Integral assessment of gas exchange during veno-arterial ECMO: accuracy and precision of a modified Fick principle in a porcine model

Assessment of native cardiac output during extracorporeal circulation is challenging. We assessed a modified Fick principle under conditions such as dead space and shunt in 13 anesthetized swine undergoing centrally cannulated veno-arterial extracorporeal membrane oxygenation (V-A ECMO, 308 measurement periods) therapy. We assumed that the ratio of carbon dioxide elimination (V̇co(2)) or oxygen uptake (V̇o(2)) between the membrane and native lung corresponds to the ratio of respective blood flows. Unequal ventilation/perfusion (V̇/Q̇) ratios were corrected towards unity. Pulmonary blood flow was calculated and compared to an ultrasonic flow probe on the pulmonary artery with a bias of 99 mL/min (limits of agreement −542 to 741 mL/min) with blood content V̇o(2) and no-shunt, no-dead space conditions, which showed good trending ability (least significant change from 82 to 129 mL). Shunt conditions led to underestimation of native pulmonary blood flow (bias −395, limits of agreement −1,290 to 500 mL/min). Bias and trending further depended on the gas (O(2), CO(2)) and measurement approach (blood content vs. gas phase). Measurements in the gas phase increased the bias (253 [LoA −1,357 to 1,863 mL/min] for expired V̇o(2) bias 482 [LoA −760 to 1,724 mL/min] for expired V̇co(2)) and could be improved by correction of V̇/Q̇ inequalities. Our results show that common assumptions of the Fick principle in two competing circulations give results with adequate accuracy and may offer a clinically applicable tool. Precision depends on specific conditions. This highlights the complexity of gas exchange in membrane lungs and may further deepen the understanding of V-A ECMO.