Substantial clinical benefit, responsiveness, and sensitivity to change of three common outcome measures following shoulder arthroplasty

OBJECTIVES: It is important for clinicians involved in the care of patients with advanced glenohumeral osteoarthritis to determine clinically significant change when using outcome measures. There is little information on the amount of substantial clinical benefit in shoulder outcomes after shoulder arthroplasty. The purpose of this study was twofold: (1) to quantify substantial clinical benefit for the American Shoulder and Elbow Surgery score, the Constant Murley Score, and the Western Ontario Osteoarthritis of the Shoulder index and (2) to provide estimates of responsiveness and sensitivity to change for these measures following shoulder arthroplasty. METHODS: The study involved a secondary analysis of previously collected data. The substantial clinical benefit and responsiveness of the measures were calculated based on external anchors related to change in pain, range of motion, and ability to carry out activities of daily living. The areas under curve and standardized response mean represented responsiveness and sensitivity to change. RESULTS: The data of 159 and 131 patients with complete follow-up at 6 months and 2 years were reviewed. The amount of substantial clinical benefit was dependent on the outcome measure and the external anchor and increased for all measures from 6 months to 2 years. Responsiveness was high (areas under curve > 0.80) at 6 months and further improved at 2 years (areas under curve > 0.88). The standardized response mean values of both time points were over 2.00, indicating high effect sizes. The standardized response means of the Constant Murley Score were statistically significantly higher than the standardized response means of the American Shoulder and Elbow Surgery and Western Ontario Osteoarthritis of the Shoulder. CONCLUSION: Amount of substantial clinical improvement in pain, range of motion, and activities of daily living following shoulder arthroplasty depends on the type of outcome measure used. All three measures, the American Shoulder and Elbow Surgery, absolute and relative Constant Murley Score, and Western Ontario Osteoarthritis of the Shoulder, demonstrated good to excellent accuracy and optimal standardized response means. LEVEL OF EVIDENCE: Level III, Retrospective Cohort study