• Christopher John Joyce, Kiran Shekar, and David Andrew Cook.
• Discipline of Anaesthesiology Critical Care, University of Queensland, Ned Hanlon Building, Royal Brisbane and Women's Hospital, Herston, QLD, 4006, Australia. chris.joyce@health.qld.gov.au.
• Intensive Care Med Exp. 2018 Aug 9; 6 (1): 25.

BackgroundVenovenous extracorporeal membrane oxygenation (vv-ECMO) is an effective treatment for severe respiratory failure. The interaction between the cardiorespiratory system and the oxygenator can be explored with mathematical models. Understanding the physiology will help the clinician optimise therapy. As others have examined O2 exchange, the main focus of this study was on CO2 exchange.MethodsA model of the cardiorespiratory system during vv-ECMO was developed, incorporating O2, CO2 and N2 exchange in both the lung and the oxygenator. We modelled lungs with shunt fractions varying from 0 to 1, covering the plausible range from normal lung to severe acute respiratory distress syndrome. The effects on PaCO2 of varying the input parameters for the cardiorespiratory system and for the oxygenator were examined.ResultsPaCO2 increased as the shunt fraction in the lung and metabolic CO2 production rose. Changes in haemoglobin and FIO2 had minimal effect on PaCO2. The effect of cardiac output on PaCO2 was variable, depending on the shunt fraction in the lung. PaCO2 decreased as extracorporeal circuit blood flow was increased, but the changes were relatively small in the range used clinically for vv-ECMO of > 2 l/min. PaCO2 decreased as gas flow to the oxygenator rose and increased with recirculation. The oxygen fraction of gas flow to the oxygenator had minimal effect on PaCO2.ConclusionsThis mathematical model of gas exchange during vv-ECMO found that the main determinants of PaCO2 during vv-ECMO were pulmonary shunt fraction, metabolic CO2 production, gas flow to the oxygenator and extracorporeal circuit recirculation.

8 keywords...

hide…