Risk of sepsis after extracorporeal cardiopulmonary resuscitation in patients with refractory cardiogenic shock from the RESCUE registry

Abstract

Abstract Objectives It is known that the incidence of sepsis is higher in patients receiving extracorporeal membranous oxygenation (ECMO). In particular, extracorporeal cardiopulmonary resuscitation (ECPR) is performed in emergency situations, which may increase the likelihood of non-aseptic management compared to routine procedures. So, we investigated the association of ECPR and incidence of sepsis in patients with refractory cardiogenic shock (CS). Background Pre- and post-ECPR sepsis are associated with increased mortality, and it is known that the occurrence of sepsis after ECPR is higher than in conventional cardiopulmonary resuscitation (CPR). However, it is not well known whether ECPR has a higher incidence of sepsis than ECMO insertion in non-CPR conditions. Method The SMART-RESCUE study is a multi-center, retrospective and prospective registry of patients that presented with CS. Between January 2014 and December 2018, 1,247 patients with CS were enrolled from 12 major centers in Korea. The incidence of sepsis was compared by dividing into two groups according to whether or not ECPR was performed. Patients with ECMO duration less than 48 hours were excluded. Result Total of 231 patients(mean age 59.46 years, men 67.70%, mead follow-up duration 136.5 days) were analyzed into ECPR group(n=95, 60.7 ± 13.0 years, men 75.8%) and non-ECPR group(n=136, 58.6 ± 15.5 years, men 65.4%) respectively. ECPR group showed lower SBP, lower DBP, lower HR, more shock caused by ischemia, more smoking, lower LVEF, higher Hb, higher serum Na, lower AST, higher Peak troponin I, higher vasoactive inotropic score, higher ICMP portion. After 1:1 propensity score matching corrected for height, SBP, DBP, Hb, initial pump flow, Na, ICMP, vasoactive inotropic score, difference remained in SBP, DBP, HR, cause of shock. Sepsis incidence was 7.4%. The risk for sepsis incidence was not different between two groups(7.4% vs 6.3%, adjusted OR 1.18, 95% CI 0.38-3.65, p=0.774). Conclusion ECPR procedure does not increase sepsis compared to ECMO insertion not in CPR state in patients with cardiogenic shock.Adjusted Baseline CharacteristicsAdjusted Outcomes