In-hospital outcomes and 6-month follow-up results of patients supported with extracorporeal membrane oxygenation for COVID-19 from the second wave to the end of the pandemic (EuroECMO-COVID): a prospective, international, multicentre, observational study

Abstract

Background: Extracorporeal membrane oxygenation (ECMO) for COVID-19 was thoroughly assessed during the first pandemic wave, but data on subsequent waves are limited. We aimed to investigate in-hospital and 6-month survival of patients with COVID-19 supported with ECMO from the second pandemic wave (Sept 15, 2020) until the end of the pandemic (March 21, 2023, announced by WHO).

Methods: EuroECMO-COVID is a prospective, observational study including adults (aged ≥16 years) requiring ECMO respiratory support for COVID-19 from 98 centres in 21 countries. We compared patient characteristics and outcomes between in-hospital survivors and non-survivors. Mixed-effects multivariable logistic regressions were used to investigate factors linked to in-hospital mortality. 6-month survival and overall patient status were determined via patient contact or chart review. This study is registered with ClinicalTrials.gov, NCT04366921, and is complete.

Findings: We included 3860 patients (2687 [69·7%] were male and 1169 [30·3%] were female; median age 51 years [SD 11]) from 98 centres in 21 countries. In-hospital mortality was 55·9% (n=2158), with 81·2% (n=1752) deaths occurring during ECMO support. More non-survivors had diabetes, hypertension, cardiovascular disease, and renal failure, and required more pre-ECMO inotropes and vasopressors compared with survivors. Median support duration was 18 days (IQR 10-31) for both groups. Factors linked to in-hospital mortality included older age, pre-ECMO renal failure, pre-ECMO vasopressors use, longer time from intubation to ECMO initiation, and complications, including neurological events, sepsis, bowel ischaemia, renal failure, and bleeding. Of the 1702 (44·1%) in-hospital survivors, 99·7% (n=1697) were alive at 6 months follow-up. Many patients at 6 months follow-up had dyspnoea (501 [32·0%] of 1568 patients), cardiac (122 [7·8%] of 1568 patients), or neurocognitive (168 [10·7%] of 1567 patients) symptoms.

Interpretation: Our data for patients undergoing ECMO support for respiratory distress from the second COVID-19 wave onwards confirmed most findings from the first wave regarding patient characteristics and factors correlated to in-hospital mortality. Nevertheless, in-hospital mortality was higher than during the initial pandemic wave while 6-month post-discharge survival remained favourable (99·7%). Persisting post-discharge symptoms confirmed the need for post-ECMO patient follow-up programmes.

Funding: None.