O nível de eficiência dos hospitais públicos do Estado da Bahia: um estudo através da análise envoltória de dados para os anos de 2017, 2018 e 2019

Abstract

This study evaluates the relative efficiency of public hospitals and maternity units in the state of Bahia, Brazil, between 2017 and 2019, using the Data Envelopment Analysis (DEA) method under the variable returns to scale (VRS) and output-oriented specification. The research uses secondary data extracted from the APURASUS and DATASUS systems, considering inputs related to total costs and installed capacity, and outputs associated with ambulatory care, surgical procedures, hospitalization indicators, and mortality rates. To strengthen the interpretation of results, the units were classified into two analytical clusters based on the median value of the Hospital Admission Authorization (AIH), which serves as a proxy for the complexity of treated cases. The findings indicate substantial heterogeneity in efficiency levels, with hospitals of lower average AIH value systematically operating closer to the efficient frontier. In contrast, high-complexity units exhibited greater variability, lower efficiency scores, and a higher prevalence of decreasing returns to scale. Additionally, the analysis identified a significant potential for cost reduction if all hospitals operated at efficient levels. The study contributes to the understanding of efficiency dynamics within the public health network of Bahia and offers managerial insights for resource allocation, operational improvement, and strategic planning. Limitations and recommendations for further research are also discussed.