Shoulder-specific disability measures showed acceptable construct validity and responsiveness

Abstract

Objective: To compare the construct validity and responsiveness of the Shoulder Pain and Disability Index (SPADI), Croft Index, and disabilities of the arm, shoulder, and hand (DASH) for patients with adhesive capsulitis and to compare these with pain on a visual analog scale, the Health Assessment Questionnaire, and the problem-elicitation technique.

Study design and setting: Two randomized double-blind placebo-controlled trials of interventions for adhesive capsulitis were performed. Both trials recruited patients from community-based physiotherapy practices. Responsiveness at 3 weeks postbaseline was assessed using four responsiveness parameters and three external criteria for improvement.

Results: Correlations between the shoulder-specific measures ranged from 0.55 to 0.65 at baseline and 0.49-0.55 for the 3-week change scores. Greater responsiveness was seen for the SPADI. The effect size for the SPADI ranged from 1.20 to 1.64, for the Croft Index from 0.87 to 1.21, and for the DASH from 0.55 to 0.83. Rankings were similar across the four responsiveness parameters and the three external criteria for improvement. Correlations between the shoulder-specific and generic measures for baseline and 3-week change scores were lower than those among the shoulder-specific measures (range: 0.17-0.60).

Conclusion: The shoulder-specific disability measures showed acceptable construct validity and responsiveness with a small but consistent overall advantage for the SPADI.