[A meta-analysis of the effects of direct hemoperfusion with polymyxin B-immobilized fiber on prognosis in severe sepsis].

OBJECTIVE

To investigate the effects of direct hemoperfusion with polymixin B-immobilized fiber (DHP-PMX) in patients with sepsis.

METHODS

We searched Pubmed, Embase, Web of Science databases and identified relevant randomized controlled trials (RCT) from January 1995 to May 2010. Meta-analysis of DHP-PMX on mortality and levels of endotoxin in patients with sepsis were conducted using the methods recommended by the Cochrane Collaboration.

RESULTS

Eleven RCTs were included. Eight of them included the mortality of patients (sample size: 211 DHP-PMX and 178 conventional medical therapy). In total, the mortalities of patients with sepsis in DHP-PMX group and conventional group were 37.4% (79/211) and 68.5% (122/178) respectively. Compared with the conventional medical therapy, DHP-PMX appeared to significantly reduce mortality (OR = 0.24, 95%CI 0.16 - 0.38, P < 0.000 01). The results were similar when two RCTs enrolling patients with methicillin resistant staphylococcus aureus (MRSA) infections were excluded (OR = 0.27, 95%CI 0.17 - 0.45, P < 0.000 01). When the analysis was limited to the nine studies that reported 28- to 30-day mortality, results were unchanged (OR = 0.29, 95%CI 0.17 - 0.48, P < 0.000 01). Six RCTs had the available data of endotoxin. The level of endotoxin decreased 31 ng/L(95%CI 22.46 - 39.55) after DHP-PMX therapy, and the decreasing was statistically significant (P < 0.000 01), while the level of endotoxin in patients of conventional group did not change (P = 0.94).

CONCLUSIONS

This study suggests a favorable effects of DHP-PMX on mortality and endotoxin decreasing in patients with sepsis. However, lack of enough cases and blinding need to be considered. Further investigation with large sample of high quality RCTs is needed.