Heart blood volume by dilution in patients on hemodialysis.

Cardiomegaly and accompanying complications contribute to morbidity and mortality in patients on hemodialysis. Our objective is to provide a measure of heart blood volume from data collected during routine measurement of cardiac output during hemodialysis. A heart model based upon the probability of transfer between chambers during a cardiac cycle generates both the distribution of transit times (dilution curve) and the blood volumes in the chambers of the heart. This heart model is combined with a compatible lung model based upon a stochastic description of microvascular flow to simulate vasomotion. By exercising the heart/lung model systematically over a wide range of probabilities, any chosen characteristic of the distribution of transit times can be correlated to total heart blood volume for a wide range of heart/lung functional states. For a bolus traversing the heart and lung vascular beds, the length of the chord drawn upon the dilution curve at half the height of the peak is identified as an efficient, robust predictor of the total blood volume in the heart at end ventricular diastole. When applied to data previously collected from patients on hemodialysis, enlarged hearts were observed in 4 of 60 patients.