Risk Prediction Model versus United States Preventive Services Task Force Lung Cancer Screening Eligibility Criteria - Reducing Race Disparities

Lung Cancer

J Thorac Oncol. 2020 Aug 17:S1556-0864(20)30638-9. doi: 10.1016/j.jtho.2020.08.006. Online ahead of print.


INTRODUCTION: Disparities exist in lung cancer outcomes between African American and White people. The current United States Preventive Services Task Force (USPSTF) lung cancer screening eligibility criteria which is based solely on age and smoking history may exacerbate racial disparities. We evaluate whether the PLCOm2012 risk prediction model more effectively selects African American ever-smokers for screening.

METHODS: Lung cancer cases diagnosed between 2010-2019 at an urban medical center serving a racially and ethnically diverse population were retrospectively reviewed for lung cancer screening eligibility based on the USPSTF criteria versus the PLCOm2012 model.

RESULTS: This cohort of 883 ever-smokers was comprised of the following racial/ethnic makeup: 258 (29.2%) White, 497 (56.3%) African American, 69 (7.8%) Hispanic, 24 (2.7%) Asian, and 35 (4.0%) other. Compared to the USPSTF criteria the PLCOm2012 model increased the sensitivity for the African American cohort at lung cancer risk thresholds of 1.51%, 1.70%, and 2.00% per 6-years (p<0.0001). For example, at the 1.70% risk threshold the PLCOm2012 model identified 71.3% African American cases whereas the USPSTF criteria only identified 50.3% (p<0.0001). In contrast, in White cases there was no difference [66.0% vs 62.4%, respectively (p=0.203)]. Of African American ever-smokers who were PLCO1.7%+/USPSTF-, the criteria missed from the USPSTF were pack-years <30 (67.7%), quit-time >15 years (22.5%), and age<55 years (13.0%).

CONCLUSIONS: The PLCOm2012 model was found to be preferable over the USPSTF criteria at identifying African American ever-smokers for lung cancer screening. Broader use of this model in racially diverse populations may help overcome disparities in lung cancer screening and outcomes.