Evaluating the accuracy of using population pharmacokinetic data to predict plasma concentrations of alfentanil

A major reason for quantitating the relationship of drug dose to plasma concentration is to design optimal drug administration schemes (i.e., those that can achieve desired target concentrations of a drug). Recently, the authors completed a population pharmacokinetic analysis of the new opioid alfentanil using the computer program NONMEM. This analysis quantified the effects of age, weight, and sex on disposition of alfentanil in 45 patients, and determined the average pharmacokinetic profile of the drug for the group. Using these population pharmacokinetic parameters, one can predict (estimate) the plasma concentration time course of alfentanil for any given dosage scheme. The present study evaluated the accuracy with which one could use these population data to predict plasma concentrations of alfentanil in a different group of surgical patients given iv boluses and a variable-rate infusion of alfentanil for induction and maintenance of anesthesia for abdominal and superficial surgery. A total of 597 plasma concentrations of alfentanil were measured for 19 patients. For each measured concentration, we used the population pharmacokinetic parameters obtained previously with NONMEM to calculate a predicted concentration. Accuracy and precision of the prediction were assessed by the mean bias of the prediction and by the mean absolute prediction error, respectively. The mean bias (+/- SE) (systematic over- or underprediction) was -7.9 +/- 5.2%. The mean absolute error (+/- SE), a measure of the precision, was 22.3 +/- 2.9%. Therefore, the authors' previously described population pharmacokinetic parameters for alfentanil appear to be "robust," and can be used to design computerized schemes for administration of alfentanil for general surgery.