A contemporary analysis of the volume–outcome relationship for extracorporeal membrane oxygenation in the United States

• Published in: Surgery Vol. 173; no. 6; pp. 1405 - 1410
• Main Authors: Verma, Arjun, Hadaya, Joseph, Williamson, Catherine, Kronen, Elsa, Sakowitz, Sara, Bakhtiyar, Syed Shahyan, Chervu, Nikhil, Benharash, Peyman
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01-06-2023
• Subjects: Adult; Extracorporeal Membrane Oxygenation; Hospital Mortality; Hospitals, High-Volume; Humans; Respiratory Insufficiency - therapy; Retrospective Studies; Shock, Cardiogenic; United States - epidemiology; United States
• ISSN: 0039-6060; 1532-7361
• DOI: 10.1016/j.surg.2023.02.004

A paradoxical increase in mortality following extracorporeal membrane oxygenation at high-volume centers has previously been demonstrated. We examined the association between annual hospital volume and outcomes within a contemporary, national cohort of extracorporeal membrane oxygenation patients.
All adults requiring extracorporeal membrane oxygenation for postcardiotomy syndrome, cardiogenic shock, respiratory failure, or mixed cardiopulmonary failure were identified in the 2016 to 2019 Nationwide Readmissions Database. Patients undergoing heart and/or lung transplantation were excluded. A multivariable logistic regression with hospital extracorporeal membrane oxygenation volume parametrized as restricted cubic splines was developed to characterize the risk-adjusted association between volume and mortality. The volume corresponding to the maximum of the spline (43 cases/year) was used to categorize centers as low- or high-volume.
An estimated 26,377 patients met the study criteria, and 48.7% were managed at high-volume hospitals. Patients at low- and high-volume hospitals had similar age, sex, and rates of elective admission. Notably, patients at high-volume hospitals less frequently required extracorporeal membrane oxygenation for postcardiotomy syndrome but more commonly for respiratory failure. After risk adjustment, high-volume hospital status was associated with reduced odds of in-hospital mortality, relative to low-volume hospitals (adjusted odds ratio 0.81, 95% confidence interval 0.78–0.97). Interestingly, patients at high-volume hospitals faced a 5.2-day increment in length of stay (95% confidence interval 3.8–6.5) and $23,500 in attributable costs (95% confidence interval 8,300–38,700).
The present study found that greater extracorporeal membrane oxygenation volume was associated with decreased mortality but higher resource use. Our findings may help inform policies regarding access to and centralization of extracorporeal membrane oxygenation care in the United States.