Impact of lung cancer screening eligibility criteria on participants' characteristics and healthcare resources

Abstract

Selection criteria for lung cancer screening (LCS) are essential for implementing resource- and cost-efficient programs. However, different selection criteria may have similar resource requirements but select dissimilar individuals. This study explores the impact of inclusion criteria on participants' characteristics and health services in Spain. A cross-sectional study was conducted using data from 21,007 individuals from the 2012 Spanish National Health Survey. LCS eligibility was evaluated for four criteria: two from the 4-IN-THE-LUNG-RUN trial (4ITLR1: ages 60-79, 35 pack-years, current smokers or <10 years since cessation; 4ITLR2: ages 60-79, PLCOm2012noRace risk ≥2.6%), the United Kingdom's Targeted Lung Health Check (TLHC: ages 55-74, PLCOm2012noRace risk ≥1.51%), and the United States Preventive Services Task Force (USPSTF: ages 50-80, 20 pack-years, current smokers or <15 years since cessation). Under the most restrictive (4ITLR1) 5.6% of ages 50-80 were eligible, requiring 9.0% additional radiological capacity. Under the broadest (USPSTF) 21.6% of ages 50-80 were eligible, requiring 34.2% additional capacity. All criteria primarily selected men (68.1-82.2% of eligibles). Most criteria favored selecting former smokers (49.4-56.9%). However, eligible women were more often current smokers than men across all criteria (men: 38.5-46.8%, women: 52.7-66.5%). Former smokers smoked more cigarettes per day than current smokers (30 vs. 20) despite having shorter smoking durations, resulting in higher median pack-years across all criteria (range: 44-66). Overall, pack-year-based eligibility decreased with age, while risk-based eligibility was stable across age groups. Different LCS criteria yield populations with highly varied characteristics. Risk-based criteria may provide more equitable eligibility across age groups.

Keywords: eligibility criteria; health services impact; lung cancer screening; risk prediction models; screening strategies.