End-tidal carbon dioxide concentration can estimate the appropriate timing for weaning off from extracorporeal membrane oxygenation for refractory circulatory failure.

Although extracorporeal membrane oxygenation (ECMO) is widely used as temporary circulation support, there are no reports of direct parameters indicating cardiac recovery to determine the timing of weaning off. Twenty-five patients supported by ECMO due to hemodynamic deterioration were divided into 2 groups according to their outcome: weaned ECMO (W: n = 18) or not (NW: n = 7). In the W group, we examined the differences in parameters between the 2 time points, ECMO introduction, and the reduction in ECMO flow to 40% of the initial setting known as the conventional recovery point (C-point). Significant differences were observed in systolic pulmonary artery pressure, the cardiac index measured by the thermodilution method, C-reactive protein, lactate, base excess, and the end-tidal CO(2) concentration (ETCO(2)). Next, by closely examining these 6 parameters measured every 12 hours, we found that only ETCO(2) had always changed steeply, like a 'flexion point' (E-point), in all W cases, but not in NW. The E-point was defined as an initial increase in ETCO(2) of >or= 5 mmHg over the preceding 12 hours with a continued rise over the next 12 hours. E-points appeared as much as 95 +/- 60 hours earlier than C-points and also preceded weaning off of ECMO. ETCO(2) can be a useful continuous parameter for predicting the adequate timing of weaning off of ECMO for circulatory failure at the bedside.