Application of Akaike's information criterion (AIC) in the evaluation of linear pharmacokinetic equations.

Abstract

According to linear pharmacokinetics, the time course of plasma concentration of a drug, Cp,is expressed by a sum of exponential functions, Cp=ei=1nai\(C_p = \sum\nolimits_{i = 1}^n {a_i e^{ - b_i t} } \).This article describes a statistical technique to estimate the number of exponential terms, n,for the time course of drug by the application of Akaike's information criterion (AIC). Plasma concentrations of ethoxybenzamide, sulfisoxazole, bishydroxycoumarin, and diazepam measured following bolus intravenous injection were used as clinical examples for this method. Selection of models is compared between the AIC method and an Ftest method at significance levels of 5% and 1%.