DETERMINANTS OF SURVIVAL ON BRAZILIAN PATIENTS WITH BREAST CANCER IN PUBLIC AND PRIVATE PRACTICE

Abstract

Objective: The Brazilian health system can be divided into public and private sectors. In the public sector, insurance is
provided by the state to all Brazilian citizens (municipal, state, and federal levels). The private sector is comprised of the
private health insurance area, with various schemes of health plans or with out-of-pocket expenses. This study proposed to evaluate the breast cancer population characteristics and treatments according to public and private health care
systems. Methods: This retrospective cohort study included patients diagnosed with invasive breast cancer, with health
insurance information, between January 2000 and June 2020, from the Fundação Oncocentro de Sao Paulo database.
Patients were described according to age, gender, level of education, histology of neoplasm, stage at diagnosis, and type
of treatment. Categorical variables are described as percentages and frequencies. The association between demographic
and treatment factors and overall survival (OS) was evaluated using a Cox proportional hazard (PH) regression model
while accounting for different lengths of participant follow-up. Uni- and multivariable Cox PH models were used to estimate hazard ratios with corresponding 95% confidence intervals for OS. The Kaplan-Meier (KM) curves were used to
visually display survival curves, and the log-rank test was used to compare the estimated KM curves. Results: A total of
65,543 patients were included. The distribution by age, gender, and histology was the same for public and private systems.
The majority of patients were diagnosed on stages I and II (77.8%) in the private system. However, in the public system, it
was found more advanced stage disease (67.8% in stages II and III). Patients with metastatic tumor were more common
in the public system (11.1% versus 5.3%). Treatment with surgery and, at least, two types of adjuvant therapy as trimodal
therapy were the same in both groups (46.6% private versus 46.2% public). KM plot shows 5- and 10-years OS differences
in all stages; 10-year OS in stages I, II, III, and IV in private and public systems were, respectively, 81.6 versus 77.5%, 74
versus 63.3%, 55.6% versus 39.6%, and 7.6 versus 6.4%. In the multivariable analyses, the significant independent predictors for OS were private system, age and stage at diagnosis, high-level education, and trimodal therapy. Conclusion: Older
patients, less intensive treatment, and lower educational levels were independent predictors for worse OS. Public health
system presented a more advanced stage at diagnosis than private care and was associated with worse survival outcomes in Brazilian breast cancer patients.