Variation in survival in patients with Coronavirus Disease 2019 supported with extracorporeal membrane oxygenation: A multi-institutional analysis of 594 consecutive patients with Coronavirus Disease 2019 supported with extracorporeal membrane oxygenation at 49 hospitals within 21 states

OBJECTIVES: We reviewed 594 consecutive patients with Coronavirus Disease 2019 supported with extracorporeal membrane oxygenation at 49 hospitals within 21 states and examined patient characteristics, treatments, and variation in outcomes over the course of the pandemic. METHODS: A multi-institutional database was used to assess all patients with Coronavirus Disease 2019 cannulated for extracorporeal membrane oxygenation between March 17, 2020, and December 20, 2021, inclusive, and separated from ECMO on or prior to January 14, 2022. Descriptive analysis was stratified by 4 time categories: group A = March 2020 to June 2020, group B = July 2020 to December 2020, group C = January 2021 to June 2021, group D = July 2021 to December 2021. A Bayesian mixed-effects logistic regression was used to assess continuous trends in survival where time was operationalized as the number of days between each patient's cannulation and that of the first patient in March 2020, controlling for multiple variables and risk factors. RESULTS: At hospital discharge, of 594 patients, 221 survived (37.2%) and 373 died. Throughout the study, median age [interquartile range] declined (group A = 51.0 [41.0-60.0] years, group D = 39.0 [32.0-48.0] years, P < .001); median days between Coronavirus Disease 2019 diagnosis and intubation increased (group A = 4.0 [1.0-8.5], group D = 9.0 [5.0-14.5], P < .001); and use of medications (glucocorticoids, interleukin-6 blockers, antivirals, antimalarials) and convalescent plasma fluctuated significantly (all P < .05). Estimated odds of survival varied over the study period with a decline between April 1, 2020, and November 21, 2020 (odds ratio, 0.39, 95% credible interval, 0.18-0.87, probability of reduction in survival = 95.7%), improvement between November 21, 2020, and May 17, 2021 (odds ratio, 1.85, 95% credible interval, 0.86-4.09, probability of improvement = 93.4%), and decline between May 17, 2021, and December 1, 2021 (odds ratio, 0.49, 95% credible interval, 0.19-1.44, probability of decrease = 92.1%). CONCLUSIONS: Survival for patients with Coronavirus Disease 2019 supported with extracorporeal membrane oxygenation has fluctuated during the stages of the pandemic. Minimizing variability by adherence to best practices may refine the optimal use of extracorporeal membrane oxygenation in a pandemic response.