SURGICAL OUTCOMES OF ACUTE EPIDURAL HEMATOMA EVACUATION: TIMING OF INTERVENTION AND PREDICTORS OF OUTCOME — A SYSTEMATIC REVIEW OF RANDOMIZED CONTROLLED TRIAL AND PRIMARY STUDIES

Abstract

Introduction: Acute epidural hematoma (EDH) is a time-critical neurosurgical emergency in which prompt evacuation can convert a potentially lethal mass lesion into a fully recoverable injury. Outcomes nevertheless vary widely, particularly in low- and middle-income countries (LMICs) where computed tomography (CT), neurosurgeons, operating theatres and referral systems are unevenly distributed. This review aimed to quantify contemporary mortality and functional outcome after surgical evacuation of acute EDH, to estimate the association between delayed surgery and outcome, to identify pre-operative predictors of outcome, and to synthesise evidence on facility- and human resource-related causes of delay.

Methods: The study strictly adhered to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 2020 guidelines. Randomized controlled trials and prospective, retrospective, cohort, case-control and cross-sectional studies of patients undergoing surgical evacuation of traumatic acute EDH were eligible. Risk of bias was assessed with the Quality In Prognosis Studies (QUIPS) tool and the Newcastle–Ottawa Scale. Proportions were pooled on the logit scale and odds ratios (ORs) with DerSimonian–Laird random-effects models; Mantel–Haenszel and Peto methods, leave-one-out analyses, subgroup analyses, meta-regression and funnel-plot tests were used for robustness. Certainty of evidence was graded with GRADE for prognosis.

Results: Thirty-one studies (3,368 patients; 24 from LMICs, two from Indonesia) met inclusion criteria; no randomized trial was identified. Pooled mortality after evacuation was 7.5% (95% CI 5.2–10.6; k = 21; I² = 74%), rising to 21.1% in cohorts restricted to comatose patients, and favourable outcome was achieved in 78.9% (95% CI 72.8–84.0; k = 15). Delayed evacuation was associated with unfavourable outcome (OR 6.17, 95% CI 1.27–29.94; k = 6) and, for a 6-hour threshold, with mortality (OR 14.23, 95% CI 2.90–69.78; k = 4), with marked heterogeneity driven by confounding by indication in high-income cohorts. Pre-operative Glasgow Coma Scale (GCS) ≤ 8 (mortality OR 9.26, 95% CI 3.68–23.33; unfavourable outcome OR 17.34, 95% CI 7.54–39.88), abnormal pupillary status (OR 5.85, 95% CI 3.10–11.05), associated intracranial lesions (OR 3.44, 95% CI 1.78–6.63) and older age (OR 2.85, 95% CI 1.33–6.08) predicted poor outcome. The proportion of comatose patients explained 38% of between-study heterogeneity in mortality. In LMIC cohorts, median injury-to-surgery intervals reached 15–78 hours, and logistical problems (28.1%) and resource constraints in staff, intensive care beds and equipment (22.5%) were the leading reported causes of delay.

Discussion: Surgical evacuation of acute EDH achieves low mortality and high rates of functional recovery, but outcome is governed chiefly by the neurological state at the time of surgery. Delay matters most where it allows neurological deterioration—a situation concentrated in referral-dependent LMIC systems lacking on-site CT, neurosurgeons and theatre capacity. In high-income systems with triage-driven prioritisation, crude time–outcome associations are attenuated or reversed.

Conclusion: Low pre-operative GCS, abnormal pupils, associated intracranial injury, older age and delayed evacuation identify patients at highest risk. Reducing facility- and workforce-related delay through regional neurotrauma networks, teleradiology, transfer protocols and credentialed damage-control craniotomy where neurosurgical transfer is not feasible is the most actionable lever for improving EDH outcomes in Indonesia and comparable settings