paradox, I found evidence in my data to suggest that one group within the AP system who are particularly vulnerable to the deleterious effects of standardisation were family members of residents. Pinder et al. (2005) showed that family’s wishes were often overridden by protocols. My data contained similar accounts, in which GPs were reported to override the wishes of family members, as recorded within ReSPECT forms. ReSPECT documentation emerged in my data as a useful vehicle for patient-specific information, often derived from those who know residents better than anyone else, their families. To this extent, this documentation is valuable for individualising care and thereby staving off the protocol paradox. My analysis found that families often found themselves on the wrong end of damagingly ambiguous, jargonistic or euphemistic language which had the effect of clouding the specificity and transparently needed for truly patient-centred care. It has been found that poor communication between care home staff and families can cause the latter emotional distress (Payne et al., 2015), as well as increase unnecessary hospital admissions (Spacey et al., 2021). There is a need for open, honest, and early discussions about AP between care home caregivers, residents, and families, with residents' medication preferences being central to individualised care. In one sense, this can be achieved by ReSPECT forms, which open up crucial lines of communication between families and care homes. o Recommendation 5: Combatting jargon and improving communication My data showed that the ambiguity around what exactly constitutes the end-of-life allows healthcare professionals to exploit the jargon of AP in order to cloak the reality of death in euphemism and thus shield themselves from the responsibility of undertaking emotionally challenging discussions, at the expense of families. I therefore recommended that clear communication with residents and families be prioritised during the end of life process. One measure which could be taken is the circulation of Pitt et al’s seven categories of medical jargon within the care space. This is a framework intended to combat ‘jargon oblivion’, a situation wherein clinicians continue to use medical jargon despite being aware of its negative effects. Chapter 7: Recommendations for tackling the paperwork paradox o Recommendation 6: Participatory governance In Chapter 7, I explored the paperwork paradox: the contradiction that AP documentation, which appeared to demand and even enforce good standards of care, took up the time needed for CHNs to 229 provide it. My analysis showed that paperwork, used as a regulatory tool, promotes a culture of fear in which fears that small administrative errors will result in harsh disciplinary treatment lead to conservative administration practices that harm residents. In order to correct for this state of affairs, it would be useful to shift CHNs from a position of compliance to paperwork, to a position of control over paperwork. The 2022 WHO report recommended participatory governance model as a way to empower those who had hitherto generally been afforded a maximum of liability and a minimum of freedom within the regulatory framework which enmeshes the care landscape (WHO, 2022). Hammersley (2005) and Smith (2013) similarly emphasise the importance of pointing out that policymaking should be the domain of those who perform the work with which that policy is associated (Guyatt et al. 1992; Melnyk, 2022). o Recommendation 7: Centralised guideline interpretation resource In Chapter 7, one of my participants expressed the need for care home managers to be able to openly discuss problems and to discuss their interpretations of confusing guidelines and policies, without fear of being blamed for said problems. I therefore recommended an online resource, perhaps a forum, for such discussions. This resource could host and centralise guidance documentation and also include a comments section where care home professionals could discuss how they actually implemented the often unclear guidance contained therein in practice, perhaps in the form of full- blown accounts. There is precedent for this in the ‘case studies’ section of the GSF Care Homes Training Programme Good Practice Guide (GSF, 2014). In the following section, I reflect on my role in the production of this knowledge and the assumptions made throughout the research process. 8.3 Reflexivity statement In this study, reflexivity served as a tool for understanding myself as an individual and researcher. Reflexivity demands an acknowledgment of biases and how they shape the research process (Berger, 2015; Creswell, 2016; Pillow, 2003), and involves the identification of the researcher's social and personal positions – factors like age, gender, ethnicity, and race play significant roles (Thurairajah, 2018). This is important in qualitative research, as the researcher's involvement enriches the findings, adding depth and authenticity to the work as they actively recognise and reflect on their role in knowledge production (Creswell, 2016). In the introductory and methodological sections (Chapters 1 and 3) of this thesis, I detailed the experiences driving this research, established my epistemological position, and shared reflections on my role as the interviewer. This section thus focuses on personal reflexivity – the interplay between my personal experiences and my research direction after the data collection period, detailed 230 in Box 8.2, and functional reflexivity – involving a re-evaluation of my research assumptions and their implications on this work, detailed in Box 8.3. Box 8.2 Personal reflexivity ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Drawing from the early stages, my volunteering in local care homes laid the initial groundwork for understanding the complexities of working with medicines in these contexts. As I explored these complexities as a PhD student, it became clear that my own life was interweaving with my research, adding layers of complexity and depth to this work. The challenges posed by the pandemic presented both logistical and emotional hurdles. On one hand, my study plan had to be reformulated (at the outset of the PhD, I intended to conduct an ethnography in a local care home, though this quickly became unfeasible); on the other, the personal toll of caring for my ailing parents provided an intimate perspective into the realities of caregiving under duress. This intersection of personal life and academic pursuits deeply impacted my research process. Returning to Cambridge as a second year PhD student, the solitude I felt - intensified during the pandemic's second wave - was a significant obstacle. While this solitude enhanced my connection to participants, it also presented potential pitfalls. The shared sentiments of isolation and disconnection may have influenced the way I framed questions or interpreted responses. For instance, if a CHN described feeling challenged by having to administer AMs alone and without community support, I occasionally followed up by asking, ‘How did facing such challenges alone influence your perspectives on administering those medicines?’ This line of questioning, while rooted in the context of the research, might have been deepened by my personal experiences of isolation. Recognising the weight of these emotions, I sought formal counselling for psychological support, demonstrating the profound impact of the data collection phase on me both personally and academically. To safeguard the integrity of my research, I prioritised regular reflections on my post- interview notes and transcripts, striving for a measured and balanced approach to data collection and analysis. Finally, the data analysis and writing-up phases of my research were profoundly influenced by personal difficulties. The physical challenges of pregnancy, followed by the emotional challenges and responsibility of caring for a newborn, profoundly deepened my connection to themes of care and patient-centredness. These experiences fostered a richer understanding through which I could explore and interpret the research themes more holistically. In essence, this research journey has seen to my personal and academic growth. Each experience and reflection has added layers of depth, and the result is a body of work that captures the nuances, challenges, and realities of AP and end-of-life care during the pandemic. ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 231 Box 8.3 Functional reflexivity ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ o Reflecting on recruitment The method of recruitment employed in this study, specifically the primary snowball strategy, undeniably affected the research sample and, by extension, the overall results. One inherent characteristic of snowball is the reliance on participants to refer other potential subjects (Patton, 2015). Consequently, this could introduce a certain degree of selection bias, wherein participants might be inclined to recommend others with similar perspectives or experiences, potentially leading to homogeneity in the sample (Patton, 2015) and capturing a particular subset of perspectives more than others. This trend was evident in my participant sample, where care home professionals constituted the majority (29/42), likely because CHMs, who initiated the recruitment chain, referred other care home professionals within their networks to participate. Thus, it is essential to recognise that while snowball sampling can yield rich insights, especially in contexts where participants are difficult to access, it does present challenges in terms of achieving a truly diverse and representative sample (Patton, 2015). The use of incentives in the recruitment process surfaced another layer of complexity. Post-interview observations indicated diverse reactions to the incentive. While some care home professionals enquired about the voucher, suggesting it might have been a motivating factor, most community- based health care professionals opted out of it. The differential acceptance and its ethical implications suggest that for some care home professionals, the incentive was a potential extrinsic consideration and motivator, possibly secondary to the intrinsic motivation to contribute to the research. For health care professionals based in the community, their refusal could point to relative financial stability or possibly a preference to participate purely for the sake of contributing to research that they care about, without external inducements. The ethical implications of these differential responses are worth considering, as they may call into question the extent to which financial incentives introduce potential bias in participation and their possible interplay with intrinsic motivation to contribute to research. o Reflecting on the interviews Reflecting on my qualitative interview design choice, I found that interviews were both appropriate and well-suited to the pandemic context. Pandemic restrictions led to many participants describing feelings of isolation from their social networks, including family, friends, and colleagues. As a result, many participants appeared to regard me as a confidant with whom they could discuss their work- related challenges candidly, leading to rich insights about their experiences related to AP. The interview presented an opportunity for human connection with a researcher eager and willing to listen. Furthermore, as a newcomer to the country without a patient care role, it appeared that the one-off experience of the interview generated a ‘strangers on a train’ effect, wherein individuals feel more comfortable sharing personal thoughts in anonymous and brief encounters with strangers they 232 do not expect to meet again (McKenna et al., 2000). The combination of my newness and the pandemic-induced isolation of my participants may have made them more comfortable in sharing their experiences with me, despite the lack of tangible presence. In conclusion, while interviewing involved its own set of challenges, especially during the pandemic, it aligned well with the core objective of this research. The depth and breadth of the dataset underscores the value of this method, supporting a holistic understanding of the experiences and challenges of health and social care professionals working in the AP system in care home settings. o Reflecting on the research questions In this research, the initial research questions stemmed from the findings of the literature review (Chapter 2), which was centred around the need to build an evidence base surrounding the processes of prescribing, dispensing, administration, storage, monitoring, and disposal of AMs in practice. These questions provided a sense of direction and consistency, though they were not the questions that penetrated to the heart of the dataset. For example, an interaction I had with a CHM beginning with a topic guide question about disposal (Appendix 8) led her to describe AM disposal as a straightforward, mechanical process. Her response resonated with the procedural portrayal commonly found in AM disposal policy: denature the drugs using a DOOP kit and promptly return them to pharmacy. However, when the conversation began to centre her personal experiences with disposing AMs, she shared an emotive account of the guilt she felt when disposing unused medicines and the frustration she felt in response to the rules that enforced it. This laid bare the human element of the process, highlighting the tension and incompatibility between her values and those of the system she worked in – largely disconnected from the human realities it governed. This understanding was important, and after adopting Braun and Clarke’s six phase approach to RTA (2006, 2019 & 2021), such accounts underscored the need to iterate and adapt my research questions to ensure that my research spoke to the heart of the dataset. Ultimately, adopting a paradox lens allowed me to do justice to such tensions and incompatibilities. o Reflecting on the analysis Both the design of the study and the analytical methods chosen shaped the findings of my empirical research. RTA was particularly suited to this work, given its alignment with my epistemological and ontological positions (constructivism and relativism) (Bryman, 2012; Gray, 2014; Braun & Clarke, 2022). Adapting Braun and Clarke’s six-phase approach to RTA (Braun & Clarke, 2006, 2019 & 2021) supported the integration of a paradox lens. This integration was heavily influenced by the work of Smith and Lewis (2011) and Suter et al. (2022), which had an important role in shaping my analysis at each stage (Chapter 3). This shift helped me develop a narrative around the inherent paradoxes in the AP system which, as I have shown, lead to various tensions and challenges. Additionally, Braun and Clarke (2013) advocate for researchers to combine, in their analytical write- ups, results and discussion – largely unorthodox in health services research. Heeding their advice brought a richness and depth to my analysis, improving its overall coherence and relevance. 233 Recognising the inherent subjectivity of the research process and importance of reflecting on my role as the researcher, I adopted the methodological tool of a research journal (Finlay, 2021). First, it captured the rich, dynamic essence of qualitative data, emphasising the relationship between researcher and participant (Braun & Clarke, 2022). Second, echoing Warr, it acted as a counterweight against inadvertent biases (Warr, 2004). Extracts from my reflexive journal can be found in Appendix 9. They demonstrate its utility throughout the research process, and how it helped me grow as a researcher. o Reflecting on the writing-up process While writing up my analysis, I noticed that the voices of certain participants, such as Mrs. Raymond, Mrs. Barnes, and Dr. Cameron were quoted more than others. This pattern is common in qualitative research where quotes are often selected for their alignment with the study’s aims and relevance to the research questions, as recommended by Braun and Clarke (2006). This selective quoting is further supported by Geertz’s principle of 'thick description', which suggests that some individuals offer richer and more detailed responses, rendering their contributions more substantial for analysis and discussion (Geertz, 1973). Similarly, Charmaz (2006) and Patton (2015) noted that participants vary in their speech—some speak in a manner that is particularly resonant, clear, or succinct. The variations of such contributions led to some participants’ words being more illustrative of the paradox themes, hence their prominence in my analysis. Therefore, while I aimed for a balanced distribution of ‘voice’ in this study, a certain degree of unevenness is inherent and serves as a methodological reflection point grounded in the analytical goals and the quality of the data. ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 8.4 Quality, strengths, and limitations Evaluating the quality of qualitative studies is important, as discussed in Chapter 3. For this thesis, I adopted Tracy's (2010) universal criteria to evaluate the quality of my research due to its comprehensive and flexible nature, as endorsed by Tracy and later by scholars like Smith and McGannon (Smith & McGannon, 2018). Tracy's (2010) framework encompasses: a worthy topic, rich rigour, sincerity, credibility, resonance, significant contribution, ethics, and meaningful coherence. Employing Tracy's criteria has enriched this research, providing a thorough critique and underscoring its strengths and limitations. Table 8.1 Strengths and weaknesses of the qualitative research with Tracy’s (2010) criteria Worthy topic A topic is considered ‘worthy’ when it is relevant, timely, significant, and engaging (Tracy, 2010). The need for this research emerges from the recognised importance yet insufficient exploration of AP in palliative and end-of-life care research. Despite its endorsement in policy, the practical application of AP and stakeholder views remain under-researched. In 2015, NICE identified a lacking evidence base regarding the use of AMs in practice (NICE, 2015). Even with advancements in the literature, there is still a need for research, particularly 234 in care home contexts. This study responds to NICE's (2015) call. My literature review found that existing studies identify the challenges professionals face in the work of AP (Majumder et al., 2022), but their root causes remained unexplored. This work extends the academic conversation by exploring these core challenges, revealing tensions and inherent paradoxes in the system. Such a landscape underscores the timeliness and relevance of this research. Rich rigour A rigorous qualitative study draws on sufficient, abundant, appropriate, and complex theoretical constructs, data sources, contexts, and samples (Tracy, 2010). Depth and breadth in these areas can ensure a comprehensive and nuanced understanding of the topic under study. I carefully consider the rigour of the empirical research below. PPI: One notable strength was the active involvement of PPI collaborators during the development of the research plan (Appendix 6). The PELiCam Working Group’s PPI insights significantly influenced various aspects of the study design, including the development of the initial research objectives, recruitment strategy, and the language of participant information sheets for CHMs. They played an essential role in identifying ethical issues specific to care home research and in shaping the interview topic guide. However, due to COVID-19, planned collaborations for refining results and guiding the analysis were disrupted. The pandemic posed significant ethical challenges for working with care home residents and their families, and their well-being took precedence over research collaborations. Recruitment strategy: Given the challenges posed by the COVID-19 context, recruitment was not straightforward. At the start of my PhD, I planned to interview care home residents and their family members; however, due to logistical and ethical issues associated with research involving exceptionally vulnerable populations during the COVID-19 pandemic, this did not happen. In future studies, it will be important to address this gap and centre the perspectives of care home residents and family. With the support of my supervisors, I decided to focus the research on the experiences of health and social care professionals working with AMs. Triadic recruitment of one CHM, one CHN, and one GP per care home was intended, and the method was heavily reliant on managers disseminating information about the study. However, this did not occur, possibly due to the pandemic context and research being deprioritised for health and social care professionals working on the frontlines. Therefore, study recruitment relied primarily on a snowball strategy, which involved leveraging my supervisors’ professional contacts and established networks such as ENRICH and QNI. This may have inadvertently restricted our reach and introduced certain biases (Atkinson and Flint, 2001). For instance, those who opt for research participation, especially in care homes, may already have positive predispositions towards their work, potentially overshadowing deeper issues faced by other staff (Collingridge et al., 2019; Griffiths et al., 1993; Hendricks et al., 1992). To counteract this potential bias, vouchers were offered to encourage participation from groups underrepresented in research, like CHNs (Collingridge et al., 2019; Thompson, 1996). However, further examination is needed to understand how this incentive might truly impact recruitment bias and diversity. Furthermore, study recruitment could have benefited from increased inclusion. For example, given that interviews were solely conducted in English, the study did not account for the experiences of professionals whose first or preferred language was English. Future research could consider involving a multicultural team to include and centre their experiences. Moreover, the study did not aim to recruit care home staff in non-nursing and non-managerial positions, such as volunteers and direct care workers. While the primary focus was on professionals central to the AP processes (primarily nurses, managers, and doctors), others affected by AP in the care home context also have important experiences that future research could explore. Study population: A multi-phase mixed recruitment strategy was used, which combined purposive and convenience sampling techniques (Patton, 2015). This allowed for a diverse range of health and social care professionals to be interviewed, evidenced by the extensive dataset from 42 interviews with a diverse range of health and social care professionals across 29 care homes in the UK, spanning various sizes, geographic environments, and financing structures. This supports my findings’ transferability to other UK care home contexts. This also supported a rigorous cross-examination of professionals’ perspectives, enabling a thorough exploration of the underlying causes of challenges they experience in the AP system. 235 The inclusion of a diverse range of professionals speaks back to the literature, responding to calls for such inclusion by Kinley et al. (2016), Bowers et al. (2022), and Teggi (2023). However, there were areas for improvement, particularly regarding participant distribution. Of the participants, 29 were from care homes (19 CHMs and 10 CHNs), while only 13 were community-based health care professionals, which included DNs, doctors, pharmacists, and a single ambulance paramedic. For instance, only three DNs were interviewed. This can be a limitation given my findings, which suggest a hierarchical structure for care home-based AP led by visiting health care professionals. Thus, future research could further study the role of visiting nurses and explore their involvement in AP, their infrequent interactions with care home patients, and any collaborative decision-making with care home staff, especially concerning AM administration and symptom identification. Data collection: The choice of semi-structured interviews, while appropriate for the study's objectives, presents important limitations. For instance, such interviews can call dependability into question and introduce biases (Saunders, Lewis, & Thornhill, 2019). A primary concern regarding dependability in interviews is the potential for interviewer bias: the researcher’s preconceived ideas and attitudes can influence both the framing of questions and their interpretation of responses (Saunders, Lewis, & Thornhill, 2019). To counteract this, I engaged in reflexivity exercises throughout the research process, by keeping a reflexive journal and engaging in routine discussions and reflections, both during supervisory sessions and with peers in the PELiCam group. Moreover, the interviews may not have captured the full spectrum of participants' experiences and views, especially given the sensitive nature of the topic. For example, care home participants occasionally grew reticent when the discussion veered towards the topic of errors or oversights with AMs. This may have been due to the possibility of professional repercussions or personal guilt or regret. Thus, while the semi-structured format offers flexibility and depth, it can also be limited by what participants are willing to share, underscoring the need for a supportive and non-judgmental interview environment. Future studies could consider combining interviews with other methods, such as anonymous surveys or focus groups, to capture a more comprehensive view of professionals' experiences and perspectives. Data analysis: For the analysis, I adopted Braun and Clarke’s (2006, 2019, & 2021) RTA, celebrated for its depth in exploring data (Braun & Clarke, 2021). However, Braun & Clarke (2022) emphasised that while the approach can offer rich insights, its interpretive power may be restricted without the use of a specific theoretical framework. Acknowledging this, I focused and strengthened the analysis by integrating a paradox lens, informed by Smith and Lewis (2011) and Suter et al. (2022), which proved invaluable in generating meaningful interpretations. A key strength in my analysis lies in the iterative development of themes from the data. This was supported by my consistent reflexivity: after each interview, I quickly recorded reflective notes, allowing for an evolving understanding of emerging patterns (Braun & Clarke, 2021). This attentive approach is aligned with Tracy's (2010) emphasis on meticulousness in data analysis as an indicator of academic rigour in qualitative research. Additionally, to evaluate the quality of the analysis, Braun and Clarke's (2020) 'Twenty questions to evaluate the quality of thematic analysis' served as a rigorous benchmark, which can be reviewed in Appendix 10. Notably, the COVID-19 pandemic unexpectedly disrupted intended PPI collaborations, designed to refine and guide my analysis. However, I sought feedback on my findings from a diverse audience, including the general public, academic peers, and health care professionals in the community during public engagement events hosted by the PELiCam department and virtual departmental presentations. This allowed me to maintain transparency and refine the study's direction in light of its broader implications. Sincerity Sincerity is a fundamental component in qualitative research, emphasising the authenticity and genuineness of the process (Tracy, 2010). This involves being forthright about one's values, biases, and objectives and understanding their influence on the chosen methods (Tracy, 2010). As discussed in the reflexivity statement, the path to completing this study was laden with personal challenges that inevitably impacted the research process. 236 My academic training and interviewing experience instilled in me the confidence to create an atmosphere where participants felt comfortable discussing sensitive subjects, which supported the authenticity of my dataset. However, it is important to point out that my personal motivations and pre-existing ideas about end-of-life care and medicines in care homes have influenced the analysis. Reflecting on the nature of subjectivity in understanding, Goffman (1974) suggests that we all 'frame' our experiences, suggesting that our perceptions, beliefs, and backgrounds inevitably shape our insights (Goffman, 1974). Given this, while my close connection to the topic may have provided nuanced depth, there is a risk it may have also limited the breadth of my perspective. To consider this potential limitation, I endeavoured to remain anchored to the dataset, consistently revisiting the raw data, cross-referencing themes, and double-checking interpretations. I sought to ensure that the findings remained rooted in participants' voices and experiences rather than being heavily swayed by my own perspectives. Furthermore, by actively seeking feedback and engaging in peer debriefing sessions, I attempted to introduce checks and balances to my interpretative process. This rigorous approach was my conscious effort to bring transparency and authenticity to the analysis and strike a balance between personal insights and data-driven conclusions. I also acknowledge that there were lapses in my reflexivity. Several concurrent challenges had a profound effect on me during the research process. The pandemic presented both professional and personal difficulties. Balancing the rigour of conducting a literature review and planning research with caring for ailing family was physically and emotionally taxing, adding layers of complexity to the backdrop of this work. Added to this was the challenge of navigating a particularly strenuous pregnancy during the analysis and write-up. The subsequent experience of caring for my child, who faces complex health challenges, has brought forth an entirely different range of emotional and practical complexities. Each of these experiences, detailed further in my reflexivity statement, stretched the bounds of my resilience and influenced my approach and ideas during this research journey. Credibility Tracy (2010) suggests that credibility is important in qualitative research, serving as a gauge for the plausibility and trustworthiness of the findings. Achieving credibility requires thick description and embracing multi-vocality (Tracy, 2010). I provided thick description by weaving direct quotes into the analysis and juxtaposing the findings against extant literature. Multi-vocality in research involves meaningful incorporation of multiple voices during the analysis (Tracy, 2010). I ensured diversity in this work by interviewing a wide range of professionals and including their direct quotes in my analysis, thus reflecting the spectrum of experiences and attitudes in my dataset. The use of RTA as my primary analytical framework aptly foregrounded participants’ experiences. Resonance Resonance, as defined by Tracy (2010), is the capacity of the research to resonate with, influence, and move its readers. This thesis strives to tell a story anchored in the lived experiences and challenges of health and social care professionals navigating the work of AP. The golden threads woven throughout are the paradoxes inherent in the system: these serve as the linchpin of my exploration. My intention was to explore participants’ experiences through this lens, aiming for both depth and clarity. Moreover, resonance extends to the concept of transferability – translating insights across diverse settings, populations, and contexts (Smith, 2018; Tracy, 2010). My findings have been contextualised in and connected to previous research, the wider literature, and relevant theory, suggesting their wider relevance to care home settings across the UK. However, it is important to consider the context of this study: in many ways, research captures a moment in time (Saunders, Lewis, and Thornhill, 2019), and this study, significantly shadowed by the COVID-19 pandemic, is no exception. Thus, critical questions emerge: how have the findings been shaped by the pandemic, and do these pandemic-informed insights have resonance beyond their temporal context? In the context of my findings, the pandemic did not create new fissures within the care home sector but rather intensified pre-existing faults. For instance, pre-existing challenges such as staff shortages (Coon et al., 2014; Hockley et al., 2010), inadequate training and preparedness in urgent end-of-life care situations (Bone et al., 2018; Brighton et al., 2016), inconsistencies in care quality across homes (Oliver et al., 2014), and a lack of interprofessional collaboration (Oliver et al., 2014; Reeves, 2017) became even more pronounced during this period. Therefore, while the pandemic's mark on the data is clear, the analysis offers a deeper understanding of the care home sector's long-standing challenges, 237 further detailed Chapter 8. In essence, the pandemic might have acted as a lens and a magnifier, highlighting persistent issues while underscoring the need for systemic reform thus reinforcing the study's wider relevance. Ethics Ethics in qualitative research is paramount, especially when the subject matter explores sensitive topics such as end-of-life care. Tracy (2010) emphasises that ethical considerations must span the entire research process, from procedural and situational aspects to relational dynamics. For this study, the foundation of procedural ethics was rooted in the guidelines for qualitative research laid out by Kvale and Brinkmann (2005 & 2009). Additionally, the principles outlined by the GMC Research Ethics Code and the British Psychological Society's codes on human research ethics and ethics and conduct (The British Psychological Society, 2014 & 2018) provided further direction, with ethical approval granted through the Cambridge University Psychology Ethics Committee. Situational ethics refers to the ethical challenges that emerge in the actual process of data collection (Ellis, 2007). Given the emotional weight of some participants' stories, I faced immediate ethical decisions during interviews, particularly when participants became emotional. My training and experience equipped me with strategies to support participants during these moments. Sharing options like pausing the interview, taking a break, or discontinuing altogether were important, reflecting my commitment to participants’ well- being, as discussed in Chapter 3. However, none of the participants accepted chose to halt the interview at any point. Central to relational ethics is the notion of mutual respect, ensuring that the research process does not marginalise or unduly burden participants (Tracy, 2010). For this reason, interviews were conducted at times and locations convenient to them. Throughout our conversation, regular breaks were proposed. Additionally, participants were given the option to review anonymised transcripts, ensuring their comfort with the content shared. Furthermore, recognising the emotional weight of the topics discussed, I offered information about organisations that could provide support if needed during the post-interview debrief discussion (Appendix 6), in accordance with post-interview recommendations by Kvale and Brinkmann (2009). In sum, each phase of this research was approached with sensitivity and responsibility, ensuring an ethical, respectful, and supportive experience for all involved. Meaningful coherence Meaningful coherence refers to the alignment of a study's various components with its stated objectives (Tracy, 2010). For this research, the core objective was to expand understanding of the experiences and challenges faced by health and social care professionals within the AP system in UK care homes. Part 1 of the thesis started this journey with a literature review that identified the known challenges faced by health and social care professionals in care home- based AP. However, prior research had not explored the reasons underlying these challenges. Importantly, Bowers (2022), Kinley (2016), and Teggi (2023) emphasised the need to broaden our understanding of the diverse experiences of health and social care professionals within the AP landscape. To build on the evidence base and respond to these calls, I set out to explore the experiences and challenges of health and social care professionals working in UK care home settings. Part 2 of the thesis presents the methods to do so. Using RTA and a paradox lens, I detail how I came into the three defining paradoxes of the AP system in UK care homes: the anticipatory prescription paradox, the protocol paradox, and the paperwork paradox. Part 3 delves into an in-depth exploration of the challenges as narrated by participants. This exploration illuminated the 'tensions between' the different facets of these challenges and the 'paradoxes within' the system that fuel these tensions. Drawing meaningful connections between these paradoxes and the wider corpus of literature, theory, and policy served to contextualise the research findings. Part 4 then offers thoughtful recommendations for policy and practice, drawn from the findings. In synthesising these components, the thesis achieves meaningful coherence. It bridges the literature review, methodology, design, and empirical findings to present a holistic understanding of the complex web of experiences and challenges faced by health and social care professionals working in the AP system in care home settings. This alignment of research components with the study's overarching aim underscores its rigour and positions it as a valuable contribution to the wider conversations on AP, care home research, and interprofessional working. 238 In reviewing my work, I found a clear alignment with Tracy’s (2010) quality criteria, which is instrumental for qualitative research that explores sensitive subjects such as end-of-life care (Ashton, 2014). The inherent complexities of end-of-life research necessitate a framework for evaluation that emphasises sensitive and ethical research (Ashton, 2014). Importantly, these complexities have been an avenue for learning for me as a researcher. My exploration of the AP system has thus yielded valuable insights. Box 8.4 outlines distinct directions for future research. 8.5 Future research directions Box 8.4 Future research directions ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ While recognising that none of my recommendations can be universally applied across all care home contexts, I emphasise that the findings from this research provide novel insights into the ways in which underlying systemic paradoxes shape the outcomes and experiences associated with AP. These insights can support health and social care professionals in navigating challenges they face in their day-to-day work, as well as policymakers in developing policy that considers the underlying paradoxes of the AP system. However, further research is needed for a fuller understanding of the AP system in care home settings. To that end, I present research recommendations that focus on its different aspects: its bureaucracy, the regulatory authorities which sustain it, and the processes of prescribing, dispensing, procurement, administration, storage, monitoring, and disposal. 1. Relationships. From my findings, a priority for future research in the AP system is to better understand ways in which to cultivate and sustain the relationships that drive each aspect of it. Further ethnographic and observational research would help to explore the factors involved in building effective intra-organisational and inter-professional relationships, and determine if and how they can support professionals in navigating the paradoxes in the AP system. Observing the relational dynamics of this complex and interconnected system could present opportunities to improve communication, coordination, and collaboration among care home staff, visiting health care professionals, residents, and families, and to foster sensitive, open conversations that support a shared understanding of AP. 2. Prescribing. Observational studies of the symptom profiles of care home residents at end-of-life will enable greater understanding of the personalisation needed in AP, instead of relying on expert consensus or prescribing standard AMs with set doses. In the light of the protocol paradox and the unpredictable nature of dying trajectories, future research could also investigate how prescribing clinicians navigate the uncertainties of prescribing and administration in decision-making. This is particularly relevant for residents with dementia, who may be unable to communicate discomfort and pain. 239 3. Dispensing. Further research is needed on the role of community pharmacists in optimising the safety and effectiveness of AP and in understanding issues of access to AMs, including the impact of pharmacist-led interventions in prescribing and administration practices, and the barriers and facilitators to pharmacists' involvement in care home-based AP. 4. Procurement. In light of the anticipatory prescription paradox, future research is needed to better understand AM supply chains in order to increase the accessibility of medicines and prevent shortages, and to identify strategies that improve communication and coordination between care home staff, nurses, pharmacists, and prescribing clinicians to streamline procurement, especially during out-of-hours periods. Potential interventions include a database of contracted pharmacies with extended open hours and medication availability tracking in real-time. 5. Administration. Further research is needed to better understand the individual, relational, and bureaucratic elements that influence AM administration decisions, and to clarify the roles of the various stakeholders involved. Future studies could also investigate the effectiveness of training and education programs, support mechanisms such as palliative care phone lines, and the GSF’s end-of- life care toolkit in improving CHNs' knowledge and confidence in administering AMs. 6. Documentation. In light of the paperwork paradox, further research could identify ways in which to streamline AP paperwork, including storage and disposal logs, to increase worker efficiency, enable care home staff to spend more time on direct resident care, and help policymakers ensure that regulatory requirements maintain appropriate focus on the needs of care home residents. 7. Monitoring and Storage. Clinical audits of monitoring documentation could harness the data recorded in CD logs and MAAR sheets to discern patterns in AM administration and flag potential issues such as over- or under-administration of AMs. Interventions could then be developed and evaluated to improve the safety and efficacy of AP for care home residents. 8. Disposal. Future research could investigate the use of DOOP kits and other strategies that support timely disposal of AMs following a resident's death. Delayed disposal leads to prolonged storage in care homes, posing risks to care home safety and limiting medication storage capacity. ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 8.6 Final remarks This study of anticipatory prescribing offered me a way into understanding the system of care as a whole – one that is living and breathing and whose best parts are its people. Among them, I feel especially grateful for Dr. Powers and Mrs. Raymond, who rekindled my belief in a system that at some points felt to me irrevocably fractured. They and many others showed me that the spirit of care can coexist with – even animate – the skeleton of regulation. The interplay between the two have profoundly shifted the way I view medicine and end-of-life caring, which hinges on the art of interpretation, skilled and empathic care, and perhaps, most crucially, the courage to adapt amid pandemic and paradox. 240 Closing this chapter of both this body of work and my life, I am left believing more deeply than I had in the goodness of people and in the possibility for change. There is hope here – not so much in sweeping overhauls, but through informed, incremental shifts initiated by those like Dr. Powers—reminders that through the bounds of paperwork and protocols, the heart of care beats. 241 References Adom, D., Osei, M. and Adu-Agyem, J., 2020. COVID-19 lockdown: A review of an alternative to the traditional approach to research. Research Journal in Advanced Social Sciences, 1, pp.1-9. Agha, Z., Roter, D.L. and Schapira, R.M., 2009. An evaluation of patient-physician communication style during telemedicine consultations. Journal of Medical Internet Research, 11(3), p.e1193. Al-Shaer, D., Hill, P.D. and Anderson, M.A., 2011. Nurses' knowledge and attitudes regarding pain assessment and intervention. Medsurg Nursing, 20(1). Alcorn, G., Murray, S.A. and Hockley, J., 2020. Care home residents who die in hospital: exploring factors, processes and experiences. Age and Ageing, 49(3), pp.468-480. Alldred, D. P., Kennedy, M. C., Hughes, C., Chen, T. F., Miller, P., 2009. Interventions to optimise prescribing for older people in care homes. Cochrane Database of Systematic Reviews. Allen, D., 1997. The nursing‐medical boundary: a negotiated order?. Sociology of Health & Illness, 19(4), pp.498-520. Allen, L., Williamson, S., Berry, E. and Alderwick, H., 2022. The cost of caring: poverty and deprivation among residential care workers in the UK. Health Foundation. Allitt, T., Kirby, E. and Hignett, S., 2017. Medication management in community care: using hierarchical task analysis to describe complex systems. Altman, I.L., 2017. Pharmacists' perceptions of the nature of pharmacy practice (Doctoral dissertation, University of Brighton). Altman, I.L., Mandy, P.J. and Gard, P.R., 2019. Changing status in health care: community and hospital pharmacists’ perceptions of pharmacy practice. International Journal of Pharmacy Practice, 27(3), pp.249-255. Alvesson, M. and Sandberg, J., 2013. Constructing research questions: Doing interesting research. Sage. Alzheimer's Association, 2022. Alzheimer's disease facts and figures. Alzheimer's & dementia, 9(2), pp.208-245. Alzheimer’s Society, 2013. Low Expectations: Attitudes on Choice, Care and Community for People with Dementia in Care Homes. Amador, S., Goodman, C., King, D., Machen, I., Elmore, N., Mathie, E. and Iliffe, S., 2014. Emergency ambulance service involvement with residential care homes in the support of older people with dementia: an observational study. BMC geriatrics, 14(1), pp.1-6. Amass, C. and Allen, M., 2005. How a just in case approach can improve out-of-hours palliative care. Pharmaceutical journal, 275(7356), pp.22-23. Amri, M., Angelakis, C. and Logan, D., 2021. Utilizing asynchronous email interviews for health research: overview of benefits and drawbacks. BMC Research Notes, 14(1), pp.1-5. Andriopoulos, C. and Lewis, M. W., 2009. Exploitation-exploration tensions and organizational ambidexterity: Managing paradoxes of innovation. Organization Science, 20(4), 696-717. Antunes, B., Bowers, B., Winterburn, I., Kelly, M.P., Brodrick, R., Pollock, K., Majumder, M., Spathis, A., Lawrie, I., George, R. and Ryan, R., 2020. Anticipatory prescribing in community end-of-life care in the UK and Ireland during the COVID-19 pandemic: online survey. BMJ supportive & palliative care, 10(3), pp.343-349. Antunes, B.C.P., Bowers, B., Barclay, S., Gallagher, J., Conci, R. and Polak, L., 2022. Community-based anticipatory prescribing during COVID-19: a qualitative study. BMJ Supportive & Palliative Care. Archer, A., 2015. Medication errors: an unacceptable gamble. Nursing and Residential Care, 17(7), pp.393-397. Atkinson and Flint, 2001. Accessing Hidden and Hard-to-Reach Populations. Attride-Stirling, J., 2001. ‘Thematic networks: an analytic tool for qualitative research’, Qualitative Research, 1(3), pp. 385– 405. Available at: https://doi.org/10.1177/146879410100100307. BMA, 2021. Moral distress and moral injury: Recognising and tackling it for UK doctors BMA, 2022. BMA Covid Review 3: Delivery of healthcare during the pandemic. Barclay, S., Froggatt, K., Crang, C., Mathie, E., Handley, M., Iliffe, S., Manthorpe, J., Gage, H. and Goodman, C., 2014. Living in uncertain times: trajectories to death in residential care homes. British Journal of General Practice, 64(626), pp.e576-e583. Barlet, M.H., Barks, M.C., Ubel, P.A., Davis, J.K., Pollak, K.I., Kaye, E.C., Weinfurt, K.P. and Lemmon, M.E., 2022. Characterizing the language used to discuss death in family meetings for critically ill infants. JAMA Network Open, 5(10), pp.e2233722-e2233722. Barron, D.N. and West, E., 2017. The quasi-market for adult residential care in the UK: Do for-profit, not-for-profit or public sector residential care and nursing homes provide better quality care?. Social Science & Medicine, 179, pp.137-146. Bateson, G., Jackson, D.D., Haley, J. and Weakland, J., 1972. Double bind. Steps to an Ecology of Mind, 276. Batkovskyte, I., Bourke, L., Ellis-Smith, C., Bone, A., Goodman, C., Barclay, S., Higginson, I., Ogden, M., Evans, C. and Sleeman, K., 2021. P-198 Care home staff experiences of palliative and end-of-life care in care homes during COVID-19. Batten, J.N., Wong, B.O., Hanks, W.F. and Magnus, D.C., 2019. Treatability statements in serious illness: The gap between what is said and what is heard. Cambridge Quarterly of Healthcare Ethics, 28(3), pp.394-404. 242 Bayliss, K. and Gideon, J., 2020. The privatisation and financialisation of social care in the UK. SOAS Department of Economics Working Paper, p.238. Beck, U., 1992. Risk Society: Towards a New Modernity Becker, M.L., Kallewaard, M., Caspers, P.W., Schalekamp, T. and Stricker, B.H., 2005. Potential determinants of drug-drug interaction associated dispensing in community pharmacies. Drug safety, 28, pp.371-378. Bedell, S.E., Graboys, T.B., Bedell, E. and Lown, B., 2004. Words that harm, words that heal. Archives of internal medicine, 164(13), pp.1365-1368. Beech, J. and Baird, B., 2020. GP funding and contracts explained. The King’s Fund, pp.1-8. Bell, E., Bryman, A. and Harley, B., 2022. Business research methods. Oxford university press. Bergen, N. and Labonté, R., 2020. “Everything is perfect, and we have no problems”: detecting and limiting social desirability bias in qualitative research. Qualitative health research, 30(5), pp.783-792. Bertalanffy, L.V., 1968. General system theory: Foundations, development, applications. G. Braziller. Bharmal A., Morgan T., Kuhn I., Wee B., Barclay S., 2022. Palliative and end-of-life care and junior doctors: a systematic review and narrative synthesis. BMJ Supportive & Palliative Care, pp.862-868. Birks, M. and Mills, J., 2014. Quality in qualitative research. Qualitative methodology: A practical guide, pp.221-236. Birt, L., Lane, K., Corner, J., Sanderson, K. and Bunn, D., 2023. Care‐home Nurses' responses to the COVID‐19 pandemic: Managing ethical conundrums at personal cost: A qualitative study. Journal of Nursing Scholarship, 55(1), pp.226- 238. Blackburn, S., McLachlan, S., Jowett, S., Kinghorn, P., Gill, P., Higginbottom, A., Rhodes, C., Stevenson, F. and Jinks, C., 2018. The extent, quality and impact of patient and public involvement in primary care research: a mixed methods study. Research involvement and engagement, 4(1), pp.1-18. Bleeker, S.E., Derksen-Lubsen, G., van Ginneken, A.M., Van Der Lei, J. and Moll, H.A., 2006. Structured data entry for narrative data in a broad specialty: patient history and physical examination in pediatrics. BMC medical informatics and decision making, 6, pp.1-7. Bone, A.E., Gomes, B., Etkind, S.N., Verne, J., Murtagh, F.E., Evans, C.J. and Higginson, I.J., 2018. What is the impact of population ageing on the future provision of end-of-life care? Population-based projections of place of death. Palliative medicine, 32(2), pp.329-336. Bowers, B. and Redsell, S.A., 2017. A qualitative study of community nurses’ decision‐making around the anticipatory prescribing of end‐of‐life medications. Journal of Advanced Nursing, 73(10), pp.2385-2394. Bowers, B., 2022. Understanding community end of life anticipatory medication care. Bowers, B., Barclay, S., Pollock, K., Barclay, S., 2018. GPs’ decisions about prescribing end-of-life anticipatory medications: a qualitative study. Br J Gen Pract, 70, pp.731-739. Bowers, B., Howard, P., Madden, B., Pollock, K. and Barclay, S., 2023. Is end-of-life anticipatory prescribing always enough?. bmj, 381. Bowers, B., Pollock, K. and Barclay, S., 2022. Unwelcome memento mori or best clinical practice? Community end of life anticipatory medication prescribing practice: A mixed methods observational study. Palliative Medicine, 36(1), pp.95-104. Bowers, B., Ryan, R., Kuhn, I. and Barclay, S., 2019. Anticipatory prescribing of injectable medications for adults at the end of life in the community: a systematic literature review and narrative synthesis. Palliative medicine, 33(2), pp.160- 177. Boyatzis, R.E., 1998. Transforming qualitative information: Thematic analysis and code development. sage. Bradbury, M., 1996. Representations of ‘good’and ‘bad’death among deathworkers and the bereaved. Contemporary issues in the sociology of death, dying and disposal, pp.84-95. Braillon, A., 2021. “Poor Medication Recall”: A Euphemism for a Broken System?. Academic Emergency Medicine, 28(2), pp.271-271. Braun, V. and Clarke, V., 2020. One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Research in Psychology. Braun, V. and Clarke, V., 2021. To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales, Qualitative Research in Sport, Exercise and Health. Routledge, pp. 201–216. Braun, V. and Clarke, V., 2022. Thematic Analysis: A practical guide. London: Sage Publications Ltd. Braun, V. and Clarke, V., 2006. Using thematic analysis in psychology. Qualitative research in psychology, 3(2), pp.77-101. Braun, V. and Clarke, V., 2019. Reflecting on reflexive thematic analysis. Qualitative research in sport, exercise and health, 11(4), 589-597. Braun, V., and Clarke, V., 2021. Can I use TA? Should I use TA? Should I not use TA? Comparing reflexive thematic analysis and other pattern‐based qualitative analytic approaches. Counselling and psychotherapy research, 21(1), pp.37-47. Braun, V., Clarke, V. & Weate, P., 2016. Using thematic analysis in sport and exercise research. Routledge handbook of qualitative research in sport and exercise, Routledge, pp. 191-205. Braun, V., Clarke, V., Boulton, E., Davey, L. and McEvoy, C., 2021. The online survey as a qualitative research tool. International journal of social research methodology, 24(6), pp.641-654. 243 Breen, L., Zaman, H., McCulloch, E. and Isaq, S., 2018. DOOP kit, domestic bin or watery Grave? A study investigating disposal practices of transdermal drug delivery products in care homes. British Sociological Association, 2017. Statement of Ethical Practice. Belmont. Bryant, A. and Charmaz, K. eds., 2007. The Sage handbook of grounded theory. Sage. Bryman, A., 2016. Social Research Methods (5th edition, 2016). Oxford: Oxford University Press. Bunting, M., 2020. Labours of Love: the crisis of care. Granta Books. Burns, M. and McIlfatrick, S., 2015. Nurses' knowledge and attitudes towards pain assessment for people with dementia in a nursing home setting. International Journal of Palliative Nursing, 21(10), pp.479-487. Burton, J., 2016. Liberating Institutions. Bynum, W.F., 2013. Nosology. In Companion encyclopedia of the history of medicine, pp. 361-382. CQC, 2023. Regulatory assessment framework. www.cqc.org.uk/assessment/quality-statements/responsive Cameron, K.S. and Quinn, R.E., 1988. Organizational paradox and transformation. Care Quality Commission, 2012. The state of health care and adult social care in England in 2011/12 (Vol. 763). The Stationery Office. Care Quality Commission, 2019. The safer management of controlled drugs. https://www.cqc.org.uk/publications/major- report/safer-management-controlled-drugs Care Quality Commission, 2020. The state of health care and adult social care in England 2019/20. Care Quality Commission, 2022. State of Care 2021/22: Workforce. Care Quality Commission, 2022b. Controlled drugs in home care. https://www.cqc.org.uk/guidance-providers/adult-social- care/controlled-drugs-home- care?fbclid=IwAR2w21RLI2_j3usa5L2hrYx3BRCgCQlJRsXh7y0O75qkmvBMSve5cRxoo3o#:~:text=Staff%20m ust%20make%20sure%20ordering,you%20are%20transporting%20controlled%20drugs Carello, J. and Butler, L. D., 2014. Practicing what we teach: Trauma-informed educational practice. Journal of Teaching in Social Work, 34(3), pp.262-278. Carney, J. and MacRobbie, A., 2008. Analysis of ‘Just in Case’anticipatory prescribing pilot. Final Report. Palliative Care Network and NHS Highland, Inverness. Chambers, N. and Tyrer, J., 2002. Long-term care. Unknown quantities. The Health Service Journal, 112(5835), pp.28-29. Chambers, N. and Tyrer, J., 2003. What kind of leaders do we need in our care homes?. Nursing And Residential Care, 5(8), pp.389-391. Chapman, S., 2005. Logic and conversation. In Paul Grice, Philosopher and Linguist, pp. 85-113. Charity Commission for England and Wales, 2012. National Skills Academy for Social Care. Chu, C.H., Ronquillo, C., Khan, S., Hung, L. and Boscart, V., 2022. Technology Recommendations to Support: Person- Centered Care in Long-Term Care Homes during the COVID-19 Pandemic and Beyond. In The COVID-19 Pandemic and Older Adults, pp. 247-262. Chung, J. and Monroe, G.S., 2003. Exploring social desirability bias. Journal of Business Ethics, 44, pp.291-302. Cialdini, R.B., Trost, M.R. and Newsom, J.T., 1995. Preference for consistency: The development of a valid measure and the discovery of surprising behavioral implications. Journal of personality and social psychology, 69(2), p.318. Clark, D., 2007. From margins to centre: a review of the history of palliative care in cancer. The lancet oncology, 8(5), pp.430-438. Clarke, V. and Braun, V., 2013. Successful qualitative research: A practical guide for beginners. Successful qualitative research, pp.1-400. Clarke, V. and Braun, V., 2021. Thematic analysis: a practical guide. Thematic Analysis, pp.1-100. Clegg, S. R., 2002. General introduction, S. R. Clegg (Ed.), Management and organization paradoxes. Amsterdam: John Benjamins B.V., pp.1–10. Clegg, S.R., da Cunha, J.V. and da Cunha, M.P., 2002. Management paradoxes: A relational view. Human relations, 55(5), pp.483-503. Collingridge Moore, D., Payne, S., Van den Block, L., Ten Koppel, M., Szczerbińska, K. and Froggatt, K., 2019. Research, recruitment and observational data collection in care homes: lessons from the PACE study. BMC research notes, 12(1), pp.1-6. Collis, E. and Al-Qurainy, R., 2013. Care of the dying patient in the community. Bmj, pp. 347. Comas-Herrera, A., Zalakaín, J., Lemmon, E., Henderson, D., Litwin, C., Hsu, A.T., Schmidt, A.E., Arling, G., Kruse, F. and Fernández, J.L., 2020. Mortality associated with COVID-19 in care homes: international evidence. Article in LTCcovid. org, international long-term care policy network, CPEC-LSE, 14. Comondore, V.R., Devereaux, P.J., Zhou, Q., Stone, S.B., Busse, J.W., Ravindran, N.C., Burns, K.E., Haines, T., Stringer, B., Cook, D.J. and Walter, S.D., 2009. Quality of care in for-profit and not-for-profit nursing homes: systematic review and meta-analysis. Bmj, 339. Competition and Markets Authority, 2022. Competition and Markets Authority Annual Report and Accounts 2021/22. Connor, S.R. and Bermedo, M.C.S. eds., 2017. Global atlas of palliative care at the end of life. Worldwide Palliative Care Alliance. 244 Cooper, E., Spilsbury, K., McCaughan, D., Thompson, C., Butterworth, T. and Hanratty, B., 2017. Priorities for the professional development of registered nurses in nursing homes: a Delphi study. Age and Ageing, 46(1), pp.39-45. Cope, D.G., 2014. Methods and meanings: credibility and trustworthiness of qualitative research. Oncology nursing forum. Corbett, A., Nunez, K.M., Smeaton, E., Testad, I., Thomas, A.J., Closs, S.J., Briggs, M., Clifton, L., Gjestsen, M.T. and Lawrence, V., 2016. The landscape of pain management in people with dementia living in care homes: a mixed methods study. International journal of geriatric psychiatry, 31(12), pp.1354-1370. Covington, M., 2013. End-of-life care: implementing the Gold Standards Framework. Nursing & Residential Care, 15(3), pp.146-149. Creswell, 2013. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. Crossan, F., 2003. Research philosophy: towards an understanding. Nurse Researcher (through 2013), 11(1), pp.46. Cunha, M.P.E., Simpson, A.V., Clegg, S.R. and Rego, A., 2019. Speak! Paradoxical effects of a managerial culture of ‘speaking up’. British Journal of Management, 30(4), pp.829-846. Cunningham, C., McClean, W. and Kelly, F., 2010. The assessment and management of pain in people with dementia in care homes. Nursing Older People, 22(7). Davies, S.L., Goodman, C., Bunn, F., Victor, C., Dickinson, A., Iliffe, S., Gage, H., Martin, W. and Froggatt, K., 2011. A systematic review of integrated working between care homes and health care services. BMC health services research, 11(1), pp.1-21 DeJonckheere, M. and Vaughn, L.M., 2019. Semistructured interviewing in primary care research: a balance of relationship and rigour. Family medicine and community health, 7(2). Dennis, A. and Martin, P.J., 2007. Symbolic interactionism and the concept of social structure. Sociological Focus, 40(3), pp.287-305. Denzin, N.K., 2001. Interpretive interactionism (Vol. 16). Sage. Department of Health and Social Care, 2016. Cutting Red Tape: review of adult social care – residential and nursing home sector. Department of Health and Social Care, 2020. Guidance on care home visiting during the COVID-19 pandemic. Department of Health and Social Care, 2020a. COVID-19: Novel coronavirus (COVID-19) standard operating procedure Running a medicines re-use scheme in a care home or hospice setting. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/881838/medicine s-reuse-in-care-homes.pdf Department of Health and Social Care, 2020b. Busting bureaucracy: empowering frontline staff by reducing excess bureaucracy in the Department of Health and Social Care, 2022. Health and Care Act 2022 (c. 31): Government Bill. https://bills.parliament.uk/bills/3022 Department of Health and Social Care, U.K, 2001. The Misuse of Drugs Regulations. Department of Health, 2004. Delivering the out of hours review: Securing proper access to medicines in the out of hours period. Department of Health, 2008. End-of-life care strategy: promoting high quality care for adults at the end of their life. Department of Health, U.K., 2007. Safer management of controlled drugs: A guide to good practice in secondary care (England). Dewar, B., Barrie, K., Sharp, C. and Meyer, J., 2019. Implementation of a complex intervention to support leadership development in nursing homes: a multimethod participatory study. Journal of Applied Gerontology, 38(7), pp.931- 958. Dickman, A. and Schneider, J., 2016. The syringe driver: continuous subcutaneous infusions in palliative care. Oxford University Press. Dickson-Swift, V., James, E. L., Kippen, S. and Liamputtong, P. (2007). Doing sensitive research: What challenges do qualitative researchers face? Qualitative Research, 7(3), 327-353. Dickson-Swift, V., James, E. L., Kippen, S. and Liamputtong, P. (2009). Researching sensitive topics: Qualitative research as emotion work. Qualitative Research, 9(1), 61-79. Dimon, C., 2005. The challenging role of the care home manager. Nursing And Residential Care, 7(12), pp.571-572. Dingwall, R. and Wilson, E., 1995. Is pharmacy really an 'incomplete profession’? Dixon-Woods. Agarwal S., Jones D., et al., 2005. Synthesising qualitative and quantitative evidence: a review of possible methods. Journal of Health Services Research & Policy, 10, pp.45–53. Dodds, S. and Hess, A.C., 2020. Adapting research methodology during COVID-19: lessons for transformative service research. Journal of Service Management, 32(2), pp.203-217. Donaldson, M.S., Corrigan, J.M. and Kohn, L.T. eds., 2000. To err is human: Building a safer health system. Donnelly, L., 2013. Liverpool Care Pathway rebranded, not axed. The Telegraph. Donyai, P., McCrindle, R., Hui, T.K. and Sherratt, R.S., 2021. Stakeholder Views on the Idea of Medicines Reuse in the UK. Pharmacy, 9(2), p.85. Doty, P., 1986. Family care of the elderly: The role of public policy. The Milbank Quarterly, pp.34-75. Dowling, E., 2022. The care crisis: What caused it and how can we end it? Verso Books. 245 Dworkin, R., 1986. Law’s Empire. Harper Collins. Dyer, C., 2018. Gosport hospital: GP prescribed opioids that shortened at least 456 lives. EU, 2018. General data protection regulation (GDPR) compliance guidelines. Eaton, G., Wong, G., Tierney, S., Roberts, N., Williams, V. and Mahtani, K.R., 2021. Understanding the role of the paramedic in primary care: a realist review. BMC Medicine, 19(1), pp.1-14. Ebert, L., Hoffman, K., Levett-Jones, T. and Gilligan, C., 2014. “They have no idea of what we do or what we know”: Australian graduates' perceptions of working in a health care team. Nurse Education in Practice, 14(5), pp.544-550. Eli, K., Hawkes, C.A., Fritz, Z., Griffin, J., Huxley, C.J., Perkins, G.D., Wilkinson, A., Griffiths, F. and Slowther, A.M., 2021. Assessing the quality of ReSPECT documentation using an accountability for reasonableness framework. Resuscitation Plus, 7, p.100145. Elliott, R. and Timulak, L., 2005. Descriptive and interpretive approaches to qualitative research. A handbook of research methods for clinical and health psychology, 1(7), pp.147-159. Ellis, P. and Abbott, J., 2016. Preparing for CQC inspection: understanding the five key questions. Journal of Renal Nursing, 8(1), pp.42-44. Elmir, R., Schmied, V., Jackson, D. and Wilkes, L., 2011. Interviewing people about potentially sensitive topics. Nurse Researcher, 19(1). Emanuel, E.J. and Emanuel, L.L., 1998. The promise of a good death. The Lancet, 351, pp.SII21-SII29. England, N.H.S. and Improvement, N.H.S., 2016. The framework for enhanced health in care homes. NHS England. England, N.H.S. and Improvement, N.H.S., 2020. The Framework for Enhanced Health in Care Homes: Version 2. London: NHS England. Erueti, C., Glasziou, P., Mar, C.D. and Van Driel, M.L., 2012. Do you think it's a disease? A survey of medical students. BMC Medical Education, 12(1), pp.1-4. Esterhuizen, T.M., Thabane, L., 2016. Con: Meta-analysis: Some key limitations and potential solutions. Nephrology Dialysis Transplantation, 31(6), pp.882-885. Etkind, S.N. and Koffman, J., 2016. Approaches to managing uncertainty in people with life-limiting conditions: role of communication and palliative care. Postgraduate Medical Journal, 92(1089), pp.412-417. Farjoun, M., 2010. Beyond dualism: Stability and change as a duality. Academy of Management Review, 35(2), 202-225. Farooq, M.B. and De Villiers, C., 2017. Telephonic qualitative research interviews: When to consider them and how to do them. Meditari Accountancy Research, 25(2), pp.291-316. Farrant, W., 1991. Addressing the contradictions: health promotion and community health action in the United Kingdom. International Journal of Health Services, 21(3), pp.423-439. Faugier, J. and Sargeant, M., 1997. Sampling hard to reach populations. Journal of Advanced Nursing, 26(4), pp.790-797. Ferrell, B.R., Twaddle, M.L., Melnick, A. and Meier, D.E., 2021. National Consensus Project Clinical Practice Guidelines for Quality Palliative Care Guidelines, 5th edition. Journal of Palliative Medicine Faull, C., Windridge, K., Ockleford, E. and Hudson, M., 2013. Anticipatory prescribing in terminal care at home: what challenges do community health professionals encounter? BMJ Supportive & Palliative Care, 3(1), pp.91-97. Finlay, L., 2002. “Outing” the researcher: The provenance, process, and practice of reflexivity. Qualitative Health Research, 12(4), pp.531-545. Finlay, L., 2021. Thematic Analysis:: The ‘Good’, the ‘Bad’and the ‘Ugly’. European Journal for Qualitative Research in Psychotherapy, 11, pp.103-116. Finucane, A.M., Stevenson, B., Gardner, H., McArthur, D. and Murray, S.A., 2014. Anticipatory prescribing at the end of life in Lothian care homes. British Journal of Community Nursing, 19(11), pp.544-547. Foucault, M., 1973. The Birth of the Clinic: An Archaend-of-lifeogy of Medical Perception. 1963. Trans. AM Sheridan Smith. New York: Vintage. Francis, R., 2013. Report of the Mid Staffordshire NHS Foundation Trust public inquiry: executive summary (Vol. 947). The Stationery Office. Freidson, E., 1988. Professional powers: A study of the institutionalization of formal knowledge. University of Chicago Press. Froggatt, K., Hockley, J., Parker, D. and Brazil, K., 2011. A system lifeworld perspective on dying in long term care settings for older people: Contested states in contested places. Health & Place, 17(1), pp.263-268. Fusch, P.I. and Ness, L.R., 2015. Are we there yet? Data saturation in qualitative research. GMC, 2021. Research ethics code of conduct. Gaim, M., Clegg, S., e Cunha, M.P. and Berti, M., 2022. Organizational paradox. Cambridge University Press. Gale, N.K., Heath, G., Cameron, E., Rashid, S. and Redwood, S., 2013. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC medical research methodology, 13, pp.1-8. Gardiner, C., Gott, M., Ingleton, C., Hughes, P., Winslow, M. and Bennett, M.I., 2012. Attitudes of health care professionals to opioid prescribing in end-of-life care: a qualitative focus group study. Journal of Pain and Symptom Management, 44(2), pp.206-214. Gaskell, G., 2004. Science policy and society: the British debate over GM agriculture. Current Opinion in Biotechnology, 15(3), pp.241-245. Geddes, 2017. The Overworked Entrepreneur. 246 General Medical Council, 2021. Research Ethics Code of Conduct: Good practice in research. Ghosh, A., Dzeng, E. and Cheng, M.J., 2015. Interaction of palliative care and primary care. Clinics in Geriatric Medicine, 31(2), pp.207-218. Giddens, A., 1984. The Constitution of Society. Giebel, C., Hanna, K., Cannon, J., Eley, R., Tetlow, H., Gaughan, A., Komuravelli, A., Shenton, J., Rogers, C., Butchard, S. and Callaghan, S., 2020. Decision-making for receiving paid home care for dementia in the time of COVID-19: a qualitative study. BMC Geriatrics, 20, pp.1-8. Giebel, Cannon, J., Hanna, K., Butchard, S., Eley, R., Gaughan, A., Komuravelli, A., Shenton, J., Callaghan, S., Tetlow, H., Limbert, S., Whittington, R., Rogers, C., Rajagopal, M., Ward, K., Shaw, L., Corcoran, R., Bennett, K., & Gabbay, M., 2021. Impact of COVID-19 related social support service closures on people with dementia and unpaid carers: a qualitative study. Aging & Mental Health, 25(7), pp.1281–1288. Gilmore-Bykovskyi, A.L. and Bowers, B.J., 2013. Understanding nurses’ decisions to treat pain in nursing home residents with dementia. Research in Gerontological Nursing, 6(2), pp.127-138. Glasgow, N. and Doebler, S., 2020. Policy and program challenges in delivering health and social care services to rural older people. Rural Gerontology, pp.117-128. Goddard, C., Stewart, F., Thompson, G. and Hall, S., 2013. Providing end-of-life care in care homes for older people: a qualitative study of the views of care home staff and community nurses. Journal of Applied Gerontology, 32(1), pp.76-95. Goffman, E., 1974. Frame analysis: An essay on the organization of experience. Harvard University Press. Goldenberg, 2006. On evidence and evidence-based medicine: Lessons from the philosophy of science. Social Science & Medicine (1982), 62(11), pp.2621–2632. Gómez-Batiste, X., Caja, C., Espinosa, J., Bullich, I. and Martínez-Muñoz, M., 2020. The Catalonia WHO demonstration project of palliative care implementation: quantitative and qualitative results at 20 years. Journal of Pain and Symptom Management Goodman, C., Dening, T., Gordon, A.L., Davies, S.L., Meyer, J., Martin, F.C., Gladman, J.R., Bowman, C., Victor, C., Handley, M. and Gage, H., 2016. Effective health care for older people living and dying in care homes: a realist review. BMC Health Services Research, 16(1), pp.1-14. Goodman, C., Woolley, R. and Knight, D., 2003. District nurse involvement in providing palliative care to older people in residential care homes. International Journal of Palliative Nursing, 9(12), pp.521-527. Gordon, A.L., Franklin, M., Bradshaw, L., Logan, P., Elliott, R. and Gladman, J.R., 2014. Health status of UK care home residents: a cohort study. Age and Ageing, 43(1), pp.97-103. Gosport Independent Panel, 2018. Gosport War Memorial Hospital: The report of the Gosport Independent Panel. Gough, D., 2007. Weight of evidence: a framework for the appraisal of the quality and relevance of evidence. Research Papers in Education, 22(2), pp.213-228. Grady, C., 2005. Payment of clinical research subjects. The Journal of Clinical Investigation, 115(7), pp.1681-1687. Graeber, D., 2015. The utopia of rules: On technology, stupidity, and the secret joys of bureaucracy. Melville House. Gray, D.E., 2014. Doing research in the real world. Doing research in the real world, pp.1-100. Gray, J.N., 2005. Notes on data base operating systems. Operating systems: An advanced course, pp.393-481. Green, A., Bramley, G., Annibal, I. and Sellick, J., 2018. Rural workforce issues in health and care. Greenberg, N., Docherty, M., Gnanapragasam, S., & Wessely, S., 2020. Managing mental health challenges faced by healthcare workers during the COVID-19 pandemic. The British Journal of Psychiatry, 217(3), pp.547-556. Grice, HP, 2004. Logic and conversation. London: Palgrave Macmillan UK. Griffiths, P. et al., 1993. Reaching hidden populations of drug users by privileged access interviewers: methodological and practical issues, Addiction, 88(12), pp.1617–1626. Griggs, C., 2010. Community nurses’ perceptions of a good death: a qualitative exploratory study. International journal of palliative nursing, 16(3), pp.140-149. Guba, E. G. and Lincoln, Y. S., 1994. Competing paradigms in qualitative research. Handbook of qualitative research, 2(163- 194), p.105. Guest, G., Bunce, A. and Johnson, L., 2006. How many interviews are enough? An experiment with data saturation and variability. Field methods, 18(1), pp.59-82. Guyatt, G., Cairns, J., Churchill, D., Cook, D., Haynes, B., Hirsh, J., Irvine, J., Levine, M., Levine, M., Nishikawa, J. and Sackett, D., 1992. Evidence-based medicine: a new approach to teaching the practice of medicine. Jama, 268(17), pp.2420-2425. Gysels, M., Evans, C.J., Lewis, P., Speck, P., Benalia, H., Preston, N.J., Grande, G.E., Short, V., Owen-Jones, E., Todd, C.J. and Higginson, I.J., 2013. MORECare research methods guidance development: recommendations for ethical issues in palliative and end-of-life care research. Palliative medicine, 27(10), pp.908-917. Habermas, J., 1992. Between facts and norms: Contributions to a discourse theory of law and democracy. Hall, J., Gaved, M. and Sargent, J., 2021. Participatory research approaches in times of Covid-19: a narrative literature review. International Journal of Qualitative Methods, 20, p.16094069211010087. 247 Hall, S., Davies, J.M., Gao, W. and Higginson, I.J., 2014. Patterns of dignity-related distress at the end of life: a cross- sectional study of patients with advanced cancer and care home residents. Palliative medicine, 28(9), pp.1118-1127. Hall, S., Goddard, C., Stewart, F. and Higginson, I.J., 2011. Implementing a quality improvement programme in palliative care in care homes: a qualitative study. BMC geriatrics, 11(1), pp.1-11. Hammersley, M. and Traianou, A., 2012. Ethics in qualitative research: Controversies and contexts. Sage. Hammond, J., Mason, T. and Sutton, M., 2018. Exploring the impacts of the 2012 Health and Social Care Act reforms to commissioning on clinical activity in the English NHS: a mixed methods study of cervical screening. BMJ Open 2019; 9: e024156. 10.1136/bmjopen-2018-024156. Health Policy, 122, pp.457-472. Handley, M., Goodman, C., Froggatt, K., Mathie, E., Gage, H., Manthorpe, J., Barclay, S., Crang, C. and Iliffe, S., 2014. Living and dying: responsibility for end‐of‐life care in care homes without on‐site nursing provision‐a prospective study. Health & social care in the community, 22(1), pp.22-29. Hardill, I. and Dwyer, P., 2011. Delivering public services in the mixed economy of welfare: perspectives from the voluntary and community sector in rural England. Journal of Social Policy, 40(1), pp.157-172. Harding, R., Defilippi, K. and Cameron, D., 2019. What palliative care-related problems do patients experience at HIV diagnosis? A systematic review of the evidence. Journal of Pain and Symptom Management, Harrington, C., Jacobsen, F.F., Panos, J., Pollock, A., Sutaria, S. and Szebehely, M., 2017. Marketization in long-term care: A cross-country comparison of large for-profit nursing home chains. Health services insights, 10, p.1178632917710533. Harrison, S. and Ahmad, W.I., 2000. Medical autonomy and the UK state 1975 to 2025. Sociology, 34(1), pp.129-146. Health Education England, 2012. Allied Health Professionals: Paramedic. Health Service, 2010. The route to success in end-of-life care; achieving quality in care homes. Healthcare Improvement Scotland, 2014. Scottish palliative care guidelines: anticipatory prescribing. Hendricks, V.M., Blanken, P. and Adriaans, N.F., 1992. Snowball sampling: Theoretical and practical considerations. Rotterdam: IVO. Hennink, M.M., Kaiser, B.N. and Weber, M.B., 2019. What influences saturation? Estimating sample sizes in focus group research. Qualitative health research, 29(10), pp.1483-1496. Hensen, B., Mackworth-Young, C.R.S., Simwinga, M., Abdelmagid, N., Banda, J., Mavodza, C., Doyle, A.M., Bonell, C. and Weiss, H.A., 2021. Remote data collection for public health research in a COVID-19 era: ethical implications, challenges and opportunities. Health policy and planning, 36(3), pp.360-368. Higginson, I.J., 2016. Research challenges in palliative and end-of-life care. BMJ supportive & palliative care, 6(1), pp.2-4. Pham, T. and Williamson, K., 2018, December. A two-way street: collaboration and information sharing in academia. A theoretically-based, comparative Australian/Vietnamese study. In ISIC-The Information Behaviour Conference 2018. University of Boras. Hofmann, B., 2001. The paradox of health care. Health Care Analysis, pp.369-386. Homa, L., Rose, J., Hovmand, P.S., Cherng, S.T., Riolo, R.L., Kraus, A., Biswas, A., Burgess, K., Aungst, H., Stange, K.C. and Brown, K., 2015. A participatory model of the paradox of primary care. The Annals of Family Medicine, 13(5), pp.456-465. Hossain, F. and Clatty, A., 2021. Self-care strategies in response to nurses’ moral injury during COVID-19 pandemic. Nursing Ethics, 28(1), 23–32. https://doi.org/10.1177/0969733020961825 Hunter, K., 1996. “Don't Think Zebras”: Uncertainty, Interpretation, and the Place of Paradox in Clinical Education. Theor Med 17, pp.225-241. Ibrahim, A.S.H., Barry, H.E. and Hughes, C.M., 2023. GPs’ and pharmacists’ views of integrating pharmacists into general practices: a qualitative study. British Journal of General Practice, 73(731), pp.e407-e417. Ingram, A.E., Lewis, M.W., Barton, S. and Gartner, W.B., 2016. Paradoxes and innovation in family firms: The role of paradoxical thinking. Entrepreneurship Theory and Practice, 40(1), pp.161-176. Jacob, R.G. and Moore, D.J., 1984. Paradoxical Interventions in Behavioural Medicine. J Behav Ther Exp Psychiatry 15, pp.205-213. Jacoby, S.M., 2004. Employing bureaucracy: Managers, unions, and the transformation of work in the 20th century. Psychology Press. Jarzabkowski, P., Lê, J. K. and Van de Ven, A. H., 2013. Responding to competing strategic demands: How organizing, belonging, and performing paradoxes coevolve. Strategic Organization, 11(3), 245-280. Jenkins, C., Smythe, A., Galant-Miecznikowska, M., Bentham, P. and Oyebode, J., 2016. Overcoming challenges of conducting research in nursing homes. Nursing older people, 28(5). Johns, M., Lockwood, R., Longlands, S. and Qureshi, A., 2020. Institute for Public Policy Research. Johnson, S. and Rasulova, S., 2017. Qualitative research and the evaluation of development impact: incorporating authenticity into the assessment of rigour. Journal of Development Effectiveness, 9(2), pp.263-276. Johnstone, L., 2017. Facilitating anticipatory prescribing in end-of-life care. Pharm J, 298(7901). Jones, R., 2016. Fundamental flaws. British Journal of General Practice, 66(647), pp.283-283. Josselson, R., 2013. Interviewing for qualitative inquiry: A relational approach. Guilford Press. Julia Neuberger, Dying Well: A Guide to Enabling a Good Death, Radcliffe, Oxford, 2004. 248 Kafka, B., 2020. The demon of writing: Powers and failures of paperwork. Mit Press. Katz, J., Komaromy, C. and Sidell, M., 1999. Understanding palliative care in residential and nursing homes. International Journal of Palliative Nursing, 5(2), pp.58-64. Kennedy, P., 2015. The contradictions of capitalist healthcare systems. Critique, 43(2), pp.211-231. Killett, A., Burns, D., Kelly, F., Brooker, D., Bowes, A., La Fontaine, J., Latham, I., Wilson, M. and O'NEILL, M.A.R.T.I.N., 2016. Digging deep: how organisational culture affects care home residents' experiences. Ageing & Society, 36(1), pp.160-188. King, N., 2012. ‘Doing Template Analysis’, in G. Symon and C. Cassell (eds) Qualitative Organizational Research: Core Methods and Current Challenges. London: Sage Publications, pp.426–450. Kingston, A., Robinson, L., Booth, H., Knapp, M., Jagger, C. and MODEM project, 2018. Projections of multi-morbidity in the older population in England to 2035: estimates from the Population Ageing and Care Simulation (PACSim) model. Age and ageing, 47(3), pp.374-380. King’s Fund, 2019. The Health Care Workforce in England: Make or Break. Kinley, J. and Hockley, J., 2010. A baseline review of medication provided to older people in nursing care homes in the last month of life. International journal of palliative nursing, 16(5), pp.216-223. Kinley, J., Hockley, J. and Stone, L., 2016. Achieving symptom control not medication burden at the end of life, Abstract No. 179.30: NP331-NP332. Palliat Med. Kinley, J., Hockley, J.O., Stone, L., Dewey, M., Hansford, P., Stewart, R., McCrone, P., Begum, A. and Sykes, N., 2014. The provision of care for residents dying in UK nursing care homes. Age and Ageing, 43(3), pp.375-379. Kitwood, T., 1997. Dementia reconsidered. Buckingham. Open University Press. Kydd A (2000) Eduction and training in dementia care. Community Nurse, 5(12), pp.15-16. Klein, R., 2010. The new politics of the NHS: from creation to reinvention. Radcliffe Publishing. Kleinman, A., 1988. Illness Narratives: Suffering, Healing, and the Human Condition. Knights, D., 2016. In defence of the ‘tick-box approach’: why end-of-life care is no exception. British Journal of General Practice, 66(647), pp.290-291. Knights, D., Wood, D. and Barclay, S., 2013. The Liverpool Care Pathway for the dying: what went wrong?. British Journal of General Practice, 63(615), pp.509-510. Koestler, A., 1964. The act of creation. Kuhn, T. S., 1996. The Structure of Scientific Revolutions. Chicago, IL: The University of Chicago Press. Kuhn, T., 2015. The politics of buzzwords: Deployed in policy battles in global development. Third World Quarterly, 36(3), 463-480. Kuhn, T., 1962. The structure of scientific revolutions. University of Chicago Press. (APA 7). Kvale, S. and Brinkmann, S., 2009. Interviews: Learning the craft of qualitative research interviewing. sage. Latter, S., Campling, N., Birtwistle, J., Richardson, A., Bennett, M.I., Meads, D., Blenkinsopp, A., Breen, L., Edwards, Z., Sloan, C. and Miller, E., 2022. Patient and carer access to medicines at end of life: the ActMed mixed-methods study. Lavi, S., 2004. Autopoiesis, Nihilism and Technique: On Death and the Origins of Legal Paradoxes. Lawton, J., 2002. The dying process: patients' experiences of palliative care. Routledge. Lawton, S., Denholm, M., Macaulay, L., Grant, E. and Davie, A., 2012. Timely symptom management at end of life using ‘just in case’boxes. British Journal of Community Nursing, 17(4), pp.182-190. LeGreco, M. and Tracy, S.J., 2009. Discourse tracing as qualitative practice. Qualitative Inquiry, 15(9), pp.1516-1543. Lewis, M. W., 2000. Exploring paradox: Toward a more comprehensive guide. Academy of management review, 25(4), 760- 776. Lewis, P., Saunders, M.N. and Thornhill, A., 2019. Research methods for business students (8th útg.). Li, N., Romera Rodriguez, G., Xu, Y., Bhatt, P., Nguyen, H.A., Serpi, A., Tsai, C. and Carroll, J.M., 2022, June. Picturing one's self: Camera use in Zoom classes during the COVID-19 pandemic. Proceedings of the Ninth ACM Conference on Learning@ Scale, pp.151-162. Lievesley, N., Crosby, G., Bowman, C. and Midwinter, E., 2011. The changing role of care homes. London: BUPA and Centre for Policy on Ageing. Lim, R.H., Anderson, J.E. and Buckle, P.W., 2016. Work domain analysis for understanding medication safety in care homes in England: an exploratory study. Ergonomics, 59(1), pp.15-26. Lincoln, Y.S. and Guba, E.G., 1985. Naturalistic Inquiry. Beverley Hill, CA: Sage Publications Ltd. Lobe, B., Morgan, D. and Hoffman, K.A., 2020. Qualitative data collection in an era of social distancing. International journal of qualitative methods, 19, p.1609406920937875. Lovis, C., Baud, R.H. and Planche, P., 2000. Power of expression in the electronic patient record: structured data or narrative text? International Journal of Medical Informatics, pp.16-32. Lundin, E. and Godskesen, T.E., 2021. End-of-life care for people with advanced dementia and pain: a qualitative study in Swedish nursing homes. BMC nursing, 20, pp.1-11. Lynch, M., 2017. The Medical Handbook: Isle of Wight NHS Trust. 249 Lynn, J. and Adamson, D.M., 2003. Living well at the end of life. Adapting health care to serious chronic illness in old age. Rand Health White Paper WP-137. Mackenzie & Knipe, S., 2006. Research dilemmas: paradigms, methods and methodology. Issues in Educational Research, 16(2), pp.193–205. Magnani et al., 2005. Review of Sampling Hard-to-Reach and Hidden Populations for HIV Surveillance. Majumder, M., Bowers, B., Pollock, K., Goodman, C., Kuhn, I. and Barclay, S., 2022. End-of-life care in UK care homes– controlled drugs: systematic review and narrative synthesis. BMJ Supportive & Palliative Care, 12(3), pp.253-261. Malley, J., Holder, J., Dodgson, R. and Booth, S., 2014. Regulating the quality and safety of long-term care in England. Regulating long-term care quality: An international comparison, 180, p.210. Maslen, S., 2017. Layers of sense: the sensory work of diagnostic sensemaking in digital health. Digital health, 3, p.2055207617709101. Matosevic, T., Knapp, M., Kendall, J., Henderson, C. and Fernández, J.L., 2007. Care-home providers as professionals: understanding the motivations of care-home providers in England. Ageing & Society, 27(1), pp.103-126. Mays, N. and Pope, C., 1995. Qualitative research: rigour and qualitative research. Bmj, 311(6997), 109-112. McCann, M., O’Reilly, D. and Cardwell, C., 2009. A Census-based longitudinal study of variations in survival amongst residents of nursing and residential homes in Northern Ireland. Age and ageing, 38(6), pp.711-717. McCrory, V., 2019. An overview of the role of the district nurse caring for individuals with complex needs. British journal of community nursing, 24(1), pp.20-26. McDowell, L., 1998. Elites in the City of London: some methodological considerations. Environment and planning A, 30(12), pp.2133-2146. McGilton, K.S., Bowers, B.J., Heath, H., Shannon, K., Dellefield, M.E., Prentice, D., Siegel, E.O., Meyer, J., Chu, C.H., Ploeg, J. and Boscart, V.M., 2016. Recommendations from the international consortium on professional nursing practice in long-term care homes. Journal of the American Medical Directors Association, 17(2), pp.99-103. McIntyre, A., 2004. Doctrine of double effect. McKenna, K.Y. and Bargh, J.A., 2000. Plan 9 from cyberspace: The implications of the Internet for personality and social psychology. Personality and social psychology review, 4(1), pp.57-75. McLellan, E., MacQueen, K.M. and Neidig, J.L., 2003. Beyond the qualitative interview: Data preparation and transcription. Field methods, 15(1), pp.63-84. Mehta, A. and Chan, L.S., 2008. Understanding of the concept of" total pain": a prerequisite for pain control. Journal of Hospice & Palliative Nursing, 10(1), pp.26-32. Melnyk, B.M. and Fineout-Overholt, E., 2022. Evidence-based practice in nursing & healthcare: A guide to best practice. Lippincott Williams & Wilkins. Meyer, K. and Willis, R., 2019. Looking back to move forward: The value of reflexive journaling for novice researchers. Journal of gerontological social work, 62(5), pp.578-585. Miller, E.A., 2002. Telemedicine and doctor–patient communication: a theoretical framework for evaluation. Journal of Telemedicine and Telecare, 8(6), pp.311-318. Mills, C.W., 1940. Situated actions and vocabularies of motive. American sociological review, 5(6), pp.904-913. Ministry of Health., 2017. Te Ara Whakapiri: Principles and guidance for the last days of life. Miron-Spektor, E., Ingram, A., Keller, J., Smith, W.K. and Lewis, M.W., 2018. Microfoundations of organizational paradox: The problem is how we think about the problem. Academy of management journal, 61(1), pp.26-45. Mitchell S, Harrison M, Oliver P, Gardiner C, Chapman H, Khan D, Boyd K, Dale J, Barclay S, Mayland CR., 2022. Service change and innovation in community end-of-life care during the COVID-19 pandemic: Qualitative analysis of a nationwide primary care survey. Palliat Med. 6(1):161–70. Mitchell S, Oliver P, Gardiner C, Chapman H, Khan D, Boyd K, Dale J, Barclay S, Mayland C., 2021. Community end-of- life care during COVID-19: Findings of a UK primary care survey. BJGP Open. 5(4):BJGPO.2021.0095. Mitchell, S., Loew, J., Millington-Sanders, C. and Dale, J., 2016. Providing end-of-life care in general practice: findings of a national GP questionnaire survey. British Journal of General Practice, 66(650), pp.e647-e653. Mockford, C., Staniszewska, S., Griffiths, F. and Herron-Marx, S., 2012. The impact of patient and public involvement on UK NHS health care: a systematic review. International journal for quality in health care, 24(1), pp.28-38. Morgan, T., Bharmal, A., Duschinsky, R. and Barclay, S., 2020. Experiences of oldest-old caregivers whose partner is approaching end-of-life: A mixed-method systematic review and narrative synthesis. PLoS One, 15(6), p.e0232401. Morris, K. and Hockley, J., 2013. Implementing stock end-of-life medication in UK nursing homes. End of life Journal, 3(3). Morris, P., Phalippou, L. and Wu, B.H., 2022. How Deadly Is Financial Leverage? Evidence from Care Homes during the COVID-19 Crisis. Evidence from Care Homes during the COVID-19 Crisis (January 25, 2022). Morse, J.M., Barrett, M., Mayan, M., Olson, K. and Spiers, J., 2002. Verification strategies for establishing reliability and validity in qualitative research. International journal of qualitative methods, 1(2), pp.13-22. Mossialos, E., Naci, H. and Courtin, E., 2013. Expanding the role of community pharmacists: policymaking in the absence of policy-relevant evidence?. Health policy, 111(2), pp.135-148. Murray, S.A., Kendall, M., Boyd, K. and Sheikh, A., 2005. Illness trajectories and palliative care. Bmj, 330(7498), pp.1007- 1011. 250 NHS, H.V. Clinical Commissioning Group, 2022. When Required Medication: PRN Guidance for Care Homes. NHS England, 2008. The Health and Social Care Act 2008 (Regulated Activities). NHS England, 2013. The Controlled Drugs (Supervision of Management and Use) Regulations 2013 Single Operating Model. https://www.england.nhs.uk/wp-content/uploads/2013/11/som-cont-drugs.pdf NHS England, 2019. NHS Long Term Plan [Online]. https://www.longtermplan.nhs.uk/publication/nhs-long-term-plan/. NHS Wales, 2015. Risk of harm from the inappropriate use and disposal of fentanyl patches. http://www.patientsafety.wales.nhs.uk/sitesplus/documents/1104/PSN022%20Risk%20of%20harm%20from%20in appropriate%20use%20and%20disposal%20of%20fentanyl%20patches.pdf NHS, 2006. NHS End-of-life care Programme. www.endoflifecareforadults.nhs.uk. NHS, 2014. Five Year Forward View. https://www.england.nhs.uk/wp-content/uploads/2014/10/5yfv-web.pdf NHS, 2015. Pharmaceutical waste reduction in the NHS. https://www.england.nhs.uk/wp- content/uploads/2015/06/pharmaceutical-waste-reduction.pdf NHS, 2018. Medicines optimisation in care homes: Programme overview. https://www.england.nhs.uk/wp- content/uploads/2018/03/medicines-optimisation-in-care-homes-programme-overview.pdf NHS, 2020. COVID-19 social-distancing guidance. www.nhs.uk/conditions/covid-19/ NHS, ENHCCG, 2017. End-of-life care anticipatory prescribing guidance: East and North Hertfordshire Clinical Commissioning Group. NHS, North Yorkshire CCG, Valley of York CCG, 2021. Controlled Drugs Recording Guidelines. NHS, SWBH Trust, 2021. Anticipatory medication guidelines. NHS, WECCG, 2018. Primary care medication guidelines for anticipatory last days of life. NHSE, 2021. Ambitions for Palliative and End-of-life care: A National Framework. NICE, 2014. Managing medicines in care homes, Social care guideline [SC1]. NICE, 2018. Dementia: assessment, management and support for people living with dementia and their carers. NICE, 2020. COVID-19 Scenario: Palliative Care. https://cks.nice.org.uk/topics/coronavirus-covid- 19/management/palliative-care/ NIHR, 2020. Safeguarding Guidance. https://www.nihr.ac.uk/documents/nihr-safeguarding-guidance/25744 National Care Homes Research and Development Forum, 2007. My Home Life, Quality of life in care homes: a review of the literature. National Careers Service, 2020. Care Worker [Online]. https://nationalcareers.service.gov.uk/job-profiles/care-worker. NICE, 2015. Care of dying adults in the last days of life: NG31. Nehls, K., Smith, B.D. and Schneider, H.A., 2015. Video-conferencing interviews in qualitative research. In Enhancing qualitative and mixed methods research with technology, pp. 140-157. Neuberger, B.J., 2013. Review of Liverpool Care Pathway for dying patients. UK: NHS. Neuberger, J., Guthrie, C. and Aaronovitch, D., 2013. More care, less pathway: a review of the Liverpool Care Pathway. Department of Health. Nickel, B., Barratt, A., Copp, T., Moynihan, R. and McCaffery, K., 2017. Words do matter: a systematic review on how different terminology for the same condition influences management preferences. BMJ open, 7(7), p.e014129. Nugus, P., 2019. Re‐structuring the negotiated order of the hospital. Sociology of Health & Illness, 41(2), pp.378-394. Nugus, P., Greenfield, D., Travaglia, J., Westbrook, J. and Braithwaite, J., 2010. How and where clinicians exercise power: Interprofessional relations in health care. Social science & medicine, 71(5), pp.898-909. Nursing and Midwifery Council, 2009. Record keeping: Guidance for nurses and midwives. Office for National Statistics, 2019. Deaths in the care sector, England and Wales: 2019. Office for National Statistics, 2019. Deaths registered in England and Wales: 2019. Oluyase, A.O., Hocaoglu, M., Cripps, R.L., Maddocks, M., Walshe, C., Fraser, L.K., Preston, N., Dunleavy, L., Bradshaw, A., Murtagh, F.E. and Bajwah, S., 2021. The challenges of caring for people dying from COVID-19: a multinational, observational study (CovPall). Journal of pain and symptom management, 62(3), pp.460-470. Ong, W., 1982. Orality and Literacy. Orellana, K., 2014. Care home managers: a scoping review of evidence. Orellana, K., Manthorpe, J. and Moriarty, J., 2017. What do we know about care home managers? Findings of a scoping review. Health & Social Care in the Community, 25(2), pp.366-377. Owen, J., Down, A. and Dhillon, A., 2016. Pharmacy team led anticipatory prescribing in nursing homes: increasing proportion of deaths in usual place of residence. Int J Pharm Pract, 24, p.12. O’Brien, M., Kirton, J., Knighting, K., Roe, B. and Jack, B., 2016. Improving end-of-life care in care homes; an evaluation of the six steps to success programme. BMC Palliative Care, 15(1), pp.1-10. O’Donnell, S.B., Bone, A.E., Finucane, A.M., McAleese, J., Higginson, I.J., Barclay, S., Sleeman, K.E. and Murtagh, F.E., 2021. Changes in mortality patterns and place of death during the COVID-19 pandemic: a descriptive analysis of mortality data across four nations. Palliative medicine, 35(10), pp.1975-1984. O’Reilly, M. and Parker, N., 2013. ‘Unsatisfactory Saturation’: a critical exploration of the notion of saturated sample sizes in qualitative research. Qualitative research, 13(2), pp.190-197. 251 Papachroni, A. and Heracleous, L., 2020. Ambidexterity as practice: Individual ambidexterity through paradoxical practices. The Journal of Applied Behavioral Science, 56(2), pp.143-165. Patton, M.Q., 2014. Qualitative research & evaluation methods: Integrating theory and practice. Sage publications. Payne, S., Turner, M., Seamark, D., Thomas, C., Brearley, S., Wang, X., Blake, S. and Milligan, C., 2015. Managing end of life medications at home—accounts of bereaved family carers: a qualitative interview study. BMJ supportive & palliative care, 5(2), pp.181-188. Penge, J., and Peter C., 2014. Appropriate prescribing in older people. Reviews in Clinical Gerontology, 24(1), pp.58-77. Penney, S. and Ryan, A., 2022. The effect of a leadership support programme on care home managers. Nursing older people, 34(5). Perez, O. and Teubner, G. eds., 2005. Paradoxes and Inconsistencies in the Law. Bloomsbury Publishing. Perkins, E., Gambles, M., Houten, R., Harper, S., Haycox, A., O’Brien, T., Richards, S., Chen, H., Nolan, K. and Ellershaw, J.E., 2016. The care of dying people in nursing homes and intensive care units: a qualitative mixed-methods study. Petticrew M, Roberts H. Systematic reviews in the social sciences: a practical guide. Oxford: Blackwell Publishing; 2006. Pierre, E.A., 2008. Decentering voice in qualitative inquiry. International Review of Qualitative Research, 1(3), pp.319-336. Pinder, R., Petchey, R., Shaw, S. and Carter, Y., 2005. What's in a care pathway? Towards a cultural cartography of the new NHS. Sociology of health & illness, 27(6), pp.759-779. Pitt, M.B. and Hendrickson, M.A., 2020. Eradicating jargon-oblivion—a proposed classification system of medical jargon. Journal of general internal medicine, 35(6), pp.1861-1864. Pitt, M.B., Hendrickson, M.A. and Marmet, J., 2022. Use of Euphemisms to Avoid Saying Death and Dying in Critical Care Conversations—A Thorn by Any Other Name. JAMA Network Open, 5(10), pp.e2233727-e2233727. Poldrugovac, M., Pot, A.M., Klazinga, N. and Kringos, D., 2023. How are regulatory oversight organisations using long-term care performance indicators: a qualitative descriptive study in 10 high-income countries. BMJ open, 13(2), p.e067495. Popay J, Roberts H, Sowden A, Petticrew M, Arai L, Rodgers M, Britten N, Roen K, Duffy S. Guidance on the conduct of narrative synthesis in systematic reviews: a product from the ESRC Methods Programme [Internet]. Lancaster: Lancaster University; 2006. Popay, J., Williams, G., Thomas, C. and Gatrell, T., 1998. Theorising inequalities in health: the place of lay knowledge. Sociology of health & illness, 20(5), pp.619-644. Popham, C. and Orrell, M., 2012. What matters for people with dementia in care homes?. Aging & mental health, 16(2), pp.181-188. Power, M., 2004. The risk management of everything. The Journal of Risk Finance, 5(3), pp.58-65. Public Health Agency, 2022. Personal and Public Involvement: ENGAGE, a central resource for involvement in health and social care. Public Health England, 2018. Statistical commentary: end-of-life care profiles, February 2018 update. Putnam, L.L., Fairhurst, G.T. and Banghart, S., 2016. Contradictions, dialectics, and paradoxes in organizations: A constitutive approach. Academy of Management Annals, 10(1), pp.65-171. Quinn, R.E. and Cameron, K.S., 1988. Paradox and transformation: Toward a theory of change in organization and management. Ballinger Publishing Co/Harper & Row Publishers. Rabionet, S.E., 2011. How I learned to design and conduct semi-structured interviews: an ongoing and continuous journey. Qualitative Report, 16(2), pp.563-566. Raleigh, V., 2020. Deaths from Covid-19 (coronavirus): how are they counted and what do they show. The King’s Fund. Razai, M.S. and Majeed, A., 2022. General practice in England: the current crisis, opportunities, and challenges. The Journal of Ambulatory Care Management, 45(2), pp.135-139. Rechel, B., Doyle, Y., Grundy, E. and McKee, M., 2009. How can health systems respond to population ageing?. Richardson, E., Aissat, D., Williams, G., Fahy, N., 2020. Keeping what works: Remote consultations during the COVID-19 pandemic. https://apps.who.int/iris/bitstream/handle/10665/336301/Eurohealth-26-2-73-76- eng.pdf?fbclid=IwAR00Oq9EpNEGxj21QESn9cCL7YGByr61VW-c4eXNoA65FPa152EsJnKTO7A Richter-Sundberg, L., Garvare, R. and Nystrom, M., 2017. Reaching beyond the review of research evidence: a qualitative study of decision making during the development of clinical practice guidelines for disease prevention in healthcare, BMC Health Services Research, 17(1), p.344. Ritchie, J. and Ormston, R., 2003. The applications of qualitative methods to social research. Qualitative research practice: A guide for social science students and researchers, p.e46. Roberts, D. and Green, L. eds., 2021. Collaborative practice in palliative care. Routledge. Roberts, J.K., Pavlakis, A.E. and Richards, M.P., 2021. It’s more complicated than it seems: Virtual qualitative research in the COVID-19 era. International Journal of Qualitative Methods, 20, p.16094069211002959. Robertson, C., 2021. The Filing Cabinet: A vertical history of information. U of Minnesota Press. Rosoff, P.M., 2012. Unpredictable drug shortages: an ethical framework for short-term rationing in hospitals. The American Journal of Bioethics, 12(1), pp.1-9. Rothstein, H., 2006. The institutional origins of risk: A new agenda for risk research. Royal Pharmaceutical Society & Royal College of Nursing, 2007. The Handling of Medicines in Social Care. 252 Royal Pharmaceutical Society of Great Britain, 2015. Medicines, ethics and practice 39: the professional guide for pharmacists. Pharmaceutical Press. Russell, S., Coombs, M. and Loney, J., 2018. The last days and hours of life. Palliative Care Nursing: Principles and Evidence for Practice, p.280. Ryan, R., Bowers, B., Spathis, A. and Barclay, S., 2020. How is anticipatory prescribing in community end-of-life care governed in the UK? A content analysis of national and local guidance documents. Rycroft-Malone, J., Fontenla, M., Bick, D. and Seers, K., 2010. A realistic evaluation: the case of protocol-based care. Implementation science, 5(1), pp.1-14. Rycroft‐Malone, J., Fontenla, M., Bick, D. and Seers, K., 2008. Protocol‐based care: impact on roles and service delivery. Journal of evaluation in clinical practice, 14(5), pp.867-873. Rycroft-Malone, J., McCormack, B., Hutchinson, A.M., DeCorby, K., Bucknall, T.K., Kent, B., Schultz, A., Snelgrove- Clarke, E., Stetler, C.B., Titler, M. and Wallin, L., 2012. Realist synthesis: Illustrating the method for implementation research. Implementation Science, 7, pp.1-10. Sacks, O., 1991. Awakenings. Pan Macmillan. Saunders, B., Sim, J., Kingstone, T., Baker, S., Waterfield, J., Bartlam, B., Burroughs, H. and Jinks, C., 2018. Saturation in qualitative research: exploring its conceptualization and operationalization. Quality & quantity, pp.1893-1907. Schneider, C.E., Bristol, A.A. and Brody, A., 2019. A scoping review of dementia symptom management in persons with dementia living in home-based settings. Current geriatrics reports, pp.291-301. Schofield, B. and McClean, S., 2022. Paramedics and health promotion. Perspectives in Public Health, 142(3), pp.135-136. Scotland, J., 2012. Exploring the philosophical underpinnings of research: Relating ontology and epistemology to the methodology and methods of the scientific, interpretive, and critical research paradigms. English language teaching, 5(9), pp.9-16. Scott-Aiton, A., 2009. Creating a tool to improve anticipatory prescribing in palliative care. Pharmaceutical journal, 283(7569), pp.283-284. Sedgwick, E.K., 1997. Paranoid reading and reparative reading, or, You're so paranoid, you probably think this introduction is about you. Seidman, I., 2013. Interviewing as qualitative research: A guide for researchers in education and the social sciences. Teachers college press. Senge, P. M., 1990. The fifth discipline: The art and practice of the learning organization. New York: Currency Doubleday. Seymour, J.E., Kumar, A. and Froggatt, K., 2011. Do nursing homes for older people have the support they need to provide end-of-life care? A mixed methods enquiry in England. Palliative Medicine, 25(2), pp.125-138. Sheaff, R., Halliday, J., Byng, R., Øvretveit, J., Exworthy, M., Peckham, S. and Asthana, S., 2017. Bridging the discursive gap between lay and medical discourse in care coordination. Sociology of health & illness, 39(7), pp.1019-1034. Silverman, D., 1993. Interpreting Qualitative Data: Methods for Analysing Talk. Skills for Care, 2018. The state of the adult social care sector and workforce in England. Skills for Care, 2020. The state of the adult social care sector and workforce in England. Skyttner, L., 2006. General systems theory. Perspectives and Practice. Small, S.S., Lau, E., McFarlane, K., Archambault, P.M., Longstaff, H. and Hohl, C.M., 2023. Research recruitment and consent methods in a pandemic. BMC Medical Research Methodology, 23(1), p.113. Smith, B. and McGannon, K.R., 2018. ‘Developing rigor in qualitative research: problems and opportunities within sport and exercise psychology’, International Review of Sport and Exercise Psychology, 11(1), pp.101–121. Available at: https://doi.org/10.1080/1750984X.2017.1317357. Smith, J.A., 2015. Qualitative psychology: A practical guide to research methods. Qualitative psychology, pp.51-80. London: Sage. Smith, W. K. and Lewis, M. W., 2011. Toward a theory of paradox: A dynamic equilibrium model of organizing. Academy of Management Review, 36(2), 381-403. Smith, W.K. and Lewis, M.W., 2011. Toward a theory of paradox: A dynamic equilibrium model of organizing. Academy of management Review, 36(2), pp.381-403. Smith, W. K. and Tracey, P., 2016. Institutional complexity and paradox theory: Complementarities of competing demands, 14(4), pp.455-466. Social Care Institute of Excellence, 2017. Introduction to person-centred care for older people in care homes. https://www.scie.org.uk/person-centred-care/older-people-care-homes/introduction Sorensen R, 2003. A brief history of the paradox: Philosophy and the labyrinths of the mind. Oxford University Press. Spacey, A., Porter, S., Board, M. and Scammell, J., 2021. Impact of the COVID-19 pandemic on end-of-life care delivery in care homes: a mixed method systematic review. Palliative Medicine, 35(8), pp.1468-1479. Spilsbury, K., Devi, R., Daffu-O’Reilly, A., Griffiths, A., Haunch, K., Jones, L. and Meyer, J., 2020. LESS COVID-19. Learning by experience and supporting the care home sector during the COVID-19 pandemic: Key lessons learnt, so far, by frontline care home and NHS staff. LTC Responses to COVID-19, pp.1-63. Spilsbury, K., Hanratty, B. and McCaughan, D., 2015. Supporting nursing in care homes. The RCN Foundation Patient Care and Professional Development for Nursing Staff in Care and Nursing Homes: A Research and Consultation Project. 253 Spreen and Zwaagstra, 1994. Personal Network Sampling, Outdegree Analysis and Multilevel Analysis: Introducing the Network Concept in Studies of Hidden Populations. Stange, K.C. and Ferrer, R.L., 2009. The paradox of primary care. The Annals of Family Medicine, 7(4), pp.293-299. Statista Research Department, 2019. Leading care home operators in the United Kingdom (UK) in 2019, by number of beds. Stebbins, R.A., 2001. Exploratory research in the social sciences (Vol. 48). Sage. Stern, M.C., Jagger, C., Clarke, M., Anderson, J., McGrother, C., Battock, T. and McDonald, C., 1993. Residential care for elderly people: a decade of change. British Medical Journal, 306(6881), pp.827-830. Stjernswärd, J., Foley, K.M. and Ferris, F.D., 2007. The public health strategy for palliative care. Journal of Pain and Symptom Management, 33(5), pp.486-493. Strauss, A., Schatzman, L., Bucher, R., Ehrlich, D. and Sabshin, M., 1963. The hospital and its negotiated order. In E. Friedson (Ed.), The hospital and modern society, pp.149-169. Strauss, A.L., 1978. Negotiations: Varieties, contexts, processes, and social order. Sudman, S. and Kalton, G., 1986. New developments in the sampling of special populations. Annual review of Sociology, 12(1), pp.401-429. Suter, J., Irvine, A. and Howorth, C., 2023. Juggling on a tightrope: Experiences of small and micro business managers responding to employees with mental health difficulties. International small business journal, 41(1), pp.3-34. Tanghe, M., Van Den Noortgate, N., Deliens, L., Smets, T., Onwuteaka-Philipsen, B., Szczerbińska, K., Finne-Soveri, H., Payne, S., Gambassi, G., Van den Block, L. and Piers, R., 2020. Opioid underuse in terminal care of long-term care facility residents with pain and/or dyspnoea: a cross-sectional PACE-survey in six European countries. Palliative medicine, 34(6), pp.784-794. Tariq, S. and Woodman, J., 2013. Using mixed methods in health research. JRSM short reports, 4(6), p.2042533313479197. Teggi, D., 2023. 25-The role of anticipatory prescribing in care homes: identifying dying, complying with regulations and preventing death in hospital–a qualitative study. Teggi, D., 2023b. End-of-life care in English Care Home: Governance, Care Homes, and the Good Death. Terry, G., Hayfield, N., Clarke, V. and Braun, V., 2017. Thematic analysis. The SAGE handbook of qualitative research in psychology, 2, pp.17-37. Teti, M., Schatz, E. and Liebenberg, L., 2020. Methods in the time of COVID-19: the vital role of qualitative inquiries. International Journal of Qualitative Methods, 19, p.1609406920920962. The Association of Directors of Social Care Services, 2021. ADASS Spring Survey 2021. https://www.adass.org.uk/media/8766/adass-spring-survey-report-2021-final-no-embargo.pdf The British Psychological Society, 2014. Code of Human Research Ethics. The British Psychological Society, 2018. Code of Ethics and Conduct. The Gold Standards Framework, 2006. Palliative Care and the GMS Contract Quality and Outcomes Framework (QOF). The Gold Standards Framework, 2014. The GSF Care Homes Training Programme Good Practice Guide. The Gold Standards Framework, 2016. GSF ‘Going for Gold’ Guidance. The Gold Standards Framework, 2019. The Gold Standards Framework (GSF) Care Homes Training Programme (flyer) [Online]. Available from: https://www.goldstandardsframework.org.uk/cd- content/uploads/files/NEW%20Care%20Homes%20Flyer%202019.pdf. The Gold Standards Framework, 2020. The Gold Standards Framework for Care Homes - Bronze, Silver and Gold Support Offers enabling attainment of EHCH [Online]. Available from: https://www.goldstandardsframework.org.uk/cd- content/uploads/files/Care%20Homes/GSF%20CH%20Support%20Offers.%20vs6%20- %20Updated%2016.10.2020.pdf The Gold Standards Framework, 2021. Gold Standard Framework - Disclaimer. The National Care Forum, 2021. The care workforce at breaking point. The National Institute for Health and Care Excellence, 2017. NICE Guideline NG67: Managing medicines for adults receiving social care in the community. The Royal Pharmaceutical Society, 2007. The Handling of Medicines in Social Care. Theys, S., 2017. Constructivism. International relations theory. Thomas, K., 2001. Out-of-hours palliative care in the community [Continuing care for the dying at home]. Macmillan Cancer Relief. Thomas, K., 2003. Caring for the dying at home: companions on the journey. Radcliffe Publishing. Thomas, K., 2004. 12 The research process as a journey. Qualitative research in tourism: ontologies, epistemologies and methodologies. Thomas, R. and Richardson, A., 2004. The NICE guidance on supportive and palliative care—implications for oncology teams. Clinical Oncology, 16(6), pp.420-424. Thomas, S.P. and Pollio, H.R., 2002. Listening to patients: A phenomenological approach to nursing research and practice. Springer Publishing Company. Thompson, S., 1996. Paying respondents and informants. Social research update, 14, 1-5. Thunberg, S. and Arnell, L., 2022. Pioneering the use of technologies in qualitative research–A research review of the use of digital interviews. International Journal of Social Research Methodology, 25(6), pp.757-768. https://www.goldstandardsframework.org.uk/cd-content/uploads/files/Care%20Homes/GSF%20CH%20Support%20Offers.%20vs6%20-%20Updated%2016.10.2020.pdf https://www.goldstandardsframework.org.uk/cd-content/uploads/files/Care%20Homes/GSF%20CH%20Support%20Offers.%20vs6%20-%20Updated%2016.10.2020.pdf https://www.goldstandardsframework.org.uk/cd-content/uploads/files/Care%20Homes/GSF%20CH%20Support%20Offers.%20vs6%20-%20Updated%2016.10.2020.pdf 254 Timmermans, S. and Berg, M., 1997. Standardization in action: achieving local universality through medical protocols. Social studies of science, 27(2), pp.273-305. Timulak, L. and Elliott, R., 2019. Taking stock of descriptive–interpretative qualitative psychotherapy research: Issues and observations from the front line. Counselling and Psychotherapy Research, 19(1), pp.8-15. Townsend, P., 1962. The last refuge: a survey of residential institutions and homes for the aged in England and Wales. London, Routledge. Tracy, S.J., 2010. Qualitative quality: Eight “big-tent” criteria for excellent qualitative research. Qualitative inquiry, 16(10), pp.837-851. Tuffour, I., 2017. A critical overview of interpretative phenomenological analysis: A contemporary qualitative research approach. Journal of healthcare communications, 2(4), p.52. Twycross R, Wilcock A and Howard P., 2017. Palliative care formulary. Nottingham: Palliativedrugs.com Ltd. pp.681. Tytgat, G.N., 2007. Hyoscine butylbromide: a review of its use in the treatment of abdominal cramping and pain. Drugs, pp.1343-1357. U.K. Department of Health and Social Care, 2010. Caldicott Guardian Manual. UKCDR, 2020. UK Collaborative on Development Research: Safeguarding Guidance During COVID-19. https://www.ukcdr.org.uk/resource/practical-application-of-ukcdr-safeguarding-guidance-during-covid-19/ UNISON, 2014. UNISON survey shows NHS staff are overworked and underpaid. Ulsperger, J.S., Knottnerus, J.D. and Ulsperger, K., 2014. Bureaucratic rituals in nursing homes: The “CARE Model” and culture change. The Journal of Aging and Social Change, 3(3), pp.21-22. Vestesson, E., Lloyd, T., Santos, F., Caunt, M. and Steventon, A., 2019. Briefing: The impact of providing enhanced support for care home residents in Wakefield. Victoria State Government, 2017. Anticipatory medications. https://www2.health.vic.gov.au/hospitals-and-health- services/patient-care/end-of-life-care/palliative-care/ready-for-community/anticipatory-medicines. Vince, R. and Broussine, M., 1996. Paradox, defense and attachment: Accessing and working with emotions and relations underlying organizational change. Organization studies, 17(1), pp.1-21. Vossius, C., Selbæk, G., Šaltytė Benth, J. and Bergh, S., 2018. Mortality in nursing home residents: A longitudinal study over three years. PloS one, 13(9), p.e0203480. Wagstaff, B. and Mistry, V., 2020. The integration of paramedics into primary care. British Journal of General Practice, 70(692), pp.123-123. Warmington, J., Afridi, A. and Foreman, W., 2014. Is excessive paperwork in care homes undermining care for older people?. York: Joseph Rowntree Foundation. Watson, J., Hockley, J., Dewar, B., 2006. Barriers to implementing an integrated care pathway for the last days of life in nursing homes. Int J Palliat Nurs, 12, pp.234-240. Weber, M., 1946. Essays in Sociology, Edited by H.H. Gerth and C. Wright Mills. New York: Oxford University Press. Weiss, C.H., 1998. Evaluation: methods for studying programs and policies. New Jersey: Prentice-Hall. Wenger, E., 1999. Communities of practice: Learning, meaning, and identity. Cambridge University Press. Whiting, L.S., 2008. Semi-structured interviews: guidance for novice researchers. Nursing Standard (through 2013), 22(23), p.35. Williams, H., Donaldson, S.L., Noble, S., Hibbert, P., Watson, R., Kenkre, J., Edwards, A. and Carson-Stevens, A., 2019. Quality improvement priorities for safer out-of-hours palliative care: lessons from a mixed-methods analysis of a national incident-reporting database. Palliative medicine, 33(3), pp.346-356. Willig, C., 2012. Qualitative interpretation and analysis in psychology. McGraw-Hill Education (UK). Willis, E., 1989. Medical dominance. Routledge. Willis, J., 2007. Foundations of qualitative research: Interpretive and critical approaches. Sage. Wilson, E. and Seymour, J., 2017. The importance of interdisciplinary communication in the process of anticipatory prescribing. International journal of palliative nursing, 23(3), pp.129-135. Wilson, E., Morbey, H., Brown, J., Payne, S., Seale, C. and Seymour, J., 2015. Administering anticipatory medications in end-of-life care: a qualitative study of nursing practice in the community and in nursing homes. Palliative Medicine, 29(1), pp.60-70. Wilson, E., Seymour, J. and Seale, C., 2016. Anticipatory prescribing for end-of-life care: a survey of community nurses in England. Primary Health Care, 26(9). Woolley, S.R., Butler, M.H. and Wampler, K.S., 2000. Unraveling change in therapy: Three different process research methodologies. American Journal of Family Therapy, 28(4), pp.311-327. World Health Organization, 2014. Global Atlas of Palliative Care at the End of Life. World Health Organization, 2018. WHO global strategy on people-centred and integrated health services: interim report (No. WHO/HIS/SDS/2015.6). World Health Organization. World Health Organization., 2020. Coronavirus disease (COVID-19) advice for the public. World Health Organization, 2020b. Integrating palliative care and symptom relief into responses to humanitarian emergencies and crises: a WHO guide. 255 World Health Organization, 2022. Beyond Consultations and Surveys: Enhancing Participatory Governance in Health Systems. Wright, M., Wood, J., Lynch, T., and Clark, D., 2008. Mapping levels of palliative care development: a global view. Journal of Pain and Symptom Management, 35(5), pp.469-485. Wyld, J. The experiences of participants in care home research. 2017. University of Sheffield. https://enrich.nihr.ac.uk/casestudies/the-experiences-of-participants-in-care-home-research-final-report/ Xafis, V., Watkins, A. and Wilkinson, D., 2016. Death talk: Basic linguistic rules and communication in perinatal and paediatric end-of-life discussions. Patient Education and Counseling, 99(4), pp.555-561. Yardley, L., 2000. ‘Dilemmas in qualitative health research’, Psychology and Health, 15(2), pp. 215–228. Zimmermann, C., Swami, N., Krzyzanowska, M., Hannon, B., Leighl, N., Oza, A., & Donner, A., 2016. Early palliative care for patients with advanced cancer. The Lancet, 387(10024), pp.1721-1730. 256 Appendices Appendix 1 Published systematic review Majumder M, Bowers B, Pollock K, Goodman C, Kuhn I, Barclay S. End-of-life care in UK care homes—controlled drugs: systematic review and narrative synthesis. BMJ Supportive & Palliative Care, 2022; 33(2): 160-177 257 258 259 260 261 262 263 264 Appendix 2 Systematic review search strategy CINAHL via Ebsco # Query Limiters/Expanders Last Run Via Results S18 S3 AND S6 AND S9 AND S17 Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 51 S17 S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 1,266,989 S16 TI ( (armagh or "armagh's" or belfast or "belfast's" or lisburn or "lisburn's" or londonderry or "londonderry's" or derry or "derry's" or newry or "newry's") ) OR AB ( (armagh or "armagh's" or belfast or "belfast's" or lisburn or "lisburn's" or londonderry or "londonderry's" or derry or "derry's" or newry or "newry's") ) OR AF ( (armagh or "armagh's" or belfast or "belfast's" or lisburn or "lisburn's" or londonderry or "londonderry's" or derry or "derry's" or newry or "newry's") ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 9,050 S15 TI ( (aberdeen or "aberdeen's" or dundee or "dundee's" or edinburgh or "edinburgh's" or glasgow or "glasgow's" or inverness or (perth not australia*) or ("perth's" not australia*) or stirling or "stirling's" ) OR AB ( (aberdeen or "aberdeen's" or dundee or "dundee's" or edinburgh or "edinburgh's" or glasgow or "glasgow's" or inverness or (perth not australia*) or ("perth's" not australia*) or stirling or "stirling's" ) OR AF ( (aberdeen or "aberdeen's" or dundee or "dundee's" or edinburgh or "edinburgh's" or glasgow or "glasgow's" or inverness or (perth not australia*) or ("perth's" not australia*) or stirling or "stirling's" ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 59,746 S14 TI ( (bangor or "bangor's" or cardiff or "cardiff's" or newport or "newport's" or st asaph or "st asaph's" or st davids or swansea or "swansea's") ) OR AB ( (bangor or "bangor's" or cardiff or "cardiff's" or newport or "newport's" or st asaph or "st asaph's" or st davids or swansea or "swansea's") ) OR AF ( (bangor or "bangor's" or cardiff or "cardiff's" or newport or "newport's" or st asaph or "st asaph's" or st davids or swansea or "swansea's") ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 21,884 265 S13 TI ( (bath or "bath's" or ((birmingham not alabama*) or ("birmingham's" not alabama*) or bradford or "bradford's" or brighton or "brighton's" or bristol or "bristol's" or carlisle* or "carlisle's" or (cambridge not (massachusetts* or boston* or harvard*)) or ("cambridge's" not (massachusetts* or boston* or harvard*)) or (canterbury not zealand*) or ("canterbury's" not zealand*) or chelmsford or "chelmsford's" or chester or "chester's" or chichester or "chichester's" or coventry or "coventry's" or derby or "derby's" or (durham not (carolina* or nc)) or ("durham's" not (carolina* or nc)) or ely or "ely's" or exeter or "exeter's" or gloucester or "gloucester's" or hereford or "hereford's" or hull or "hull's" or lancaster or "lancaster's" or leeds* or leicester or "leicester's" or (lincoln not nebraska*) or ("lincoln's" not nebraska*) or (liverpool not (new south wales* or nsw)) or ("liverpool's" not (new south wales* or nsw)) or ((london not (ontario* or ont or toronto*)) or ("london's" not (ontario* or ont or toronto*)) or manchester or "manchester's" or (newcastle not (new south wales* or nsw)) or ("newcastle's" not (new south wales* or nsw)) or norwich or "norwich's" or nottingham or "nottingham's" or oxford or "oxford's" or peterborough or "peterborough's" or plymouth or "plymouth's" or portsmouth or "portsmouth's" or preston or "preston's" or ripon or "ripon's" or salford or "salford's" or salisbury or "salisbury's" or sheffield or "sheffield's" or southampton or "southampton's" or st albans or stoke or "stoke's" or sunderland or "sunderland's" or truro or "truro's" or wakefield or "wakefield's" or wells or westminster or "westminster's" or winchester or "winchester's" or wolverhampton or "wolverhampton's" or (worcester not (massachusetts* or boston* or harvard*)) or ("worcester's" not (massachusetts* or boston* or harvard*)) or (york not ("new york*" or ny or ontario* or ont or toronto*)) or ("york's" not ("new york*" or ny or ontario* or ont or toronto*))))) ) OR AB ( (bath or "bath's" or ((birmingham not alabama*) or ("birmingham's" not alabama*) or bradford or "bradford's" or brighton or "brighton's" or bristol or "bristol's" or carlisle* or "carlisle's" or (cambridge not (massachusetts* or boston* or harvard*)) or ("cambridge's" not (massachusetts* or boston* or harvard*)) or (canterbury not zealand*) or ("canterbury's" not zealand*) or chelmsford or "chelmsford's" or chester or "chester's" or chichester or "chichester's" or coventry or "coventry's" or derby or "derby's" or (durham not (carolina* or nc)) or ("durham's" not (carolina* or nc)) or ely or "ely's" or exeter or "exeter's" or gloucester or "gloucester's" or hereford or "hereford's" or hull or "hull's" or lancaster or "lancaster's" or leeds* or leicester or "leicester's" or (lincoln not nebraska*) or ("lincoln's" not nebraska*) or (liverpool not (new south wales* or nsw)) or ("liverpool's" not (new south wales* or nsw)) or ((london not (ontario* or ont or toronto*)) or ("london's" not (ontario* or ont or toronto*)) or manchester or "manchester's" or (newcastle not (new south wales* or nsw)) or ("newcastle's" not (new south wales* or nsw)) or norwich or "norwich's" or nottingham or "nottingham's" or oxford or "oxford's" or peterborough or "peterborough's" or plymouth or "plymouth's" or portsmouth or "portsmouth's" or preston or "preston's" or ripon or "ripon's" or salford or "salford's" or salisbury or "salisbury's" or sheffield or "sheffield's" or southampton or "southampton's" or st albans or stoke or "stoke's" or sunderland or "sunderland's" or truro or "truro's" or wakefield or "wakefield's" or wells or westminster or "westminster's" or winchester or "winchester's" or wolverhampton or "wolverhampton's" or (worcester not (massachusetts* or boston* or harvard*)) or ("worcester's" not (massachusetts* or boston* or harvard*)) or (york not ("new york*" or ny or ontario* or ont or toronto*)) or ("york's" not ("new york*" or ny or ontario* or ont or toronto*))))) ) OR AF ( (bath or "bath's" or ((birmingham not alabama*) or ("birmingham's" not alabama*) or bradford or "bradford's" or brighton or "brighton's" or bristol or "bristol's" or carlisle* or "carlisle's" or (cambridge not (massachusetts* or boston* or harvard*)) or ("cambridge's" not (massachusetts* or boston* or harvard*)) or (canterbury not zealand*) or ("canterbury's" not zealand*) or chelmsford or "chelmsford's" or chester or "chester's" or chichester or "chichester's" or coventry or "coventry's" or derby or "derby's" or (durham not (carolina* or nc)) or ("durham's" not (carolina* or nc)) or ely or "ely's" or exeter or "exeter's" or gloucester or "gloucester's" or hereford or "hereford's" or hull or "hull's" or lancaster or "lancaster's" or leeds* or leicester or "leicester's" or (lincoln not nebraska*) or ("lincoln's" not nebraska*) or (liverpool not (new south wales* or nsw)) or ("liverpool's" not (new south wales* or nsw)) or ((london not (ontario* or ont or toronto*)) or ("london's" not (ontario* or ont or toronto*)) or manchester or "manchester's" or (newcastle not (new south wales* or nsw)) or ("newcastle's" not (new south wales* or nsw)) or norwich or "norwich's" or nottingham or "nottingham's" or oxford or "oxford's" or peterborough or "peterborough's" or plymouth or "plymouth's" or portsmouth or "portsmouth's" or preston or "preston's" or ripon or "ripon's" or salford or "salford's" or salisbury or "salisbury's" or sheffield or "sheffield's" or southampton or "southampton's" or st albans or stoke or "stoke's" or sunderland or "sunderland's" or truro or "truro's" or wakefield or "wakefield's" or wells or westminster or "westminster's" or winchester or "winchester's" or wolverhampton or "wolverhampton's" or (worcester not (massachusetts* or boston* or harvard*)) or ("worcester's" not (massachusetts* or boston* or harvard*)) or (york not ("new york*" or ny or ontario* or ont or toronto*)) or ("york's" not ("new york*" or ny or ontario* or ont or toronto*))))) ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 897,810 266 S12 TI ( (gb or "g.b." or britain* or (british* not "british columbia") or uk or "u.k." or united kingdom* or (england* not "new england") or northern ireland* or northern irish* or scotland* or scottish* or ((wales or "south wales") not "new south wales") or welsh*) ) OR AB ( (gb or "g.b." or britain* or (british* not "british columbia") or uk or "u.k." or united kingdom* or (england* not "new england") or northern ireland* or northern irish* or scotland* or scottish* or ((wales or "south wales") not "new south wales") or welsh*) ) OR AF ( (gb or "g.b." or britain* or (british* not "british columbia") or uk or "u.k." or united kingdom* or (england* not "new england") or northern ireland* or northern irish* or scotland* or scottish* or ((wales or "south wales") not "new south wales") or welsh*) ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 496,909 S11 ( TI ( (english not ((published or publication* or translat* or written or language* or speak* or literature or citation*) N5 english) ) OR AB ( (english not ((published or publication* or translat* or written or language* or speak* or literature or citation*) N5 english) ) ) OR ( TI ( national health service* or nhs ) OR AB ( national health service* or nhs ) OR AF ( national health service* or nhs ) ) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 134,827 S10 (MH "United Kingdom+") Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 322,557 S9 S7 OR S8 Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 109,517 S8 (MH "Terminal Care+") OR (MH "Terminally Ill Patients+") OR (MH "Hospice and Palliative Nursing") OR (MH "Palliative Care") OR (MH "Dying Care (Iowa NIC)") Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 73,510 S7 ((end N life) or ((final* or last*) N (hour* or day* or minute* or week* or month* or moment*)) or palliat* or terminal* or (end adj stage) or dying or (body N (shutdown or shut* down or deteriorat*)) or deathbed) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 97,742 S6 S4 OR S5 Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 70,766 S5 (MH "Controlled Substances") OR (MH "Analgesics, Opioid+") OR (MH "Fentanyl+") OR (MH "Buprenorphine") OR (MH "Oxycodone") OR (MH "Codeine") OR (MH "Tramadol") OR (MH "Methadone") OR (MH "Dihydromorphinone") OR (MH "Phenobarbital+") Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 38,601 S4 (''Controlled drug*'' or ''Controlled medic*'' or ''End of life drug*'' or ''End of life medic*'' or anticipatory or pre-emptive or preemptive or ''Just in case'' or ''Pro re nata'' or PRN or ''Core 4'' or ''Core four'' or opioid* or morphine or diamorphine or oxycodone or methadone or midazolam or phenobarbital or tramadol or buprenorphine or codeine or fentanyl) Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 68,614 S3 S1 OR S2 Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 85,368 S2 (MH "Nursing Home Patients") OR (MH "Nursing Homes+") OR (MH "Residential Care+") Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 44,662 S1 (''Care home*'' or ''Nursing home*'' or ''Residential home*'' or ''Long term care facilit*'' or ''Extended care facilit*'' or ''Care facilit*'' or ''Rest home*'' or ''Homes for the aged'' or ''Residential care'' or ''Geriatric home*'' or ''Geriatric facilit*'' or ''Geriatric institution*'' or ''Nursing center*'' or ''Nursing centre*'') Expanders - Apply equivalent subjects Search modes - Boolean/Phrase Interface - EBSCOhost Research Databases Search Screen - Advanced Search Database - CINAHL 81,071 Searches in MEDLINE, Embase, PsycINFO, Web of Science, Cochrane Library, and Social Care Online were adapted from this strategy. 267 Appendix 3 Systematic review data extraction template Data Extraction Tool Details of publication • First author • Reference Introduction • Aims Study participants • Country of study • Recruitment • Characteristics of participants: - Number - Setting (home, hospital etc) - Age / sex / social class / ethnicity Methods • Date of fieldwork • Research methods • Analysis Key findings relevant to review 1) For whom and by whom are End-of-life CDs prescribed? 2) How are CDs stored and monitored? 3) For whom and by whom are CDs administered? 4) How are CDs disposed of after a resident’s death? 5) What are the views of care home staff and other health and social care professionals concerning these issues? 6) What are the views of residents and their family members concerning these issues? Author(s) conclusion(s) 268 Reviewers’ quality assessment of research (Gough WoE) Weight of Evidence A Coherence and integrity of the evidence in its own terms Weight of Evidence B Appropriateness of form of evidence for answering review question Weight of Evidence C Relevance of the evidence for answering review question Weight of Evidence D Overall assessment of study contribution to answering review question 269 Appendix 4 Email text to potential participants for the interview study Dear Colleague, I am undertaking a research study with care home managers, staff, visiting clinicians and pharmacists concerning their views and experiences concerning Anticipatory prescribing (AP) in care homes. AP, the prescription and dispensing of injectable medications for a named patient, in advance of clinical need, for administration if symptoms arise in the last days of life, is widespread practice across the UK, but has been little researched to date. This research seeks to inform AP practice and policy in care homes in light of changes made during the COVID-19 pandemic. I would like to invite you to take part in this research. This would involve an interview lasting between 30 and 60 minutes over Zoom (or by telephone) at a time that is convenient for you. I am looking to talk with a range of health and social care workers who are involved in anticipatory prescribing for care home residents at the end of life. I am seeking to understand the processes and perceptions surrounding the prescribing, storage, monitoring, administration, and disposal of anticipatory medicines. I attach the study participant information sheet which describes the study in more detail. Please feel very free to get in touch if anything is not clear or if you need more information. I look forward to hearing from you. Warm wishes, Megha Majumder, PhD student [Email address redacted] Supervisor: Professor Stephen Barclay [Email address redacted] Palliative and End-of-life care Research Group University of Cambridge 270 Appendix 5 Participant information sheet Primary Care Unit, Department of Public Health and Primary Care, East Forvie Building, University Forvie Site, Cambridge. CB2 0SR Anticipatory prescribing in UK Care Homes Study PARTICIPANT INFORMATION SHEET Introduction and study aims Anticipatory prescribing (AP) is the prescription and dispensing of injectable medications to a named patient, in advance of clinical need, for administration by suitably trained individuals if symptoms arise in the last days of life and they are unable to take medications orally. AP is intended to ensure rapid access to medications, particularly out-of-hours when there can be added delays in sourcing medication from professionals who are unfamiliar with the resident and their situation. AP is NICE-recommended practice despite having a limited evidence base to date. This research study aims to understand the processes, views and experiences of care home managers, staff, clinicians, and pharmacists in relation to AP at the end of life. Specifically, this study will explore: 1. For whom and by whom AP medicines are prescribed 2. How AP medicines are stored and monitored 3. For whom and by whom AP medicines are administered 4. How AP medicines are disposed of after a resident’s death? 5. The views of care home staff and other health and social care professionals concerning these issues We would like to invite you to take part in this research. Please read this information sheet before deciding: we would be happy to answer any questions you might have and our contact details are at the end of this leaflet. We are hoping to recruit care home managers, staff, and visiting professionals with a variety of views and experiences. There are no “right” or “wrong” answers! If you decide to take part, we will contact you to arrange a convenient time for a one to one Zoom or telephone interview. Your interest in participating in this study is very much appreciated. Who is taking part? We are approaching care home managers, staff, and visiting clinicians and pharmacists working in a range of care homes. Why have I been invited? You have been invited because we understand that you are involved in decisions and processes related to anticipatory prescribing in care homes. Do I have to take part? No. Whether you decide to take part or not is entirely up to you. Your contribution would be very valuable. 271 What will happen if I decide to take part? If you decide you would like to join the study, the researcher, Megha Majumder, will contact you to discuss this. We are aiming to gather perspectives and experiences from a variety of workers with differing levels of experience. Megha will then arrange a time and place convenient for you to have a one-to-one interview over a Zoom video call or by phone. After providing your verbal consent, we will then hold an informal interview, lasting between 30 to 60 minutes, concerning your experiences and views around anticipatory prescribing at the end of life for care home residents. With your permission we will record the interview. Will my taking part in this research be kept confidential? All the information that you give will be kept strictly confidential. You will be given a unique study number and your name and contact details will be kept in a locked filing cabinet in the East Forvie Building, University of Cambridge. The recording of your interview will be sent to a typist who has signed a confidentiality agreement. The transcript will be anonymised to remove your name and the name of anyone else mentioned. Neither you nor your team or care home will be identifiable in any presentations of the findings or written reports. Interview recordings will be transcribed, anonymised, and checked. The original interview recordings will be hosted on the University of Cambridge Secure Data Hosting Service (SDHS) for the duration of the study period and then destroyed, according to the policy of the University of Cambridge. General information about how the University uses personal information can be found at https://www.information-compliance.admin.cam.ac.uk/data-protection/research-participant- data What will happen to the information I give? Your views and experiences are central to this research. We will analyse the interviews to better understand the experiences of care home managers, staff, and professionals involved in anticipatory prescribing at the end of life. All data will be kept confidential: confidentiality will only be broken in the unlikely situation that information is disclosed that causes concern for the welfare of residents or others. Should this arise, the researcher will discuss the situation with senior clinical colleagues including the project supervisor, Professor Barclay. Professional practice guidelines concerning disclosure will be followed. What will happen to the results of the study? We will present the findings to service providers, at professional and academic conferences and in journals. Neither you nor your respective care home will be identifiable in any report or publication. If you would like to receive a summary of the study report and journal publications, we would be pleased to send them to you. What are the possible disadvantages and risks of taking part? We expect that participating in an interview will be a supportive and positive experience. We are interested in finding out about your views about and experiences with AP, so there are no “right” or “wrong” answers in this important but challenging area of resident care. Discussions around experiences of caring for and communicating with dying patients and their families may raise particular instances that have been challenging or distressing for you. If this is the case, we will encourage you to discuss these feelings with your line manager, a trusted colleague, or with the interviewer after recording has finished. What are the possible benefits of taking part? https://www.information-compliance.admin.cam.ac.uk/data-protection/research-participant-data https://www.information-compliance.admin.cam.ac.uk/data-protection/research-participant-data 272 In recognition of your time, we will offer an Amazon voucher of £30 for taking part in the research. Moreover, people often find it helpful to talk about their experiences and perspectives. The research aims to improve the care of dying residents in the future by obtaining a better understanding of issues concerning anticipatory prescribing and related processes. What if I don’t want to carry on with the study? You are free to withdraw from the study at any time. What if there is a problem? If you have any concern about any aspect of this study, you should speak to a member of the research team, whose contact details are at the end of this sheet. Who has reviewed the study? In order to protect the safety, rights, wellbeing and dignity of participants, this study has been reviewed by the Cambridge University Psychology Research Ethics Committee. Who is organising and funding the research? Megha Majumder is the lead researcher: she is a PhD student in the Palliative and End-of- life care Group at the University of Cambridge. Professor Stephen Barclay (Principal Investigator) is supervising the research and is an experienced clinical academic at the University of Cambridge. If you would like more information or have any questions, please feel free to contact either Megha Majumder at [Email address redacted] or via phone at [Phone number redacted], or Professor Barclay at the East Forvie Building, Cambridge at [Email address redacted] or via phone at [phone number redacted]. 273 Appendix 6 Study protocol Anticipatory Prescribing of End-of-Life Care Drugs in U.K. Care Homes. A Qualitative Interview Study Study protocol for University Ethics application 1) Background Anticipatory prescribing (AP) is the prescription and dispensing of injectable medications to a named patient, in advance of clinical need, for administration by suitably trained individuals if symptoms arise in the last days of life and they are unable to take medications orally. Injectable medications are typically prescribed with authorisation to administer them for four common symptoms: pain, nausea / vomiting, agitation, and respiratory secretions. In the United Kingdom (UK), Australia, New Zealand and a number of other countries, AP has been promoted to optimise last days of life symptom control in the community and prevent crisis hospital admissions. The intervention is intended to ensure rapid access to medications, particularly out-of-hours when there can be added delays in sourcing medication from professionals who are unfamiliar with the patient. AP is recommended practice by the National Institute of Health and Care Excellence (NICE) despite having a limited evidence base for its clinical effectiveness or cost-effectiveness. During the COVID-19 pandemic, national guidance has been introduced to permit the repurposing of end-of-life care drugs in care homes, such that drugs prescribed and stored for one resident can be repurposed for, and administered to, another resident if urgently needed. This recent and significant change in national medicines management guidance is one aspect of AP in care homes to be investigated in this study. This study will use semi-structured interviews with health and social care professionals to investigate the processes and relational dynamics of AP in UK care homes. This study will address several knowledge gaps in the evidence base for AP in U.K. care homes, specifically the processes and perceptions surrounding the prescribing, storage, monitoring, repurposing, administration, disposal of end of life anticipatory medicines. Over 430,000 elderly and vulnerable people live in U.K. care homes.1 In 2018 119,629 people died in a care home, 22.1% of all deaths in England and Wales.2 Given the COVID- 19 pandemic, this number of deaths within care homes is expected to have doubled during 2020.3 Care home residents in the final phase of their lives have particular clinical care and psychosocial needs, including a focus on symptom management and comfort, which may involve a suite of AP medicines. Care requires a focus on palliation; the management of the terminal decline and its accompanying symptoms. Care home staff members undertake a critical role in delivering this form of palliative and end-of-life care to residents, determining their needs for pain and symptom control and enabling their timely access to and administration of appropriate medications. A review of the small body of international research suggests that staff often face difficulty in relation to AP because of limited training 1 Hargreaves C et al. "Specialist healthcare services for UK care home residents: a latent class analysis." Primary Health Care Research & Development (2019); 20: e132 2 NHS London. “Care Home Data Pack - NHS England.” NHS England, NHS, 2019, www.england.nhs.uk/london/wp-content/uploads/sites/8/2020/03/NEL-care-home-data-pack-2019.pdf. 3 Iacobucci G. "Covid-19: Care home deaths in England and Wales double in four weeks." British Medical Journal (2020); 369: m1612 274 and experience with certain medications, a lack of resources, and power differentials between themselves and medical practitioners.4 In the UK there has been a widespread introduction of AP in care home-based palliative care. This involves general practitioners (GPs) writing prescriptions in anticipation of them being needed by a certain resident and the development of systems to ensure that medications can be readily administered ‘out-of-hours.’ AP medicines give considerable responsibility to care home nurses, who must make decisions about their administration. To date there is little evidence about the surrounding process in this area. Our study will provide nuanced accounts of health and social care professionals’ understanding, views and experiences of AP medicines. This will improve quality of care in spaces that will be recovering from the COVID-19 pandemic, which has produced substantial shifts in care home guidelines and day-to-day practice. These shifts, compounded by the social transformations that accompanied the danger of close and meaningful interaction, have complicated the fabric of elder care across the country. Thus, it is important to unfold the impact of accelerated changes on the cultivation of care home networks and on the processes of AP in order to redefine and achieve best practice in the elder care space. This study will examine the AP process and associated experiences of health and social care professionals involved in AP; consider care home-specific nuances embedded in the giving of care throughout the AP process; and explore the intricate connections between the AP medicines and the experiences of pain and other symptoms towards the end of life. This exploratory study aims to understand the views and experiences of health and social care professionals who have experience of caring for or communicating with dying people and their families, who are likely to have had held conversations with patients and/or family caregivers about symptom management near the end of life. Residents and family members are not to be included in this study. The new knowledge from our study will be helpful for key stakeholders including care home staff, residents, families, and healthcare services especially in the community and the wider public. 2) Time frame The proposed start date for participant recruitment is March 2021 and it is anticipated that data collection will continue until 30th September 2021. The following timeline outlines the timeframes for the different phases of the study: 1. Recruitment of care homes via care home managers: March to April 2021 2. Interviews with care home managers: April to July 2021 3. Interviews with care home staff (nurses, practitioners, carers): May 2021 to July 2021 4. Data collection and concurrent data analysis (transcription, reading, initial commenting): April 2021 to August 2021 5. Further data analysis (narrative analysis, phenomenological analysis) following completion of interviews: August to October 2021 6. Writing and publication of journal papers: October to December 2021 7. Writing of PhD dissertation and conference presentations: January to June 2022 4 Huisman, Bregje AA, et al. "Role of nurses in medication management at the end of life: a qualitative interview study." BMC Palliative Care (2020); 19: 1-11. 275 A brief final report will be available for participating sites and participants, with copies of publications arising and the PhD thesis for those who would like a copy. 3) Overall design Qualitative research methods enable researchers to gather data on participants’ unique, complex, situational, personal experiences and behaviours, using observational methods, interviews, ethnography and documentary analysis. Specifically, MM has selected the qualitative inductive approach of one-to-one interviews to explore and describe one particular facet of care home culture: Anticipatory Prescribing. As the research team is located in Cambridge, a purposive sample of the care homes in the Cambridgeshire area will be contacted in order to develop a holistic understanding of AP as it functions ‘on the ground.’ Should local recruitment in Cambridgeshire prove challenging, we will identify potentially suitable care homes through our wider professional networks. Semi-structured qualitative interviews will be held with care home managers, care home staff, and visiting health and social care professionals (doctors, nurses, pharmacists, etc.) who have perspectives and experiences related to the processes surrounding anticipatory prescribing and controlled medicines, specifically surrounding their prescribing, storage, monitoring, repurposing, administration, and disposal. We will purposively sample care homes in Cambridgeshire (or more widely if necessary), seeking to recruit at least 15 care homes. A multi-phase study is planned. Phase 1 involves interviews with care home managers (CHMs). As their contact details lie within the public domain, MM will send an introductory email to CHMs about the study as well as the Participant Information Sheet attached to the correspondence, titled “Anticipatory Prescribing in UK Care Homes.” See Appendix 1 for email text and Appendix 2 for the Participant Information Sheet. If CHMs are willing, one or two interviews will be conducted over the course of one to three months. Following the first interview, MM will begin Phase 2, in which the CHMs will be asked to inform staff members, visiting doctors, pharmacists, and health and social care professionals via email and / or in-person about the study. MM’s contact information and Participant Information Sheets (Appendices 1 and 2) will be provided for CHMs to distribute on behalf of MM. Those who are interested in and able to participate will be invited to respond to MM directly or to ask the CHM to inform MM of their interest in participating. These data will generate case-studies of the AP process in care homes at two levels: a) Processes of AP within care homes. Accounts of how AP operates in each of the participating care homes, specifically; by and for whom AMs are prescribed, stored, monitored, repurposed, administered and disposed of. b) Perspectives about AP within care homes. How AP is perceived and experienced by care home managers, staff, clinicians, and / or pharmacists in the care of residents approaching the end of their lives. 4) Our research questions are: With regard to AP in U.K. care homes: 1. For whom and by whom are AP medicines prescribed? 2. How are AP medicines stored and monitored? 276 3. For whom and by whom are AP medicines administered? 4. How are AP medicines disposed of after a resident’s death? 5. What are the views of care home staff and other health and social care professionals concerning these issues? Qualitative research questions can be broad, and literature suggests that without a firm underlying methodology, this may cause a lack of direction when conducting the research 5. To articulate this study and provide the clarity it needs, the research questions will shape the interview schedule and the interviews. The resulting data will inform the development of cases comprising the care homes and their communities. This study an inductive approach so as to commit to holistic, in-depth understandings of AP phenomena and participant perspectives. 5) Participant eligibility criteria Nursing home managers, care home staff, visiting health and social care workers (including clinicians and pharmacists) in Cambridgeshire care homes (and potentially more widely) who are: ● Adults (aged 18 years and over) ● Able to give informed consent to participate in interviews in English ● Involved in the care of residents for whom AP of end-of-life care medications has occurred or is considered 5) Site and Participant Recruitment Semi-structured qualitative interviews will be held with care home managers, care home staff, visiting health and social care workers (including doctors, nurses and pharmacists) in Cambridgeshire care homes (and potentially more widely) who are: 1) Adults (aged 18 years and over); 2) Able to give informed consent to participate in interviews in English and 3) Involved in the care of residents for whom AP of end-of-life care medications has occurred or is potentially considered. We will identify care homes in Cambridgeshire (or more widely) with good or outstanding Care Quality Commission inspection reports, purposively sampling by size and provider (national provider or individual owner). We will seek to recruit at least one CHM and one staff member and / or visiting health or social care worker from 15 Cambridgeshire nursing homes (of which there are 56) and 5 Cambridgeshire residential care homes (of which there are 111). Should local recruitment prove challenging we will recruit more widely, identifying potential care homes through our regional and national professional contacts. For each care home, the CHMs, or their nominated deputy, will be invited via email to take part in the research (see Appendix 1 for email text and Appendix 2 for the Participant Information Sheet, to be attached to the introductory email, developed in collaboration with PPI colleagues). These individuals will be invited to give informed consent for an interview with MM, in order to better understand the AP process, from decision-making to disposal, as perceived by the CHMs. As their contact details lie within the public domain, MM will email CHMs with information about the study, “Anticipatory Prescribing in UK Care Homes,” and 5 Bradley E, Curry L, Devers K. Qualitative data analysis for health services research: developing taxonomy, themes, and theory. Health Serv Res (2007); 42(4): 1758-72. 277 inquire about their willingness to partake in a series of one to two one-to-one interviews with MM via Zoom. At least one interview will be requested: if willing, MM will follow up with a second interview approximately two months later. MM will confirm verbally prior to all subsequent interviews that the participant is willing to continue in the study, working on the principle that ongoing consent stands unless the participant indicates otherwise, or the researcher identifies that their cognition is impaired. These Phase 1 interviews with CHMs will continue until information saturation is reached, which is anticipated to occur after approximately 10 to 15 interviews. If information saturation was not reached by this point, we will continue to purposively recruit further participants, to a maximum of 25 in total. Following the first interview, the CHM will be asked to inform other care home staff members and visiting clinicians, via email and / or in-person about the study. A similar approach will be taken during these Phase 2 interviews with care home staff members and visiting health or social care professionals. It is anticipated that information saturation will be reached after approximately 25 interviews, increasing to a maximum of 35 participants if necessary. CHMs will be asked to identify suitable participants on behalf of the research team and will send to potential participants (via email and / or paper) the letter of invitation outlining the study (Appendix 1) and the participant information sheet (Appendix 2). These documents will also be provided to CHMs to distribute on behalf of the research team to care home staff and visiting health and social care professionals. Those who are interested will be invited to contact MM directly or to ask the CHM to inform MM of their interest in participating. Prior to the interview, all participants will be sent the participant information sheet (Appendix 2) and consent form (Appendix 3) to ensure robust informed consent. For those wishing to discuss the study further, MM will be available for a phone or Zoom call prior to a possible interview. MM will then set up a Zoom (or telephone) interview for a time when the staff member or health professional is available. These interviews will explore the experiences and views surrounding end-of-life AP of these health and social care professionals (see interview topic guide in Appendix 4). MM will obtain verbal informed consent prior to each participant’s first phone or Zoom interview. Given the COVID-19 pandemic and the remote interviews, consent will be verbal rather than written. MM will record the conversation in which the consent form (Appendix 3) is reviewed and the participant verbally agrees to proceed. GPs, nurses, and staff participants may find that exploring their views and experiences of providing end-of-life care may raise emotionally charged issues. Participants will be given the option of pausing during interviews if distressed, returning to the interview at a later date if preferable, or withdrawing at any point. Distressed participants will also be directed to sources of support in the form of a toolkit curated by the Mind Organization, supported by The Royal Foundation. Their offered services include free 24/7 text support for social care and health staff, and a dedicated confidential support line for staff to speak with trained counsellors who are experts in bereavement, grief and trauma, provided by Hospice UK and Just B (https://www.mentalhealthatwork.org.uk/toolkit/ourfrontline-socialcare/) 6) Consent All participants will be given at least 24 hours to consider their involvement in the study. Avoidance of coercion is a priority, and we will take steps to minimise any risks of coercion https://www.mentalhealthatwork.org.uk/toolkit/ourfrontline-socialcare/ 278 to take part in the research. MM will explain the study in detail and gain informed consent. MM will utilize a three‐fold approach to acquiring informed consent: 1. First, by discussing the participant information sheet and any questions that may arise, before asking for verbal consent 2. Second, by verifying with the interviewee, and continuously operating via process consent, re‐verifying throughout the study, their agreement to continue 3. Third, by ensuring that interviewees are aware of MM’s role as a researcher. It will be made clear throughout the research process that participation is wholly voluntary and that not taking part will not adversely affect participants’ care or legal rights. MM will check that participants are willing to continue taking part prior to interviews. MM’s contact information will be made available if they want to withdraw at any point and will be made aware that a reason to do so is not required. MM will also practice process consent, remaining alert to any signs of discomfort or resistance, and checking in to ensure that participants are comfortable being interviewed. Participants will also be given the option of pausing or opting out of interviews, or returning at a later date if they wish. In recognition of the time given to the interview, we will offer participants an Amazon voucher of £30 in recognition of their time. The researcher (MM) will obtain verbal informed consent prior to each participant’s first interview (see Appendix 3). These interviews will be held over Zoom (or telephone) in order to ensure the safety of participants in light of the COVID-19 pandemic. MM will subsequently confirm that the participant is willing to continue prior to all subsequent interviews, working on the principle that ongoing consent stands unless the participant indicates otherwise. As part of the informed consent process, at the start of each Zoom call, MM will talk through the study using the Participant Informant Sheet and consent form to give participants the opportunity to ask any questions they have before asking them for verbal consent. The interview will then formally begin. The Participant Information Sheet outlines the purpose of the research, the benefits and potential burdens of participating in the research, the research contacts and commitments, and how personal data will be stored and used (see Appendix 2). MM will make it clear before each interview that it is up to the participant if they want to continue to take part, using her communication skills and capacity assessment training to confirm that ongoing consent still stands. 7) Confidentiality No confidential information will be passed between MM and other health or social care professionals without their consent. Data will be collected securely in order to maintain confidentiality and work within the General Data Protection Regulation (GDPR) and Data Protection Act 2018. MM will follow the Caldecott Principles when using participants’ identifiable data. James Brimicombe, Data Manager of the Primary Care Unit at the University of Cambridge, has advised on the secure management and storage of study data. Recordings will be removed from Zoom (or the digital recorder used for telephone interviews when applicable) and transferred to the University of Cambridge Secure Data Hosting Service after each interview. Only MM will have access to participants’ personally identifiable information, which will be stored University of Cambridge Secure Data Hosting Service for the duration of the study. All the applicants have received GDPR training. All study transcripts and personally identifiable data will be stored securely on secure University 279 premises for two years and then destroyed, in accordance with the policy of the University of Cambridge. Participants’ information will be treated confidentially and held securely, with study data only identifying participants as a study ID number, as explained below in Section 11. Pseudonyms will be used in all reported findings. Confidentiality will be respected at all times unless consent has been given to disclose the information to specified persons. No person who is not directly and professionally involved with the research project shall have access to the study data without prior written permission of the study’s primary Academic Supervisor, Professor Stephen Barclay. The only situation in which we foresee that confidentiality might be broken would be if information is discovered during the research that causes concern for the welfare of the participant or others. It is anticipated that this is unlikely to occur. Should this arise, the situation will be discussed with the participant and Professor Barclay as a practicing clinician and the study’s Chief Investigator, concerning whether and how to raise the concerns with appropriate individuals and care home senior staff. In exceptional circumstances, the research team may need to inform the relevant professional regulatory bodies: this will only be following discussion with the Head of the Primary Care Unit, Professor Jonathan Mant. Professional practice guidelines concerning patient safety will be followed and all those involved in the research will act within the codes of professional conduct stipulated by the bodies governing their professions. This will be made explicit to organizations when they are recruited to take part in the study and to participants as part of the informed consent process. 8) Data collection Interviews will focus on the relationships between individuals and processes inherent to the various stages of AP within care homes. Sample interview questions in the topic guide were developed in collaboration with the supervisory team and PPI colleagues (Appendix 4). They will be centred, for example, on the perceptions, prescribing, dispensing, managing, and administration events of anticipatory medicines. Interviews will be undertaken by MM at a time and location to suit participants: Zoom interviews will be sought wherever possible, with telephone interviews an alternative should the participant prefer. The interviews are anticipated to last between 30 to 60 minutes and will be digitally recorded. The interviews will be facilitated by MM to provide flexibility to explore the participants’ experiences and perspectives. The recording will be transcribed verbatim by a professional transcriber or the researcher, checked and anonymised and then entered into NVivo software for qualitative interpretive phenomenological analysis. Professional transcribers will sign a confidentiality agreement, with digital recordings and transcripts emailed in encrypted form. Data analysis will be concurrent with data collection to permit examination of emergent themes. Further interviews will continue until information saturation is reached, anticipated to occur after approximately 40 interviews in total; if information saturation was not reached by this point, we will purposively further participants, to a maximum of 60 in total. Towards the end of the data collection period, PPI collaborators will be asked to provide feedback on the results to ensure that the analysis and write-up are relevant for care home residents and family carers. 9) Interview content 280 The Participant Information Sheet will make it clear to participants that they will be discussing issues surrounding AP and their own experiences, guided by the research questions of the study. Participants will be assured that they will not be under pressure to discuss subjects they would prefer not to discuss and that they can withdraw from the study at any point. Participants may find it helpful to have the opportunity to reflect on, and share, their experience of pain and pain and symptom management with someone who is not directly involved in a resident’s care. The potential risks and benefits of participation in the research will be highlighted as part of the informed consent process. MM will gauge participants’ willingness to discuss the nature of their respective experiences during interviews and will be led by what they wish to discuss. GPs, nurses, and staff participants may find that exploring their views and experiences of providing end-of-life care may raise emotionally charged issues. Participants will be given the option of pausing during interviews if distressed, returning to the interview at a later date if preferable or withdrawing at any point. Distressed participants will also be directed to sources of support in the form of a toolkit curated by the Mind Organization, supported by The Royal Foundation. Their offered services include free 24/7 text support for social care and health staff, a dedicated confidential support line for staff to speak with trained counsellors who are experts in bereavement, grief and trauma, provided by Hospice UK and Just B (https://www.mentalhealthatwork.org.uk/toolkit/ourfrontline-socialcare/) 10) Addressing potential issues with the methods Participant safety. In the light of the COVID-19 pandemic, the researcher (MM) acknowledges that she poses a potential infection risk to care home managers, staff, visiting health and social care professionals, and residents. Therefore, interviews will be conducted remotely over Zoom or telephone, so as not to endanger members of the care home community. Researcher boundaries. It will be made clear as part of the ongoing informed consent process that the researcher is external to the care home and clinical team and is unable to provide social care or clinical advice. Researcher effects. Recruiting participants and asking about their views and experience of the AP process could induce social desirability biases. Over the course of discussion, MM will aim to elicit a more searching exploration of AP practices by phrasing questions in a manner that will enable participants to feel comfortable and to relay holistic, representative accounts of their experiences with AP. Researcher safety. MM will conduct the interviews virtually. Privacy measures such as password protection of the Zoom application will be taken in order to protect MM’s Zoom account. Researcher self-care. MM recognizes that extensive periods of time becoming acquainted with and interviewing participants about their experiences and care work in an end-of-life context is emotionally taxing. Self-care strategies will be used throughout data collection and data analysis. These strategies include debriefing with peers and academic supervisors, access to monthly clinical supervision with a trained counsellor via team funds if MM finds it helpful, and the scheduling of regular rest breaks throughout the project. https://www.mentalhealthatwork.org.uk/toolkit/ourfrontline-socialcare/ 281 11) Data protection This study has been discussed with the Primary Care Unit Data Protection Officer James Brimicombe. Any paper participant contact information with their associated study ID number will be kept in a locked cabinet, in an access restricted building (double doors with card swipe access and reception area) in the University of Cambridge East Forvie Building. Recordings will be removed from Zoom (or the digital recorder used for telephone interviews) and the audio transferred to the University of Cambridge Secure Data Hosting Service after each interview. This is an ISO:27001 certified safe haven for research data. All video parts of the recording will be immediately destroyed. As is standard practice, each participant will be allocated a unique study ID created for this project. Audio files and corresponding transcripts will be assigned a study ID number before analysis. Only MM will have access to participants’ personally identifiable information, which will be stored within University of Cambridge Secure Data Hosting Service (SDHS). Original interview recordings will be stored in the University of Cambridge SDHS, which is an ISO- 27001 certified secure data safe haven within the Clinical School. All study transcripts and personally identifiable data will be stored securely on secure University premises for two years and then destroyed. All the applicants have received GDPR and Good Clinical Practice training. The voice recordings will be sent in encrypted form to a professional transcriber based in England and returned in encrypted form. The transcriber will sign a confidentiality agreement and will destroy any voice recording they have on completion of the transcriptions. The transcripts will then be checked and made fully link-anonymised, identified only by a unique study ID number in de-identified transcripts by MM. Coding of transcripts and data analysis will be supported through the qualitative data management program NVivo 12. Research memos and an audit trail of decisions during analysis will be kept for transparency and dependability. Researcher interpretations of data, and key decisions in the stages of data analysis, will be discussed with Professors Stephen Barclay, Claire Goodman and Kristian Pollock to achieve a comprehensive and reflective analysis. Link-anonymised data (referring to participants by their study ID number) by MM who is Good Clinical Practice trained using a custom built electronic Case Report Form (eCRF) will be stored on a University of Cambridge encrypted (AES-256), strong password-protected laptop computer. No participant-identifiable data will be recorded on the eCRF. The data recorded locally on the eCRF will be transferred to and stored securely on a password protected University of Cambridge computer on MM's return to the office base at the next available opportunity. MM will have access to the full dataset. Supervisors Professors Stephen Barclay, Claire Goodman and Kristian Pollock will have access to the anonymised dataset. We propose to keep the fully anonymised research data indefinitely on the University of Cambridge Data Repository, if participants are willing for their anonymised data to be used this way, as indicated during the informed consent process. The fully anonymised dataset will be made 282 available for secondary analysis, upon formal request to and approval of the research team, through the University of Cambridge Data repository: https://osc.cam.ac.uk/repository. Should participant consent have been given, fully anonymised data may be made available to other researchers involved in relevant future projects. Data will only be shared with organisations and researchers conducting research in accordance with the UK Policy Framework for Health and Social Care Research. The data custodian will be MM. If MM leaves the Department, the Data Manager for the Department of Public Health will become the data custodian. Details on how the Department of Public Health and Primary Care, part of the University of Cambridge School of Clinical Medicine, on how we use research data is supplied here: https://www.medschl.cam.ac.uk/research/privacy-notice-how-we-use-your- research-data/ Personal details. Each participant will be given a unique study ID number. References to places or case details that might reveal the identity of participants will be anonymised to protect their identity. Personal contact details and their study ID number will be stored separately from the study data and kept in a locked cabinet, in an access restricted building (double doors with card swipe access and reception area) in the University of Cambridge East Forvie Building. Digital audio recordings will be transcribed verbatim by a professional transcriber or MM, and will be sent in encrypted form to a professional transcriber and returned in encrypted form. Transcripts will then be checked and anonymised. The original non-anonymised digital recordings and transcripts will be stored within University of Cambridge Secure Data Hosting Service (SDHS) for two years and then destroyed, in accordance with the policy of the University of Cambridge. Only MM will have access to the original interview recordings. Anonymised transcripts will be stored in password-protected files on a secure University server. Anonymised transcripts will refer to participants by their role and a unique study- specific ID number only, to maintain anonymity. At times, MM may hold anonymised transcripts on an encrypted laptop computer. 12) Data analysis and dissemination plans Interviews will have an order of questions that are flexible and will be recorded. A broadly thematic analysis will look for patterns and differences in views, experiences and understanding to identify themes and capture the complexity of the phenomena under investigation. This analysis will involve several steps, including familiarisation with the data, coding, identification of themes, reviewing and refining themes, and writing up. MM will conduct the data analysis concurrently with data collection. Perspectives and experiences of AP processes will be compared on an individual basis and between professionals and their respective groups. This will inform areas of interest and divergence for further exploration in subsequent interviews and enrich the analysis. The results of the analysis will be the main body of the research project. A summary of the findings and analysis will be sent to care homes in the UK, submitted for publication in an academic journal, published in a doctoral thesis, and disseminated in collaboration with PPI colleagues and at conferences. https://osc.cam.ac.uk/repository https://www.medschl.cam.ac.uk/research/privacy-notice-how-we-use-your-research-data/ https://www.medschl.cam.ac.uk/research/privacy-notice-how-we-use-your-research-data/ 283 Appendix 7 Ethics Approval [Signature redacted] 284 Appendix 8 Interview topic guide Professionals’ views and experiences with anticipatory prescribing at the end-of-life in UK care homes. TOPIC GUIDE Date __________ Introduction ● Introduce yourself. Check participant’s name. ● Discuss the purpose of the study (exploring CHM, staff, and health and social care professionals’ views and experiences about anticipatory medicines prescribed to residents at end of life in UK care homes) ● Everything discussed will be confidential except if concerns raised for welfare of patient or others, when professional guidelines will be followed ● Provide structure of the interview (audio recording, taking notes) ● Emphasise there are no right / wrong answers: your view is valid! ● Ask if participant has any questions ● Outline who participant can talk to if the interview causes distress, or if they have concerns about the study ● Obtain verbal consent ● Test audio recording equipment Opening question 1) This is a study of anticipatory prescribing (AP) at the end of life for care home residents. We are especially interested in the processes of prescribing, dispensing, storage, administration, and disposal of anticipatory medicines. Can you tell me a bit about yourself and your professional role? Prompts - What role do you have in the various stages of Anticipatory Prescribing for residents? 2) Can you tell me about an experience with one or more residents for whom you were involved in the anticipatory prescribing processes? Encourage the participant to talk at length about that individual and others they mention. Prompts - For whom are AMs commonly prescribed? - When are AMs likely to be prescribed / how far in advance of anticipated death? - Who is involved in the decision to prescribe? - Who prescribes the medication? - Who is involved in the decision to administer AMs? - When are anticipatory medicines commonly administered? - Who administers the medication? - Have there been any cases in which it was difficult to prescribe or administer AP medicines? 285 - Have you had experiences of disagreements with others in relation to AP? 3) How does your care home store and dispose of anticipatory medicines? Prompts - How are they stored? - How are they monitored? - What are your thoughts about the regulations regarding AP? - What forms of guidance and support do you have for AP in your care home? 4) How has the COVID-19 pandemic impacted your care home and your practices in relation to end-of-life care medications? During the COVID-19 pandemic, it has been permitted to repurpose end of life medicines that were prescribed for one resident should they be urgently needed for another resident. What are your views of such medicine repurposing? Have you had any experiences with medicines repurposing during the pandemic? Prompts - Have things changed in light of the pandemic? - Can you tell me about a recent experience with a resident who was administered repurposed medicines? - Are there any resources that you have used or found helpful to guide you and your staff concerning medicine repurposing? 5) How do you find talking about AP with residents and their families and friends? Prompts - How do families and friends respond to the prescribing of anticipatory medicines? - How much do they know about AP medicines being put in place? - Who discusses AP with the families, and to what extent? - Are family members involved in the AP decision-making process? - Is there ever conflict within families? - How do you feel discussing AP with residents, family members? - Are there any “go-to phrases” that you use when discussing AP? 6) Would you say that you have an overall view on the use of AP for residents approaching the end of their lives? Prompts - Do you feel there are some groups of residents for whom AP might be more or less important? Are certain kinds of residents more likely to receive AMs? - Or more or less appropriate? - What sort of experiences have informed your views on AP? 7) Closing question: Have you any thoughts on how we might improve end-of-life AP for care home residents? Prompts - What do you feel might helpfully change to make this better? Closing the discussion ● Thanks for participation 286 ● Ask if there is anything that has been missed, that the interviewee wants to add to what has been discussed ● Reiterate that some issues can be challenging ● Remind that if needed they can talk to a trusted colleague, line manager or the interviewer ● Ask who would like a copy of the results when they are published After the interview ● Record field-notes: observations, feelings, thoughts, reactions about the discussion 287 Appendix 9 Reflexive journal extracts Reflexive Journal Extract 1: Research Planning Phase during COVID (2020) [A plan to add to the evidence, in light of findings from the literature review] Writing up my findings from the literature review, I’ve noticed that the limited research on health and social care professionals’ views on AP highlights the challenges they face, but doesn’t probe into the factors underpinning them. Essentially, current research has identified what many of the challenges of AP in care homes are, but not why they exist. I had assumed that the literature would answer the ‘whys’ alongside presenting the ‘whats’, but that hasn’t been the case. This is something I can explore further in my research. My supervisors have recommended a range of texts and articles to help guide my research design so I can explores AP processes and figure out why they work the way they do, but they cautioned against using rigid, pre-existing frameworks that limit my analysis. Smith & Lewis' 2011 work on paradoxes can be an interesting lens. Their theory on paradoxes can be used as a ‘way in’ to understanding why professionals face the challenges I described in the literature review. Given the pandemic and the focus on health care practitioners’ challenges, I feel this exploration feels both timely and essential. Data collection and analysis can engage with this perspective and tease out the complexities of care home AP in the current landscape for a better understanding of the system, its faults and strengths, and see what can help professionals better navigate it. Reflexive Journal Extract 2: Post-interview (2021) [Interviewee: Nursing Home Manager Location: Nursing Home, via Zoom call] Interviewed [participant], a Nursing Home Manager, via a Zoom call, with her office, visibly cluttered with paperwork. [Participant] highlighted the complexity of medication administration paperwork. With staff stretched thin due to pandemic pressures, [participant] described challenges where they found themselves torn between administrative responsibilities (filling forms, reading guidance, implementing policy) and immediate patient care. This tension sometimes bred anxiety in her about potentially failing to administer AMs in a timely manner and that leading to a ‘bad death’. [Participant] expressed that much of her staff found AM administration to be both emotionally and practically challenging, believing that the medicines could ‘speed things along’ and so they largely avoided administering them. She indicated that this was misaligned with her own principles, to administer the medicines so that her residents ‘can have a good death’ if their dying process is marked by distressing symptoms. [Participant’s] narrative re-oriented me in the real-life bureaucratic and decisional challenges of administering AMs. It also made me feel disoriented, returning to my own worldReflecting on my assumptions going into the interview, I thought she, like some of the other managers I’ve interviewed, would be very open about her experiences with AP. However, when she started talking about individual residents, she did not elaborate as much as she did when discussing the processes of storage, monitoring, and disposal. I can’t be certain as to why, but there are several 288 possibilities: concerns about confidentiality of her residents or feelings of sadness or distress when talking about people she said she came to see as family. As she did appear to become emotional when describing an experience with a ‘bad death’ in her care home, I refrained from asking directly about her experiences with respect to individual residents, and focused on her experiences working with AP processes (storage, monitoring, disposal, reuse). However, whenever she did provide a patient- focused example to illustrate a point, I encouraged her using what Josh calls my ‘active listening toolkit’. Moving forward, in future interviews, I intend to delve deeper into how interviewees view and manage the bureaucracy of AP – so far, all participants have expressed frustration and/or anger in response to the amount of paperwork for AP (if only to satisfy the regulators). What is the cost of doing this work? Why is the bureaucracy set up as such? [Participant’s] insights highlight the need for more efficient systems that reduce administrative burdens while keeping the patient’s well-being at the forefront. Reflexive Journal Extract 3: During Analysis (2022) [After coding an interview transcript with a GP participant] Engaging in the coding process of the transcript from my conversation with [Participant], an experienced GP, has been both a rigorous and enlightening journey. My coding process is very thorough because I am concerned about oversimplifying, and that has made it very time-consuming. Comparing the codes in this interview with codes from other prescribers’ interviews, I’ve identified a recurring theme: the tension between standardised and individualised prescribing in AP, or standardised versus individualised prescribing in AP. ‘Standardised prescribing protocols’, ‘time constraints’, and ‘difficulties individualising prescribing’ and ‘ideals of individualisation’ were common codes across these interviews, painting a picture of their day-to-day challenges. [Participant]'s accounts captured this duality. She described the helpfulness of standard protocols offer, especially when time is of the essence. This preference materialised in clear-cut codes like ‘prescribing routine’ and ‘protocol benefits’. These structures seemed to offer a semblance of order, a tool to help ensure timely anticipatory prescribing. But then, the ‘ideals of individualised care’ and ‘principle of individualisation in guidance’ codes underscore the tension between the principles in guidance and the actual prescribing protocols, with both sets of documentation created by national health agencies. Reflecting on this transcript and others with health care professionals, I’ve found that when I follow up about the guidance or protocols they mention using to help them make prescribing decisions, they often default to describing how important it is to individualise care. This coding exercise made me think about and engage with wider discussions on the limitations of protocols. The recurrent codes of ‘time constraints’ and ‘prescribing routine’ taken with ‘difficulties individualising prescriptions’ and ‘ideals of individualised care’ suggest conflict between practice and policy. With GPs like [Participant] constantly racing against time, could the regular use of these standard protocols, especially the routine prescribing of end-of-life drugs in set doses, result in a blanket approach? Considering the pressures and time restrictions, this is a concern. And how do we reconcile the ideals of individualised care in AP guidance with the seemingly contradictory inclusion of standardised prescribing protocols? This coding revelation has emphasised a need to dissect these tensions. In the write- 289 up, I’ll explore the national and local policies that [Participant] referenced that guide her prescribing decisions, to better understand the interplay between protocols, policy, and the practical realities of prescribing decisions. But given my personal experiences as my grandmother’s carer and having witnessed how standardised care could – especially if pressed for time and resources – write patient’ preferences out of the clinical equation, I should remain cognisant of any biases that affect my analysis. It will be important to consider all of the factors, individual and systemic, that go into the prescribing decisions described by participants. 290 Appendix 10 Results of Braun & Clarke (2021) ‘20 questions to evaluate the quality of thematic analysis’ Do the authors explain why they are using TA, even if only briefly? Yes. Do the authors clearly specify and justify which type of TA they are using? Yes (RTA). Is the use and justification of the specific type of TA consistent with the research questions or aims? Yes, the use and justification of using RTA is consistent with the research question and aims. Is there a good ‘fit’ between the theoretical and conceptual underpinnings of the research and the specific type of TA (i.e. is there conceptual coherence)? Yes, there is conceptual coherence between the constructivist, relativist worldview, paradox perspective, and RTA. Is there a good ‘fit’ between the methods of data collection and the specific type of TA? Yes, RTA ‘fits’ with the qualitative interview method (interviews with health and social care professionals working in care homes). Is the specified type of TA consistently enacted throughout the paper? Yes, reflexive thematic analysis is consistently implemented. Is there evidence of problematic assumptions about, and practices around, TA? These commonly include: ● Treating TA as one, homogenous, entity, with one set of – widely agreed on – procedures. ● Combining philosophically and procedurally incompatible approaches to TA without any acknowledgement or explanation. ● Confusing summaries of data topics with thematic patterns of shared meaning, underpinned by a core concept. ● Assuming grounded theory concepts and procedures (e.g. saturation, constant comparative analysis, line-by-line coding) apply to TA without any explanation or justification. ● Assuming TA is essentialist or realist, or atheoretical. ● Assuming TA is only a data reduction or descriptive approach and therefore must be supplemented with other methods and procedures to achieve other ends. ● No, the researcher explains how they have applied reflexive thematic analysis to the data and reflected on how their own assumptions and preconceptions have influenced the construction of themes. ● Philosophical and procedurally compatible approaches to thematic analysis have been implemented. ● No, the themes are patterns of shared meaning that underlie or are united by a core concept, consistent with TA. The themes are not just organised around a shared topic in that they are trying to capture the diversity of meaning in relation to the topic (which may be reflective of data collection questions with theme titles that start with ‘type of’ or ‘benefits of’ or ‘drawbacks of’ (Evans et al., 2016)). ● No, thgrounded theory concepts and processes have not been used. ● No, it has not been assumed that thematic analysis is essentialist or realist, or atheoretical. ● No, it has not been assumed that TA is only a data reduction or descriptive approach. Are the theoretical underpinnings of the use of TA clearly specified (e.g. ontological, epistemological assumptions, guiding theoretical framework(s)), even when using TA inductively? Yes. Do the researchers strive to ‘own their perspectives’ (even if only very briefly), their personal and social standpoint and positioning? Yes. Do the researchers strive to ‘own their perspectives’ (even if only very briefly), their personal and social standpoint and positioning? Yes. Are the analytic procedures used clearly outlined, and described in terms of what the authors actually did, rather than generic procedures? Yes. 291 Is there evidence of conceptual and procedural confusion? For example, reflexive TA (e.g. Braun and Clarke 2006) is the claimed approach but different procedures are outlined such as the use of a coding frame, multiple independent coders and consensus coding, inter-rater reliability measures, and/or themes are conceptualised as analytic inputs rather than outputs and therefore the analysis progresses from theme identification to coding (rather than coding to theme development). No. Do the authors demonstrate full and coherent understanding of their claimed approach to TA? Yes. Is it clear what and where the themes are in the report? Would the manuscript benefit from some kind of overview of the analysis: listing of themes, narrative overview, table of themes, thematic map? Yes, a thematic map is available (List of themes, Table 3.3). Are the reported themes topic summaries, rather than ‘fully realised themes’ – patterns of shared meaning underpinned by a central organising concept? No, the themes are not topic summaries, and ‘fully realised themes’ have been reported and discussed at length. Is non-thematic contextualising information presented as a theme? (e.g. the first 'theme' is a topic summary providing contextualising information, but the rest of the themes reported are fully realised themes). No. In applied research, do the reported themes have the potential to give rise to actionable outcomes? Yes. Are there conceptual clashes and confusion in the paper? No. Is there evidence of weak or unconvincing analysis, such as: Too many or two few themes? Too many theme levels? Confusion between codes and themes? Mismatch between data extracts and analytic claims? Too few or too many data extracts? No. Do authors make problematic statements about the lack of generalisability of their results, and or implicitly conceptualise generalisability as statistical probabilistic generalisability (see Smith, 2017)? No. Understanding anticipatory prescribing at the end of life in UK care homes Megha Majumder Declaration Abstract Between paradox and prescribing: Understanding anticipatory prescribing at the end of life in UK care homes Acknowledgements Abbreviations Figures Chapter 1 1.1 Introduction 1.2 Background: Death and dying in UK care homes 1.3 Defining care homes 1.4 Death trends in UK care homes 1.5 The role of AP in a ‘good death’ 1.6 The AP system at the end of life Box 1.1 An overview of the anticipatory prescribing system at the end of life ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ o Anticipatory prescribing as a system o The legislative landscape of AP in UK care homes o Geographic variation across UK care homes o Changes in UK care home finances o Care home staff recruitment challenges ____________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 1.7 Health and social care professionals supporting the AP system 1.8 Personal motivations 1.9 Impact of the COVID-19 pandemic 1.10 Rationale 1.11 Challenges, tension, and paradox 1.12 Core objective & research questions 1.13 A brief overview of the thesis Chapter 2: Reviewing the evidence regarding End-of-Life Controlled Drugs in UK care homes 2.1 Introduction 2.2 Aim 2.3 Review questions 2.4 Methods 2.5 Results 2.6 Discussion and narrative synthesis of content 2.7 Conclusion Part two: Methodology & Methods Chapter 3 3.1 Introduction and research questions 3.2 Ontology & epistemology: Ontology of subjective reality and epistemology of constructivism 3.3 Theoretical perspective: Paradox 3.4 Methods: Research design 3.4.1 Introduction 3.4.2 Public and Patient Involvement 3.4.3 Setting The study took place in care homes across the UK. While interviews were conducted remotely (given the pandemic context), the participants were based in care homes which varied in size, organisational structure, accreditation, and location, ranging fro... 3.4.4 Population Participants eligible for this study were care home managers, care home nurses, and visiting health and social care professionals (including doctors, community nurses, pharmacists, and ambulance personnel) working in UK care homes. The inclusion crite... 3.4.5 Sampling Following principles of purposive sampling (Creswell, 2013), I identified care homes across the UK with good or outstanding Care Quality Commission inspection reports by searching online. I obtained the contact information of the respective managers f... This sampling strategy focused on achieving meaningful variation across stakeholders and care home organisations, based on factors such as professional roles and care home types, to ensure that the dataset could support robust interprofessional and in... 3.4.6 Recruitment The recruitment for the study was carried out in two main phases. In Phase 1, CHMs were the initial point of contact. An introductory email was sent to these managers, providing an overview of the research 'Anticipatory Prescribing in UK Care Homes' w... To initiate Phase 2, care home managers who completed the interview were asked to share the study details with potential participants: care home nurses, visiting doctors, pharmacists, and other professionals. My contact information and Participant Inf... 3.5.1 Interviews Part three: Findings & Discussion Chapter 4 Outcomes of recruitment 4.1 Participant recruitment strategy 4.2 Recruitment outcomes 4.2.1 Reflecting on recruitment outcomes 4.3 Participant and care home characteristics 4.4 Conclusion Chapter 5 5.1 Background 5.2 Introduction 5.3 Wastage vs. Shortage 5.4 Nurse-boundary vs. Doctor-boundary 5.5 Training vs. Support 5.6 Anticipation vs. Reaction 5.7 Uncertainty vs. Certainty 5.8 Holding on vs. Pushing away 5.9 Chapter summary 5.10 Discussion o Contributions to theory 5.11 Conclusion Chapter 6 6.1 Background 6.2 Introduction 6.3 Individualization vs. Standardisation 6.4 Micro-decisions vs. Algorithmic decisions 6.5 Individualized language vs. Jargon oblivion 6.1 6.1 6.6 Complex cases vs. Bureaucratic shortcuts 6.7 Chapter summary 6.8 Discussion 6.9 Conclusion Chapter 7 The Paperwork Paradox 7.1 Background 7.2 Introduction 7.3 Clock-based Time vs. Task-based time 7.4 Granular Certainty vs. Structural Stupidity 7.5 Risk Management vs. Error Multiplication 7.6 Tick-Box Compliance vs. Duty of Candour 7.7 Guidelines vs. Orders 7.8 Chapter summary 7.9 Discussion 7.10 Conclusion Part four: Reflections & Conclusion Chapter 8: Conclusion References Appendices Appendix 1 Published systematic review Appendix 2 Systematic review search strategy CINAHL via Ebsco Appendix 3 Systematic review data extraction template Appendix 4 Email text to potential participants for the interview study Appendix 5 Participant information sheet Appendix 6 Study protocol Appendix 7 Ethics Approval Appendix 8 Interview topic guide Appendix 9 Reflexive journal extracts Appendix 10 Results of Braun & Clarke (2021) ‘20 questions to evaluate the quality of thematic analysis’