Body mass index is not associated with time on veno-venous extracorporeal membrane oxygenation or in-hospital mortality

Abstract

Morbid obesity, as characterized by BMI, is often utilized as an exclusion criterion for VV-ECMO because of presumed poor prognosis and technically complex cannulation. However, the “obesity paradox” suggests obesity may be protective during critical illness, and BMI does not capture variations in body type, adiposity, or fluid balance. This study examines relationships between BMI and patient outcomes. Adult VV-ECMO patients with BMI ≥ 35 kg/m² admitted January 2012 to June 2021 were identified from an institutional registry. BMI and outcomes were analyzed with Mann-Whitney U tests and Pearson correlations with Bayesian post-hoc analyses. 116 of 960 ECMO patients met inclusion criteria. Median (Q1, Q3) BMI was 42.3 (37.3, 50.8) and min, max of 35.0, 87.8 with 9.0 (5.0, 15.5) ECMO days. BMI was not significantly correlated with ECMO days (r = −0.102; p = .279). Bayesian analyses showed moderate evidence against BMI correlating with ECMO days. In-hospital mortality (27%) was significantly associated with ECMO days (p = .014) but not BMI (p = .485). In this cohort of high-BMI patients, BMI was not associated with survival or time on ECMO. BMI itself should not be used as an exclusion criterion for VV-ECMO.