Browse the corpus

Initial reports suggested a high mortality for COVID-19 patients supported with extracorporeal membrane oxygenation (ECMO).1 However, subsequent cohort studies reported ECMO outcomes did not differ between COVID-19 and other types of acute respiratory distress,2, 3 encouraging many to provide ECMO when deemed appropriate, including coordinating its delivery at national level in England.4 In The Lancet, Ryan P Barbaro and colleagues5 report an analysis of an international cohort of 4812 patients (aged ≥16 years) with COVID-19 who were supported with ECMO. Patients were divided into three groups based on the time and centre at which ECMO was started. The median age of patients was 49–51 years across the three groups, and 3523 (73%) were male. Data on race were not reported. The primary outcome was in‑hospital mortality, assessed 90 days after ECMO started. The findings suggest that mortality rates for ECMO-supported patients have increased by 15% between an arbitrarily set early and late stage of the pandemic (on or before May 1 vs after May 1, 2020); the mortality rate was 36·9% (95% CI 34·1–39·7) for patients who started ECMO on or before May 1 versus 51·9% (50·0–53·8) for patients who started ECMO after May 1. Mortality was even higher at 58·9% (55·4–62·3) for patients treated at centres that only offered ECMO after May 1. This analysis shows that the majority of patients supported with ECMO will now die.

9·7) for patients who started ECMO on or before May 1 versus 51·9% (50·0–53·8) for patients who started ECMO after May 1. Mortality was even higher at 58·9% (55·4–62·3) for patients treated at centres that only offered ECMO after May 1. This analysis shows that the majority of patients supported with ECMO will now die. ECMO is a simple concept allowing gas exchange and can be used to support patients with acute lung injury. It is complex in its execution, requiring trained staff and specialist equipment, and has a high burden of complications.6 ECMO is not a treatment or a therapy, it simply supports patients with failing lungs, providing time for lung recovery or—exceptionally—to be bridged to lung transplantation.7 The premise of ECMO is that it decreases the added physiological insult caused by mechanical ventilation, by permitting clinicians to use a less damaging mode of lung ventilation. This hypothesis is elegant but ECMO has never been proven to be superior to not using ECMO in randomised controlled trials.8

ECMO is a simple concept allowing gas exchange and can be used to support patients with acute lung injury. It is complex in its execution, requiring trained staff and specialist equipment, and has a high burden of complications.6 ECMO is not a treatment or a therapy, it simply supports patients with failing lungs, providing time for lung recovery or—exceptionally—to be bridged to lung transplantation.7 The premise of ECMO is that it decreases the added physiological insult caused by mechanical ventilation, by permitting clinicians to use a less damaging mode of lung ventilation. This hypothesis is elegant but ECMO has never been proven to be superior to not using ECMO in randomised controlled trials.8 Providing ECMO is fraught with difficulties and experience brings benefits. Barbaro and colleagues report that centres with more ECMO experience—defined as treating at least nine patients on or before May 1, 2020—have better outcomes (risk-adjusted mortality rate of 0·56 (95% CI 0·43–0·75) relative to centres with less experience).5 This phenomenon has been seen in many other clinical scenarios,9 and is thought to be a key explanation for the superiority of being allocated to an ECMO centre (but not always receiving ECMO) in the landmark randomised controlled trial that supported using ECMO in adults with severe respiratory failure.10 It also explains why countries such as England have chosen to concentrate all ECMO-supported patients in a few centres.11

he superiority of being allocated to an ECMO centre (but not always receiving ECMO) in the landmark randomised controlled trial that supported using ECMO in adults with severe respiratory failure.10 It also explains why countries such as England have chosen to concentrate all ECMO-supported patients in a few centres.11 The selection of the right patient at the right time is guided by clinical principles and experience, as well as availability, which might perversely delay commencement of ECMO. Clinicians balance carefully and subjectively the risk and benefit of starting ECMO for every patient. The inflexion point at which one is better than the other is unknown, even if criteria are well defined and regularly reviewed.12 Clinicians sometimes select patients who have so much lung damage by the time ECMO is started that there is no possible recovery. Barbaro and colleagues show that, later in the pandemic, patients more often had non-invasive ventilation before ECMO and were given steroids or other specific COVID-19 therapies. The latter is probably reflected in the increased burden of co-infections reported in the post-May 1 cohorts.5 Clinicians rely on an experience they do not yet have to allocate patients to ECMO and might miss important characteristics. It is not surprising that outcomes are changing as the pandemic is progressing. © 2021 STR/Getty Images2021

Barbaro and colleagues show that, later in the pandemic, patients more often had non-invasive ventilation before ECMO and were given steroids or other specific COVID-19 therapies. The latter is probably reflected in the increased burden of co-infections reported in the post-May 1 cohorts.5 Clinicians rely on an experience they do not yet have to allocate patients to ECMO and might miss important characteristics. It is not surprising that outcomes are changing as the pandemic is progressing. © 2021 STR/Getty Images2021 In the report by Barbaro and colleagues, no-one knows if some patients survived despite ECMO they did not need, or if some died just because of ECMO, or what happened to those who were denied ECMO—this is still the conundrum to clarify before we decide if ECMO is worth using or not. ECMO cannot be blamed for the increased mortality; it is merely a tool and clinicians still need to understand when to use it for the greatest benefit. Barbaro and colleagues should be commended for scratching the surface and reporting honestly and openly. To date, thousands of patients with COVID-19 have been ventilated in intensive care and a proportion had access to ECMO. Collating and analysing their clinical journeys might help to clarify if ECMO is a tool to keep.