Prediction of dose-response relations based on patient characteristics.

The recent discussion of evidence-based, adaptive treatment planning highlights the need for models for the prediction of courses of treatment response. We combine a dose-response model with growth curve modeling to determine dose-response relations for well-being, symptoms, and functioning. Hierarchical linear modeling was used to model each patient's expected course of improvement. The resulting predictions were cross-validated on two samples of psychotherapy outpatients. The results give further empirical support for the dose-response model and the phase model of psychotherapy as well as for the usefulness of patient treatment response profiling for individual treatment management. Copyright 2001 John Wiley & Sons, Inc.