Risk model could reduce disparities in USPSTF lung cancer screening guidelines
Click Here to Manage Email Alerts
Draft U.S. Preventive Services Task Force guidelines that aimed to increase the number of minorities eligible for lung cancer screening may instead exacerbate racial and ethnic disparities, study results showed.
However, augmenting the guidelines to include individuals most likely to benefit according to the Life-Years From Screening with Computed Tomography (LYFS-CT) model — which calculates the gain in life expectancy for an individual if they undergo screening — could reduce disparities in screening eligibility, according to research presented at International Association for the Study of Lung Cancer World Conference on Lung Cancer.
“In July 2020, the USPSTF updated their lung screening guidelines and released draft guidelines that expanded the eligibility criteria with the explicit aim of reducing racial and ethnic disparities, after receiving criticism on their previous guidelines,” Rebecca Landy, PhD, researcher in the division of cancer epidemiology and genetics at the NCI, told Healio. “We aimed to evaluate whether the new draft guidelines reduced racial and ethnic disparities as intended.”
The 2013 USPSTF eligibility criteria for lung cancer screening are based solely upon age and smoking history, which may worsen racial disparities.
Landy and colleagues sought to examine whether the 2020 USPSTF draft guidelines for lung cancer screening would reduce disparities in eligibility, as intended, compared with the 2013 guidelines.
Researchers used data from the 2015 National Health Interview Survey to compare the performance of National Lung Screening Trial-like screening, which consists of three annual CT screens with 5 years of follow-up, among three cohorts of ever-smokers aged 50 to 80 years, including:
- those eligible for screening under the 2013 guidelines (aged 55-80 years, 30 or more pack-years of smoking history, 15 or fewer quit-years);
- those eligible under the 2020 draft guidelines (aged 50-80 years, 20 or more pack-years of smoking history, 15 or fewer quit-years); and
- an augmented 2020 guideline cohort that included individuals with 12 or more days of life gained under the LYFS-CT model.
Results showed that compared with the 8 million ever-smokers (13% minorities) eligible for screening under the 2013 guidelines, 14.5 million ever-smokers (16% minorities) were eligible under the 2020 draft guidelines. Nearly half of those who became eligible under the new guidelines were aged 50 to 54 years.
Overall, the 2020 draft guidelines appeared to increase lung cancer screening eligibility by 97.1% among minorities compared with 78.3% among white individuals. Disparities in lung cancer death sensitivity were similar between the 2020 draft guidelines and the 2013 guideline, with an absolute difference of 13% for Black individuals, 19% for Asian American individuals and 27% for Hispanic individuals.
In the cohort that included augmentation of the 2020 draft guidelines with high-benefit individuals chosen by the LYFS-CT model, lung cancer screening disparities among Black individuals appeared to be nearly eliminated, according to the researchers. The lung cancer death sensitivity among Blacks was 0% and the life-year gained sensitivity was 1%.
“A key principle for fairness in screening eligibility is to ensure that people at equal net-benefit from screening are equally eligible for screening, regardless of race and ethnicity,” Landy said. “Neither the 2020 draft guideline nor the 2013 USPSTF lung cancer screening guidelines are based on an explicit measure of benefit and, thus, will tend to induce disparities. In contrast, use of the LYFS-CT model could be a fairer basis for screening eligibility.”
In addition, augmentation of the 2020 USPSTF criteria to include those most likely to benefit from lung cancer screening showed Black individuals were more likely to be selected than white individuals (22% vs. 9%). However, researchers observed only a slight reduction in disparities among Hispanic individuals (23%) and disparities remained unchanged among Asian individuals (19%).
“The 2020 draft USPSTF guidelines explicitly reject basing guidelines on individualized risk models due to a perception that use of risk models will not increase life expectancy in a population, and because they were judged to be impractical for clinical use today due to a lack of automated approaches to collecting information and calculating risk or life-years gained,” Landy said. “However, National Comprehensive Cancer Network guidelines and new American College of Chest Physician guidelines endorse use of models to augment USPSTF screening eligibility.
“It is imperative to develop EHR systems that easily collect smoking information and automatically calculate risk estimates and life-years gained estimates to present to clinicians and patients,” Landy added. “These automatic systems could promote simple, consistent, fair and effective application of eligibility criteria based on models.”
References:
Landy R, et al. J Natl Cancer Inst. 2021;doi:10.1093/jnci/djaa211.
Landy R, et al. Abstract 3564. Presented at: World Conference on Lung Cancer (virtual meeting); Jan. 28-31, 2021.
For more information:
Rebecca Landy, PhD, can be reached at National Cancer Institute, 9609 Medical Center Drive, Room 7E620, Bethesda, MD 20892; email: rebecca.landy@nih.gov.
Collapse
Landy R, et al. Abstract 3564. Presented at: World Conference on Lung Cancer (virtual meeting); Jan. 28-31, 2021.
Click Here to Manage Email Alerts