Análise de intervenção de uma ação do programa de rastreamento do câncer de mama nas séries de mortalidade no Brasil e regiões

Abstract

This study proposes to use an intervention analysis method to assess the changes in the mortality from breast cancer in Brazil and in its regions, after the establishment of the Information System on Breast Cancer in June 2009. The time series of the monthly mortality rates were used for the period from January 1996 to March 2016. The preintervention period of the series was fitted to a Box-Jenkins model, which was used to analyse the existence or not of changes in the mortality and therefore to build a global model. Changes in the tendencies of the mortality series were observed in Brazil as a whole, and in the South and Southeast regions, but no effect was observed in the North, Northeast and Central West regions. In the first case, the Box-Tiao approach was used to build an intervention model and to estimate the intervention response times, which were 24 months for Brazil and the Southeast region, and 12 months for the South region. In the second case, a model was built to fit the series composed of the pre- and post-intervention periods. The predictive accuracy of all models was tested. The intervention analysis proved to be a useful tool to identify changes in the mortality series and to identify the different regional effects caused by the same action, which is important in the monitoring of public health actions with a national scope.