Eligibility and Prognostic Performance of Smoking Duration-Based Versus Pack-Year-Based U.S. National Lung Cancer Screening Criteria Across Racial and Ethnic Groups

Abstract

Background: The U.S. Preventive Services Task Force expanded lung cancer (LC) screening eligibility in 2021 (USPSTF-2021) by decreasing the minimum number of smoking pack-years from 30 to 20. Underrepresented minorities still experience disparities in screening eligibility.

Objective: To evaluate screening eligibility and prognostic performance of alternative smoking duration-based criteria versus USPSTF-2021 (primary outcome) and risk-based screening using the recalibrated Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial 2012 (PLCOm2012_update) model (secondary outcome) across diverse racial and ethnic groups.

Design: Prospective, population-based Multiethnic Cohort linked to SEER (Surveillance, Epidemiology, and End Results) registries.

Setting: California and Hawai'i, with recruitment from 1993 to 1996.

Participants: 105 261 adults aged 45 to 75 years with a history of smoking.

Measurements: Hypothetical eligibility and prognostic performance (sensitivity and specificity) in detecting 6-year LC.

Results: Under USPSTF-2021, 24.0% of the cohort would be eligible for screening; a 30-year smoking duration yielded the closest eligibility rate (27.5%). Compared with USPSTF-2021, the 30-year duration criteria would reduce eligibility gaps across all races relative to Whites, most notably in African Americans (30.4% vs. 28.8% for Whites under duration-based; 21.4% vs. 30.2% for Whites under USPSTF-2021) and Latinos (25.1% vs. 28.8% for Whites under duration-based; 15.7% vs. 30.2% for Whites under USPSTF-2021). Prognostic sensitivity to identify LC within 6 years increased across all races under the 30-year duration criteria, although specificity decreased commensurately. At matched overall eligibility (27.5%), a risk-based PLCOm2012_update 6-year threshold of 1.1% improved both sensitivity and specificity in the overall cohort. However, it widened the eligibility gap between Latinos and Whites (14.4% vs. 31.3%) and demonstrated lower sensitivity in Latinos than duration-based criteria (59.7% vs. 69.8%).

Limitations: Cohort geography and enrollment period may limit generalizability. Overdiagnosis was not measured.

Conclusion: Compared with USPSTF-2021, the 30-year duration-based criteria could reduce the eligibility gaps among African Americans and Latinos relative to Whites while improving 6-year LC detection sensitivity across all races.

Primary funding source: National Institutes of Health.