Aim The purpose of our study was to define and validate a population-pharmacokinetic model including the influence of patients' characteristics on the pharmacokinetics of cefepime. Patients and methods A total of 55 patients were randomized in Group 1 (34 patients, 320 cefepime concentrations) for the model building and Group 2 (21 patients, 196 cefepime concentrations) for the validation group. They received cefepime as 2 g A 2 or as 4 g continuously. The population pharmacokinetic analysis was carried out using NONMEM and a baseline model was constructed for studying the influence of demographic and biological variables. The model was then validated by a comparison of the predicted and observed concentrations in Group 2. A final model was elaborated from the whole population. Results Total clearance (CL) was significantly correlated with the serum creatinine (CREA) and the central volume of distribution (V1) was correlated with the body weight (WT). The final model was: CL = 7.14 + (-0.0133 A CREA). V1 = (-16.8) + (0.475 A WT). Q (intercompartmental clearance) = 10.5. V2 = 18.1. The mean pharmacokinetic parameters and their individual variability were: CL (8.24 l/h, 45%), V1 (20.89 l, 60%), V2 (17.95 l, 49%), total volume (38.85 l, 42%) and Q (10.56 l/h, 9%). The bias (1.07 mg/l, IC 95% = -40.46 -+42.60), precision (21.19%) and AFE (1.15) demonstrated the performance of the model. Conclusion We have developed and validated a pharmacokinetic model to estimate cefepime concentrations. We showed that serum creatinine and body weight are factors that may influence the standard dose of cefepime. Our model enabled us to predict cefepime concentrations in other patients.
