Resource allocation and patient care: a two-stage network-DEA model for optimising rural hospital operations

We propose an evaluation framework to enhance U.S. rural hospital efficiency and sustainability amid rising costs and staffing shortages. Moving beyond ‘black-box’ evaluations, we utilise a two-stage Network-Data Envelopment Analysis (N-DEA) Slack-Based Measure model - Resource Allocation and Patient Care Delivery – to analyse internal inefficiencies. To ensure validity, 57 rural hospitals were segmented by size. To capture performance evolution, we integrate the Hicks-Moorsteen Productivity Index to evaluate annual Total Factor Productivity (TFP) shifts. We model hospitals as production systems converting resources into health outcomes. Applied to a 2018–2023 dataset, the model reveals that Patient Care Delivery consistently outperforms Resource Allocation. Longitudinal analysis indicates that while 48 of 57 hospitals recorded net TFP increases, volatility persists, particularly among small and medium facilities. This suggests performance gains are achievable by addressing Resource Allocation bottlenecks. Slack analysis reveals divergent profiles: small hospitals frequently operate with zero slack, whereas the broader sample shows 20% average excess for physicians and certified nurse practitioners. Scenario testing confirms modest reductions (25%) in these roles would enhance efficiency by 1.8%. Conversely, we recommend protecting registered nurse staffing due to minimal slack (6%), as potential efficiency gains are outweighed by patient care risks. This study offers data-driven insights for healthcare decision-makers.