Sobrevida e fatores prognósticos para o câncer de mama em Juiz de Fora, Minas Gerais, na coorte diagnosticada entre 1998 e 2000

Este estudo teve como objetivo analisar a sobrevida e fatores prognósticos para o câncer de mama feminino, com ênfase no papel de variáveis relacionadas aos serviços de saúde. Foram analisadas 782 mulheres com câncer invasivo da mama e submetidas à cirurgia curativa, que realizaram tratamento cirúrgico e/ou terapia complementar no município de Juiz de Fora, Minas Gerais, com diagnóstico da doença entre 1998 e 2000. O recrutamento dos casos foi efetuado a partir de busca ativa nos registros médicos de todos os serviços de saúde que prestam atendimento em oncologia na cidade. O seguimento foi realizado mediante retorno aos prontuários e complementado por busca no banco do Sistema de Informação sobre Mortalidade (SIM), contato telefônico e consulta de situação cadastral no Cadastro de Pessoas Físicas (CPF). As principais variáveis analisadas foram: cor da pele, idade ao diagnóstico, tamanho do tumor, comprometimento de linfonodos, estadiamento, cidade de residência e variáveis relativas aos serviços de saúde, entre outras. As funções de sobrevida foram calculadas por meio do método de Kaplan-Meier. Foi utilizado o modelo de riscos proporcionais de Cox para avaliação dos fatores prognósticos. No primeiro artigo, foi observada taxa de sobrevida específica em cinco anos de 80,9%, sendo identificados tamanho tumoral e comprometimento de linfonodos axilares como importantes fatores prognósticos associados de forma independente à sobrevida pela doença na população de estudo (HR=1,99, IC95%: 1,27-3,10 e HR=3,92, IC95%: 2,49-6,16, respectivamente). No segundo artigo, quando consideradas as variáveis relativas aos serviços de saúde, foi verificada pior sobrevida para as mulheres assistidas no serviço público, embora com significância limítrofe (p=0,05), e para aquelas que não tinham plano privado de saúde, apesar de não significativa (p=0,1). As funções de sobrevida não ajustadas e estratificadas por natureza do serviço de saúde exibiram sobrevida mais desfavorável para as mulheres não brancas e para os casos com nenhum ou menos de 10 linfonodos isolados somente no serviço público, embora com significâncias estatísticas apenas marginais (p=0,09 e p=0,07, respectivamente). Na análise multivariada, foi observado maior risco de óbito nas mulheres que não fizeram uso de hormonioterapia apenas para os casos assistidos no serviço público (HR=1,56; IC95%: 1,01-2,41), independentemente do tamanho tumoral e do status ganglionar axilar. Tais achados sugerem desigualdades sociais e disparidades na prevenção primária e secundária da doença na região estudada, com desvantagem para o serviço público de saúde, enfatizando a maior necessidade de esclarecimento sobre a doença, assim como de diagnóstico e tratamento dos casos em estádios mais precoces. Destacou-se, ainda, o valor de se trabalhar com informações produzidas pelos serviços de saúde responsáveis pela assistência oncológica no país, possibilitando a caracterização do perfil e da sobrevida das pacientes assistidas e fornecendo subsídios aos órgãos competentes do setor saúde local para nortear a adoção de estratégias de prevenção direcionadas ao controle da doença na população avaliada.

This study aimed to evaluate the breast cancer survival and prognostic factors mainly concerning health service-related variables. A total of 782 women with invasive disease, who had undergone curative surgery, were analysed. All of them received surgical treatment and/or complementary therapy in the city of Juiz de Fora, State of Minas Gerais, Brazil, and were diagnosed between 1998 and 2000. The cases were recruited from all local oncology services, and followed up through medical records, the Mortality Information System (SIM), contact by phone and the National Identification Number (CPF) investigation. Main variables studied included skin color, age at diagnosis, tumour size, lymph node involvement, disease stage, city of residence, health service-related variables, and others. Survival functions were calculated by the Kaplan-Meier method, and multivariate analysis was performed by the Cox proportional hazard model. In the first article, the 5-year disease-specific survival was 80.9%, and tumour size and lymph node involvement were the main prognostic factors independently associated to disease-specific survival in the study population (HR=1.99, IC95%: 1.27-3.10 and HR=3.92, IC95%: 2.49-6.16, respectively). In the second article, when considered the health service-related variables, a worst survival was observed among women assisted in public health services in spite of bordering significance (p=0.05), and among women without private health coverage, despite not significant (p=0.1). Non-adjusted survival functions stratified by the health service nature identified a lower survival among non This study aimed to evaluate the breast cancer survival and prognostic factors mainly concerning health service-related variables. A total of 782 women with invasive disease, who had undergone curative surgery, were analysed. All of them received surgical treatment and/or complementary therapy in the city of Juiz de Fora, State of Minas Gerais, Brazil, and were diagnosed between 1998 and 2000. The cases were recruited from all local oncology services, and followed up through medical records, the Mortality Information System (SIM), contact by phone and the National Identification Number (CPF) investigation. Main variables studied included skin color, age at diagnosis, tumour size, lymph node involvement, disease stage, city of residence, health service-related variables, and others. Survival functions were calculated by the Kaplan-Meier method, and multivariate analysis was performed by the Cox proportional hazard model. In the first article, the 5-year disease-specific survival was 80.9%, and tumour size and lymph node involvement were the main prognostic factors independently associated to disease-specific survival in the study population (HR=1.99, IC95%: 1.27-3.10 and HR=3.92, IC95%: 2.49-6.16, respectively). In the second article, when considered the health service-related variables, a worst survival was observed among women assisted in public health services in spite of bordering significance (p=0.05), and among women without private health coverage, despite not significant (p=0.1). Non-adjusted survival functions stratified by the health service nature identified a lower survival among non white women and among cases with no nodes, or less than 10 nodes examined, both just in public service in spite of marginal significance (p=0.09 and p=0.07, respectively). In multivariate analysis the higher death risk observed among women who did not undergo hormonal therapy was detected only in cases assisted in the public service (HR=1.56; IC95%: 1.01-2.41) independently on tumour size and axillary node status. These results suggest social disparities and inequalities in primary and secondary disease prevention in the studied region with disadvantage to public service. The necessity of improving the disease information was emphasized, as well as earlier diagnosis and treatment. The importance of working with available oncology services data in our country was enhanced making it possible for the identification of the assisted patients profile and disease survival, and offering information to the local health government to establish preventive strategies for disease control in the studied population.