The project aimed at developing and testing a new payment system which provides financial incentives for rehabilitation centers to achieve the best outcomes possible for their patients but does not create additional costs for the insurance funds. The system is conceived as a "quality competition" organized by the centers among themselves with a scientific institute acting as a "referee". Centers with outcomes above average receive a bonus financed by a corresponding malus from the centers below average. In a stepwise process which started in 2001 and was continually accompanied by a scientific institute, we developed the methodological and organizational prerequisites for the new payment system and tested them in two multicentric studies with large case numbers (n=1,058 and n=700, respectively). As a first step, a new assessment instrument (SINGER) was developed and validated in order to measure the outcomes in a reliable, valid, and change-sensitive way. In the second phase, we developed a regression analytic model which predicted the central outcome variable with >84% variance explained. With this model, the different case-mix in the participating centers can be controlled, so that comparisons of outcomes across centers can take place under fair conditions. In the recently completed third phase, we introduced an internet-based programme SINGER-online into which the centers can enter all relevant data. This programme ensures a high quality of all data and makes comparisons of outcomes across all centers possible at any chosen time. The programme contains a special module accessible to the medical services of the health insurance only, which allows sample checks of the data entered by the clinics and helps to ensure that all centers keep to the principles of a fair competition for better quality for their patients. After successful testing of these elements, a functioning model of pay-for-performance in rehabilitation after stroke is now available.
(c) Georg Thieme Verlag KG Stuttgart, New York.