Measuring shoulder function: a systematic review of four questionnaires

Abstract

Objective: To conduct a systematic review of the quality and content of the psychometric evidence relating to 4 shoulder disability scales: the Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire, the Shoulder Pain and Disability Index (SPADI), the American Shoulder and Elbow Surgeons (ASES) score, and the Simple Shoulder Test (SST).

Methods: We conducted a structured search using 3 databases (Medline, CINAHL, EMBase). In total, 71 published primary studies were analyzed. A pair of raters conducted data extraction and critical appraisal using structured tools. A descriptive synthesis was performed.

Results: Quality ratings of 55% of the studies reviewed reached a level of > or =75%. Most studies suggest that all 4 questionnaires have excellent reliability (intraclass correlation coefficient > or =0.90). The 4 questionnaires are strongly correlated (r >0.70) with each other and with a number of similar indices, and the questionnaires were able to differentiate between different populations and disability levels. The minimal detectable change (MDC) is approximately 9.4 for the ASES, 10.5 for the DASH, and 18 for the SPADI; the minimal clinically important difference (MCID) is approximately 6.4 for the ASES and 10.2 for the DASH, and ranges between 8 and 13 for the SPADI. MDC and MCID have not been defined for the SST.

Conclusion: The psychometric properties of the ASES, DASH, and SPADI have been shown to be acceptable for clinical use. Conversely, some properties of the SST still need be evaluated, particularly the absolute errors of measurement. Overall, validation studies have focused on less clinically relevant properties (construct validity or group reliability) than estimates of MDC and MCID.