A tale of two RCTs: using randomized controlled trials to benchmark routine clinical (psychological) treatments for...

This article reports the development of natural history and active treatment benchmarks for psychological treatments of chronic pain. The benchmarks were derived from randomized controlled trials (RCTs) reported in a published meta-analysis. In two preliminary studies we surveyed small samples of active clinicians working in U.K. pain management programs. Study 1 assessed the fit between routine clinical treatment and the selected RCTs. In study 2 Delphi methodology was used to determine a set of outcome domains to be used in the development of benchmarks. In study 3 we extracted data from a set of RCTs where both pre- and post-treatment data were reported. Measures were allocated to 1 of 5 outcome domains (cognitive coping and appraisal, pain experience, pain behavior, emotional functioning, and physical functioning). Pre-treatment to post-treatment effect sizes (Cohen's d) were computed and, where necessary, aggregated within trial so that each trial contributed a single estimate to outcome domain. Effect size (ES) benchmarks were computed for all trials and those trials with an explicit cognitive behavior therapy protocol. In no case did the ES estimates for the untreated control deviate from 0. The average ES across outcome domains for the treatment arms was approximately 0.35. These benchmarks may be used to assess the effectiveness of routine clinical treatments for chronic pain. The application of these data and the limitations of the study are discussed.