Geo Data in Healthcare: A Methodological Framework for Analyzing Spatial Inequalities in Out-of-Hospital Cardiac Arrest

Introduction Cardiac arrest is the loss of mechanical cardiac function and systemic circulation; when it occurs outside hospital settings, it is defined as out-of-hospital cardiac arrest (OHCA). OHCA is influenced by both individual and contextual factors, including population structure, socioeconomic disadvantage, and access to emergency care. Geographic approaches offer a framework to investigate spatial inequalities and support place‑based decision making. Objectives This study assesses spatial inequalities in OHCA incidence in relation to demographic (ageing index, population density) and socioeconomic (low income) determinants in Lombardy between 2021 and 2024. Methods A municipal‑level retrospective ecological study was conducted in Lombardy (2021–2024). The outcome was an OHCA event. Data were obtained from the Lombardia Cardiac Arrest Registry (CARe), covering 924 municipalities across seven provinces. The eligible population included residents of municipalities in the CARe registry (cumulative at‑risk population: 17,612,296 inhabitants). Due to incomplete registry coverage in 296 municipalities, analyses focused on 628 municipalities forming the South‑East (SE) macro‑area (Pavia, Lodi, Cremona, Mantua, Brescia). Determinants included: proportion of taxpayers with annual income <€10,000 (proxy for socioeconomic vulnerability), ageing index (proxy for demographic imbalance), and population density (proxy for rurality/urbanization). Contextual covariates were: bystander cardiopulmonary (CPR) rate (proxy for community response) and EMS response time (minutes from emergency call to first vehicle arrival, proxy for geographic isolation).The analytical strategy combined global and local spatial autocorrelation (Global Moran’s I; Local Indicators of Spatial Association, LISA), hotspot analysis (Getis‑Ord Gi*), and geographically weighted regression (GWR) to identify clustering patterns and spatially varying associations between OHCA and determinants, adjusting for covariates. Results The cumulative OHCA incidence was 126 per 100,000 inhabitants. OHCA incidence was not randomly distributed, but showed significant spatial heterogeneity. Global Moran’s I confirmed clustering (I = 0.27; p < 0.001). Municipalities characterized by high- and low-incidence aggregation patterns were identified (Getis-Ord Gi* z-scores > |1.96|): a hot-spot cluster located in the south-western hilly area of the region and a cold-spot cluster in the south-eastern flat area. GWR showed that the strength and direction of associations varied across the region, highlighting spatial non-stationarity in the relationship between OHCA incidence and the area-level determinants, independently of community response, response time, and geographic isolation. Specifically, increasing the ageing index and the proportion of low-income individuals significantly increased OHCA incidence, whereas population density was inversely correlated with OHCA incidence. Conclusions OHCA incidence hot spots were concentrated in mountainous and hilly areas, especially in remote areas distant from hospitals. Indicators of social inequality showed spatially heterogeneous associations with OHCA, confirming that territorial disparities extend beyond demographic composition alone. Considering both structural determinants and contextual covariates improved interpretation of spatial patterns and highlighted the contribution of place‑based factors to OHCA burden. The integration of geographic data with health information represents a robust methodological approach for studying spatial inequalities in healthcare. Cardiac arrest findings may be translated into public health impact and guide equity-oriented planning by identifying priority areas and informing geographically targeted interventions.