Fig. 3

Our process of weaning V–V ECMO. A Sequence for V–V ECMO weaning. Throughout, monitor for stopping criteria. B As SGF is weaned, the proportion of metabolically produced CO₂ cleared by the natural lung (V’CO₂NL/V’CO₂tot) increases. In patients who are not yet suitable for decannulation, this load can only be managed with excessive effort (see next panels). Ventilatory efficiency can be expressed as the ratio of the effort (or minute ventilation) to the V’CO_2NL. In patients who fail a weaning trial ventilatory efficiency is usually worse, and may deteriorate as the demands on the respiratory system rise. If they are not able to clear all of the metabolically produced CO₂ then hypercapnia ensues. C Monitoring drive and effort during a weaning trial. Waveforms during a pressure supported breath, an expiratory occlusion throughout an inspiratory cycle and an end inspiratory occlusion. D Targets to maintain during a trial off V–V ECMO including measures of drive (P0.1), effort (Pocc or ∆P_es if available), stress (∆P_L or DP if not available) and native lung ventilator efficiency (∆P_es/V’CO_2NL, end tidal CO₂ to arterial CO₂ ratio (ETCO₂:PaCO₂) or the ratio of minute ventilation to clearance (VE/ V’CO_2NL). At our centre, volumetric capnography from the ventilator and pre- and post- oxygenator blood gases are used to calculate the V’CO_2NL and V’CO_2ML, respectively. Unfortunately, during weaning the V’O_2NL is not routinely measured. However, this is done in individual patients who have a pulmonary artery catheter for measurement of C_VmixedO₂ or a receive calorimetric measurements. FdO₂: the fraction of oxygen of the sweep gas flow (SGF), TV: tidal volume, P0.1: pressure deflection during 100 ms of occlusion, Pocc: maximal pressure deflection during occlusion, ∆P_es oesophageal pressure swing, ∆P_aw: plateau after inspiratory inclusion, including the PMI: rebound pressure from relaxing inspiratory muscles, ∆P_L: transpulmonary pressure