The research team drew on the Brazilian Mortality Information System, a nationwide database that records the circumstances of every registered death in the country. From 1999 to 2023, the system documented 339,425 deaths related to breast cancer. The investigators classified each death by location, distinguishing those that occurred in hospitals or other healthcare establishments from those recorded at home, in a public place, or in some other location outside the formal health system. This distinction matters because place of death is widely used internationally as a population-level indicator of end-of-life care pathways: it reflects how palliative services are organized, how care is coordinated in a patient’s final months, and, in resource-constrained settings, whether structural barriers prevent people from reaching or remaining in institutional care.

The headline numbers are striking in their asymmetry. Of the 339,425 deaths analyzed, 286,134, or 84.3 percent, occurred within healthcare facilities, while 53,291, or 15.7 percent, took place outside them. But that national average conceals enormous regional variation. Deaths outside healthcare facilities were most common in the Northeast region, where they accounted for 25.2 percent of breast cancer deaths, and in the North, at 21.0 percent. In the Southeast, the country’s wealthiest and most densely serviced macroregion, the figure fell to 11.6 percent, a difference the authors reported as statistically significant. The South and Central-West regions fell between these poles, tracing a gradient that maps closely onto Brazil’s well-documented socioeconomic geography.

That gradient was not merely suggestive; it was quantitatively dramatic. The researchers examined the relationship ecologically, comparing regional rates of death outside healthcare facilities with each macroregion’s Human Development Index, a composite measure combining income, education, and life expectancy. The correlation was strongly inverse, with a coefficient of r = −0.943 and a p-value of 0.016. In practical terms, the lower a region’s human development score, the higher the proportion of breast cancer patients who died outside hospitals and clinics. A correlation this strong across just five macroregions should be interpreted with ecological caution, since it describes regions rather than individuals, but it aligns with a broader body of evidence showing that palliative care infrastructure in Brazil is concentrated in the more developed South and Southeast.

Individual-level associations told a consistent story. Using univariate analysis, the team found that women with lower educational attainment had higher crude odds of dying outside a healthcare facility, suggesting that schooling, a proxy for socioeconomic position throughout life, shapes even the circumstances of death. Racial and ethnic disparities were also evident. Black and Mixed-race individuals, who together form the majority of Brazil’s population but face persistent barriers in healthcare access, had 12 percent higher crude odds of dying outside healthcare facilities compared with white individuals, an odds ratio of 1.12. The disparity was far larger for Indigenous women, whose odds were nearly double those of white decedents, at an odds ratio of 1.94. These figures echo longstanding inequities in Brazilian oncology, where Indigenous populations in particular face geographic isolation, limited service availability, and documented delays in diagnosis and treatment.

Time added another dimension to the analysis. Using joinpoint regression, a statistical technique that identifies points at which long-term trends change direction, the researchers found that the modeled annual percentage of breast cancer deaths occurring outside healthcare facilities declined steadily over the 25-year period, with an annual percent change of −1.0 percent, statistically significant at p < 0.001. On its face, this suggests gradual progress: more Brazilian women over time reached institutional care before death, plausibly reflecting the expansion of the Unified Health System (SUS), wider availability of oncology services, and improved access to hospital care in historically underserved regions.

Yet the story was more complicated than a simple downward slope. When the researchers examined age-standardized rates, which adjust for the changing age structure of the population and serve as a measure of population burden rather than raw proportion, the series followed a distinct trajectory from the modeled percentages. Most notably, the age-standardized rate showed an elevation during the COVID-19 pandemic period. This spike likely reflects the profound disruption the pandemic caused to Brazilian healthcare, when hospitals were overwhelmed, elective and non-urgent oncology care was curtailed, and many patients avoided or could not reach healthcare facilities. The pandemic-era signal is a reminder that progress in end-of-life care access can be fragile and reversible under systemic shock.

The authors are careful, and rightly so, about what these findings do and do not mean. Death outside a healthcare facility is a heterogeneous outcome. In high-income countries with mature home-based palliative care programs, dying at home is often the expressed preference of patients and may indicate high-quality, preference-concordant end-of-life care. In settings where such services are scarce, however, dying at home may instead signal that a patient never received adequate symptom control, that families lacked support, or that distance and cost made institutional care unreachable. Mortality records, the study notes, cannot determine whether a recorded location matched the patient’s preference. The researchers also lacked information on health service use in the final period of life, so the data cannot distinguish a well-supported home death from one born of abandonment or structural exclusion.

Even with those caveats, the pattern of associations is difficult to explain away as preference. If home deaths in Brazil primarily reflected patient choice, one would not expect them to cluster so tightly in the least developed regions, nor among those with the least education, nor among Black, Mixed-race, and Indigenous women, groups that consistently experience worse access to cancer care across the Brazilian system. The far more parsimonious reading, and the one the authors advance, is that these deaths mark inequitable end-of-life care pathways: populations facing the greatest barriers to timely diagnosis, treatment, and palliative support are also those most likely to spend their final days outside the institutions designed to care for them.

The implications reach beyond Brazil’s borders. Breast cancer is the most commonly diagnosed cancer in women worldwide, and as oncology advances in low- and middle-income countries, the question of how patients die, not merely whether they survive, is becoming a central equity concern. Brazil offers a particularly instructive case because its unified public health system has expanded coverage dramatically over the study period, yet territorial and social hierarchies continue to shape outcomes at every stage of the cancer trajectory. The 25-year dataset shows both sides of that ledger: a slow national improvement in access to facility-based death, and a stubborn residual of inequality concentrated in the North and Northeast and among the country’s most marginalized communities. Closing that gap, the study suggests, will require more than hospital beds. It will demand investment in community-based and home palliative care services in underserved regions, so that dying at home can become a genuine, supported choice rather than the default consequence of a system that never arrived.

Subject of Research: Social determinants and regional disparities in place of death among breast cancer decedents in Brazil

Article Title: Social determinants and regional disparities in place of death among breast cancer decedents in Brazil: a 25-year population-based analysis

Article References: da Silva, J. L., de Oliveira, L. C., de Albuquerque, L. Z., Araújo de Souza, P. H., Santos Thuler, L. C., & de Melo, A. C. (2026). Social determinants and regional disparities in place of death among breast cancer decedents in Brazil: a 25-year population-based analysis. International Journal for Equity in Health. https://doi.org/10.1186/s12939-026-03009-4

Image Credits: AI Generated

DOI: 10.1186/s12939-026-03009-4

Keywords: breast cancer, place of death, palliative care, health equity, Brazil, social determinants of health, regional disparities, mortality data, Human Development Index, end-of-life care, Indigenous health, health disparities