Mathematic modeling of drug disposition and postdialysis rebound in patients requiring hemodialysis

We developed a model to describe intradialysis and postdialysis serum concentration changes and, in concert with nonlinear regression, to estimate drug removal during hemodialysis. A Lotus-based computer simulation program was developed that describes drug disposition in this therapy. The hemodialysis removal rate constant and beta were two important factors identified that influence postdialysis rebound. Also, a nonlinear regression routine using PCNONLIN was developed that enables simultaneous fitting of predialysis, intradialysis, and postdialysis drug concentration data. Data from four patients who received intravenous sulbactam were analyzed using this model. The mean measured amount of sulbactam removed by hemodialysis was 446 +/- 32 mg versus predicted amounts of 456 +/- 92 mg from model estimates, and 449 +/- 33 mg when dialysate recovery of drug was included in the input. This program will be useful in characterizing pharmacokinetic constants and predicting hemodialysis drug removal.