Vol.:(0123456789) Neurosurgical Review (2024) 47:878 https://doi.org/10.1007/s10143-024-03089-2 RESEARCH Traumatic posterior fossa extradural hematoma in children: a meta- analysis and institutional experience of its clinical course, treatment and outcomes Keng Siang Lee1,2,3  · Shi Hui Ong3 · Conor S Gillespie4 · Lee Ping Ng3,5 · Wan Tew Seow3,5 · Sharon YY Low3,5,6,7 Received: 18 August 2024 / Revised: 23 September 2024 / Accepted: 4 November 2024 / Published online: 30 November 2024 © The Author(s) 2024 Abstract Posterior fossa extradural hematoma (PFEDH) is rare but has a greater incidence amongst children. It is also associated with a rapid deterioration. The aim of this study was to present the management of PFEDH through our institutional experience and a meta-analysis. A retrospective single institution review of all children from 2004 to 2024 who underwent craniotomy for PFEDH was undertaken. The collected variables included: demographics, type of trauma, clinical findings, computed tomography findings, and clinical course. A systematic review using Ovid Medline, Ovid Embase, and Cochrane Central Register of Controlled Trials (CENTRAL), and meta-analysis were performed. Nineteen children with PFEDH who under- went surgery were identified. All 19 (100%) patients benefited from good Glasgow Outcome Scale (GOS) score 4–5, and there were no incidences of in-hospital mortality. From the systematic review, 391 patients, across twenty-four studies and our series, were included. A total of 308 were treated with surgery, whereas 83 patients were treated conservatively. A comparative meta-analysis was not performed as the two groups were deemed too heterogeneous in clinical characteristics. Instead, single-arm meta-analyses were performed. The pooled incidence of patients initially under conservative manage- ment requiring surgery was 9.90% (95%CI 1.61;22.21%, I2 = 35.2). The incidence of good functional outcomes in patients managed surgically and conservatively were 93.68% (95%CI: 88.69;97.57%, I2 = 0.0%), and 99.99% (95%CI: 96.53;100%, I2 = 0.0%), respectively. Overall pooled of mortality in patients managed surgically and conservatively were 0.57% (95%CI: 0.00;2.87%, I2 = 0.0%) and 0.00% (95%CI: 0.00;1.18%, I2 = 0.0%). Overall, our study reiterates that pediatric PFEDH is uncommon, and patients often present atypically. Based on our institutional experience and extrapolating data from our meta-analysis of the wider literature, neurosurgical intervention is a reliable therapeutic option with good clinical outcomes. Keywords Extradural · Hematoma · Pediatric · Posterior fossa · Traumatic Introduction Traumatic brain injury (TBI) in the pediatric population is a global cause of mortality and chronic disabilities; the latter imposing significant healthcare burden costs [1, 2]. Within the spectrum of TBI, posterior fossa extradural hematomas (PFEDHs) are rare and account for up to 3% of all extra- dural hematomas (EDH) [3–7]. PFEDH are more common in children [8, 9], and of significance, untreated cases are associated with rapid deterioration and demise [6]. However, * Keng Siang Lee mrkengsianglee@gmail.com 1 Department of Neurosurgery, King’s College Hospital, London, UK 2 Department of Basic and Clinical Neurosciences, Maurice Wohl Clinical Neuroscience Institute, Institute of Psychiatry, Psychology and Neuroscience (IoPPN), King’s College London, London, UK 3 Neurosurgical Service, KK Women’s and Children’s Hospital, Singapore, Singapore 4 Department of Neurosurgery, Department of Clinical Neurosciences, University of Cambridge, Cambridge, UK 5 Department of Neurosurgery, National Neuroscience Institute, Singapore, Singapore 6 SingHealth Duke-NUS Neuroscience Academic Clinical Program, Singapore, Singapore 7 SingHealth Duke-NUS Paediatrics Academic Clinical Program, Singapore, Singapore https://orcid.org/0000-0003-2308-0579 http://crossmark.crossref.org/dialog/?doi=10.1007/s10143-024-03089-2&domain=pdf Neurosurgical Review (2024) 47:878878 Page 2 of 20 PFEDH tends to present atypically – leading to the risks of dire consequences of a late or missed diagnosis [10]. None- theless, the widespread availability of computed tomogra- phy (CT) brain scans has led to better diagnostic capability, earlier neurosurgical intervention and overall, improved prognosis [5, 8, 11–14]. To date, most cases of traumatic PFEDH in children have been limited to case reports and small series [3–5, 7, 8, 11, 12, 15–31]. The aims of this study were hence twofold: (1) to present on the management of PFEDH through our institutional experience and (2) to corroborate our findings with the first meta-analysis of the published literature. Methods Data collection This is an ethics-approved, retrospective study conducted at the KK Women’s and Children’s Hospital (SingHealth CIRB Reference: 2020/2632) with data obtained from 2008 to 2024. Patients less than 18 years old, with a diagnosis of traumatic PFEDH confirmed on neuroimaging that required intervention by the Neurosurgical Service, KK Women’s and Children’s Hospital, were included. Patients aged above 18 years old, diagnosed with supratentorial EDH, non-traumatic posterior fossa bleeds, and those with incomplete medical information were excluded. In addition, patients with known bleeding disorders and/or on long-term anticoagulation medication were also excluded. Individual patient informa- tion was either obtained from electronic data or hardcopy notes using a standardized data collection form. Imaging details for each patient were obtained from the radiology archives and assessed for completeness. The following vari- ables were collected: patient demographics, Glasgow Coma Scale (GCS) and symptoms on admission, details of injury, the Pediatric Emergency Care Applied Research Network (PECARN) rule [32], radiological features on CT brain scans, and perioperative details of the intervention of choice (Table 1). Overview of neurosurgical workflow and approaches The decision for intervention is in accordance with a stand- ardized workflow for PFEDH. Briefly, a correlation between the clinical condition of the patients and CT scan images is confirmed. The main radiological feature of concern is an expansile clot—that is, PFEDH volume > 10ml, > 15 mm in thickness, with a midline shift of > 5 mm. Associated findings include obliteration of peri-mesencephalic cisterns, fourth ventricular displacement and obstructive hydrocepha- lus (Fig. 1). The approach of choice is either a midline or paramed- ian suboccipital craniectomy or craniotomy [6]. For selected cases, the operating neurosurgeon may perform a single, large burrhole over the epicenter of the clot, especially if the CT scan demonstrates the clot to be more liquefied. Post- operatively, all patients are monitored overnight in either a high dependency or intensive care unit. We do not routinely perform postoperative CT scans if the patient has clinical improvement and adequate evacuation of the hematoma has been observed intraoperatively. Outcomes (case series) The primary outcome measure was functional outcome assessed using the Glasgow Outcome Scale (GOS), GOS- Extended (GOS-E Peds) Score and 90-day modified Rankin Scale (mRS). The secondary outcome was in-hospital mor- tality [33–35]. Statistical analysis Numerical variables were described as median (interquartile range [IQR]) for non-normal distributions. Due to the mod- est sample size of our case series, descriptive analysis rather than statistical analyses to investigate associations, were per- formed to avoid the risk of a Type 2 false negative error. Data were collated in Microsoft Excel (Microsoft, Red- mond, WA, USA). All statistical analyses were performed using R software version 4.2.1 (R Foundation for Statistical Computing, 2022). P-values less than 0.05 were considered statistically significant. Systematic review and meta‑analysis This review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [36]. The protocol was registered on the PROSPERO international prospective register of sys- tematic reviews (registration number CRD42024557754) [37, 38]. Outcomes (systematic review and meta analysis) The primary outcome was good functional outcome defined as GOS 4–5. In studies where other outcome measures were used, such as the modified Rankin scale (mRS) or GOS-E Peds Score, they were translated into good or poor outcomes as defined above. For example, mRS 0–2 or GOS-E Peds 7–8 were translated into good functional outcomes. The second- ary outcome was in-hospital mortality. Neurosurgical Review (2024) 47:878 Page 3 of 20 878 Table 1 Summary characteristics, clinical and radiological features and outcomes of children with PFEDH at our institution *Effect size reported as a median and interquartile range Abbreviations: CT = computed tomography, GCS = Glasgow Coma Scale, IVH = intraventricular hema- toma, LOC = loss of consciousness, RTA = road traffic accident, SAH = subarachnoid hemorrhage, SDH = subdural hematoma Variables of interest Number of patients (%) Demographics Age*, years 5 (IQR 4–7) Male 10 (52.6%) Mechanism of head injury Fall 100 (100.0%) Road traffic accident 0 (0.0%) Interval between trauma and admission ≤ 24 h of injury 11 (57.9%) > 24 to 48 h of injury 3 (15.8%) > 48 to 72 h of injury 5 (26.3%) Presenting Glasgow Coma Scale (GCS) GCS 13 to 15 18 (94.7%) GCS 9 to 12 1 (5.3%) GCS 3 to 8 0 (0.0%) Symptoms reported upon admission Headache 14 (73.7%) Nausea 18 (94.7%) Vomiting 18 (94.7%), Transient loss of consciousness 0 (0.0%) Drowsiness 4 (21.1%) Seizures 0 (0.0%) Asymptomatic 1 (5.3%) Hematoma dimensions on CT brain scan Hematoma thickness > 15 ml 5 (26.3%) Hematoma thickness 5 to 15 ml 14 (73.7%) Hematoma thickness < 5 ml 0 (0.0%) Hematoma volume (ml) 8.5 (IQR 5.7–13.2) Other accompanying intracranial pathology on CT brain scan Occipital bone fracture 18 (100.0%) Intraparenchymal contusion (ICH) 6 (31.6%) Subdural hematoma (SDH) 3 (15.8%) Subarachnoid hemorrhage (SAH) 1 (5.3%) Supratentorial extension of primary hematoma 2 (10.5%) Intraventricular hemorrhage (IVH) 0 (0.0%) Pneumocephalus 2 (10.5%) Mass effect and associated sequalae on CT brain scan Compression of fourth ventricle 15 (78.9%) Obstructive hydrocephalus 16 (84.2%) Obliteration of perimesencephalic cisterns 14 (73.7%) Outcomes Good functional outcome GOS 4-5 19 (100.0%) Good functional outcome GOS-E Peds 7–8 19 (100.0%) Modified Rankins Score 0 (at 90 days post-surgery) 19 (100.%) Death 0 (0.0%) Neurosurgical Review (2024) 47:878878 Page 4 of 20 Search strategy Three electronic databases – Ovid Medline, Ovid Embase, and Cochrane Central Register of Controlled Trials (CEN- TRAL) – were searched. Searches were performed in each electronic database from its inception until 4th June 2024. In addition to its synonyms and related terms, concepts of “pediatric”, “posterior fossa” and “extradural hemorrhage” were used. Supplementary Table 1 presents the full search strategy used for each database. Eligibility criteria Articles were selected for inclusion if they were either a primary interventional or observational study evaluating the management of PFEDH in an exclusively pediatric population (aged less than 18 years old). Additionally, we included mixed age population studies that provided analyses exclusive to the pediatric cohort. Supplementary Table 2 provides the full list of inclusion and exclusion criteria. Study selection Titles and abstracts were independently screened against the pre-defined eligibility criteria developed by two reviewers (KSL and CSG). Any disagreements were resolved through discussion between the reviewers or further adjudication by a third reviewer (SYYL). When encountered with multiple publications analyzing the same cohort over overlapping study periods, the publi- cation that reported the largest patient data of relevant out- comes was used for evaluation. Data extraction To ensure standardization and consistency, a proforma was developed and piloted to extract data on the following vari- ables: (1) study details, (2) study design, (3) country and dataset, (4) selection criteria, (5) patient demographics, (6) treatment/control, (7) indication for treatment, and (8) results. Risk of bias assessment The quality of included studies was assessed using the Joanna Briggs Institute (JBI) checklist for cohort studies and case series [39]. Data synthesis Meta-analyses were performed via the random effects model due to heterogeneity within and between individual studies as well as sampling variabilities across studies [39]. To obtain risk ratios (RRs) from reported binary out- comes, pairwise meta-analysis was conducted using the Mantel-Haenszel method without continuity correction, with the Paule-Mandel estimator. Overall pooled proportions of demographic comorbidities of included patients were com- puted with the generalized linear mixed model (GLMM) Fig. 1 Case example of a pediatric patient with PFEDH from our series. Representative pre-operative CT brain images in axial direc- tion depicting: (A) a left-sided, biconvex hematoma with mixed den- sity causing local mass effect on the underlying parenchyma; (B) the corresponding bone window of (A) showing an occipital bone fracture adjacent to the PFEDH; (C) another image slice that shows radiological evidence of obstructive hydrocephalus, fourth ventricular distortion and effacement of the peri-mesencephalic cisterns. (Abbre- viations: PFEDH = posterior fossa extradural hematoma; CT = com- puted tomography.) Neurosurgical Review (2024) 47:878 Page 5 of 20 878 method using a random intercept logistic regression model via logit transformation [39]. Knapp-Hartung adjustments were used to reduce the chance of a false positive and to control the estimate uncertainties of between-study hetero- geneity [39]. For pooling of means of numerical variables, we com- puted missing means and standard deviations (SDs) from medians, ranges (minimum to maximum) and interquartile ranges (IQRs) using the methods proposed by Hozo et al. and Wan et al. [40, 41]. The I2 statistic was employed to assess inter-study hetero- geneity. I2 provides an estimate of the percentage of variabil- ity in results across studies that is due to real differences and not due to chance. I2 ≤ 30%, 30–50%, 50–75%, and ≥ 75% indicated low, moderate, substantial, and considerable het- erogeneity, respectively. The quality of evidence for each outcome was evaluated using the GRADE framework [39, 42]. Results Overview of study population and patient characteristics A total of 157 pediatric head injury cases underwent neu- rosurgical intervention at our institution. Within this het- erogeneous group, there were 19 (12.1%) cases of PFEDH (63.2% left side). There was no gender predilection (52.3% males) and the median age was 5 years (IQR 4–7 years). A fall from standing height was the most common mechanism of injury (57.9%). Twelve (63.2%) patients presented to hospital only after 24 h of injury. Most patients had full GCS scores (84.2%) at the time of admission. The most common symptoms pre- sented were nausea and vomiting (94.7%), headache (73.7%) and drowsiness (21.1%). No patients experienced a lucid interval. The median volume of the hematoma was 8.5 ml (IQR 5.7–13.2 ml). Occipital bone fractures were observed in all 19 (100.0%) patients. Compression of the fourth ventricles, obliteration of the peri-mesencephalic cisterns and ventricu- lomegaly were observed in 15 (78.9%), 14 (73.7%), and 16 (84.2%) cases respectively. Five (26.3%) and 14 (73.7%) patients underwent a bur- rhole or craniotomy respectively. At the time of surgery, all patients (100.0%) were seen to have bleeding from the occipital fracture lines. No mortality was observed in our series. All 19 (100.0%) patients had good recovery after a median hospital stay of 4 days (IQR 3.5–5 days), with good functional outcome at discharge. There were no surgical complications. The length of follow-up ranged from 3 months to 8 years, and the median follow-up duration was 3 months (IQR 1.5–6.75 months). All patients reported good functional out- comes (GOS-E Peds8 and mRS0) on their last documented follow-up. Systematic review and meta‑analysis Characteristics of included studies The systematic review search yielded 496 unique publica- tions. Twenty-four studies including 372 pediatric patients met the eligibility criteria for inclusion in our meta-analysis (Fig. 2) [3–5, 7, 8, 11, 12, 15–31]. Including our institutional experience, a total of 391 pediatric patients were included. All included studies were retrospective observational studies – five cohort studies and 19 case series. Table 2 pre- sents a detailed summary of the characteristics and find- ings of the included studies. (Supplementary Tables 3 and 4 detail the risk of bias assessment.) Patient baseline characteristics, treatment approach and radiological features Pooled prevalence of baseline characteristics, stratified according to treatment arm, is summarized in Table 3. Of the 391 patients, 308 were treated with surgery, whereas 83 patients were treated conservatively. The pooled proportion of males were 61.51% (95%CI: 52.61; 70.11%, I2 = 0.0% [p = 0.518]) % and 54.13% (95%CI: 22.29; 84.56%, I2 = 33.7% [p = 0.221]) in the surgical and conservative groups, respectively. Overall pooled mean age across the surgical and conservative groups were 7.78 years (95%CI: 6.15;9.41 years, I2 = 94.8%, p < 0.001) and 6.61 years (95%CI: 1.98;11.24 years, I2 = 53.4%, p = 0.117), respectively. The most common symptoms encountered in the surgi- cal group were headache (63.55% (95%CI: 46.96;78.89%, I2 = 84.5% [p < 0.001])) and vomiting (77.69% (95%CI: 67.76;86.51%, I2 = 54.7% [p = 0.002])). The most common symptoms encountered in the conservatively managed group were headache (57.02% (95%CI: 34.27;78.60%, I2 = 58.6% [p = 0.025])) and vomiting (46.59% (95%CI: 29.04;64.50%, I2 = 33.2% [p = 0.174])). Symptoms of transient loss of consciousness (LOC) (36.85% (95%CI: 19.59; 55.73%, I2 = 78.2% [p < 0.001]) versus 12.70% (95%CI: 0.00; 42.51%, I2 = 78.6% [p = 0.001])), drowsiness (29.49% (95%CI: 16.72; 43.71%, I2 = 51.9% [p = 0.018]) versus 7.24% (95%CI: 0.00;35.66%, I2 = 77.8% [p = 0.001])), and seizures (7.95% (95%CI: 0.89;18.91%, I2 = 50.3% [p = 0.060]) versus 8.25% (95%CI: 0.00;40.90%, I2 = 0.0% [p = 0.519])) were more common in the surgical group than the conservative group. Neurosurgical Review (2024) 47:878878 Page 6 of 20 The pooled proportion of mass effective was greater in the surgical group than the conservative group – compression of the fourth ventricle (49.95% (95%CI: 34.77;65.13%, I2 = 52.1% [p = 0.022]) versus 8.25% (95%CI: 0.00;40.90%, I2 = 0.0% [p = 0.519])), ventricu- lomegaly (36.10% (95%CI: 15.57;59.06%, I2 = 79.7% [p < 0.001]) versus 8.25% (95%CI: 0.00;40.90%, I2 = 0.0% [p = 0.519])), obliteration of the perimesencephalic cisterns (43.42% (95%CI: 21.53;66.48%, I2 = 77.0% [p < 0.001]) versus 8.25% (95%CI: 0.00;40.90%, I2 = 0.0% [p = 0.519])). Following that, the pooled proportion of associated pathology was greater in the surgical group – occipital bone fracture (76.30% (95%CI: 64.20;86.89%, I2 = 72.5% [p < 0.001]) versus 66.88% (95%CI: 52.61;79.93%, I2 = 0.0% [p = 0.474])), contusion (17.64% (95%CI: 10.62;25.72%, I2 = 34.6% [p = 0.131]) versus 3.80% (95%CI: 0.01;11.39%, I2 = 0.0% [p = 0.723])) and pneumocephalus (9.41% (95%CI: 3.86;16.42%, I2 = 1.3% [p = 0.408]) versus 3.18% (95%CI: 0.00;17.75%, I2 = 0.0% [p = 0.340])). Fig. 2 PRISMA flow dia- gram for studies included and excluded from the systematic review and meta-analysis Neurosurgical Review (2024) 47:878 Page 7 of 20 878 Ta bl e 2 S um m ar y of in cl ud ed st ud ie s Fi rs t a ut ho r a nd ye ar C ou nt ry St ud y de si gn St ud y pe rio d In di ca tio ns fo r su rg er y To ta l pa tie nt s in cl ud ed , n Su rg ic al , n C on se rv at iv e, n Im pu te d m ea n ag e at su rg er y (S D ), ye ar * M al e, n (% ) G oo d fu nc tio n ou tc om es G O S4 - 5, n (% ) D ea th , n (% ) A m m ira ti M et  a l. 19 84 U SA C as e se rie s 19 76 to 1 98 1 N R 4 4 N R 5. 62 (3 .4 9) 1 (2 5% ) 4 (1 00 % ) 0 (0 .0 % ) B el lo tti C e t a l. 19 87 Ita ly C as e se rie s N R La rg e he m at o- m as a nd p at ie nt is sy m pt om at ic 5 4 1 10 .0 0 (6 .4 0) 4 (8 0% ) 5 (1 00 % ) 0 (0 .0 % ) B er ke r M e t a l. 20 03 Tu rk ey C as e se rie s 19 71 to 2 00 0 W he n th e di ag no si s i s es ta bl is he d, su rg er y sh ou ld be p er fo rm ed as so on a s p os - si bl e ev en if th e ne ur ol og ic al ex am in at io n is no rm al . 16 16 N R 7. 74 (4 .2 9) 8 (5 0% ) 14 (8 7. 5% ) 1 (6 .2 5% ) B oz bu ga M e t a l. 19 99 Tu rk ey C as e se rie s 19 82 to 1 99 2 O bl ite ra tio n of th e pe rim es - en ce ph al ic ci ste rn s, co m - pr es si on a nd /o r di sp la ce m en t of th e fo ur th ve nt ric le , a nd th e pr es en ce o f hy dr oc ep ha lu s 9 9 N R N R 9 (1 00 % ) 0 (0 .0 % ) B ra m bi lla G et  a l. 19 86 Ita ly C as e se rie s 19 79 to 1 98 5 N R 4 4 N R 10 .2 5 (5 .3 1) N R 3 (7 5% ) 1 (2 5. 0% ) C ha og uo Y e t a l. 20 19 C hi na C oh or t s tu dy 20 01 to 2 00 5 C lo t v ol um e of ≥ 15 m l o r pr es en ce o f fo ur th v en tri cl e co m pr es si on o r di sp la ce m en t an d/ or o bs tru c- tiv e ve nt ric ul o- m eg al y 48 17 31 6. 86 (4 .5 5) 24 (5 0% ) 46 (9 5. 83 % ) 0 (0 .0 % ) Neurosurgical Review (2024) 47:878878 Page 8 of 20 Ta bl e 2 (c on tin ue d) Fi rs t a ut ho r a nd ye ar C ou nt ry St ud y de si gn St ud y pe rio d In di ca tio ns fo r su rg er y To ta l pa tie nt s in cl ud ed , n Su rg ic al , n C on se rv at iv e, n Im pu te d m ea n ag e at su rg er y (S D ), ye ar * M al e, n (% ) G oo d fu nc tio n ou tc om es G O S4 - 5, n (% ) D ea th , n (% ) C iu re a AV e t a l. 19 93 Ro m an ia C as e se rie s 19 87 to 1 99 2 Su rg ic al tr ea t- m en t m us t b e ca rr ie d ou t im m ed ia te ly w he n th e di ag - no si s h as b ee n es ta bl is he d 9 9 N R 7. 31 (2 .8 9) 6 (6 6. 67 % ) 8 (8 8. 89 % ) 0 (0 .0 % ) C os ta C la ra JM et  a l. 19 96 Sp ai n C as e se rie s 19 89 to 1 99 4 O nc e th e di ag no si s o f a PF ED H is m ad e, su rg er y is m an da to ry 3 3 N R 12 .6 6 (5 .7 7) 1 (3 3. 33 % ) 3 (1 00 % ) 0 (0 .0 % ) Ec ha ra M e t a l. 20 23 In di a C as e se rie s 20 16 to 2 02 1 PF ED H ≥ 15 m l, or th e pr es en ce of m as s e ffe ct (c om pr es si on of th e fo ur th ve nt ric le a nd /o r hy dr oc ep ha lu s) 40 40 N R 15 .2 3 (3 .9 6) 32 (8 0% ) 35 (8 7. 5% ) 1 (2 .5 % ) Er sa hi n Y e t a l. 19 93 Tu rk ey C as e se rie s 19 79 to 1 99 1 O nc e th e di ag no - si s o f a P FE D H is m ad e, th e sy m pt om at ic he m at om as ar e ev ac ua te d im m ed ia te ly 9 9 N R 8. 46 (4 .3 3) 5 (5 5. 56 % ) 8 (8 8. 89 % ) 1 (1 1. 1% ) G ar za -M er ca do R e t a l. 19 83 M ex ic o C as e se rie s N R N R 6 6 N R 8. 79 (6 .1 4) 4 (6 6. 67 % ) 5 (8 3. 33 % ) 0 (0 .0 % ) G up ta P K e t a l. 20 02 O m an C as e se rie s 19 92 to 2 00 0 H em at om a > 10 m l i n vo lu m e, > 15 m m in th ic kn es s, w ith a m id lin e sh ift of > 5 m m , m ed ul la ry co m pr es si on an d as so ci at ed in tra cr an ia l le si on s 18 18 N R N R 14 (7 7. 78 % ) 15 (8 3. 33 % ) 0 (0 .0 % ) H an K e t a l. 20 18 C hi na C as e se rie s 20 12 to 2 01 5 H em at om a > 15 m l i n vo lu m e 8 5 3 4. 63 (1 .0 6) 4 (5 0% ) 8 (1 00 % ) 0 (0 .0 % ) Neurosurgical Review (2024) 47:878 Page 9 of 20 878 Ta bl e 2 (c on tin ue d) Fi rs t a ut ho r a nd ye ar C ou nt ry St ud y de si gn St ud y pe rio d In di ca tio ns fo r su rg er y To ta l pa tie nt s in cl ud ed , n Su rg ic al , n C on se rv at iv e, n Im pu te d m ea n ag e at su rg er y (S D ), ye ar * M al e, n (% ) G oo d fu nc tio n ou tc om es G O S4 - 5, n (% ) D ea th , n (% ) Ja m ou s M A e t a l. 20 21 Jo rd an C oh or t s tu dy 20 10 to 2 02 1 La rg e he m at om a th ic kn es s ( ≥ 15 m m ) o r t he pr es en ce o f fo ur th v en tri cl e co m pr es si on or o bs tru ct iv e ve nt ric ul om eg - al y or a G C S of < 14 16 7 9 7. 70 (6 .0 0) 12 (7 5% ) 14 (8 7. 5% ) 1 (6 .2 5% ) Ja ng JW e t a l. 20 10 K or ea C oh or t s tu dy 20 01 to 2 00 8 La rg e he m at o- m as w ith m as s eff ec ts , a s w el l as a ss oc ia te d in tra cr an ia l le si on s 19 9 10 7. 94 (3 .9 0) 10 (5 2. 63 % ) 18 (9 4. 74 % ) 0 (0 .0 % ) K oç R K e t a l. 19 98 Tu rk ey C as e se rie s 19 83 to 1 99 4 N R 9 9 N R 9. 44 (4 .8 7) 5 (5 5. 56 % ) 7 (7 7. 78 % ) 2 (2 2. 22 % ) Lu i T N e t a l. 19 93 Ta iw an C as e se rie s 19 77 to 1 98 9 O nc e th e di ag no - si s o f a P FE D H is m ad e, a ll ar e ev ac ua te d im m ed ia te ly 29 29 N R N R N R N R 3 (1 0. 34 % ) M ia o Z et  a l. 20 23 C hi na C as e se rie s 20 16 to 2 02 1 Tr ea tm en t w ith su rg ic al e va cu - at io n de pe nd ed on th e cl in ic al st at us o f t he pa tie nt s a s w el l as th e pr es en ta - tio n of C T im ag e 40 40 N R 5. 83 (1 .0 8) 24 (6 0% ) 32 (8 0% ) N R M or i K e t a l. 19 83 Ja pa n C as e se rie s N R H em at om as > 20 m l a nd p at ie nt is sy m pt om at ic 3 3 N R 5. 66 (1 .1 5) 1 (3 3. 33 % ) 3 (1 00 % ) 0 (0 .0 % ) Neurosurgical Review (2024) 47:878878 Page 10 of 20 Ta bl e 2 (c on tin ue d) Fi rs t a ut ho r a nd ye ar C ou nt ry St ud y de si gn St ud y pe rio d In di ca tio ns fo r su rg er y To ta l pa tie nt s in cl ud ed , n Su rg ic al , n C on se rv at iv e, n Im pu te d m ea n ag e at su rg er y (S D ), ye ar * M al e, n (% ) G oo d fu nc tio n ou tc om es G O S4 - 5, n (% ) D ea th , n (% ) Pa ng D e t a l. 19 83 U SA C as e se rie s 19 78 to 1 98 2 C lin ic al si gn s of lo ca l b ra in co m pr es si on or h er ni a- tio n (s uc h as in cr ea si ng dr ow si ne ss , pu pi lla ry ab no rm al ity , he m ip le gi a, de ce re br at e po stu rin g) , o r ca rd io re sp ira - to ry a bn or m al i- tie s 3 N R 3 7. 33 (5 .8 5) N R 3 (1 00 % ) 0 (0 .0 % ) Pe te r J C e t a l. 19 90 So ut h A fr ic a C as e se rie s 19 79 to 1 98 9 N R 5 5 N R 7. 00 (1 .5 8) 2 (4 0% ) 5 (1 00 % ) N R Pr as ad G e t a l. 20 15 In di a C oh or t s tu dy 20 08 to 2 01 4 H em at om a w ith a vo lu m e of > 20 m l o r ra di ol og ic al ev id en ce o f po ste rio r f os sa m as s e ffe ct in th e fo rm o f fo ur th v en tri cl e di sto rti on o r co m pr es si on an d/ or o bs tru c- tiv e ve n- tri cu lo m eg al y, irr es pe ct iv e of th e G C S sc or e 22 16 6 10 .8 1 (4 .6 4) 17 (7 7. 27 % ) 20 (9 0. 91 % ) 0 (0 .0 % ) Neurosurgical Review (2024) 47:878 Page 11 of 20 878 Ta bl e 2 (c on tin ue d) Fi rs t a ut ho r a nd ye ar C ou nt ry St ud y de si gn St ud y pe rio d In di ca tio ns fo r su rg er y To ta l pa tie nt s in cl ud ed , n Su rg ic al , n C on se rv at iv e, n Im pu te d m ea n ag e at su rg er y (S D ), ye ar * M al e, n (% ) G oo d fu nc tio n ou tc om es G O S4 - 5, n (% ) D ea th , n (% ) Se nc er A e t a l. 20 12 Tu rk ey C oh or t s tu dy 19 95 to 2 01 1 H em at om a th ic k- ne ss > 15 m m or h em at om a th ic kn es s be tw ee n 5– 15 m m w ith in ju - rie s/ c on di tio ns (c on tu si on , pn eu m o- ce ph al us , o r SA H ) c au si ng ad di tio na l m as s e ffe ct o r a G C S < 15 40 29 11 7. 88 (4 .6 7) 22 (5 5% ) 40 (1 00 % ) 0 (0 .0 % ) Sh en g H S et  a l. 20 17 C hi na C as e se rie s 20 10 to 2 01 5 C rit er ia fo r st an da rd cr an ie ct om y: he m at om a vo lu m e > 30 m l o r p re se nc e of se ve re m as s eff ec ts li ke c er - eb el la r t on si lla r he rn ia tio n C rit er ia fo r tre ph in at io n m in i-c ra ni ec - to m y: h em a- to m a vo lu m e 10 -3 0 m l, w ith m ild m as s eff ec ts (e .g . fo ur th v en tri cl e co m pr es si on or d is pl ac e- m en t a nd /o r ob str uc tiv e ve nt ric ul om eg - al y) a nd st ab le ne ur ol og ic al co nd iti on s 7 7 N R 5. 70 (2 .3 2) 3 (4 2. 86 % ) 7 (1 00 % ) 0 (0 .0 % ) A bb re vi at io ns : N R = N ot re po rte d, U K = U ni te d K in gd om , U SA = U ni te d St at es Neurosurgical Review (2024) 47:878878 Page 12 of 20 Ta bl e 3 S um m ar y of p oo le d ba se lin e de m og ra ph ic s, et io lo gi es , c lin ic al fi nd in gs o f i nc lu de d pa tie nt s, C T br ai n fin di ng s an d as so ci at ed p at ho lo gy b et w ee n th e tw o gr ou ps (s ur gi ca l v er su s co n- se rv at iv e) Fa ct or Su rg ic al c oh or t ( n = 30 8) C on se rv at iv e co ho rt (n = 83 ) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) D em og ra ph ic s A ge * 16 18 3 7. 78 [6 .1 5; 9. 41 ] 94 .8 < 0. 00 1 Lo w 3 16 6. 61 [1 .9 8; 11 .2 4] 53 .4 0. 11 7 Lo w G en de r 17 20 4 61 .5 1 [5 2. 61 ; 70 .1 1] 0. 0 0. 51 8 Lo w 3 19 54 .1 3 [2 2. 29 ; 84 .5 6] 33 .7 0. 22 1 Lo w M ec ha ni sm o f i nj ur y F al l 17 21 5 75 .4 4 [6 1. 51 ; 87 .4 8] 75 .0 < 0. 00 1 Lo w 4 49 92 .7 0 [8 1. 25 ; 99 .6 1] 0. 0 0. 89 8 Lo w R TA 14 18 2 31 .7 3 [1 9. 79 ; 44 .7 4] 60 .9 0. 00 2 Lo w 4 49 7. 30 [0 .3 9; 18 .7 5] 0. 0 0. 89 8 Lo w Ti m in g W ith in 2 4 h of in ju ry 7 12 5 79 .6 6  [6 9. 68 ; 88 .3 1] 24 .3 0. 24 4 Lo w 4 26 72 .2 3 [5 0. 65 ; 90 .3 0] 11 .5 0. 33 5 Lo w 2 4– 48 h o f in ju ry 4 60 28 .6 9 [0 .0 0; 78 .9 8] 85 .1 0. 00 1 Lo w 2 20 19 .4 8 [3 .8 3; 41 .0 5] 0. 0 0. 42 0 Lo w 4 8– 72 h o f in ju ry 3 55 12 .0 1 [0 .0 0; 43 .0 9] 84 .0 0. 00 2 Lo w 2 20 9. 94 [0 .0 1; 28 .8 1] 0. 0 0. 87 0 Lo w Pr eo pe ra tiv e G C S G C S1 3- 15 14 19 6 65 .8 4 [5 0. 76 ; 79 .6 4] 69 .9 < 0. 00 1 Lo w 6 70 10 0. 00 [9 8. 38 ; 10 0. 00 ] 0. 0 0. 98 7 Lo w G C S3 -8 8 13 2 17 .9 7 [8 .6 1; 29 .3 1] 49 .1 0. 05 6 Lo w 4 56 0. 00 [0 .0 0; 2. 67 ] 0. 0 0. 94 6 Lo w Sy m pt om s H ea da ch e 20 26 6 63 .5 5 [4 6. 96 ; 78 .8 9] 84 .5 < 0. 00 1 Lo w 7 73 57 .0 2 [3 4. 27 ; 78 .6 0] 58 .6 0. 02 5 Lo w V om iti ng 20 26 6 77 .6 9 [6 7. 76 ; 86 .5 1] 54 .7 0. 00 2 Lo w 7 73 46 .5 9 [2 9. 04 ; 64 .5 0] 33 .2 0. 17 4 Lo w T ra ns ie nt LO C 12 18 5 36 .8 5 [1 9. 59 ; 55 .7 3] 78 .2 < 0. 00 1 Lo w 5 60 12 .7 0 [0 .0 0; 42 .5 1] 78 .6 0. 00 1 Lo w D ro w si ne ss 12 13 1 29 .4 9 [1 6. 72 ; 43 .7 1] 51 .9 0. 01 8 Lo w 5 58 7. 24 [0 .0 0; 35 .6 6] 77 .8 0. 00 1 Lo w S ei zu re s 7 11 1 7. 95 [0 .8 9; 18 .9 1] 50 .3 0. 06 0 Lo w 2 9 8. 25 [0 .0 0; 40 .9 0] 0. 0 0. 51 9 Lo w Neurosurgical Review (2024) 47:878 Page 13 of 20 878 Ta bl e 3 (c on tin ue d) Fa ct or Su rg ic al c oh or t ( n = 30 8) C on se rv at iv e co ho rt (n = 83 ) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) A sy m pt o- m at ic 6 12 8 5. 26 [0 .0 0; 16 .3 8] 73 .9 0. 00 2 Lo w 4 57 22 .0 4 [1 1. 19 ; 34 .7 9] 0. 0 0. 97 9 Lo w Si gn s C er eb el la r si gn s 7 74 39 .8 8 [1 9. 87 ; 61 .4 2] 55 .3 0. 03 7 Lo w 2 14 19 .4 2 [0 .8 3; 47 .8 8] 0. 0 0. 55 0 Lo w B ab in sk i’s si gn 4 22 40 .6 4 [1 8. 57 ; 64 .3 7] 0. 0 0. 90 9 Lo w N A N A N A N A N A N A P ap ill oe de m a 4 25 24 .8 1 [2 .4 2; 55 .8 1] 46 .1 0. 13 5 Lo w N A N A N A N A N A N A A ni sc or ia 3 74 7. 75 [2 .1 4; 15 .6 5] 0. 0 0. 86 1 Lo w N A N A N A N A N A N A A bd uc en s pa ls y 5 31 24 .2 7 [7 .1 9; 45 .5 4] 10 .5 0. 34 6 Lo w N A N A N A N A N A N A O cc ip ita l sw el lin g 5 60 44 .2 2 [2 3. 39 ; 65 .9 7] 46 .9 0. 11 0 Lo w N A N A N A N A N A N A A ss oc ia te d pa th ol og y O cc ip ita l bo ne fr ac - tu re 20 25 7 76 .3 0 [6 4. 20 ; 86 .8 9] 72 .5 < 0. 00 1 Lo w 5 67 66 .8 8 [5 2. 61 ; 79 .9 3] 0. 0 0. 47 4 Lo w C on tu si on 10 20 1 17 .6 4 [1 0. 62 ; 25 .7 2] 34 .6 0. 13 1 Lo w 5 67 3. 80 [0 .0 1; 11 .3 9] 0. 0 0. 72 3 Lo w S D H 8 17 6 6. 85 [2 .7 2; 12 .2 0] 5. 0 0. 39 1 Lo w 3 26 8. 80 [0 .0 0; 31 .5 1] 41 .8 0. 18 0 Lo w S A H 5 11 4 9. 01 [3 .8 5; 15 .6 1] 0. 0 0. 90 5 Lo w 3 52 4. 49 [0 .0 1; 13 .2 6] 0. 0 0. 62 6 Lo w IV H 3 66 2. 51 [0 .0 0; 9. 18 ] 19 .4 0. 28 9 Lo w N A N A N A N A N A N A P ne um o- ce ph al us 6 11 8 9. 41 [3 .8 6; 16 .4 2] 1. 3 0. 40 8 Lo w 2 21 3. 18 [0 .0 0; 17 .7 5] 0. 0 0. 34 0 Lo w C T fin di ng s H em at om a th ic k- ne ss > 15 m l 6 85 52 .8 4 [3 2. 12 ; 73 .1 4] 65 .1 0. 01 4 Lo w 4 27 3. 47 [0 .0 0; 32 .1 9] 57 .8 0. 06 9 Lo w H em at om a th ic kn es s 5- 15 m l 6 85 12 .9 5 [0 .0 0; 51 .0 1] 88 .3 < 0. 00 1 Lo w 4 27 19 .3 3 [0 .0 0; 65 .7 8] 81 .5 0. 00 1 Lo w Neurosurgical Review (2024) 47:878878 Page 14 of 20 Ta bl e 3 (c on tin ue d) Fa ct or Su rg ic al c oh or t ( n = 30 8) C on se rv at iv e co ho rt (n = 83 ) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) N um be r of st ud ie s re po rti ng th is va ria bl e N um be r o f pa tie nt s/ an al ys ed Po ol ed p ro - po rti on [9 5% C I] I2 (% ) p- va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) H em at om a th ic kn es s < 5 m l 6 85 56 .9 4 [3 6. 54 ; 76 .3 3] 64 .5 0. 01 5 Lo w 3 24 49 .8 3 [0 .0 0; 10 0. 00 ] 91 .0 < 0. 00 1 Lo w M ea n he m a- to m a vo lu m e (m l)* 10 16 9 23 .7 4 [1 6. 19 ; 31 .3 0] 93 .5 < 0. 00 1 Lo w 6 62 10 .4 6 [6 .1 6; 14 .7 6] 87 .9 < 0. 00 1 Lo w M as s e ffe ct C om pr es si on of fo ur th ve nt ric le 11 12 2 49 .9 5 [3 4. 77 ; 65 .1 3] 52 .1 0. 02 2 Lo w 3 54 0. 00 [0 .0 0; 1. 60 ] 0. 0 0. 90 0 Lo w V en tri cu lo - m eg al y 10 11 6 36 .1 0 [1 5. 57 ; 59 .0 6] 79 .7 < 0. 00 1 Lo w 4 54 0. 00 [0 .0 0; 2. 93 ] 25 .7 0. 25 7 Lo w O bl ite ra tio n of p er si - m en ce ph al ic ci ste rn s 9 11 0 43 .4 2 [2 1. 53 ; 66 .4 8] 77 .0 < 0. 00 1 Lo w 4 54 0. 00 [0 .0 0; 2. 93 ] 25 .7 0. 25 7 Lo w *E ffe ct s iz e is re po rte d as a m ea n. A bb re vi at io ns : C T = co m pu te d to m og ra ph y, G C S = G la sg ow C om a Sc al e, IV H = in tra ve nt ric ul ar h em at om a; L O C = lo ss o f c on sc io us ne ss , R TA = ro ad tr affi c ac ci de nt , S AH = su ba ra ch no id h em or rh ag e, S D H = su bd ur al h em at om a Neurosurgical Review (2024) 47:878 Page 15 of 20 878 The mean hematoma volume was greater in the surgi- cal group (23.74 ml (95%CI: 16.19;31.30 ml, I2 = 93.5%, p < 0.001)) compared with the conservatively managed group (10.46 ml (95%CI: 6.16;14.76  ml, I2 = 87.9%, p < 0.001)) (Table 3). Conversion to surgery when initially managed conservatively Conversion to surgery from conservative management was reported in five studies across 77 patients initially man- aged conservatively. Reasons cited were enlargement of the hematoma and/or clinical deterioration. The overall pooled incidence of conversion was 9.90% (95%CI: 1.61;22.21%, I2 = 35.2%, p = 0.187). Rates of good functional outcomes and mortality A comparative meta-analysis of outcomes was not per- formed as the two groups were deemed too heterogeneous in terms of baseline clinical characteristics. Instead, single- arm meta-analyses were performed (Table 4). The overall pooled proportion of good functional outcomes across 279 patients between these studies, and our series, managed surgically was 93.68% (95%CI: 88.69;97.57%, I2 = 0.0% [p = 0.46]) (Fig. 3). The incidence of good functional out- comes was reported in eight studies assessing 82 patients managed conservatively. Overall pooled incidence was 99.99% (95%CI: 96.53,100%, I2 = 0.0% [p = 0.62]) (Sup- plementary Fig. 1). The overall pooled incidence of mor- tality across 263 patients between these studies, and our series, managed surgically was 0.57% (95%CI: 0.00;2.87%, I2 = 0.0% [p = 0.71]). The incidence of mortality was reported in eight studies assessing 82 patients managed con- servatively. Overall pooled incidence was 0.00% (95%CI: 0.00,1.18%, I2 = 0.0% [p = 0.99]). Discussion Overview of PFEDH in the pediatric population Amidst the spectrum of head injuries treated by neurosur- geons, the entity of PFEDHs is relatively uncommon [11, 14, 15, 17]. Nonetheless, a late diagnosis is associated with a high mortality rate [14]. Despite published studies report- ing a higher incidence of PFEDH in children in comparison to adults, children tend to fare better with prompt interven- tion [8, 9, 30]. PFEDHs are usually associated with occipital bone fractures. Here, the anatomical disadvantage is that children tend to have larger occipital sinuses; hence, tor- rential venous bleeding may be encountered in fractures crossing the midline with tearing of adjacent venous sinuses. Under such circumstances, bleeding leads to progressive stripping of the dura from the inner table of the skull [43]. This expanding PFEDH could be rapidly fatal due to the compromise of a smaller posterior cranial fossa, leading to brainstem compression, tonsillar herniation and obstructive hydrocephalus [3, 8, 44]. Specifically for PFEDH, one of the main challenges is that affected patients are likely to present in an insidious fashion and may not manifest the classic textbook descrip- tion of an initial loss of consciousness followed by a lucid interval, subsequently developing an ipsilateral pupillary dilation and progressive obtundation [10, 45]. Clinical assessment in young children is often difficult due to their variable, non-specific complaints, especially in the preverbal age group [46–48]. Owing to their smaller posterior fossa, precipitous drops in the neurological status may occur with- out forewarning [6, 43]. Congruently, we observed that most patients had full GCS and no focal neurological deficit in our cohort, at the time of admission. Following that, headache, and or vomiting are the most common presenting complaints in our study, aligning with other publications [3, 5, 11, 15, 23]. Therefore, a diagnosis of PFEDH should be considered if there is a history of head- ache, vomiting and trauma to the occipital region, even for relatively stable GCS on admission. In corroboration with existing studies, we practiced a low threshold for neuro- surgical intervention. This is to mitigate the risk of sudden and rapid clinical deterioration, especially for symptomatic patients with existing radiological evidence of mass effect [7, 17]. Imaging characteristics in PFEDH The current gold standard for neuroimaging in TBI is a CT brain to reliably exclude any life-threatening intracranial hemorrhage [11, 49]. The impact of an early CT upon the treatment of PFEDH is easily recognized—that is, diagnosis can be established quickly to initiate intervention; hence, affording better prognosis [5]. As mentioned previously, occipital linear fractures are common in up to 80% of cases of PFEDH. Occasionally, the hematoma may be undetected on the initial CT scan if the scan is performed too early because PFEDHs tend to appear less dense [13]. This is due to the fact that the bleed tends to be venous and liquefies earlier than supratentorial hematomas [50]. Extension of the PFEDH to the supratentorial compartment is another com- mon radiological finding [4]. In our cohort, we noted two (10.5%) had such extensions, comparable to other reports [4, 23, 51]. The presence of obstructive hydrocephalus is a poor prognosticator [3, 8, 14, 15, 23, 26, 52–54]. Presence of other accompanying lesions such as contusions, subdural hematomas, subarachnoid hemorrhage, intraventricular bleeds and pneumocephalus have been reported [3–5, 7, 10, Neurosurgical Review (2024) 47:878878 Page 16 of 20 Fig. 3 Forest plots, with random-effects model, of pooled percentage of (A) good functional outcome (GOS 4–5), and (B), in-hospital mortality, in patients PFEDH who underwent surgery, respectively Neurosurgical Review (2024) 47:878 Page 17 of 20 878 11, 23]. In our study, 16 patients (84.2%) had radiological features of early obstructive hydrocephalus and six (31.6%) had other accompanying bleeds and/or pneumocephalus. Nonetheless, we did not observe a poorer outcome for the subgroup of patients. Towards an evidence‑based management strategy As reflected in our population, time lapse from injury may be delayed up to a few days from initial injury [17]. Early diagnosis relies on astute clinical observations and failure to recognize the more subtle clinical presentation in children is a genuine concern [10]. To mitigate this, some authors have suggested serial control scans at various time-points post-TBI, for cases where the clinical history is suspicious of PFEDH even if no bleed is obvious on earlier imaging [5, 11]. We are cognizant that most clinicians are mindful of exposing young patients to excessive ionizing radiation in view of the risks of future malignancies [55–58]. Other pertinent challenges associated with CT scans include the possible need for sedation during the scan and parental preferences [58]. Levelling on these opposing points, we advocate that an early CT scan is a reasonable option for selected patients versus a more conservative approach, con- sidering the potential for rapid deterioration. Interestingly, we also observe the limitations of the PECARN rule in our series. This is a prediction score used to assess the need for CT scans children with a history of minor head trauma that has been validated in large, multi-centred studies – the PECARN rule provides the highest level of internationally valid guidance available [59, 60]. However, here, only four patients (21%) of our children underwent initial CT scan based on PECARN predictions. We hypothesize that the role of PECARN may have a limited negative predictive value in this specific cohort of patients with PFEDH, due to the initial subtle nature of PFEDH presentation. This lim- ited negative predictive value may be attributed to the fact that the PECARN rule was derived from a wider population that included a spectrum of traumatic head injuries of vary- ing severity [59]. With the limited utility of the PECARN in PFEDH, the infrequent occurrence of PFEDH and its extended history of evolution from comparatively low- impact fall mechanisms, affected children may initially be cleared with false reassurance – leading to false negative cases with dire consequences. A revised sub-criteria to rec- ognize their diagnosis early may be more suitable for them. However, this requires input from international workgroups and external validation in future studies. The traditional treatment of PFEDH is surgical evacu- ation. Balancing the of risk of a rapid deterioration with conservative management versus performing a relatively safe prophylactic surgery, surgery has conventionally been recommended even if the child is initially asymptomatic [4, Ta bl e 4 P oo le d ou tc om es o f i nc lu de d pa tie nt s b et w ee n th e tw o gr ou ps (s ur gi ca l v er su s c on se rv at iv e) O ut co m es Su rg ic al (n = 30 8) C on se rv at iv e (n = 83 ) N o. o f s tu di es re po rti ng v ar i- ab le N o. o f pa tie nt s/ an al yz ed Po ol ed p ro po r- tio n [9 5% co nfi de nc e in te rv al ] I2 (% ) P va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) N o. o f s tu di es re po rti ng v ar i- ab le N o. o f pa tie nt s/ an al yz ed Po ol ed p ro po r- tio n [9 5% co nfi de nc e in te rv al ] I2 (% ) P va lu e of I2 (f ro m χ 2 te st) Q ua lit y of Ev id en ce (G R A D E) G oo d fu nc tio na l ou tc om e 22 27 9 93 .6 8 [8 8. 69 ; 97 .5 7] 0. 0 0. 46 1 Lo w 8 82 99 .9 9 [9 6. 53 ; 10 0. 00 ] 0. 0 0. 62 4 Lo w D ea th 21 26 3 0. 57 [0 .0 0; 2 .8 7] 0. 0 0. 70 6 Lo w 8 82 0. 00 [0 .0 0; 1 .1 8] 0. 0 0. 99 7 Lo w Neurosurgical Review (2024) 47:878878 Page 18 of 20 43]. However, no surgery is risk-free and torrential venous bleeding may be encountered intraoperatively especially with associated fractures crossing the midline with tearing of adjacent venous sinuses. To date, there is no clear, defined criteria for deciding between conservative and surgical treat- ment [5, 13, 17]. We observe the following CT brain charac- teristics are often used: hematoma > 10 ml in volume or > 15 mm in thickness, with a midline shift of > 5 mm, or mass effect of peri-mesencephalic cisterns obliteration, fourth ventricle displacement, and or presence of hydrocephalus [5, 14, 21, 23, 52, 53, 61]. Generally, most studies emphasize that enlarging hematomas and neurological deterioration are urgent indications for surgery due to the risk of rapid com- promise in the posterior fossa. In our meta-analysis, a fifth of patients were managed conservatively with close neuro- monitoring and serial imaging, whilst the rest were treated surgically. The prognosis of PFEDH is reported to be good [3–5, 7, 8, 11, 12, 15–31]. Regardless of either approach, studies generally report good prognosis in pediatric PFEDH [3–5, 7, 10, 11, 15, 19, 21, 23]. In addition, the rate of good postoperative outcomes imply that surgery does not further morbidity or mortality to the existing neurological state. For the purposes of our study, we did not perform a com- parative meta-analysis between the conservative and surgical cohorts as the clinical characteristics were too heterogene- ous between both groups. In addition, an established criteria for either of the treatments remain to be elucidated, which we have summarized in Table 2 [3–5, 7, 8, 11, 12, 15–31]. Future research could address this current paucity in evi- dence with direct head-to-head comparisons. Limitations and future directions Owing to its rare entity, the pediatric cases of traumatic PFEDH published in the literature are limited to case reports and small series. Henceforth, the limitations of our meta- analysis stem from inclusion of small retrospective obser- vational studies and the heterogeneity across them [62–64]. Restricted by the modest number of suitable studies and small number of events, our ability to delineate risk factors for poor outcomes was also limited, largely due to the risk of Type 2 false negative errors. Nonetheless, congruent find- ings from our study and meta-analysis do demonstrate the effectiveness of neurosurgical intervention in children with PFEDH, as long as they are interpreted judiciously with the aforementioned limitations. Separately, we want to highlight that patients in this age group tend to live longer, and significant brain trauma leads to long-term consequences that may permanently compro- mise their motor development and cognitive abilities [2]. The acquired injury in a developing brain can disrupt subsequent its development, with delayed consequences emerging over time, causing significant short- and long-term alterations in a wide range of functional abilities [65–67]. Longitudinal studies have reported that most children experience persis- tent disability at one year after severe TBI, especially in the cognitive, behavioural, emotional and social domains [67, 68]. Building on this knowledge, we recommend future stud- ies to consider the long-term quality of life and the major factors limiting the resumption of daily activities in PFEDH patients [14]. Conclusion Overall, our study reiterates that pediatric PFEDH is uncom- mon, and patients often present atypically. Based on our institutional experience and extrapolating data from our meta-analysis of the wider literature, neurosurgical inter- vention is a reliable therapeutic option with good clinical outcomes. Emphasis is on early recognition of their subtle symptoms by mindful clinicians to initiate early neuroimag- ing. Future multi-centre collaborations may be warranted to propose a standardized criterion for conservative manage- ment of PFEDH. As the way forward, we advocate contin- ued efforts to educate the public on pediatric head injury prevention. Supplementary Information The online version contains supplemen- tary material available at https:// doi. org/ 10. 1007/ s10143- 024- 03089-2. Acknowledgements NIL. Author contributions All authors listed have made substantial, direct, and intellectual contribution to the work and approved it for publi- cations. Keng Siang Lee: Conceptualization, Methodology, Formal analysis and investigation, Writing - original draft preparation, Writ- ing - review and editing, Visualization, Funding acquisition. Shi Hui Ong: Data curation. Conor S Gillespie: Formal analysis and investiga- tion. Lee Ping Ng: Data curation. Wan Tew Seow: Writing - review and editing, Supervision. Sharon YY Low: Conceptualization, Data curation, Writing - review and editing. Funding Open Access funding provided to Dr Keng Siang Lee by King’s College London. Data availability No datasets were generated or analysed during the current study. Declarations Ethical approval Not applicable. Competing interests The authors declare no competing interests. Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are https://doi.org/10.1007/s10143-024-03089-2 Neurosurgical Review (2024) 47:878 Page 19 of 20 878 included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References 1. Bowman SM et al (2008) Trends in hospitalizations associated with pediatric traumatic brain injuries. Pediatrics 122(5):988–993 2. Chong SL et al (2016) A prospective surveillance of paediat- ric head injuries in Singapore: a dual-centre study. BMJ Open 6(2):e010618 3. Berker M, Cataltepe O, Ozcan OE (2003) Traumatic extradural hematoma of the posterior fossa in childhood: 16 new cases and a review of the literature. Br J Neurosurg 17(3):226–229 4. Prasad GL et al (2015) Traumatic pediatric posterior Fossa rxtra- dural hematomas: a tertiary-care Trauma Center experience from India. Pediatr Neurosurg 50(5):250–256 5. Sencer A et al (2012) Posterior fossa extradural hematomas in children: clinical experience with 40 cases. J Neurosurg Pediatr 9(2):139–143 6. Balik V et al (2010) Posterior fossa extradural haematomas. Cent Eur Neurosurg 71(4):167–172 7. Jamous MA et al (2021) Management of traumatic posterior fossa extradural hematomas in pediatrics: our experience and review of the literature. Childs Nerv Syst 37(9):2839–2846 8. Ciurea AV et al (1993) Posterior fossa extradural hematomas in children: report of nine cases. Childs Nerv Syst 9(4):224–228 9. Mohanty A et al (1995) Prognosis of extradural haematomas in children. Pediatr Neurosurg 23(2):57–63 10. Pillay R, Peter JC (1995) Extradural haematomas in children. S Afr Med J 85(7):672–674 11. Chaoguo Y et al (2019) Traumatic posterior Fossa Extradural hematomas in children: experience with 48 cases and a review of the literature. J Korean Neurosurg Soc 62(2):225–231 12. Jang JW et al (2011) Traumatic extradural hematomas of the pos- terior cranial fossa. Br J Neurosurg 25(1):55–61 13. Karasu A et al (2008) Traumatic extradural hematomas of the pos- terior cranial fossa. Surg Neurol 69(3):247–51; dicussion 251–2 14. Bor-Seng-Shu E et al (2004) Extradural hematomas of the poste- rior cranial fossa. Neurosurg Focus 16(2):ECP1 15. Ammirati M, Tomita T (1984) Posterior fossa extradural hema- toma during childhood. Neurosurgery 14(5):541–544 16. Bellotti C et al (1987) Chronic extradural hematomas of the pos- terior cranial fossa. Surg Neurol 27(6):580–584 17. Bozbuğa M et al (1999) Posterior fossa extradural hematomas: observations on a series of 73 cases. Neurosurg Rev 22(1):34–40 18. Brambilla G et al (1986) Extradural haematoma of the posterior fossa: a report of eight cases and a review of the literature. Acta Neurochir (Wien) 80(1–2):24–29 19. Costa Clara JM et al (1996) Traumatic extradural hematomas of the posterior fossa in children. Childs Nerv Syst 12(3):145–148 20. Echara M et al (2023) Prognostic factors and outcome of surgically treated supratentorial versus infratentorial extradural hematoma in pediatrics: a comparative study of 350 patients at a Tertiary Center of a developing country. World Neurosurg 171:e447–e455 21. Ersahin Y, Mutluer S (1993) Posterior fossa extradural hematomas in children. Pediatr Neurosurg 19(1):31–33 22. Garza-Mercado R (1983) Extradural hematoma of the posterior cranial fossa. Report of seven cases with survival. J Neurosurg 59(4):664–672 23. Gupta PK, Mahapatra AK, Lad SD (2002) Posterior fossa extra- dural hematoma. Indian J Pediatr 69(6):489–494 24. Han K et al (2018) Liquid posterior fossa extradural hematoma in pediatric trauma: a single-center case series. J Neurol Surg Cent Eur Neurosurg 79(5):380–385 25. Koç RK et al (1998) Extradural hematoma of the posterior cranial fossa. Neurosurg Rev 21(1):52–57 26. Lui TN et al (1993) Extradural hematomas in the posterior cranial fossa. J Trauma 34(2):211–215 27. Miao Z et al (2023) Axial convex-shaped hematoma was associ- ated with poor curative effect of surgical treatment for traumatic posterior fossa extradural hematoma in Children. World J Surg 47(11):2932–2939 28. Mori K et al (1983) Extradural hematomas of the posterior fossa in children. Childs Brain 10(2):130–140 29. Pang D et al (1983) Nonsurgical management of extradural hema- tomas in children. J Neurosurg 59(6):958–971 30. Peter JC, Domingo Z (1990) Subacute traumatic extradural hae- matomas of the posterior fossa: a clinicopathological entity of the 5- to 10-year-old child. Childs Nerv Syst 6(3):135–138 31. Sheng HS et al (2017) Trephination mini-craniectomy for trau- matic posterior fossa extradural hematomas in selected pediatric patients. Chin J Traumatol 20(4):212–215 32. Kuppermann N et al (2009) Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet 374(9696):1160–1170 33. Rankin J (1957) Cerebral vascular accidents in patients over the age of 60. II. Prognosis. Scott Med J 2(5):200–215 34. Swieten JCv et al (1988) Interobserver agreement for the assess- ment of handicap in stroke patients. Stroke 19(5):604–607 35. Jennett B et al (1981) Disability after severe head injury: observa- tions on the use of the Glasgow Outcome Scale. J Neurol Neuro- surg Psychiatry 44(4):285–293 36. Page MJ et al (2021) The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71 37. Lee KS, Higgins JP, Prevedello DM (2024) Systematic reviews and meta-analyses in neurosurgery part I: interpreting and criti- cally appraising as a guide for clinical practice. Neurosurg Rev 47(1):339 38. Lee KS, Prevedello DM (2024) Systematic reviews and meta- analyses in neurosurgery part II: a guide to designing the protocol. Neurosurg Rev 47(1):360 39. Lee KS et al (2022) Tenets for the proper conduct and use of meta- analyses: a practical guide for neurosurgeons. World Neurosurg 161:291–302e1 40. Hozo SP, Djulbegovic B, Hozo I (2005) Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 5(1):13 41. Wan X et al (2014) Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol 14(1):135 42. Guyatt GH et al (2008) GRADE: an emerging consensus on rat- ing quality of evidence and strength of recommendations. BMJ 336(7650):924–926 43. Kaushik S, Sandip C (2018) Posterior Fossa Acute Extradural Hematoma in Children: Review and Management guidelines. J Pediatr Neurosci 13(3):289–293 44. Verma SK et al (2018) Traumatic posterior Fossa Extradural Hematoma: experience at Level I Trauma Center. Asian J Neuro- surg 13(2):227–232 http://creativecommons.org/licenses/by/4.0/ Neurosurgical Review (2024) 47:878878 Page 20 of 20 45. Deen HG et al (2007) CHAPTER 103 - HEAD TRAUMA. In: Schapira AHV (ed) Neurology and Clinical Neuroscience. Mosby, Philadelphia, pp 1386–1396 46. Thiam DW, Yap SH, Chong SL (2015) Clinical decision rules for paediatric minor Head Injury: are CT scans a necessary evil? Ann Acad Med Singap 44(9):335–341 47. Simon B et al (2001) Pediatric minor head trauma: indications for computed tomographic scanning revisited. J Trauma 51(2):231– 237 (discussion 237-8) 48. Griffin ES et al (2014) A matched-cohort study of pediatric head injuries: collecting data to inform an evidence-based triage assess- ment. J Emerg Nurs 40(1):98–104 49. Schweitzer AD et al (2019) Traumatic Brain Injury: imaging pat- terns and complications. Radiographics 39(6):1571–1595 50. Winter RC (2015) Posterior Fossa Extradural Hematoma. Pediatr Emerg Care 31(11):808–809 51. Wang EC, Lim AY, Yeo TT (1998) Traumatic posterior fossa extradural haematomas (PFEDH). Singap Med J 39(3):107–111 52. Malik NK et al. (2007) Posterior fossa extradural hematoma: our experience and review of the literature. Surg Neurol 68(2):155– 158. discussion 158. https:// doi. org/ 10. 1016/j. surneu. 2006. 10. 051 53. Pozzati E et al (1989) Extradural hematomas of the posterior cra- nial fossa. Observations on a series of 32 consecutive cases treated after the introduction of computed tomography scanning. Surg Neurol 32(4):300–303 54. Roda JM et al (1983) Posterior fossa extradural hematoma: a review and synthesis. Surg Neurol 19(5):419–424 55. Brenner DJ, Hall EJ (2007) Computed tomography–an increasing source of radiation exposure. N Engl J Med 357(22):2277–2284 56. Brenner D et al (2001) Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 176(2):289–296 57. Ishida Y et al (2013) Association between parental preference and head computed tomography in children with minor blunt head trauma. JAMA Pediatr 167(5):491–492 58. Niele N et al (2022) Young children with a minor traumatic head injury: clinical observation or CT scan? Eur J Pediatr 181(9):3291–3297 59. Atabaki SM et al (2016) Comparison of prediction rules and clinician suspicion for identifying children with clinically important brain injuries after blunt head trauma. Acad Emerg Med 23(5):566–575 60. Holmes JF et al (2024) PECARN prediction rules for CT imaging of children presenting to the emergency department with blunt abdominal or minor head trauma: a multicentre prospective vali- dation study. Lancet Child Adolesc Health 8(5):339–347 61. Otsuka S et al (1990) Study on cases with posterior fossa extra- dural hematoma–clinical features and indications for operation. Neurol Med Chir (Tokyo) 30(1):24–28 62. Lee KS et al (2023) Accuracy of robot-assisted stereotactic MRI- guided laser ablation in children with epilepsy. J Neurosurg Pedi- atr 32(2):214–222 63. Lee KS et al (2024) Endovascular coiling versus neurosurgical clipping in the management of aneurysmal subarachnoid haemor- rhage in the elderly: a multicenter cohort study. Neurosurg Rev 47(1):100 64. Lee KS et al (2023) Surgical revascularizations for pediatric moy- amoya: a systematic review, meta-analysis, and meta-regression analysis. Childs Nerv Syst 39(5):1225–1243 65. Babikian T et  al (2015) Chronic aspects of Pediatric Trau- matic Brain Injury: review of the literature. J Neurotrauma 32(23):1849–1860 66. Ryan NP et al (2014) Predictors of very-long-term sociocogni- tive function after pediatric traumatic brain injury: evidence for the vulnerability of the immature social brain. J Neurotrauma 31(7):649–657 67. Fay TB et al (2009) Predicting longitudinal patterns of functional deficits in children with traumatic brain injury. Neuropsychology 23(3):271–282 68. Babikian T, Asarnow R (2009) Neurocognitive outcomes and recovery after pediatric TBI: meta-analytic review of the litera- ture. Neuropsychology 23(3):283–296 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. https://doi.org/10.1016/j.surneu.2006.10.051 Traumatic posterior fossa extradural hematoma in children: a meta-analysis and institutional experience of its clinical course, treatment and outcomes Abstract Introduction Methods Data collection Overview of neurosurgical workflow and approaches Outcomes (case series) Statistical analysis Systematic review and meta-analysis Outcomes (systematic review and meta analysis) Search strategy Eligibility criteria Study selection Data extraction Risk of bias assessment Data synthesis Results Overview of study population and patient characteristics Systematic review and meta-analysis Characteristics of included studies Patient baseline characteristics, treatment approach and radiological features Conversion to surgery when initially managed conservatively Rates of good functional outcomes and mortality Discussion Overview of PFEDH in the pediatric population Imaging characteristics in PFEDH Towards an evidence-based management strategy Limitations and future directions Conclusion Acknowledgements References