Institute of Public Health, Department of Health Economics, Faculty of Social Sciences, University of Southern Denmark, Windsløwparken 9B, 5000 Odense C., Denmark.
Health Care Manag Sci. 2010 Dec;13(4):334-45. doi: 10.1007/s10729-010-9133-8. Epub 2010 Jul 6.
The Danish hospital sector faces a major rebuilding program to centralize activity in fewer and larger hospitals. We aim to conduct an efficiency analysis of hospitals and to estimate the potential cost savings from the planned hospital mergers. We use Data Envelopment Analysis (DEA) to estimate a cost frontier. Based on this analysis, we calculate an efficiency score for each hospital and estimate the potential gains from the proposed mergers by comparing individual efficiencies with the efficiency of the combined hospitals. Furthermore, we apply a decomposition algorithm to split merger gains into technical efficiency, size (scale) and harmony (mix) gains. The motivation for this decomposition is that some of the apparent merger gains may actually be available with less than a full-scale merger, e.g., by sharing best practices and reallocating certain resources and tasks. Our results suggest that many hospitals are technically inefficient, and the expected "best practice" hospitals are quite efficient. Also, some mergers do not seem to lower costs. This finding indicates that some merged hospitals become too large and therefore experience diseconomies of scale. Other mergers lead to considerable cost reductions; we find potential gains resulting from learning better practices and the exploitation of economies of scope. To ensure robustness, we conduct a sensitivity analysis using two alternative returns-to-scale assumptions and two alternative estimation approaches. We consistently find potential gains from improving the technical efficiency and the exploitation of economies of scope from mergers.
丹麦医院部门面临着一个重大的重建计划，旨在将活动集中在数量更少、规模更大的医院中。我们旨在对医院进行效率分析，并估计计划中的医院合并可能带来的成本节约。我们使用数据包络分析（DEA）来估计成本前沿。在此分析的基础上，我们为每家医院计算一个效率得分，并通过将个别效率与合并后的医院效率进行比较，估计拟议合并的潜在收益。此外，我们应用分解算法将合并收益分为技术效率、规模（规模）和协调（混合）收益。这种分解的动机是，一些明显的合并收益实际上可能在不完全合并的情况下就可以实现，例如通过分享最佳实践以及重新分配某些资源和任务。我们的结果表明，许多医院在技术上效率低下，而预期的“最佳实践”医院效率相当高。此外，一些合并似乎并没有降低成本。这一发现表明，一些合并后的医院变得过大，因此出现规模不经济。其他合并则导致了相当大的成本降低；我们发现通过学习更好的实践和利用规模经济可以获得潜在收益。为了确保稳健性，我们使用两种替代的规模报酬假设和两种替代的估计方法进行了敏感性分析。我们一致发现，通过提高技术效率和利用合并带来的规模经济，可以获得潜在收益。
