ABSTRACT
Objectives: Treatment of severe respiratory distress syndrome (ARDS) due to COVID-19 by veno-venous extracorporeal membrane oxygenation (VV-ECMO) had a mortality of up to 70% in Germany. Many patients with COVID-19 need VV-ECMO support longer than 28 days (long-term VV-ECMO). Evidence on mortality, complications during intensive care, functional status after discharge and mortality-predictors for patients supported with long-term VV-ECMO is lacking. Method(s): Retrospective study of 137 consecutive patients treated with VV-ECMO for ARDS due to COVID-19 at University Hospital Regensburg from March 2020 to March 2022. Result(s): 38% (n=52;87% male) of patients needed longterm VV-ECMO support. In these, SOFA score (median [IQR]) at ECMO initiation was 9 [8-11], age 58.2 [50.6- 62.5] years, PaO2/FiO2-ratio 67 [52-88] mmHg, pCO262 [52-74] mmHg, Murray-Score 3.3 [3.0-3.6] and PEEP 15 [13 - 16] cmH2O. Duration of long-term support was 45 [35-65] days. 26 (50%) patients were discharged from the ICU. Only one patient died after hospital discharge. At VVECMO initiation, baseline characteristics did not differ between deceased and survivors. Complications were frequent (acute kidney injury: 31/52, renal replacement therapy: 14/52, pulmonary embolism: 21/52, intracranial hemorrhage 8/52, major bleeding 34/52 and secondary sclerosing cholangitis: 5/52) and more frequent in the deceased. Karnofsky index (normal 100) after rehabilitation was 70 [57.5-82.5]. Twelve of the 18 patients discharged from rehabilitation had a satisfactory quality of life according to their own subjective assessment. Four patients required nursing support. Mortality-predictors within the first 30 days on VV-ECMO only observed in those who deceased later, were: Bilirubin >5mg/dl for > 7 days, pulmonary compliance <10ml/mbar for >14 days, and repeated serum concentrations of interleukin 8 >150ng/L. Conclusion(s): Long-term extracorporeal lung support in patients with COVID-19 resulted in 50 % survival and subsequently lead to a satisfactory quality of life and functionality in the majority of patients. It should preferably be performed in experienced centers because of a high incidence of complications. Several findings during the early course were associated with late mortality but need validation in large prospective studies.