To evaluate the effectiveness of risk-based lung cancer screening strategies compared to current USPSTF eligibility criteria.
Approach:
Study Design: A large cohort study assessing 16 lung cancer risk prediction models among over 641,000 US adults aged 50 to 80 with a smoking history.
Population: Participants included diverse racial and ethnic groups: Asian, Hispanic, non-Hispanic Black, and non-Hispanic White individuals.
Performance Evaluation: Comparison of risk-model thresholds to USPSTF criteria to assess screening efficiency and racial/ethnic disparities.
Key Findings:
Existing prediction models significantly underestimated lung cancer risk in non-Hispanic Black participants.
Lower discrimination was observed in Asian participants compared to all other groups.
Risk-based strategies improved average screening efficiency and reduced differences across racial and ethnic groups compared to USPSTF criteria.
Some models did not improve estimated screening efficiency among non-Hispanic Black participants compared to USPSTF criteria.
No single strategy optimized all performance measures equally well.
Interpretation:
Risk-based strategies were superior to USPSTF criteria in optimizing efficiency and minimizing variation across racial and ethnic groups.
Limitations:
Underrepresentation of racial and ethnic minority groups, particularly Asian and Hispanic participants.
Inability to evaluate model performance among Native Hawaiian or other Pacific Islander and American Indian or Alaska Native participants due to small sample sizes.
Conclusion:
While risk-based approaches enhance screening efficiency, disparities in performance across racial and ethnic groups persist.
