Early surgery vs. conservative management for traumatic intracerebral hematoma: A systematic review and Bayesian meta-analysis

Abstract

Introduction The management of traumatic intracerebral hematoma (TICH) remains controversial. This Bayesian meta-analysis evaluated the effects of early surgery versus conservative treatment on mortality, unfavourable functional outcome, and hospital length of stay. Methods PubMed, Scopus, and the Cochrane Library were searched from inception to 15 December 2025 (PROSPERO CRD420251248144) for studies comparing early surgery with conservative treatment in patients with TICH. Outcomes included mortality reported at the study-specific available assessment time, ranging from the in-hospital period to 6 months, unfavourable functional outcome assessed at 3 or 6 months, and hospital length of stay. Bayesian random-effects models were fitted using vague priors for overall effects and informative or weakly informative priors for between-study heterogeneity. Sensitivity analyses with alternative prior specifications were performed. Posterior probabilities (PPs), 95% credible intervals (CrIs), and posterior predictive distributions were estimated. Results Four studies, including two randomized controlled trials and two observational studies comprising 664 patients were included. Early surgery showed no clear reduction in mortality (OR 0.90; 95% CrI [0.50; 1.56]; PP OR < 1: 65%) or unfavourable functional outcome (OR 1.16; 95% CrI [0.57; 2.43]; PP OR < 1: 31%). Early surgery was associated with a low probability of reducing hospital length of stay (MD 3.00 days; 95% CrI [-0.79; 7.18]; PP MD < 0: 5%). Subgroup analyses suggested potential differences according to study design and risk of bias, although predictive distributions indicated substantial uncertainty. Conclusion Current evidence does not demonstrate a clear benefit of routine early surgery over conservative treatment for TICH and shows an uncertain trend toward longer hospital stay, which should be interpreted cautiously given imprecision and potential confounding by indication. Early surgery should therefore be considered selectively pending further high-quality randomized evidence.