Estimation of total arterial compliance: an improved method and evaluation of current methods.

Determination of arterial compliance in vivo has long interested physiologists. Most current methods for estimating this parameter assume that compliance is constant, i.e., that arterial pressure-volume (P-V) relations are linear, and they also assume that diastolic aortic pressure decay is an exponential function of time. Both of these assumptions, however, are questionable. This study proposes improved methods of estimating compliance based on a Windkessel model of the arterial system but which utilize the area under the pressure tracing rather than the waveform itself. Formulations accounting for both linear and three hypothetical nonlinear arterial P-V relations (exponential, logarithmic, and parabolic) are presented. Data from patients with congestive heart failure and hypertension are used for illustration. Compliances assuming linear P-V relations are reasonably close to those assuming nonlinear P-V relations only at mean aortic pressure. At end-diastolic pressure the linear assumption underestimates and at peak systolic it overestimates the compliances obtained assuming nonlinear P-V relations. The simpler linear assumption still allows a first approximation to compliance, but we show that existing methods for obtaining compliance under this assumption have severe theoretical as well as practical shortcomings. Our proposed method avoids these shortcomings primarily because deviations from an exact exponential form of the pressure wave have less influence on these compliance estimates than currently used methods.