Comparison of mediational selected strategies and sequential designs for preventive trials: Comments on a proposal by pillow et al.

Compared the strengths and weaknesses of a novel intervention design strategy called mediational selection, proposed by Pillow et al. (1991), with designs based on a sequence of trials, that is, those which build on prior work and at each stage are administered to subjects from subpopulations who have high risk or as yet poorly quantified risk. Both designs would be most useful in short-term preventive trials. The goals behind these two approaches, however, are quite different. Mediational selection strategies attempt to improve the chances that an intervention developed to modify theoretically important mediational factors will have significant impact by selecting a subject subpopulation that is expected to benefit the most. Sequential designs attempt to identify the extent to which an intervention succeeds or fails to work in subpopulations and allows for adjustment of the intervention to these subpopulations as necessary.