An inquiry into the treatment of sepsis using plasma exchange therapy: A systematic review and meta‐analysis

Abstract

AbstractSepsis is a potentially lethal condition that occurs when the body's response to infection damages tissue and organs. The production of inflammatory mediators typically assists in defending the body against infection; however, an overreaction to inflammation can cause coagulation problems, vascular endothelial damage, and organ hypoperfusion. Blood purification methods, such as plasmapheresis, can effectively remove inflammatory mediators from plasma. The purpose of this meta‐analysis was to explore the efficacy of plasma exchange for sepsis treatment as noted in recent studies. The authors searched the Pubmed (Medline), Cochrane Central Register of Controlled Trials (The Cochrane Library), Embase (Ovid), and Scopus databases and included controlled clinical studies that compared plasmapheresis or plasma filtration with conventional treatment in patients with severe sepsis. The Newcastle–Ottawa Scale literature quality assessment tool was used to assess the risk of bias. The primary study outcome was all‐cause mortality. The random effects model was adopted for conducting the meta‐analysis. Among the 1013 records found, the study included 5 trials, all of which carried a low risk of bias. The use of plasmapheresis was associated with a longer stay in the intensive care unit (odds ratio [OR], 0.85, 95% confidence interval [CI], 0.39–1.32, heterogeneity [I2] = 0%), a significant reduction in all‐cause mortality (OR, 0.54, 95% CI, 0.33–0.89, I2 = 70%), and reduced mortality (OR, 0.29, 95% CI, 0.13–0.67, I2 = 0%) in adults; the results for children differed from this (OR, 0.79, 95% CI, 0.36–1.72, I2 = 89%). Four trials reported no adverse events; one trial reported an adverse event related to plasma exchange, including an instance of hypotension in one patient. Plasmapheresis appeared to be an effective treatment for patients suffering from sepsis. A large number of additional randomised controlled trials are needed to confirm this finding.