A data-centric approach for optimizing multi-objective healthcare service accessibility and equity

The optimal location of healthcare facilities is a critical challenge in public health, directly influencing service accessibility and efficiency. This study presents a data-driven optimization framework for healthcare facility location, leveraging real-world data from three Brazilian cities within the Unified Health System (SUS). The main contribution of this work is a multiobjective healthcare facility location model, solved using an ɛ-constraint approach, which explicitly analyzes trade-offs among demographic and socioeconomic criteria to support decision-making in the SUS. We conducted five experiments to evaluate different location–allocation strategies: (i) Facility Relocation, where all facilities are relocated to optimize coverage; (ii) Expanding the Facility Network, increasing the number of facilities by 20%; (iii) Hybrid Optimization, keeping the top 80% of existing facilities fixed while adding new ones to maintain the original total; (iv) Sensitivity Analysis on the Number of Facilities, assessing how coverage varies with different facility counts; and (v) Sensitivity Analysis on the Capacity Limit, examining the effects of capacity constraints on service efficiency. The optimization process incorporates key social metrics, including population density, households in precarious conditions, social deprivation index, the population below the poverty line, and household income. The results demonstrate that integrating these socioeconomic indicators into the facility location–allocation process significantly enhances healthcare accessibility and equity, even under budgetary and logistical constraints. These findings provide valuable insights for policymakers seeking to improve healthcare service distribution through data-driven decision-making.