Effect of Extracorporeal Blood Purification on Mortality in Sepsis: A Meta-Analysis and Trial Sequential Analysis

Abstract

Objective: The objective of this study was to conduct a meta-analysis and trial sequential analysis (TSA) of published randomized controlled trials (RCTs) to determine whether mortality benefit exists for extracorporeal blood purification techniques in sepsis. Data Sources: A systematic search on MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials for RCTs was performed. Study Selection: RCTs investigating the effect of extracorporeal blood purification device use on mortality among critically ill septic patients were selected. Data Extraction: Mortality was assessed using Mantel-Haenszel models, and I2 was used for heterogeneity. Data are presented as odds ratios (OR); 95% confidence intervals (CIs); p values; I2. Using the control event mortality proportion, we performed a TSA and calculated the required information size using an anticipated intervention effect of a 14% relative reduction in mortality. Data Synthesis: Thirty-nine RCTs were identified, with 2,729 patients. Fourteen studies used hemofiltration (n = 789), 17 used endotoxin adsorption devices (n = 1,363), 3 used nonspecific adsorption (n = 110), 2 were cytokine removal devices (n = 117), 2 used coupled plasma filtration adsorption (CPFA) (n = 207), 2 combined hemofiltration and perfusion (n = 40), and 1 used plasma exchange (n = 106). On conventional meta-analysis, hemofiltration (OR 0.56 [0.40–0.79]; p < 0.001; I2 = 0%), endotoxin removal devices (OR 0.40 [0.23–0.67], p < 0.001; I2 = 71%), and nonspecific adsorption devices (OR 0.32 [0.13–0.82]; p = 0.02; I2 = 23%) were associated with mortality benefit, but not cytokine removal (OR 0.99 [0.07–13.42], p = 0.99; I2 = 64%), CPFA (OR 0.50 [0.10–2.47]; p = 0.40; I2 = 64%), or combined hemofiltration and adsorption (OR 0.71 [0.13–3.79]; p = 0.69; I2 = 0%). TSA however revealed that based on the number of existing patients recruited for RCTs, neither hemofiltration (TSA-adjusted CI 0.29–1.10), endotoxin removal devices (CI 0.05–3.40), nor nonspecific adsorption devices (CI 0.01–14.31) were associated with mortality benefit. Conclusion: There are inadequate data at present to conclude that the use of extracorporeal blood purification techniques in sepsis is beneficial. Further adequately powered RCTs are required to confirm any potential mortality benefit, which may be most evident in patients at greatest risk of death.