Entropy-weighted assurance region DEA: an application to healthcare system efficiency assessment

Abstract The paper proposes a methodological framework for performance efficiency assessment that integrates entropy weighting with data envelopment analysis (DEA) and assurance region constraints. Entropy-derived lower and upper bounds for input and output weight ratios are constructed, providing a more objective, data-driven way to limit weight flexibility without relying on additional information or expert judgment. The proposed model addresses key limitations of classical DEA - particularly its tendency to assign extreme or zero weights that can artificially overestimate the efficiency of low-performing units. By incorporating a scalable parameter, the model enables a controlled narrowing of the admissible weight space. The methodology is applied to the assessment of healthcare systems in European countries. Four aggregated input variables cover the staff providing healthcare, the scope and availability of healthcare services and the population behaviour. The survival to 65 years of age and the expected healthy life years at age 65 as significant health outcomes covering the whole human life are considered outputs. Efficient healthcare systems with satisfactory outcomes are identified, and the healthcare systems are ranked based on adjusted efficiency scores. The results reveal that, under stricter assurance region constraints, the DEA frontier becomes more robust, and efficiency scores better reflect acceptable performance. The approach provides a flexible tool for analysts and policymakers seeking to evaluate efficiency with greater realism and policy alignment.