Bone and Joint Institute
Rasch analysis indicates that the Simple Shoulder Test is robust, but minor item modifications and attention to gender differences should be considered
Document Type
Article
Publication Date
7-1-2017
Journal
Journal of Hand Therapy
First Page
348
Last Page
358
URL with Digital Object Identifier
10.1016/j.jht.2017.01.005
Abstract
© 2017 Hanley & Belfus Study Design Repeated cross-sectional study. Introduction Multiple studies have evaluated the psychometric properties of the Simple Shoulder Test (SST) through traditional methods supporting it as valid and reliable. Since the evidentiary pool supporting the use of the SST has only partially addressed key measurement properties and the development of SST pre-dates the common use of Rasch model, validation of SST has become a necessity to establish as a reliable and valid PRO for shoulder conditions. To date, no study has analysed SST through Rasch, a modern method for analyzing properties of measurement tools. Purpose of the Study The purpose of this study was to perform a Rasch analysis of the SST to assess the overall fit to the Rasch model, individual item fit, gender-based DIF, local dependency of items and the unidimensionality of the scale. A secondary purpose was to determine the stability of fit to the Rasch model when captured pre-operatively or post-operatively. Methodology Patients completed SST before surgery and between 6 months and 1 year after surgery. Rasch analysis was performed to analyse the carious properties of SST through the Rasch model. Results SST appears to be robust when tested against the Rasch model. Rasch analysis has highlighted potential areas for to improve in the SST questionnaire. The potential areas to improve are to consider questions that measure the ability of a person to lift the arm above shoulder level and to consider gender differences when measuring the ability to carry weights with the affected arm. Discussion This study adds to previous body of empirical evidence arising classical measurement approaches that have suggested that the SST has robust measurement properties, by providing evidence of adequate fit to the Rasch model after minor adjustments. Conclusion The results of this study should provide confidence to clinicians on SST who wish to use a brief shoulder-specific measure in their practice. The SST appears to be robust when tested against the Rasch model despite some potential areas for improvement. The potential areas that should be explored in future Rasch analyses are the questions that measure the ability of a person to lift the arm above shoulder level and the potential for gender differences when measuring the ability to carry weights with the affected arm.
