Improving Manpower Allocation at Primary Healthcare Facilities: Development and Validation of a Machine Learning Quadratic Model to Strengthen Public Health Service Availability

The paper addresses the critical issue of human resource shortages in healthcare, particularly in developing countries like India. It emphasizes the importance of adequate and well-distributed health professionals for achieving desirable health outcomes. The focus is on quadratic modelling as a tool for optimal planning in healthcare systems, considering the composition and distribution of health personnel.The methodology involves creating a comprehensive unit optimization framework using a Demand Index, Supply Index, and Distance Matrix to derive a Site Suitability Index. The Demand Index integrates demographic, mortality, and footfall data, while the Supply Index considers medical, para-medical, and non-medical staff weights. A Gravity model is employed for accessibility scoring, and a three-by-three matrix is used for rule-based classification. The quadratic optimization model aims to maximize human resource allocation based on physical accessibility, minimizing the cost of deviation from desired healthcare standards. Data mining and cleaning involve secondary data from various sources, subjected to screening using geographical and population data.Results include a Gravity model-derived Baseline Accessibility and Burden Index, categorizing villages and PHCs. The quadratic optimization model allocates resources according to IPHS standards, using a weekly timetable for staff sharing to address shortages. The model successfully optimizes human resource allocation based on demand burden and accessibility, presenting a systematic approach to address healthcare workforce challenges.