Social Inequality and Access to Healthcare: An Analysis Using Unsupervised Machine Learning in Greater São Paulo

Abstract

The social determinants of health are demographic and socioeconomic characteristics that influence the population’s lifestyle and access to healthcare, including the comprehensive and universal care provided by Brazil’s Unified Health System (SUS). This study aims to understand how population characteristics are associated with the availability of health services in Greater S˜ao Paulo. We applied an
unsupervised machine learning technique to analyze how municipalities within the S˜ao Paulo Metropolitan Region cluster based on demographic and socioeconomic indicators. We then used the resulting clusters to evaluate the number of healthcare facilities available in each. Our findings indicate that peripheral cities in the region tend to share certain features, such as a higher proportion of Black and Brown populations and lower average income (up to one minimum wage). These clusters tend to have fewer healthcare facilities. In contrast, the cluster consisting solely of S˜ao Caetano do Sul, which has the highest proportion of White residents and a large share of individuals earning more than two minimum wages, has, relative to its population, the highest number of healthcare facilities, tripling the number found in the second-ranking cluster (Cluster 2). Through a machine learning approach, our results highlight the structural inequality present in the S˜ao Paulo Metropolitan Region and the disparities in access to healthcare, particularly in its peripheral areas.