Estimating changes in cardiac output using an implanted hemodynamic monitor in heart failure patients

Objectives. The aim of this study was to evaluate an algorithm that estimates changes in cardiac output (CO) from right ventricular (RV) pressure waveforms derived from an implantable hemodynamic monitor (IHM) in heart failure patients. Design. Twelve heart failure patients (NYHA II-III, EF 32%) with an implantable hemodynamic monitor (Chronicle (R)) were included in this study. Changes in cardiac output were provoked by body position change at rest (left lateral supine, horizontal supine, sitting, and standing) and a steady state bicycle exercise at 20 watts. Estimated CO derived from the IHM (CO(IHM)) was compared to CO measured with inert gas rebreathing (CO(RB)), echocardiography (COE(CHO)) and impedance cardiography (CO(ICG)). CO(RB) was considered the reference method. Results. The median intra-patient correlation coefficient comparing CO(RB) and CO(IHM) was 0.83 (range: 0.63-0.98). Comparing CO(RB) with CO(ECHO) and CO(ICG) resulted in mean intra-patient correlation coefficients of 0.73 (-0.29-0.94) and 0.63 (-0.29-0.96). In a statistical model where slope and intercept was considered random between patients the coefficient of determination (R(2)) comparing CO(RB) and CO(IHM) was 0.91. Mean bias was -0.39 L/min (11%). Limits of agreement were +/- 1.56 L/min and relative error was 21%. Conclusions. A simple algorithm based on RV pressure wave form characteristics derived from an IHM can be used to estimate changes in CO in heart failure patients. These findings encourage further research aiming to improve and validate the algorithm.