DESIGNING EFFICIENT LUNG CANCER SCREENING PROGRAMS: ROLES OF ADHERENCE, ELIGIBILITY, AND SCREEN FREQUENCY

Purpose

The characteristics of the targeted population and the adherence of high-risk individuals to screening guidelines will affect the efficiency of lung cancer screening.  Risk perception and intention to screen vary by smoking status.  We simulated the potential efficiency of various CT lung cancer screening eligibility criteria and quantified losses in effectiveness due to low adherence using an existing lung cancer model.

Methods

Using a microsimulation model, we estimated the gains in life-years possible with helical CT screening for lung cancer, under varying assumptions of screening eligibility and adherence.  We determined the optimal age for a one-time lung cancer screen in programs with no minimum smoking history (universal screening); screening of moderate and heavy smokers (>20 pack-years); and of heavy smokers (>30 pack-years).  We also evaluated the potential benefit of second and third screens.  We compared the life-years gained per screened individual (screening efficiency) in sub-populations stratified by smoking status and pack-years of smoking.  Using published rates of willingness to consider lung cancer screening, we evaluated the efficiency loss from non-adherence.

Results

Regardless of the minimum smoking history for screening eligibility, the optimal age of a one-time screen was 58 years.  In a universal screening program of 58-year olds, 57.6% of the gains accrued to those with ≥ 30 pack-years of smoking history (31.8% of the cohort, with 20.2 days per screened individual).  Individuals with a moderate smoking history (22% of 58-year olds) generated 11.9 days per screened individual, compared to 3.2 days per never-smoker (39% of 58-year olds) screened.  In heavy smokers, life expectancy gains were increased by 61% by screening twice compared to screening once; a third screen in this population increased life expectancy gains by an additional 27% compared to screening twice-ever.  Twenty percent of the benefits of universal screening and approximately 25% of benefits of screening targeted at heavy smokers were lost when imperfect adherence to screening guidelines was considered.

Conclusions

Lung cancer screening of smokers with greater than 30 pack-years was most efficient at generating additional life-years per screened individual.  Imperfect adherence among current smokers substantially decreased the effectiveness of lung cancer screening.  Models like this one can be used to determine optimal screening regimens and the cost-effectiveness of programs aimed at increasing screening adherence.