In a new study, researchers developed an alternative set of simple criteria using a prediction model to identify groups of people who would most benefit from lung cancer screening (LCS) but are disproportionately excluded by current eligibility criteria.
The study found that these high-benefit groups include those who smoked for a long duration but at a lower intensity, people with heavy smoking who quit over 15 years ago, and high-benefit persons from racial and ethnic minority groups. The findings are published in Annals of Internal Medicine.
Researchers used over 58 million smoking adults’ responses to the annual National Health Interview Survey, dated from 1997 to 2018, to study whether an alternative criteria would work to better identify high-benefit persons and whom this would include. They used Fast and Frugal Tree algorithms and the same information as is included in the U.S. Preventative Services Task Force (USPSTF) criteria to develop an alternative set of simple criteria. They compared the alternative criteria to USPSTF criteria in their sensitivity, and specificity to identify high-benefit people. Researchers also determined where the selected populations overlapped and where the alternative criteria overcame USPSTF’s limitations. Results suggest that the alternative criteria has higher sensitivity and specificity in identifying high-benefit people and is substantially more sensitive in identifying high benefit people in racial and ethnic minority populations than the USPSTF criteria. Analysis confirmed the inequity in current USPSTF criteria, which has a poor sensitivity for identifying high-benefit Black people for LCS. The alternative criteria would produce more effective screening, lead to better support of ethical risk management, and substantially reduce racial disparities in LCS eligibility. The alternative criteria is easy to remember and represents a feasible way to improve current LCS eligibility criteria. Beyond the LCS context, this approach could also be useful for other cancer screening and prevention modalities.
