We read with great interest the article by Zhang and colleagues : a single-center experience showing that although 48.8% of all extracorporeal membrane oxygenation (ECMO) patients are successfully decannulated and survive for 30 days or until discharge, 44.5% of patients die while on support or in the following 30 days, and 6.7% are bridged to advanced heart failure therapies.

This study highlights once again the presence of the so-called ECMO gap: the discrepancy between patients who are successfully weaned from ECMO and patients who survive to discharge. This concept has been previously investigated by our groups.

 We demonstrated that many published series do not discriminate death occurring during or after support. Furthermore, we showed that 15.3% of all patients receiving venoarterial ECMO die after initially successful weaning. Our groups also described the ECMO gap within the postcardiotomy population, a group of patients that is also represented in the study by Zhang and colleagues.

 Through single- or multicenter investigations, we showed that 23% to 36% of patient receiving postcardiotomy ECMO die despite initially successful weaning, defined as survival >24 hours from ECMO discontinuation.

Finally, we demonstrated that postoperative complications (bleeding requiring rethoracotomy, cardiac arrest, bowel ischemia, acute kidney injury, or septic shock) are associated with postweaning mortality. Postoperative renal failure, respiratory failure, and rethoracotomy are associated with 12-month postdischarge mortality.

The study by Zhang and colleagues confirms the above-mentioned concepts but also highlights the compelling need for a universal definition of successful weaning and ECMO gap. Indeed, Zhang and colleagues considered all cases of death within 30 days from decannulation as unsuccessful weaning. However, within 30 days, many ECMO-unrelated complications can occur and lead to death. We therefore question whether it is accurate to consider all 30-day deaths to be directly caused by an unsuccessful weaning.