Designing studies to investigate client-treatment matching

Abstract

The process of designing studies of client-treatment matching involves a number of methodological complexities. Besides the large number of potential client characteristics and a range of possible outcome variables, a variety of treatment attributes can be matched including modality, intensity, duration, format, setting, therapist and goal. Hindsight matching designs (which can include hypothesis testing) examine interactions between client characteristics and treatments to which they were not intentionally matched. Foresight matching designs involve a prospective experimental test of a matching hypothesis, comparing the outcomes of cases matched by specific decision rules with other cases not so matched. Under certain conditions, these hindsight and foresight matching designs are logically equivalent. Mathematical modeling of clinical judgment offers yet another promising avenue for future studies of matching schemas. Glaser's core-shell strategy provides a practical model for implementing clinical systems and research on matching within ongoing services.