Timing of spontaneous breathing in veno-venous extracorporeal membrane oxygenation: An exploratory analysis of intensive care outcomes

Background:

The clinical impact of early spontaneous breathing during veno-venous extracorporeal membrane oxygenation (V-V ECMO) for acute respiratory distress syndrome (ARDS) is unclear, balancing potential benefits (reducing sedation and preserving diaphragm function) against risks of patient self-inflicted lung injury (P-SILI). Research Question Explore whether early transition to sustained spontaneous breathing (≤5 days after ECMO initiation) may be associated with improved ICU outcomes. 

Study design and methods:

Multicenter cohort study across eight Dutch ECMO centers, including adult ARDS patients on mechanical ventilation supported with V-V ECMO. Sustained spontaneous breathing was defined as the first episode of spontaneous breathing sustained across two consecutive days. ICU mortality was analyzed using multistate and landmark Cox models; liberation from ventilation, ECMO, and ICU discharge by day 60 were assessed with competing risk regression, adjusted for Respiratory ECMO Survival Prediction (RESP) score. 

Results:

Among 123 patients, median time to sustained spontaneous breathing was 5.0 days (IQR 1–12). Across landmark times from ECMO day 1 through 14, transitioning on any given days was not associated with a difference in ICU mortality. After RESP adjustment, early sustained spontaneous breathing was associated with higher rates of extubation and ICU discharge by day 60.

Interpretation:

Early spontaneous breathing was associated with higher adjusted rates of extubation and ICU discharge in an exploratory analysis; however, these findings should be interpreted cautiously because of residual confounding, selection bias, and immortal-time bias. This study highlights the need for randomized trials incorporating physiological markers of safe transition to help identify patients that could benefit.