Predictive performance at different concentrations

From: "Martin, Tomas" <tmj@cptcnt2.cvm.ncsu.edu> Subject: Predictive performance at different concentrations Date: 23 Feb 1998 20:19:25 -0500 Dear NMusers: We have developed and validated (using data-splitting) a population pharmacokinetics model using NONMEM. What are the different statistical approaches that we can use in order to verify the adequacy of the model at all levels of predicted concentrations?. To test the predictive performance of the model we have used the approach suggested in "Sheiner, L.B. & Beal, S.L. Some Suggestions for Measuring Predictive Performance. J. Pharmacok. and Bioph. 9, 4:503-512, 1981. It would be correct to test the predictive performance for different concentration groups in order to verify the adequacy of the model at all levels of predicted concentrations? Thanks very much.