Level of patient–physician agreement in assessment of change following conservative rehabilitation for shoulder pain

Background: Assessment of health-related status has been shown to vary between patients and physicians, although the degree of patient–physician discordance in the assessment of the change in status is unknown. Methods: Ninety-nine patients with shoulder dysfunction underwent a standardized physician examination and completed several self-reported questionnaires. All patients were prescribed the same physical therapy intervention. Six weeks later, the patients returned to the physician, when self-report questionnaires were re-assessed and the Global Rating of Change (GROC) was completed by the patient. The physician completed the GROC retrospectively. To determine agreement between patient and physician, intra-class correlation (ICC) coefficient and Pearson’s r using the 15-point GROC and weighted kappa using a consolidated three-point GROC were calculated. Results: Utilizing the 15-point GROC, complete agreement was observed in 37 of 99 patients (37%). ICC and Pearson’s r between patient and physician were 0.62 and 0.63, respectively. Utilizing a consolidated three-point GROC, complete agreement was observed in 76 of 99 patients (77%). Weighted kappa was 0.62. Conclusions: Assessment of change reported by the patient demonstrates moderate to good agreement with physician assessment. These findings indicate that the GROC does reflect and represent similar assessment of change in health status by patients and physicians. This can aid discussion of both past treatment results and future treatment plans.