Measuring shoulder function with the Shoulder Pain and Disability Index

Abstract

Objective: To extend the validity of the Shoulder Pain and Disability Index (SPADI) by (1) making it suitable for telephone administration; (2) determining its convergent validity with other health status measures; and (3) assessing the responsiveness of the SPADI to clinical change.

Methods: Consecutive primary care patients with shoulder discomfort were followed for 3 months. At enrollment, a detailed shoulder specific history was obtained by a trained research assistant, and the Health Assessment Questionnaire (HAQ), the Medical Outcomes Study SF-20 (SF-20), and numeric and visual analog versions of the SPADI were completed by the patient. At 2, 4, and 12 weeks the numeric scaled SPADI was administered by telephone and patients rated globally the change in shoulder discomfort.

Results: One hundred and two subjects were enrolled; 96 completed at least one followup assessment and 75 completed all followup assessments. Subjects were men (98%), predominantly white (73%), with a median age of 60 years, and the majority had experienced shoulder discomfort for > 3 months (66%). At baseline the visual analog (VAS) and numeric scaled SPADI were highly concordant (intraclass correlation coefficient = 0.86), and the SPADI correlated substantially with the HAQ (r = 0.61) and the physical functioning (r = -0.50) and pain (r = -0.43) domains of the SF-20. The SPADI delta (baseline-followup) discriminated accurately between subjects who improved versus those who stayed the same or worsened [receiver operating characteristic cure, (ROC) = 0.91, likelihood ratio for improvement = 34].

Conclusion: The numerically scaled SPADI is highly correlated with the original VAS version of the SPADI and other measures of health status. The SPADI is responsive to change and accurately discriminates among patients who are improved or worsened.