Current extracorporeal membrane oxygenation (ECMO) monitoring methods for circuit obstructions have limitations. In this study, we test the hypothesis that changes in flow in a pre-existing shunt tubing of the ECMO circuit can be used as a predictor of obstructions for both venovenous (V-V) and venoarterial ECMO (V-A ECMO) circuits using a computational hemodynamic model.

A custom computational hemodynamic model (CHM) representing the blood flow within the V-A ECMO and V-V ECMO circuit and native cardiovascular system was developed for a hemodynamic parametric analysis. Several degrees of obstructions within the oxygenator and outlet cannula were simulated, and the flow in the ECMO circuit and shunt was monitored to assess its sensitivity to circuit obstructions. Over 500 cases were analyzed to investigate various clinically-relevant levels of ECMO support (2 – 5.5 L/min).

Our model demonstrated that shunt flow changed linearly with increasing oxygenator obstruction for all oxygenator obstruction scenarios tested, which matched experimental data. Similar relationships were also seen for all outflow cannula obstructions.

The findings are consistent with our hypothesis, indicating the sensitivity of the shunt flow rate as a predictor of ECMO circuit obstructions for both V-A and V-V ECMO circuits. The data supports that shunt flow responds to hemodynamic changes throughout the ECMO-patient circuit and thus can be used as an early indicator for ECMO system obstructions regardless of V-A or V-V ECMO setup, indicating its robustness as an early warning indicator of ECMO system obstructions.