Pharmacodynamics of warfarin at steady state

We studied the pharmacodynamics of warfarin in chronically treated patients. Two methods were used to estimate the pharmacodynamic parameters M/Kd and Cmax (mg/L). In Method 1 the prothrombin time response was modeled directly without use of warfarin concentrations and Method 2 used warfarin concentrations and prothrombin time response to estimate M/Kd and Cmax. The mean Cmax and M/Kd for Method 1 (n = 88) were 5.5 +/- 2.3 mg/L and 51 +/- 46 and for Method 2 (n = 27) 6.3 +/- 2.8 mg/L and 35.4 +/- 13. When Method 1 was applied to the same 27 patients in Method 2, the mean Cmax and M/Kd were 5.7 +/- 3.3 mg/L and 36.1 +/- 14.9. These differences were not significant. Multiple regression analysis revealed that the value of Cmax and M/Kd varied between medical centers. No other patient characteristics were found to be significant. We conclude that modeling steady-state prothrombin time response directly adequately describes pharmacodynamic response to warfarin.