[Conformance, Completeness and Plausibility of the Service Classification of Vocational Rehabilitation for...

In 2017, the service classification of vocational rehabilitation (LBR) became mandatory for vocational rehabilitation centres as part of the German Pension Insurance's (DRV) quality assurance process. This joint study aims to determine the quality of LBR data by analysing routine data on conformance, completeness and plausibility, using vocational training services in vocational retraining centers as an example.

This analysis is based on a scientific use file 'Completed rehabilitation in the insurance history 2015-2022', as well as data on the LBR and the assignment to rehabilitation centres. Following the selection of vocational training services (i. e., qualification and integration measures in vocational retraining centres), 35,878 measures from rehabilitants who ended their rehabilitation from 2018 to 2022 were evaluated. The documented LBR codes were analysed descriptively, as well as using logistic and linear regression models.

Regarding the conformance of the LBR, the analyses showed that the documentation met the structural and content-related requirements. A review of the completeness of the LBR documentation revealed that the documentation increased over the years, but to date (2022), approximately 25% of the measures were still undocumented. The decision regarding documentation depends largely on the rehabilitation centre. Plausibility analyses revealed that the documented duration varied considerably between institutions. Furthermore, it was revealed that expectations regarding the association between personal characteristics (e. g., age, sex) or measure (e. g., termination of measure, scope of rehabilitative assistance) and the documented duration of LBR were only confirmed in part in an exemplarily selected subset of the data (with services relating to health literacy and therapeutic services).

Overall, the results of the analyses allow a positive conclusion to be drawn about the quality of the LBR data with regard to conformance and completeness. In terms of plausibility, however, the results are inconclusive. It has not yet been possible to take any external evaluation standards into account. These should be developed in the next step.