A meta-analysis of the impact of sample, kind of outcome measurement and time of follow up on occupational re-integration after vocational retraining

Objective: Vocational Rehabilitation (VR) is an essential element of intervention to rehabilitate people with work disability due to chronic diseases. The activities are heterogeneous; of particular importance in the rehabilitation process is vocational retraining. These interventions are cost-intensive, take very long and have a decisive impact on the main goal, the return to work (RTW). However, vocational retraining is conducted under different settings: The question is to what extent this leads to specific methodical implications of RTW measurement that have an impact on the level of RTW.

Methods: The analysis was concentrated on the main outcome RTW after vocational retraining. A structured review was conducted of all German-language publications from 2005 to 2010 which report an RTW quota after vocational retraining. The main methodical conditions were: kind of RTW measurement (point in time rate vs. cumulative course rate), time of follow-up in years, and sample definition (all participants vs. participants with regular completion of retraining, RC). The impact of these conditions on the level of RTW was predicted by using a meta-regression model. The time of follow-up was standardized on the mean (1 year).

Results: 20 publications from 10 studies were included in the analysis. In all, 23 RTW quota were observed, from which 22 were included in the regression model. A positive impact on the level of RTW was identified for reduction of the sample to participants with RC (b=10.04; p=0.001), the time of follow-up (b=5.47; p=0.052) and, by trend, the interaction of the kind of RTW measurement and time of follow-up (b=6.32; p=0.090). There was no significance given for the main effect of kind of RTW measurement (p=0.787). The model fit was calculated with an adjusted R2=75.17%.

Conclusions: The level of RTW given in the publications is very heterogeneous, but a major part of the variance was explained by the 3 methodical conditions we examined. Based on this, a classification system for RTW measurement after VR is presented, which could be a model for designing future studies as well as inter-actor communication of success after VR.