Validation of performance indicators for rehabilitation of workers with mental health problems
TNO Kwaliteit van Leven
de Boer, A.G.E.M.
van Dijk, F.J.H.
Objective: The objective of this study was to evaluate content validity and predictive validity of 11 performance indicators for the rehabilitation of workers with mental health problems. Research Design and Subjects: Content validity was assessed by experts who evaluated the validity of the performance indicators in 2 Delphi rounds. Predictive validity was evaluated by relating the performance indicators and their sum score to the outcome of a cohort of 191 employees absent from work as a result of mental health problems. Scores for the performance indicators were obtained by processing registration forms of consultations filled out by occupational physicians. Main Outcome Measures: Three outcome measures were used: time to return to work, change in level of fatigue, and patient satisfaction. Results: Ten of the 11 performance indicators showed adequate content validity according to the expert panel. The evaluation of predictive validity yielded mixed results. One performance indicator did not show sufficient variability and was excluded from further analysis. The sum score of 9 performance indicators and performance on the evaluation of work disabilities were significantly related to a shorter time to return to work (hazard ratio [HR], respectively, 0.7; confidence interval [CI], 0.7-0.9 and 0.5; CI, 0.2-0.9). Adequate care regarding interventions aimed at providers of care in the curative sector was related to a longer time to return to work (HR, 1.8; CI, 1.1-3.0). The linear regression revealed that the sum score was not significantly related to a change in level of fatigue. However, lower quality of overall care was significantly related to moderately higher patient satisfaction (β = 0.18; P < 0.05). Conclusions: The evaluated performance indicators showed sufficient content validity and overall predictive validity, but no clear relation could be established between individual performance indicators and outcome. Copyright © 2005 by Lippincott Williams & Wilkins.
To reference this document use:
Health care quality
Proportional hazards model
Mental Health Services
Proportional Hazards Models
Quality Indicators, Health Care
Medical Care, 43 (43), 1034-1042