When Matching Fails: Understanding the Process of Matching Pain-Disability Treatment to Risk Profile

A previous study (Bergbom et al. in J Occup Rehabil, 2013) showed that matching people at risk for pain-related disability to an intervention aimed at targeting their psychological problem profile did not, as hypothesized, improve the effect of the intervention. Methodological issues were suggested to explain the lack of differential effect. It was questioned whether the profiles used to allocate people to treatment were adequate. The aim of this study was to investigate if the risk profiles used to determine matching were sufficiently stable and valid by comparing the original profiles with profiles constructed using other methods.

Ninety-five people suffering musculoskeletal problems were screened, profiled, and matched to workplace based early interventions according to profiles. We studied stability and validity of their psychological risk profiles by investigating their concordance at different time points. People were originally assigned to profiles at inclusion, using a brief screening questionnaire. Then, they were profiled just before treatment start, using the same items. Finally, they were profiled again at treatment start, using extensive questionnaires. Concordance among the three sets of profiles was investigated.

Profiles at inclusion were unstable until treatment start. People moved from profiles with more severe elevations in psychological variables, to a profile with moderate elevations. Concordance between the two means of profiling at treatment start was better; the brief screening and the extensive questionnaires assigned people to similar profiles.

Risk level may be determined with brief instruments at an early stage of problem development. However, profiles and targets for interventions should be determined immediately prior to treatment start, preferably using full questionnaires.