Abstract
OBJECTIVES: To estimate lung cancer mortality rates adjusted by age distribution for the country's 32 states between 1998 and 2017; to contrast the territorial distribution of demand for oncological services with the availability of specialists to provide care; and to determine the predictive capacity of three different supervised classification algorithms in the context of automated learning techniques. METHODS: An exploratory analysis and data modeling were conducted, considering death records from the national health information system. RESULTS: Deaths from lung cancer in Mexico dropped by 14.5% between the period prior to implementation of the General Law on Tobacco Control and the subsequent period. A 22% reduction was observed in the male population by the end of the entire period. There is evidence of an imbalance between the demand for oncological services and the availability of specialists. The modeling phase demonstrated the usefulness of the country's electronic death records. CONCLUSIONS: Despite reductions in lung cancer mortality patterns in Mexico in the last two decades, the analysis showed persistent areas of opportunity for improvement, mainly in the female population of Guerrero, Oaxaca, and Puebla states. Based on this research, the main recommendation for focusing efforts to manage this oncological disease in Mexico is to determine whether these patterns are associated with smoking habits or with other social determinants.