Network meta-analysis finds PMX and HA 330 blood purification may reduce mortality in severe septic shock
A Bayesian network meta-analysis evaluated 15 distinct blood purification treatment approaches for severe septic shock, incorporating data from 31 randomized controlled trials involving a total of 2678 patients. The comparator was standard treatment, and the primary outcome was mortality. The analysis found that the TORAYMYXIN™ filter (PMX) and the HA 330™ device were the sole interventions that notably decreased mortality relative to standard care. In contrast, convection at any prescribed convective dose was associated with increased mortality compared to standard therapy. The study reported summary relative risks but did not provide the specific numerical values, absolute mortality rates, or confidence intervals for these findings.
Safety and tolerability data for the devices were not reported in the analysis. The risk of bias for the included trials was evaluated using the Cochrane Handbook methodology, but the overall certainty of the evidence is tempered by the indirect nature of the comparisons.
Key limitations stem from the network meta-analysis design. The 15 treatment approaches were compared indirectly, and direct head-to-head comparisons between all devices were not detailed. The clinical relevance is restrained because the analysis did not report specific effect sizes (relative risk values), absolute mortality numbers, or confidence intervals, making precise clinical interpretation difficult. The findings suggest a potential signal for PMX and HA 330 but require confirmation in future direct comparative trials with more complete outcome reporting.