| Popis: |
General practitioners (GPs) play a crucial role as initial outpatient care providers and serve as coordinators of subsequent treatment pathways, ensuring efficient and effective care. Moreover, they provide many minor treatments themselves at low cost. Between 2008 and 2018, however, initial care provision by GPs decreased from 59% to 55% of accidents in Switzerland. At the same time, initial care by emergency departments (EDs) increased from 29% to 35%. Previous research has shown that health care costs of trauma patients are lower when treatment is provided by GPs instead of hospitals. Therefore, concerns arise that the shift away from GPs providing initial care is driving up costs. We test this hypothesis and analyse not only the average treatment costs per case, but the entire distribution. Based on claims data from the largest Swiss accident insurer (approx. 250,000 cases per year, from 2008 to 2018), we create individual treatment sequences to determine when patients were treated and by whom. We then analyse to what extent the observed increase in costs-per-case over this period can be attributed to a change in sample composition and to structural factors. Our cost decomposition analyses are based on distribution regression and non-parametric reweighting. Unlike prevailing methods such as Oaxaca-Blinder, this allows us to analyse the entire distribution and to rely on less restrictive assumptions. In decomposing the total cost-per-case increase, we can distinguish the contribution of the following factors: 1) demographic characteristics, 2) injury-related characteristics, 3) initial care provider, and 4) involved care providers. The cost-per-case increase from 2008 to 2018 is mainly, and especially for expensive cases, due to a change in the composition of the cases and not to structural factors. Changes in the type of involved providers along the treatment pathway is by far the biggest contributor to the observed cost-per-case increase. Supporting providers, such as radiologists or physiotherapists, are contributing significantly to all areas of the cost distribution. The involvement of specialists has a cost-driving effect especially for expensive cases. The type of initial care provider, however, makes no relevant contribution to the observed cost-per-case increase. We find no evidence that the shift from GPs to EDs as initial care providers is associated with an increase in cost-per-case. Nevertheless, GPs treat patients at lower costs compared to EDs - especially in less complex cases. However, it is rather the number of different health care providers involved that contributes to the increase in costs. Accordingly, especially more expensive, and thus more complex, cases are being treated by an increasing number of different providers. |
