COVID-19 ARDS: getting ventilation right - Authors' reply.

We thank Xosé Pérez-Fernández and colleagues for their thoughtful Correspondence regarding our study of extracorporeal membrane oxygenation (ECMO) in COVID-19. We agree that our study does not provide evidence that forms of non-invasive ventilation (NIV), such as high-flow nasal cannula and mask or helmet ventilation, might be deleterious compared with other strategies. Our observational study was not designed to make causal inferences regarding the potential superiority of ECMO or any pre-ECMO support strategy. We showed that the more recent cohort with higher mortality had increased use of NIV and decreased duration of pre-ECMO invasive mechanical ventilation (IMV). We did not measure the initiation time of NIV, however, and so could not test for an association between duration of pre-ECMO NIV and the relative risk of mortality.

Although many patients with severe COVID-19 might benefit from the use of NIV, the subset of patients who ultimately do not respond to NIV and require IMV are precisely those who are likely to have high work of breathing, high transpulmonary pressures, and who are therefore at risk of developing patient self-inflicted lung injury. This situation might select for more severely ill patients receiving IMV and ultimately ECMO. It is one hypothesis out of a number we put forward to help explain the association with increased mortality in those who ultimately do not respond to these levels of support. However, this is not an argument for or against the use of NIV in this setting. Even if the hypothesis is correct, NIV might still be the appropriate therapy for any given patient. A randomised clinical trial is required to fully address this question.

To date, there are no prospective clinical trials evaluating the effect on outcomes of the timing of initiating ECMO support. However, in accord with the suggestion of Pérez-Fernández and colleagues, post-hoc analysis of the ECMO to rescue lung injury in severe ARDS trial suggests that patients with greater risk of developing ventilator-induced lung injury might be more likely to benefit from ECMO than those who were enrolled because of severe hypoxaemia.

Our study showed that the mortality rate of ECMO-supported patients with COVID-19 worsened and the duration of ECMO support lengthened later in the pandemic. We encourage centres to consider these factors when creating policies to guide ECMO allocation. Moreover, during a pandemic, the use of resource-intensive interventions such as ECMO must also be informed by the needs of local health-care systems.

RPB is the ELSO Registry chair. GM, and DB are on the ELSO board of directors. DB is the president-elect of ELSO. DB also chairs the executive committee for the International ECMO Network. ASS chairs the Scientific Oversight Committee of the International ECMO Network. RPB reports grants from the US National Institutes of Health (R01 HL153519, R01 HD015434, and K12 HL138039). ASS reports consulting fees from Baxter and Xenios in relation to ECMO. DB reports grants from ALung Technologies, and medical advisory board relationships with Xenios, Abiomed, Cellenkos, and Medtronic. JS declares no competing interests.