Surgical strategies for spontaneous intracerebral hemorrhage: a Bayesian network meta-analysis of randomized controlled trials.

Study Design: A Systematic Review and Bayesian Network Meta-Analysis. Objective: To compare the efficacy and perioperative outcomes of conservative medical treatment (CMT) and surgical interventions-including decompressive craniectomy (DC), craniotomy (CC), endoscopic surgery (ES), and minimally invasive puncture surgery (MIPS)-in patients with spontaneous intracerebral hemorrhage (ICH). Background: Spontaneous intracerebral hemorrhage (ICH) is a severe neurological emergency associated with substantial mortality and long-term disability. Although several surgical strategies have been developed to reduce hematoma burden and secondary brain injury, the comparative effectiveness and perioperative trade-offs among different interventions remain controversial. Direct head-to-head randomized evidence comparing multiple surgical strategies is limited, complicating evidence-based clinical decision-making. Methods: We conducted a systematic review and network meta-analysis (NMA) in accordance with PRISMA 2020 and PRISMA-NMA guidelines. PubMed, Web of Science, and Cochrane Library were searched from inception to January 2026 for randomized controlled trials comparing CMT, DC, CC, ES, and MIPS in patients with spontaneous ICH. Primary outcomes included 6-month mortality and good functional outcome at 6 months. Secondary outcomes included hematoma clearance rate, operative time, intraoperative blood loss, and length of hospital stay. Pairwise meta-analysis was performed using Stata 18.0, and Bayesian NMA was conducted in R 4.3.1 using the gemtc and BUGSnet packages. Surface under the cumulative ranking curve (SUCRA) values were used to rank interventions. Results: Eighteen randomized controlled trials involving 4,497 patients were included. For good functional outcome at 6 months, MIPS (SUCRA = 87.0) and ES (SUCRA = 84.6) ranked highest and were significantly superior to CC and CMT, whereas no significant difference was observed between MIPS and ES. For 6-month mortality, DC probabilistically ranked highest (SUCRA = 81.5), although most pairwise comparisons did not reach statistical significance. Regarding perioperative outcomes, both ES and MIPS significantly reduced operative time and intraoperative blood loss compared with CC, with MIPS showing the largest reductions. ES achieved higher hematoma clearance rates and shorter hospital stay, whereas MIPS demonstrated lower hematoma clearance. Conclusion: MIPS and ES may provide advantages in functional recovery and perioperative burden in spontaneous ICH, whereas DC may offer potential survival benefit, although current evidence remains uncertain. Clinical decision-making should balance long-term outcomes against perioperative trade-offs and be individualized according to disease severity and patient-specific risk factors. Additional high-quality multicenter randomized trials are needed to clarify the role of DC and define optimal indications for each surgical strategy.