Nonlinear models for the analysis of longitudinal or repeated measures data have been an area of intense biostatistical research. Biological, clinical trial, and especially pharmacometric applications have largely driven this research. Typically the individual serves as the sampling unit, and repeated measurements of the same or related outcome measures are gathered at successive time points. Variability within an individual is modeled using fixed effects and random errors, while random coefficients are used to model variability between individuals. Often only a few repeated measurements on a large number of individuals are measured, so pooled estimation methods are required. In the pharmaceutical industry concern is primarily with the nonlinear compartmental modeling of drug concentrations. We propose to incorporate recent and ongoing advances in nonlinear mixed effects modeling into comprehensive software integrated into S-PLUS. Specializing this software to the pharmacometric arena. We propose a comprehensive module for the analysis of pharmacometric data containing a library of predefined models, an easy to use iconic graphical user interface designed for compartmental modeling, and extensive model diagnostics statistics. In the process of building this software, we propose to commercially implement new models from several areas of current research in nonlinear mixed effects modeling.
A large percentage of all graduates with degrees in statistics or biostatistics are employed in the pharmaceutical industry where pharmacometric models are of great importance. A much larger number of pharmacologists work in the same area. The proposed module offers exciting new dimensions and it has many ease-of-use features that will make it quite popular with users. MathSoft intends to target the biomedical and biopharmaceutical markets. Each of these markets represents tens of thousands of potential users.
