Estimating Optimal Rules for Individually Tailored Treatments (Dynamic Treatment Regimes)

Dynamic treatment regimes are individually tailored treatments that provide treatment to individuals only when and if they need the treatment and adjust the level of treatment to the individual's need. In a dynamic regime, decision rules for how the treatment level and type should vary with time are specified prior to the beginning of treatment; these rules are based on time varying measurements of subject-specific status. These regimes hold the promise of maximizing treatment efficacy by avoiding ill effects due to over-treatment and by providing increased treatment levels to those who can benefit.

The ascertainment or estimation of optimal decision rules belongs to the class of sequential or multistage decision problems. This talk discusses methodology for estimating optimal decision rules, rules that when implemented over a time period will produce the highest mean response at the end of the time period. The proposed methodology will use experimental or observational longitudinal data to construct estimators of the optimal decision rules.

Although there has been very little design and analysis work, dynamic treatments are beginning to appear in preventive-interventions for drug abuse and conduct disorders and in the treatment of chronic, relapsing disorders such as substance abuse and depression.

This work highlights underlying statistical issues and assists in the design.