Allocating treatment options to patient profiles:: clinical art or science?

Background For many researchers, the disappointing results of Project MATCH were the death blow for any further activities in the field of patient-treatment interactions. On the other hand, we have an increased knowledge of patient heterogeneity and a greater variety of treatment options than before, and allocation guidelines for an ongoing process of patient-placement decisions are of high practical relevance. Aims To analyse deficits in the current research and to provide suggestions for future action. Findings It is argued that (1) certain major design aspects of Project MATCH and other research studies-such as stringent patient exclusion criteria and low treatment 'dosage'-minimize the chances of detecting possible patient-treatment interactions and (2) Project MATCH obscures our view of previous treatment-allocation research findings. Conclusions Several research strategies and specific research topics are suggested for (1) improving the theoretical and methodological basis for detecting possible patient-treatment interactions and (2) stimulating research on major treatment decision needs, such as site, setting, time in treatment (extensiveness and intensity), service components and specific treatment modules. More international research cooperation is needed to clarify the inconsistent findings.