Purpose: To develop an integrated model for microdialysis data that incorporates all data including the recovery measurements in one model, and to compare this model to a previous model and the results from a noncompartmental analysis.
Methods: The models were developed in NONMEM. The modes of analysis were compared with respect to parameter estimates, model structures, gained mechanistic insight, and practical aspects.
Results: Both modeling approaches resulted in similar model structures. The parameter estimates in blood and brain from the models and the results from the noncompartmental analysis were comparable. Using the integrated model all data, that is, the total arterial concentrations, the venous and brain dialysate concentrations, and the recovery measurements, were analyzed simultaneously.
Conclusion: The theoretical benefits of the integrated model are related to the inclusion of the recovery in the model and the use of all collected data as it was observed. Thus, all data are described in a single model, corrections for the recovery and the protein binding are done within the model, and the dialysate observations are described by the integral over each collection interval. Thereby, the variability and the uncertainty in the model parameters are handled correctly to give more reliable parameter estimates.