(601) Incidence and Clearance of Bloodstream Infections for Patients Receiving Venovenous ECMO

Introduction: Consensus guidelines recommend against routine antimicrobial prophylaxis for patients receiving veno-venous extracorporeal membrane oxygenation (VV ECMO) due to lack of improvement in clinical outcomes, yet routine prophylaxis remains common. The purpose of this study is to characterize incidence, management of and outcomes of patients with bloodstream infections (BSI) who do not receive routine antimicrobial prophylaxis while cannulated for VV ECMO.

Methods: This single-center, retrospective cohort included patients identified by institutional data repository cannulated for VV ECMO from January 2014 to February 2024 in a medical intensive care unit. Exclusion criteria were BSI known or present at time of cannulation, cannulation for greater than 48 hours at an outside hospital, or death within 24 hours of cannulation. No antimicrobial prophylaxis was routinely administered to any patient in this cohort.

Results: Of 396 ECMO courses, 70 patients (17.7%) were identified with a new BSI. When commensal organisms were removed, the incidence of BSI was 14.3%. The median ECMO duration was 22 days (IQR 14.8, 40.5) with a primary indication for ECMO of ARDS (85.7%) due to viral pneumonia (68.3%). Median time from cannulation to BSI was 12.2 days (IQR 3.1, 18.3). The most common organisms identified were coagulase-negative staphylococci (19.9%), methicillin-susceptible and methicillin-resistant S. aureus (18.5% and 14.3%, respectively) and Candida spp (17.1%). The most common BSI source was line (47.1%) or pulmonary (34.2%). A persistent BSI occurred in 21 patients (30%), with 14 of these patients (71.4%) eventually clearing their BSI. Clearance of the BSI occurred in 52 patients (74.3%), with 48 (92.3%) cleared without ECMO circuit exchange. Of the 18 patients who did not clear their BSI, 17 patients died and one was transferred to another institution. Overall, in-hospital mortality was 64.3% and BSI-attributed mortality was 31.8%.

Conclusions: In this cohort of patients receiving VV ECMO without routine antimicrobial prophylaxis, the incident of BSI was lower than observed in other studies and BSIs were routinely cleared without ECMO circuit exchange.