< Back to previous page

Project

Optimal treatment regimes: Tracing truly optimal regimes, drawing valid conclusions, and application to ADHD and depression

For many health problems, multiple treatment alternatives are available. A major challenge then is to identify optimal decision rules or treatment regimes that specify the preferable treatment alternative for each individual client, based on that client’s pattern of pretreatment characteristics. An optimal decision rule is one leading to the best possible result when applied to the whole target population of clients under study. While everybody agrees that personalized treatment decisions are very important, there currently is a considerable risk that decisions based on estimated treatment regimes are suboptimal, in that many clients would not receive the best possible treatment. This may imply a high human and financial cost, and may therefore be detrimental for both the individual client and society. The present project will study this problem in depth and will propose solutions to it. In particular, it will contribute: (1) a better understanding of whether and why the rules that result from state-of-the-art methods to look for optimal decision rules are far from truly optimal; (2) new, effective tools (along with an accessible and user-friendly software package) for tracing truly optimal rules, for drawing correct conclusions about whether the truly best rule is a one-size-fits-all or a personalized one, and for assessing the quality of estimated decision rules; (3) an application of the newly developed tools to data from randomized clinical trials on ADHD and depression.

Date:1 Jan 2019 →  31 Dec 2021
Keywords:Psychotherapie
Disciplines:Mathematical psychology