
Abstract
Background
Predictors of death resulting from extracorporeal membrane oxygenation (ECMO) withdrawal, in-hospital death after ECMO liberation, and survival to hospital discharge have been evaluated incompletely, despite the prognostic insight they provide.
Research Question
What are the predictors of 3 vital outcomes after venoarterial ECMO: (1) death in the context of ECMO withdrawal, (2) ECMO liberation followed by in-hospital death, and (3) survival to hospital discharge?
Study Design
This retrospective observational study using Extracorporeal Life Support Organization registry data included adults supported by venoarterial ECMO from 2018 through 2022 at 325 North American sites. Three generalized linear mixed models (each comparing 2 outcomes) measured associations between predictors and outcomes, using random intercepts to address data clustering by site.
Results
Of 23,177 patients, 10,122 patients (43.7%) died in the context of ECMO withdrawal, 3,510 patients (15.1%) died in the hospital after ECMO liberation, and 9,545 patients (41.2%) survived to hospital discharge. Statistical analysis was performed for 16,277 patients supported for ≥ 24 hours with complete data available (32.5% female; mean age, 55.7 years; and 62.7% White, 16.4% Black, 6.1% Hispanic, and 3.2% Asian). Older age, higher BMI, cardiac arrest before ECMO initiation, and renal failure were associated with increased odds of death in the context of ECMO withdrawal and death after liberation compared with survival. Higher pH and male sex also were associated with increased odds of survival relative to withdrawal. Among decedents, death in the context of ECMO withdrawal was less common than death after ECMO liberation when patients were male and pH was higher.
Interpretation
Patients who were older, had higher BMI, or experienced cardiac arrest or renal failure before ECMO initiation seemed to have increased risk of in-hospital death, both in the context of ECMO withdrawal and after ECMO liberation. Male individuals were shown to be less likely to experience ECMO withdrawal. These findings offer prognostic associations that may inform how to support patients and families after ECMO initiation.
We use cookies to provide you with the best possible user experience. By continuing to use our site, you agree to their use. Learn more