Câncer colorretal – carga da doença no estado do Mato Grosso

Abstract

Introduction: Colorectal cancer presents increasing mortality trend in Brazil from 1979 to 2015. Early screening may reduce the mortality of the disease by removing initial lesions or before they evolve into malignant ones. In spite of the disease’s importance, there are no organize screening actions in the country. Temporal and spatiotemporal trends of colorectal cancer mortality in the state of Mato Grosso and the burden of the disease in Cuiabá and Várzea Grande municipalities are assessed here through the analysis of mortality, incidence and survival. Material and Methods: To analyze the temporal and spatiotemporal trends of mortality in Mato Grosso State DATASUS data were used and linear regression modeling was applied, having the rates been later log- transformed to calculate the yearly percentage change using the statistical packet SPSS15.0. For the spatiotemporal trends in meso-regions and micro-regions in the state, a conditional autoregressive model (CAR) from the package CARBayesST of software R was used. The adjusted models are hierarchical bayesian with inference based on Monte Carlo simulation via Markov chain. In Cuiabá and Várzea Grande municipalities, DATASUS data were also used in the analysis of mortality trends, as well as linear regression modeling and log-linear transformation. Data from the Registro de Câncer de Base Populacional (RCBP) were also used to analyze colorectal cancer incidence and survival. The database was confronted with the data of the Sistema de Informação sobre Mortalidade (SIM) using software R and Reclink. By the same token, linear regression modeling for the incidence trends was used. Kaplan-Meier estimates, Cox regression models and net survival estimates were used to analyze survival. Results: The linear regression modeling exercise showed significant increase of global mortality for both genders (p<0.05), with an increase in the yearly percentage especially in the age range <45 years (6,5% for men and 6,61% for women). The spatiotemporal modeling, particularly the CARlinear and CARar models, showed better adjustment and more consistent results for smaller areas (micro-regions). The increase in mortality in all strata analyzed was mainly observed in the micro-regions of Cuiabá e Rondonópolis. As for Cuiabá and Várzea Grande municipalities, there wasn’t a statistically significant trend in incidence for colorectal cancer. Mortality, in turn, showed a trend of increase for both genders. Utilizing the Kaplan-Meier method, survival in five years was greater among women, in the 45-54 age range and for residents in Cuiabá, but the differences observed were not statistically significant. The net survival estimated for the RCBP data was of 47.7% (IC95%: 43,4-51,8%) in five years. Discussion: The use of secondary data and the application of statistical models find a favorable opportunity to characterize some pathologies, especially cancer, a multifactor-influenced disease that must be observed in time and space. Studies with biomedical approach of the burden of the disease provide a global vision to enable public policies that may benefit prevention and deliver better services for the reduction in colorectal cancer, an increase in survival and the reduction in mortality. Conclusion: The association of techniques in the study of incidence, mortality and survival may result in further crucial information to direct decentralization and pact upon understanding the reality of small areas. Besides, it enhances the importance of implantation, maintenance, updating and availability of the data of the Registros de Câncer to better understand the disease’s profile in our country.