Aim 1 is to use PLCO data to estimate the joint distribution of risk factors (race, education,BMI, family history of lung cancer, personal history of cancer, and COPD) by conditioning on smoking history data (obtained by the NCI Smoking History Generator). A series of conditional probabilities will be derived using the chain rule of probabilities, which will be estimated using various statistical methods including logistic regression, polytomous regression, and linear regression.

Aim 2 is to calibrate the intercept parameters of the estimated joint distributions using the data for prevalence of risk factors in the U.S. 1950 birth cohort. Specifically, we will use the data from the U.S. Census to calibrate for race and education variables, the National Health and Nutrition Examination Survey (NHANES) program of the National Center for Health Statistic for BMI, the National Center for Health Statistics for COPD, and National Health Interview Survey for FHLC.

Sylvia Plevritis, Stanford University