Tobacco-related neoplasms: survival analysis and risk of death of population data from Florianópolis, SC

Abstract

ABSTRACT OBJECTIVE To estimate the probability of survival and prognostic factors for tobacco-related neoplasms in a population-based cohort. METHODS This is a cohort with data from the Population-Based Cancer Registry of Florianópolis, southern Brazil, from 2008 to 2012. The Stata 16.0 software was used to estimate the probabilities of survival in five years after diagnosis, by the Kaplan Meier method, and the risk of death, by the Cox regression. RESULTS A total of 2,829 cancer records related to smoking were included, more prevalent among males, over 70 years of age, nine years or more of schooling, white, with a partner and metastatic diagnosis. The most frequent groupings were colon and rectum (28.7%), trachea, bronchi and lungs (18.6%) and stomach (11.8%). At follow-up, 1,450 died. Pancreatic cancer had the worst probability of survival (14.3%), followed by liver cancer (19.4%). CONCLUSION Risk factors for death and survival rates differ across the 13 types of tobacco-related cancers. Early diagnosis and primary prevention are strategies that must be improved to improve survival and decrease the burden related to these types of cancer.

Keywords

• Tobacco Use Disorder
• Neoplasms, epidemiology
• Survival Analysis
• Risk Factors
• Mortality
• Proportional Hazards Models