A simple in silico model as a support in the clinical practice: automatic setting of a VA-ECMO system

The use of peripheral VA-ECMO remains controversial despite its increased use in clinical practice. Complications are still very common, and their underlying causes, possibly related to CS (Cardiogenic Shock) etiology and VA-ECMO setting, are not comprehensively understood yet. A mathematical model of the complete native circulation supported by VA-ECMO is proposed, which can help in automatic setting the mechanical pump and monitoring patient’s hemodynamics. The real system is represented as a combination of functional compartments (e.g. the heart chambers, described by the time-varying elastance), where pressure and flow rate (or volume) waves are calculated. The CS condition is simulated by reducing the maximum elastance of the left ventricle of an appropriate grade of impairment, k. Different criteria for CS identification i.e., thresholds on SBP, MAP UO, EF and CI, are adopted. Depending on the CS criterion the impairment k ranges from 50% up to 84%, and the VA-ECMO pressure head that guarantees a flow of 3.6 l/min falls in the range 100-80 mmHg. Hemodynamics in the supported patient changes accordingly, with respect to both the positive (e.g., perfusion increase) and negative (e.g. LV distension) effects of ECMO. According to the model, the setting of the VA-ECMO pump depends on the CS hemodynamic criterion and the selected pump flow rate. This could lead to different management of the patient and different outcomes. The model proves its potential as a bed-side tool for the automatic control of the mechanical circuit.