Predicting muscular pain outcomes

A UK study of 650 patients with non-inflammatory musculoskeletal pain across five general practices showed combining information on the duration of the episode, pain interference in daily activities, the presence of pain at multiple sites and overall judgement of the GP allowed them to correctly classify patient outcomes in 69% of cases.

The authors attributed improved prognostic accuracy with the combined model to it correcting GPs’ tendency towards overoptimistic expectations of outcome.

Unfavourable outcomes were predicted by GPs 37% of the time, compared with an actual rate of 48%.