Hospital reform in Germany: Population-based calculation and simulation model for planning and impact assessment

Abstract

Background: Due to the increasing shortage of specialists, disproportionately rising costs and inadequate quality of care, Germany is planning a fundamental hospital reform. The government commission has developed the central principles for this reform. The aim is to achieve a centralization of hospital services in oversupplied metropolitan areas by planning according to groups of services based on defined requirements for structural quality. The economic pressure to increase the number of cases shall be reduced through the introduction of a reserve payment and by dividing hospitals into levels - at least by transparent information for the general public. So far, however, there has been no generic model to determine the significance of specific care providers considering population needs (reachability), capacity of a hospital location, and the number and severity of treated patients. Methods: We developed a generic model to determine the significance of hospital location-related care and the need for the various groups of services. The model may also be used to assess reserve financing. For the model, the groups of services were divided into four categories according to degree of specialization and urgency with limits of reachability of 30, 45, 90 and 180 minutes. Gravity models were used to simulate the population-based significance of care providers for each group of services. In the basic model, the allocation of the significance was based exclusively on the reachability for the population living within the respective reachability corridor. In extended models, the historical number of cases and their case severity were also taken into account in order to account for current care capacities and existing choices of the population regarding, among other things, the quality of care. The model was implemented on the basis of nationwide data provided in accordance with § 21 data (of the German law for hospital remuneration). We then determined the effects of weighting the three influencing variables (i) population, (ii) number of cases and (iii) case severity on the significance of the hospital sites. Using the example of "endoprosthetics of the knee" (service group 14.2 in North Rhine-Westphalia) in the federal state of Saxony, the effect of concentration on the distribution of the reserve budgets and on the accessibility of the population was illustrated. The importance of care determined in this way is contrasted with a key value that reflects the need for hospitals in terms of securing care in rural areas. Results: From the approximately 16.5 million somatic treatment cases in 2021, 98.5% could be assigned to one of the 60 somatic groups of services according to North Rhine-Westphalia model. The simulation models show a differentiated picture for the various service groups. For the majority of service groups, a moderate concentration of services does not lead to relevant restrictions in terms of reachability. Exclusively considering the population to be cared for (basic model) would lead to significant shifts in the financing of running costs, which are, however, very well mitigated by considering the historic number and severity of cases (extended models). As an example, we show the effects for the service group "endoprosthetics of the knee" in the federal state of Saxony compared to the extrapolation at location level. Discussion: The empirically based simulation model proposed here takes into account reachability, patient preference and hospital capacity and offers a scientific way of comparing the regional significance of care providers as well as the necessity of hospital locations for each service group against the background of the state's obligation to provide care and to make economical use of resources. The simulation model is meant to support hospital planning (here: the allocation of groups of services) in the federal states and to guide rational planning. It is also suitable for the management of financial resources in the context of hospital reform. It also enables an impact analysis. The model is not intended to automate planning or otherwise make rigid specifications. The calculations should be regarded as exemplary. The weights of all parameters can be varied. However, the reachability thresholds and the parameterization of the simulation model should be defined jointly by federal and state governments in terms of a common set of objectives.