246 – PK and QTc Analysis
Minimum-Chi-Square-Estimation of Pharmacokinetic Parameters
Karl-Ernst Biebler
Ernst-Moritz-Arndt University
Parameter estimation in the pharmacokinetics requires a probability theoretical model. One often uses a regression formula with a random error term. Another approach is developed here. Random variable is the residence time of a drug molecule in the individual. Its probability distribution is derived from the pharmacokinetic model via the related ordinary differential equations. The related parameter estimation is calculated according to the varied minimum-chi-square-method from the measured concentration values. The applicability of this procedure was examined for certain classic compartment models. Examples and counterexamples are given. If the varied minimum-chi-square-estimation exists, it is uniquely determined and consistent. The test of Pearson then can be used for the model choice.