ObjectiveTo investigate the effect of polymyxin B hemoperfusion on the prognosis of patients with sepsis and septic shock by meta-analysis.MethodsSupplemented by manual search and document traceability, the US National Library of Medicine Pubmed, the Dutch Medical Abstracts Embase database, and the Cochrane clinical trial database were searched. Randomized controlled trials (RCTs) were collected from January 1998 to October 2018 for the treatment of sepsis and septic shock with polymyxin B hemoperfusion, only limited to English publications. The collected RCTs were evaluated and the prognosis of patients with sepsis and septic shock was analyzed by the Cochrane Collaboration.ResultsFinally six RCTs were included, and a total of 926 patients were analyzed, with 471 patients in the polymyxin B hemoperfusion group and 455 patients in the control group. The mortality rate was 36.3% (171/471) in the polymyxin B hemoperfusion group and 39.1% (178/455) in the control group. Hemoperfusion with polymyxin B could not reduce the patient mortality (RR=0.80, 95% CI 0.56 to 1.15, P=0.233). A subgroup analysis was taken on the patients with moderate to severe septic shock. Four RCTs were included in total and 418 patients were analyzed, with 207 patients in the polymyxin B hemoperfusion group and 211 in the control group. The mortality rate was 38.65% (80/207) in the polymyxin B hemoperfusion group and 50.71% (107/211) in the control group were. The hemoperfusion of polymyxin B could significantly reduce the mortality of patients with moderate to severe septic shock (RR=0.70, 95% CI 0.52 to 0.96, P=0.025).ConclusionsOlymyxin B hemoperfusion can not improve the prognosis of patients with sepsis and septic shock. However, compared with conventional treatment, polymyxin B hemoperfusion can improve the 28-day mortality rate of patients of severe septic shock. Due to the limit number of randomized controlled trials, more high-quality trials are needed to a further confirmation.