Operational assessment of nursing homes at times of pandemic: an integrated DEA and machine learning approach

Assessing the performance of nursing homes during pandemics such as COVID-19 is critically important, particularly in light of an aging global population and the heightened need for long-term care. This urgency has led to a heightened global emphasis on optimizing nursing home resources. To address this objective, we developed a hybrid method that integrates Data Envelopment Analysis (DEA) with Machine Learning (ML) techniques to improve and predict the performance of these facilities. We applied this innovative approach to over 500 nursing homes across Pennsylvania. Given the complex regulatory and funding environments, with significant variations across regions, we performed a comparative efficiency analysis using DEA across three Pennsylvania regions: West, East, and Central. Once we identified the sources of inefficiency, we suggested actionable solutions to improve these facilities. We further utilized ML techniques to predict efficiency of nursing homes. Our results showed that the number of citations, complaints, COVID-19 cases, and COVID-19 related deaths as critical factors affecting nursing home efficiency. Comprehensive approaches to address these factors include refining staff training programs, adopting regular feedback mechanisms, enhancing regulatory compliance, strengthening infection control practices, and managing resources effectively. These measures are vital for improving the quality of care and operational efficiency in nursing homes.