A dynamic programming approach to the efficient design of clinical trials

If the prospective evaluation of all feasible strategies of patient managem ent is not possible or efficient then this poses a number of questions: (i) which clinical decision problems will be worth evaluating through prospect ive clinical research; (ii) if a clinical decision problem is worth evaluat ing which of the many competing alter-natives should be considered "relevan t" and be compared in the evaluation; (iii) what is the optimal (technicall y efficient) scale of this prospective research; (iv) what is an optimal al location of trial entrants between the competing alternatives; and (v) what is the value of this proposed research? The purpose of this paper is to pr esent a Bayesian decision theoretic approach to the value of information wh ich can provide answers to each of these questions. An analysis of the valu e of sample information was combined with dynamic programming and applied t o numerical examples of sequential decision problems. The analysis demonstr ates that this approach can be used to establish: optimal sample size; opti mal sample allocation; and the societal payoff to proposed research. This a pproach provides a consistent way to identify which of the competing altern atives can be regarded as "relevant" and should be included in any evaluati ve study design. Bayesian decision theory can provide a general methodologi cal framework that can ensure consistency in decision making between servic e provision, research and development, and the design, conduct and interpre tation clinical research. (C) 2001 Elsevier Science B.V. All rights reserve d.