Annual hospital procedural volume and outcome in extracorporeal membrane oxygenation for respiratory failure

Abstract

Background The hospital mortality of patients suffering from pulmonary failure requiring venovenous extracorporeal membrane oxygenation (V‐V ECMO) or extracorporeal carbon dioxide removal (ECCO2R) is high. It is unclear whether outcome correlates with a hospital's annual procedural volume. Methods Data on all V‐V ECMO and ECCO2R cases treated from 2007 to 2019 were retrieved from the German Institute for Medical Documentation and Information. Comorbidities and outcomes were assessed by DRG, OPS, and ICD codes. The study population was divided into 5 groups depending on annual hospital V‐V ECMO and ECCO2R volumes (<10 cases; 10–19 cases; 20–29 cases; 30–49 cases; ≥50 cases). Primary outcome was hospital mortality. Results A total of 25 096 V‐V ECMO and 3607 ECCO2R cases were analyzed. V‐V ECMO hospitals increased from 89 in 2007 to 214 in 2019. Hospitals handling <10 cases annually increased especially (64 in 2007 to 149 in 2019). V‐V ECMO cases rose from 807 in 2007 to 2597 in 2019. Over 50% of cases were treated in hospitals handling ≥30 cases annually. Hospital mortality was independent of the annual hospital procedural volume (55.3%; 61.3%; 59.8%; 60.2%; 56.3%, respectively, p = 0.287). We detected no differences when comparing hospitals handling <30 cases to those with ≥30 annually (p = 0.659). The numbers of ECCO2R hospitals and cases has dropped since 2011 (287 in 2007 to 48 in 2019). No correlation between annual hospital procedural volume and hospital mortality was identified (p = 0.914). Conclusion The number of hospitals treating patients requiring V‐V ECMO and V‐V ECMO cases rose from 2007 to 2019, while ECCO2R hospitals and their case numbers decreased. We detected no correlation between annual hospital V‐V ECMO or ECCO2R volume and hospital mortality.