Background. Studies indicate that the recovery from coronavirus disease 2019 (COVID-19)-associated acute respiratory distress syndrome may be slower than other viral pneumonia. There are limited data to guide decisions among patients who need extracorporeal membrane oxygenation (ECMO) support, especially the expected time of recovery and considering lung transplantation (LT). Methods. This was a retrospective chart review of patients with COVID-19-associated acute respiratory distress syndrome placed on ECMO between March 1, 2020, and September 15, 2021 (n = 20; median age, 44 y; range, 22-62 y; male:female, 15:5). We contrasted the baseline variables and clinical course of patients with and without the need for ECMO support >30 d (ECMO long haulers, n = 10). Results. Ten patients met the criteria for ECMO long haulers (median duration of ECMO, 86 d; range, 42-201 d). The long haulers were healthier at baseline with fewer comorbidities but had worse pulmonary compliance and higher partial pressure of CO2. They had a significantly higher number of membrane oxygenator failures, changes to their cannulation sites, and suffer more complications on ECMO. One of the long hauler was bridged to LT while another 6 patients recovered and were discharged. Overall survival was better among the ECMO long haulers (70% versus 20%; 9.3, 1.2-73; P = 0.03). Conclusions. Despite worse pulmonary physiology, frequent complications, and a tortuous hospital course that may appear to portend a poor prognosis, ECMO long haulers have the potential to recover and be weaned off ECMO without the need for LT. A customized approach comprising a more conservative timeline for the consideration of LT may be prudent among these patients.

Original languageEnglish (US)
Pages (from-to)E202-E211
Issue number4
StatePublished - Apr 1 2022

ASJC Scopus subject areas

  • Transplantation


Dive into the research topics of 'ECMO Long Haulers: A Distinct Phenotype of COVID-19-Associated ARDS With Implications for Lung Transplant Candidacy'. Together they form a unique fingerprint.

Cite this