Measuring Satisfaction With Upper Limb Prostheses: Orthotics and Prosthetics User Survey Revision That Includes Issues of Concern to Women

Abstract

ObjectiveTo (1) modify the Orthotics and Prosthetics User Survey (OPUS) Client Satisfaction with Device (CSD) instrument to incorporate issues of concern to women and (2) evaluate measure's structural and concurrent validity and reliability in persons with upper limb amputation (ULA). DesignCross-sectional survey study with retest after 2 weeks. Exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and Rasch analyses were used to select items and examine differential item functioning, range of coverage, and person and item reliability. Test-retest reliability was evaluated with intraclass correlation coefficients. Pearson correlations were used to estimate associations with other prosthesis satisfaction measures. SettingTelephone administered survey. ParticipantsConvenience sample of 468 participants in the US (N=468; 19.9% women) with ULA, including a 50-person retest subsample (4% female). InterventionsNot applicable. Main Outcome MeasuresModified OPUS CSD. ResultsEFA suggested 3 subscales: Comfort, Appearance, and Utility. CFA found acceptable model fit. After dropping items with poor fit and high pairwise correlations in Rasch partial credit models, CFA model fit indices were acceptable (comparative fit index=0.959, Tucker-Lewis Index=0.954, root mean square error of approximation=0.082). Rasch person reliability was 0.62 (Utility), 0.77 (Appearance), and 0.82 (Comfort). Cronbach α was 0.81, 87, and 0.71 for Comfort and Appearance, and Utility subscales, respectively. Correlations between the modified CSD, the original CSD, and the Trinity Amputation and Prosthesis Experience Satisfaction Scale were 0.54-0.94. ConclusionsWe identified 3 subscales: Comfort (6 items), Appearance (8 items), and Utility (4 items) with 7 new items identified as important to women. The subscales demonstrate evidence of sound concurrent structural and test-retest reliability and concurrent validity. The Appearance and Comfort subscales have good reliability for group-level use in clinical and research applications, whereas the Utility subscale had poor to fair person reliability but excellent item reliability.