Measurement Properties of the Full and Brief Version of the Work Rehabilitation Questionnaire in Persons with Physical Disabilities

• Published in: Journal of occupational rehabilitation 2021-12, Vol.31 (4), p.886-894
• Main Authors: Roels, Ellen H., Reneman, Michiel F., Post, Marcel W. M.
• Format: Article
• Language: English
• Subjects: Adult; Agreements; Clinical Psychology; Consistency; Convergence; Correlation analysis; Criteria; Cross-Sectional Studies; Disabilities; Disabled Persons; Environmental Health; Health Psychology; Humans; Male; Measurement; Medicine; Medicine & Public Health; Middle Aged; Occupational Medicine/Industrial Medicine; Orthopedics; Physically disabled persons; Questionnaires; Rehabilitation; Rehabilitation, Vocational; Reliability; Reproducibility of Results; Social service; Spinal cord injuries; Stroke; Surveys; Surveys and Questionnaires; Validity; Vocational rehabilitation
• ISSN: 1053-0487; 1573-3688
• DOI: 10.1007/s10926-021-09973-8

Purpose The Work Rehabilitation Questionnaire (WORQ) is a self-report vocational rehabilitation assessment. A comprehensive (WORQ-FULL) and a brief version (WORQ-BRIEF) are available. The purpose of this study was to investigate measurement properties of both versions in persons with physical disabilities. Methods Cross sectional and test–retest design. Adults with physical disabilities in vocational rehabilitation were included. Internal consistency (Cronbach’s alpha), test–retest reliability (intra-class correlation; ICC), agreement between sessions (Bland–Altman Plots), criterion validity (ICC and agreement with Bland–Altman Plots between WORQ-FULL and WORQ-BRIEF) and convergent validity with the Work Ability Index -Single item (WAS) and the EuroQOL 5D-5L were analyzed. Results Out of the 91 individuals who agreed to participate, 74 (81%) returned questionnaire T1 and 49 (54%) participants returned questionnaire T2 within the maximum time interval (= 27 days). At T2, 28 (57%) participants reported no medical changes compared to T1. Median age was 49 (IQR 40–60), 57% were male, 47% had experienced a stroke and 27% a spinal cord injury (n = 49). Internal consistency was good: 0.95/0.95/0.94 for the WORQ-FULL and 0.88/0.89/0.85 for the WORQ-BRIEF (n = 74/n = 48/n = 28, respectively). Test–retest reliabilitywas good: ICC = 0.86/0.85 for the WORQ-FULL and ICC = 0.87/0.86 for the WORQ-BRIEF (n = 49/ n = 28). Bland Altman plots suggested a higher score at T1. As for criterion validity of the WORQ-FULL versus the WORQ-BRIEF, ICC was good (ICC = 0.84; n = 74), however Bland Altman plots indicated potential bias. Correlations with the WAS/EuroQOL 5D-5L were variable: r = -0.24/r = -0.57 (WORQ-FULL) and r = -0.28/-0.65 (WORQ-BRIEF). Conclusions The WORQ showed good internal consistency and test–retest reliability. Agreement demonstrated large score differences are needed to indicate change beyond random chance at individual level, whereas small changes are sufficient at group level. Criterion validity of the WORQ-FULL versus the WORQ-BRIEF was supported, however, agreement demonstrated moderate to large score differences are needed to indicate change beyond random chance at individual level, whereas small changes are sufficient at group level. This indicates the WORQ-FULL and WORQ-BRIEF are better not used interchangeably. Correlation analyses provided better insight in the validity of the WORQ. Convergent validity was supported for the WORQ-BRIEF with the EuroQoL