Responsiveness of the Care Dependency Scale for Rehabilitation (CDS-R)

Around 10% of Western Europe's population suffer from a disability which can entail a decrease of independency and quality of life. However, the lives of these people can be improved by rehabilitative treatment and care. Changing the degree of dependency from dependent to independent is essential in rehabilitation, as is the assessment of these changes. To perform such kind of measurements, assessment instruments have to be responsive. In spite of this concern, responsiveness of assessment instruments is studied to a small extent only. This also applies to the Care Dependency Scale for Rehabilitation (CDS-R), a short assessment instrument measuring the care dependency of patients regarding physical and psychosocial aspects. In this longitudinal-study, the responsiveness of the CDS-R, in general and related to different disease-groups, should be determined. Therefore, a convenience sample of 1564 patients was assessed in an Austrian rehabilitation centre with the scale after admission and before discharge. Responsiveness was determined by descriptive analysis, calculation of effect-sizes and significance tests. Differences between admission and discharge occurred on a statistically significant level for patients who changed. Kazis' effect-sizes can be considered as of small/medium effect for patients who changed (0.24/0.49), and as of large effect according to Liang (0.86/1.46). Eta squared was 0.10/0.19 which can be interpreted as of moderate/large effect for patients who changed. Responsiveness-analyses related to different disease-groups showed constantly large effect-sizes for patients with musculoskeletal-disorders. These results indicate that the CDS-R can detect patient-changes over time and discriminate between patients who change under rehabilitation or not. These aspects argue for the responsiveness of the scale, wherefore the CDS-R seems to be appropriate for the assessment of treatment/health-care effectiveness and the evaluation of individual patient-changes. Nevertheless further research is recommended to confirm the level of responsiveness of the scale in general and for different disease-groups.