Background: The Probable Cochrane Behavioral Medicine (BM) Field has been involved in preliminary activities to adapt evidence-based methods to fully consider behavioral medicine topics and to consider ways that these behavioral intervention-specific methodologies may be useful to others. For example, we have proposed additional CONSORT items to be reported when detailing behavioral interventions. We have also begun to consider how factors empirically related to behavioral intervention treatment effects, such as patient and provider preferences, can be considered within trial designs. We are working on external validity issues as these are particularly relevant to this literature, including design requirements for conducting practical behavioral trials. Finally, we are exploring important areas of interface with other Cochrane activities (e.g., use of non-randomized designs) that are particularly relevant to this Probable Field.

Objectives: To outline methodologies being developed or adapted for the Probable Cochrane BM Field which may be useful to others.

Methods: We will review the reporting conventions for trial reporting that have been adapted from CONSORT and TREND, approaches to data abstraction tables for unique suggested behavioral medicine components, and the empiric evidence for external validity issues being considered by this Probable Field. We will present case examples from behavioral medicine systematic reviews.

Results: The evidence supporting reporting elements that are (perhaps) specific to behavioral interventions such as the training level required of the provider, the treatment allegience of the provider, and the receipt of the intervention by the patient will be presented. Patients and providers may have strong preferences for therapy and not medication treatments and we will present some of the randomization schemes that consider this variable within their designs. In considering potential policy implications, we have proposed design elements required to conduct a practical behavioral trial. We will also discuss lessons learned from conducting initial reviews of behavioral medicine topics.

Conclusions: Behavioral medicine trials appear to have specific reporting and perhaps design issues that influence reporting and ability to synthesize across studies. The presenters will highlight these issues and recommend strategies to improve the evidence, the reporting of behavioral intervention evidence, the use of this evidence, and its synthesis.