Predicting oral anticoagulant response using a pharmacodynamic model

Abstract

We developed a pharmacokinetic and pharmacodynamic model of warfarin absorption, metabolism, and anticoagulant action appropriate for guiding anticoagulant therapy. The model requires only two independently adjustable parameters to describe warfarin's effect on individual patients. For any given individual, these parameters are rapidly and inexpensively identified using a computer program based on the model. Test data were generated by superimposing Gaussian noise on dose-response curves calculated with the model. Then the computer program was applied to the test data. Future prothrombin complex activities (PCA's) and maintenance doses were predicted accurately early in the course of drug administration. In addition, the program accurately predicted PCA response in two groups of normal volunteers.