2012 Conference Presentation
Background: Deficits in quality, a lack of professional process management and especially the neglect of outcome quality are criticized in long-term care in Germany. A two-armed cluster randomized controlled trial was conducted to assess whether the Resident Assessment Instrument (RAI) can contribute to an improvement or stabilization of functional abilities (ADL, IADL) and cognitive skills (MMST), to an improvement of quality of life (EQ-5D) and a reduction of institutionalization and thereby increase outcome quality.
Objective: The aim of the study was (i) to investigate, whether the introduction of RAI leads to an improvement of outcomes and (ii) to additionally investigate in how far the effect of RAI depends on the degree of implementation.
Data and Methods: 69 home care services throughout Germany were included and randomized. The treatment group (n=36) got training in RAI and was support by the research team during the study (13 months). Before data analysis a factor analysis was conducted to distinguish between those home care services that implemented RAI intensively (optimal users) and those that did not (suboptimal users). The analysis then compares the control group with the treatment group and its two subgroups.
Results: When comparing mean differences between treatment and control group (n=33), no significant effect was detected. Although in multilevel regression the clients in the treatment group show a better development (smaller decline) according to ADL and IADL, fewer moves to nursing homes and a smaller hospitalization rate, none of these effects is significant. The clients of home care service providers working intensively with RAI were significantly less likely to be hospitalized (p=0.0284) and fared slightly better in assessments using ADL, IADL, MMST and EQ-5D (smaller decline or greater improvement, but not significant) compared to the control group. In contrast, those working not intensively with RAI have slightly worse outcomes (IADL, MMST, EQ-5D) than the control group (not significant).
Discussion: The overall comparison between treatment and control group shows no positive effect of RAI. The optimal users, however, seem to fare better, though effects are only significant for avoidance of hospitalization. If implemented suboptimal, RAI causes more harm than good. If implemented properly, it seems to lead to improvements. The (partial) lack of significance might result from a reduced number of clients included in the study and a power calculation based on the comparison of control and treatment group only. Moreover, there was no entire implementation of RAI and the implementation lasted much longer than expected. The observation period of 1 year might therefore be too short to detect long-term effects. All in all, it is important to guarantee a successful implementation. Therefore, the effort, work load and requirements of qualification involved should not be underestimated and implementation should be supported intensively.