Sobrevida de mulheres com câncer do colo do útero: desigualdades entre as macrorregiões de Minas Gerais

Abstract

Objectives: Cervical cancer-related morbimortality is marked by inequalities across different geographical areas, as observed among the regions of Brazil and the state of Minas Gerais. We investigated the association between macroregion of residence and survival in women diagnosed with cervical cancer who underwent cancer treatment within the Unified Health System (SUS) between 2002 and 2010 in Minas Gerais. Methods: Non-concurrent prospective study with data from the National Database in Oncology (Base Onco), developed through deterministic-probabilistic pairing of SUS information systems. We identified in this database women aged over 18 years, diagnosed with cervical cancer between 2002 and 2010, who underwent radiotherapy and/or chemotherapy and who were living in Minas Gerais. We estimated five-year overall and cancer-specific survival probabilities by the Kaplan-Meier method and compared these probabilities according to chosen variables using the log-rank test. Extended Cox models were employed to assess the association between macroregion of residence and overall and cancer-specific death risk adjusting for age, clinical variables, and place of treatment. Results: A total of 5,613 patients were included in the study, with median age of 55.0 years (IQR: 22,0). Median follow-up time was 70.0 months (IQR: 89.0). Five-year overall and cancer-specific survival were 56.3% (95%CI: 55.0; 57.6) and 63.6% (95%CI: 62.3; 64.9), respectively. Among the 13 macroregions, five-years survival probabilities ranged from 46.5% to 64.2% for overall survival and from 52.0% to 72.0% for cancer-specific survival. In the multivariate analysis, the risk of global death was higher than the reference macroregion (Norte, chosen reference due to its high survival and large study population) for seven macroregions, nearly doubling in Jequitinhonha (HR: 1.97; 95%CI: 1.33; 2.93). Nine macroregions had a higher risk of cancer-specific death, and Jequitinhonha also presented the highest magnitude in this association (HR: 2.29; 95%CI: 1.44; 3.62). Adjustments by age and clinical variables (stage, comorbidities and treatment), in general, had little effect on the main association investigated. Conclusion: We identified regional inequalities in survival that persisted after adjustments. Our hypothesis is that these disparities should be related mainly to the heterogeneity in the quality of cancer care in Minas Gerais and, in addition, to sociodemographic and economic inequalities observed across the state. In this sense, the improvement of regionalized and hierarchical cancer care networks, aimed to ensure comprehensive care for these women, especially with regard to access to timely and appropriate treatments, is crucial to improve this scenario of inequalities. In addition, sociodemographic and economic issues need to be considered in cervical cancer control policies, as these factors may be implicated in regional disparities in cervical cancer survival.