Nonparametric maximum likelihood estimation for population pharmacokinetics, with application to cyclosporine

Abstract

A new method, nonparametric maximum likelihood (NPML), for statistical analysis of population kinetic data is proposed. NPML provides a discrete estimate of the whole probability density function of the pharmacokinetic parameters. This permits a straightforward derivation of usual population characteristics. To illustrate the application of the NPML method, a population analysis of cyclosporine RIA measured plasma levels in 188 bone marrow transplant patients after intravenous infusion, is presented. The capability of NPML to extract population information from sparse individual data is also outlined.