Chapter 10: A Macro-Model of Smoking and Lung Cancer: Examining Aggregate Trends in Lung Cancer Rates Using the CPS-I and CPS-II and Two-Stage Clonal Expansion Models


David T. Levy, Ph.D., Population Sciences, Department of Oncology, Georgetown University, 3300 Whitehaven St. NW, Ste. 4100, Washington, DC 2007, USA;


Past studies have examined the relationship of lung cancer to smoking using longitudinal data for select samples. This study applies the two-stage clonal expansion (TSCE) model to U.S. +xsmoking data over a 25-year period. Smoking Base Case (SBC) data on actual smoking duration and intensity from the years 1975–2000 are applied by gender to separate TSCE models, which are then calibrated to historical trends in lung cancer death rates using regression analysis. The uncalibrated and calibrated TSCE models are also applied to SBC data for two scenarios: (1) no tobacco control and (2) complete tobacco control. The results are used to develop estimates of the number of lives saved as a result of tobacco control and how many lives would be saved if cigarette use had ceased in 1965. Predictions of lung cancer from the TSCE models with CPS-II and the CPS-I data for males and especially females are considerably below historical rates with the deviations from historical rates increasing over time. Residual trends unrelated to the smoking models were also found. Tobacco control activities saved approximately 625,000 lives between the years 1975 and 2000. An additional 2,110,000 lives would have been saved if all smoking was stopped in 1965. Tobacco control has successfully prevented lung cancer deaths, but many more lives could be saved with further reductions in smoking rates. Systematic biases were observed from TSCE models using CPS-I and CPS-II data to estimate smoking-related lung cancer deaths.