Evaluating the unintended uptake of lung cancer screening
Nikki M Carroll, Kris F Wain, Brian Hixon, Roger Y Kim, Mahesh Maiyani, Julie S Steiner, Debra P RitzwollerAbstract
Background
U.S. Preventive Services Task Force (USPSTF) lung cancer screening (LCS) eligibility relies on smoking history data that are often incomplete or inaccurate in Electronic Health Records (EHRs). Inaccurate data combined with clinicians’ unfamiliarity with LCS recommendations may lead to guideline-discordant screening, potentially reducing the benefit-risk ratio and misallocating resources. Limited data exist that describe USPSTF-ineligible patients who receive and complete LCS orders.
Methods
We conducted a retrospective cohort study of adults receiving an LCS order at Kaiser Permanente Colorado (5/1/2014-12/31/2023). Using structured EHR data to determine eligibility, we assessed USPSTF-ineligible LCS orders, screening completion rates, and lung cancer yield.
Results
Of 15,624 patients receiving LCS orders, 2,048 (13%) were classified as USPSTF-ineligible. Most ineligibility resulted from incomplete (28%) or out-of-range smoking history data (54%). USPSTF-ineligible patients were more likely to be female and affluent and less likely to be White, aged 55-80, or currently smoke. Orders were more common during program rollout and the COVID-19 pandemic. Among ineligible patients, 47% completed screening; 3% were diagnosed with lung cancer, 67% at an early stage. Lung cancer detection was 8% among Positive screens (Lung-RADS 3-4) and 1% among Negative screens (Lung-RADS 1-2).
Conclusions
A substantial proportion of patients receiving lung cancer screening orders were classified as USPSTF-ineligible using structured EHR data, primarily because of incomplete or out-of-range smoking histories. The unexpectedly high lung cancer detection rate suggests that many may have been misclassified rather than truly ineligible. These findings highlight the challenges of utilizing EHR data to identify guideline-discordant screening rates.