Object To investigate the efficacy and safety of four interventions of spontaneous intracerebral hemorrhage simultaneously.

Methods PubMed, EmBase, Web of Science, and the Cochrane Central Register of Controlled Trials were searched for randomized controlled trials (RCTs) investigating endoscopic surgery (ES), minimally invasive puncture surgery (MIPS), conventional craniotomy (CC), and/or conservative medical treatment (CMT). Good functional outcome, death, and hemorrhage recurrence rates were evaluated by a network meta-analysis.

Results 20 RCTs with 3603 patients were included. Compared with CMT, a higher rate of good functional outcome was found after ES (RR=2.21, 95% CI 1.37 to 3.55) and MIPS (RR=1.47, 95% CI 1.24 to 1.73). Both ES (RR=0.62, 95% CI 0.44 to 0.86) and MIPS (RR=0.72, 95% CI 0.58 to 0.90) markedly reduced the rate of death. However, there was no significant difference in efficacy and safety between ES and MIPS. The top ranked P score for the efficacy outcome was for ES (P score=0.9810). ES (P-score=0.0709) ranked lowest for the primary safety outcome. There was a higher risk of hemorrhage recurrence after CC (RR=3.80, 95% CI 1.90 to 7.63) and MIPS (RR=2.86, 95% CI 1.70 to 4.82) compared with CMT whereas no significant difference was found for ES (RR=1.46, 95% CI 0.53 to 4.02).

Conclusions The results suggest that both ES and MIPS significantly improve neurological function and reduce the risk of death compared with CMT, and there is no significant difference between ES and MIPS. Ranking of P scores revealed that ES may be the most optimal intervention to improve functional outcome and prevent death. This needs to be evaluated further.