ASPO Abstracts
Characteristics of lung cancer screening eligible population in the U.S and prediction of the eligibility with simplified criteria
Category: Early Detection & Risk Prediction
Conference Year: 2023
Abstract Body:
Purpose: In 2021, the guidelines for lung cancer screening (LCS) were updated, expanding LCS eligible population to younger people and those with lower lifetime smoking intensity. Lifetime smoking intensity is essential to determine LCS eligibility, however it is usually unavailable in electronic health records, which is a barrier of LCS implementation. With the updated guidelines, this study aimed to 1) estimate the proportion and examine characteristics of individuals eligible for LCS among people aged 50-80 years in the U.S.; 2) evaluate the performance of five simplified criteria in estimating LCS eligibility among population aged 50-80 years and subpopulations defined by sex, race/ethnicity, and education. Methods: Participants of National Health and Nutrition Examination Survey (NHANES) 2013-2018 aged 50-80 years who answered smoking questionnaire were included. The 2021 LCS guidelines were used, i.e., age 50-80 years, current or former smoker quitting smoking within 15 years, with ≥20 pack-years. Five simplified criteria were: 1) ever smoker; 2) current or former smoker quitting smoking within 15 years; 3) current smoker; 4) current smoker with >0.5 pack per day (ppd); 5) current smoker with >1 ppd. Sensitivity, specificity, positive predictive value (PPV), negative predictive value, and accuracy were calculated for each simplified criterion. Complex survey design was considered. Results: About 16.7 million individuals (representing 16.0% of population aged 50-80 years) were eligible for LCS in the U.S. The percentage of LCS eligibility was higher among people who were younger, male, non-Hispanic White, less educated, single, not insured, with poorer health status and lower socioeconomic status. The five criteria had sensitivity ranging from 0.08 to 1.00, PPV from 0.35 to 1.00, and accuracy from 0.70 to 0.91. The performance of the five criteria did not vary significantly among subpopulations. Conclusions: Individuals with less favorable social and clinical characteristics are associated with higher chances of being eligible for LCS, which could exacerbate disparities in LCS utilization. Simplified criteria can be used as prescreening tools to identify target populations, which could facilitate LCS implementation at population level.
Keywords: lung cancer screening, NHANES