Avaliação de tratamento à paciente com câncer de mama nas unidades oncológicas do sistema único de saúde no Rio de Janeiro

This study investigates the association between the structure and interventions of fifteen oncological health units and breast cancer survival in Rio de Janeiro state, Brazil, controlling for demographic and clinical variables. It relies on data from the Brazilian information systemof high complexity outpatient procedures and a sample of 310 medical records of women prevalent with breast cancer. To analyze the data, we applied the Kaplan-Meier technique and the Cox hazard model (pseudo likelihood). The Kaplan-Meier analyses pointed out significant associations (α= 0,05), between survival and age at treatment beginning, age at diagnosis, time between diagnosis and treatment beginning, tumor stage, histological grade, surgery, adjuvant hormonal therapy, type of adjuvant hormonal therapy, multiple therapy, hormone receptor tumor status, lower complexity health units without health insurance, volume of breast cancer patients and juridical nature of the health care units. Estimates from the Cox model indicated positive associations between the hazard of death and time between diagnosis and treatment greater than 6 months (HR=3,129), tumor stage 3 or 4 (HR=3,861), high histological grade (HR=1,863), being attended in lower complexity health units without health insurance (HR=2,049), and being attended in units with low (HR=2,674) or medium(HR=1,910) breast cancer care volume. They also indicated negative associations between the dependent variable and breast cancer surgery (HR=0,448) and non-classified adjuvant hormonal therapy (HR=0,285). This study indicates that part of breast cancer survival differences reflect variations in health care provided by credentialed services. These differences remain despite some health policies oriented towards reducing them, and the availability, in Brazil, of breast cancer treatment technologies compatible with the international state-of-art knowledge. The findings show associations of breast cancer survival with characteristics of health care units and interventions applied with practical implication to formulate policies to control the cancer in Brazil.