Clinical Benefit of Extubation in Patients on Venoarterial Extracorporeal Membrane Oxygenation

OBJECTIVES: Although patients on venoarterial extracorporeal membrane oxygenation for refractory cardiogenic shock are usually supported with mechanical ventilation, it is not clear whether sedation cessation and extubation might improve outcomes. DESIGN: Retrospective cohort study with propensity score overlap weighting analysis. SETTING: Three ICUs in a 1,500-bed tertiary university hospital. PATIENTS: From an overall cohort of 641 patients with venoarterial-extracorporeal membrane oxygenation support, the primary analysis was performed in 344 patients who had been successfully decannulated in order to reduce immortal time bias. MEASUREMENTS AND MAIN RESULTS: Seventy-five patients (22%) were extubated during extracorporeal membrane oxygenation support and were subsequently decannulated alive. Forty-nine percent received noninvasive ventilation, and 25% had emergency reintubation for respiratory, neurologic, or hemodynamic reasons. Higher Simplified Acute Physiology Score II at admission (odds ratio, 0.97; 95% CI [0.95-0.99]; p = 0.008) was associated with a lower probability of extubation, whereas cannulation in cardiac surgery ICU (odds ratio, 3.14; 95% CI [1.21-8.14]; p = 0.018) was associated with an increased probability. Baseline characteristics were well balanced after propensity score overlap weighting. The number of ICU-free days within 30 days of extracorporeal membrane oxygenation decannulation was significantly higher among extubated patients compared with nonextubated patients (22 d [11-26 d] vs 18 d [7-25 d], respectively; p = 0.036). There were no differences in other outcomes including ventilator-associated pneumonia (odds ratio, 0.96; 95% CI [0.51-1.82]; p = 0.90) and all-cause mortality within 30 days of extracorporeal membrane oxygenation decannulation (5% vs 17%; hazard ratio, 0.54; 95% CI [0.19-1.59]; p = 0.27). As a secondary analysis, outcomes were compared in the overall cohort of 641 venoarterial extracorporeal membrane oxygenation-supported patients. Results were consistent with the primary analysis as extubated patients had a higher number of ICU-free days (18 d [0-24 d] vs 0 d [0-18 d], respectively; < 0.001) and a lower risk of death within 30 days of extracorporeal membrane oxygenation cannulation (hazard ratio, 0.45; 95% CI [0.29-0.71]; p = 0.001). CONCLUSIONS: Extubation during venoarterial-extracorporeal membrane oxygenation support is safe, feasible, and associated with greater ICU-free days.