Use of the DEA method to verify the performance model for hospitals

Abstract

This paper employs the method of data envelopment analysis (DEA) to compare the relative efficiency of Polish hospitals in meso perspective. It means that the indicators selected to measure the performance of hospitals were aggregated at the regional level (a level of 16 Polish provinces). As the hospitals are a critical part of the healthcare system, they are increasingly the subject of analyses aimed at defining, measuring, and improving their performance. Therefore, in the methodology part, we present the DEA as the method for efficiency measurement together with its advantages and limitations. The study attempts to find out which provinces can be used as models and illustrated the areas where inefficient units need to be improved. The hospital input measures included are the Average time of hospitalisation (in days), Average costs of day hospital treatment. The output measures included are the Average number of patients per bed per year, the Share of accredited hospitals as a proportion of the number of all hospitals, net profit per physician. The DEA models are solved using the computer program Frontier Analyst, Version 4. We found that five provinces were efficient and eleven were not, where the efficiency score varied from 76.2% to 100%. Provinces such as Lower Silesia Province, Lublin Province, Lubuskie Province, Świętokrzyskie Province, and Warmia-Masuria Province were the best performers in that they maximised both quantitative and qualitative outcomes. The identification of the strongest and the weakest within provinces could be beneficial in improving the efficiency and performance of the hospitals. The result identifies the inefficient provinces that can improve their efficiency by making the efficient provinces as their role model. This paper is the first published study that benchmarks the performance of healthcare services in Poland.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 25 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.