Cardiac output monitoring with pulmonary versus trans-cardiopulmonary thermodilution in left ventricular assist devices: Interchangeable methods?

Cardiac output (CO) measurement is mandatory in patients with left ventricular assist devices (LVADs). Thermodilution with pulmonary artery catheter (PAC) remains the clinical gold standard to measure CO in these patients, however it is associated with several complications. Therefore, the agreement between PAC and new, minimally invasive monitoring methods in LVAD needs to be further investigated. The aim of this study was to assess the accuracy and reliability of transpulmonary thermodilution with a PiCCO2 monitor compared with pulmonary artery thermodilution with PAC in a LVAD. Continuous-flow LVADs were implanted in six mini-pigs to assist the left ventricle. We studied two methods of measuring CO—intermittent transpulmonary thermodilution (CO(TPTD)) by PiCCO2 and intermittent pulmonary artery thermodilution by CAP, standard technique (CO(PTD))—obtained in four consecutive moments of the study: before starting the LVAD (basal moment), and with the LVAD started in normovolemia, hypervolemia (fluid overloading) and hypovolemia (shock hemorrhage). A total of 72 paired measurements were analysed. At the basal moment, CO(TPTD) and CO(PTD) were closely correlated (r (2) = 0.89), with a mean bias of −0.085 ± 0.245 L/min and percentage error of 16%. After 15 min of partial support LVAD, CO(TPTD) and CO(PTD) were closely correlated (r (2) = 0.79), with a mean bias of −0.040 ± 0.417 L/min and percentage error of 26%. After inducing hypervolemia, CO(TPTD) and CO(PTD) were closely correlated (r (2) = 0.78), with a mean bias of −0.093 ± 0.339 L/min and percentage error of 13%. After inducing hypovolemia, CO(TPTD) and CO(PTD) were closely correlated (r (2) = 0.76), with a mean bias of −0.045 ± 0.281 L/min and percentage error of 28%. This study demonstrates a good agreement between transpulmonary thermodilution by PiCCO monitor and pulmonary thermodilution by PAC in the intermittent measurement of CO in a porcine model with a continuous-flow LVAD.