Real-World Study Tracks Lung Cancer Screening Trends and Overdiagnosis
A retrospective study of a community-based program shows a significant drop in late-stage cases alongside signs of potential overdiagnosis.
Key takeaways
- A retrospective study of the KPCO program from 2014 to 2023 analyzed 1,006 incident lung cancers and 198 deaths.
- Late-stage lung cancer incidence decreased significantly by 5.2% annually.
- Early-stage incidence (non-significant 4.9% annual increase) and mortality (non-significant 6.0% annual decrease) did not show statistically significant changes.
- Between 11.3% and 35.9% of early-stage cases could represent potential overdiagnosis, with only a portion reflecting true overdiagnosis.
- Continued monitoring is required to refine overdiagnosis estimates and evaluate implications for resources, survivorship care, and net benefit.
While lung cancer screening enables early cancer detection, this retrospective study provides critical data from a community-based healthcare system. The implementation of lung cancer screening raises concerns about overdiagnosis, particularly in community-based settings where evidence has remained limited.
To understand these dynamics, a retrospective study evaluated the trends of a community-based lung cancer screening program. Published in the journal Cancer Epidemiology in September 2026, the research analyzed data from the Kaiser Permanente Colorado (KPCO) healthcare system to measure shifts in cancer incidence, stage distribution, and mortality rates.
Program Outcomes
Researchers analyzed outcomes from the KPCO lung cancer screening program between 2014 and 2023. Out of 1,006 incident lung cancers identified during this period, 619 cases (62%) were classified as early stage, 367 cases (37%) were late stage, and 20 cases (2%) had an unknown stage. A total of 198 lung cancer deaths were recorded during the study timeframe.
Incidence and Mortality Trends
The analysis of the screening program's data revealed distinct trends in cancer incidence and mortality over time:
Early-stage incidence increased, showing an annual percent change (APC) of 4.9% (95% confidence interval [CI], -1.6-1.8%), representing a statistically non-significant increase.
Late-stage incidence decreased significantly, with an APC of -5.2% (95% CI, -9.8% to -0.3%).
Lung cancer-specific mortality decreased with an APC of -6.0% (95% CI, -12.3-0.8), representing a statistically non-significant decline.
Additionally, calculations of excess incidence suggested that between 11.3% and 35.9% of early-stage lung cancer cases could represent potential overdiagnosis, though only a portion of these cases reflect true overdiagnosis.
Future Monitoring
Continued monitoring of lung cancer incidence and mortality remains essential to refine estimates. Ongoing tracking is needed to evaluate the implications of these trends for healthcare resources, survivorship care, and the overall net benefit of screening.
Questions readers ask
- What did the study find regarding late-stage lung cancer?
- Late-stage lung cancer incidence decreased significantly, with an annual percent change of -5.2%.
- How did the study estimate overdiagnosis?
- The study compared post-screening early-stage incidence with pre-program implementation and predicted incidence, estimating that 11.3% to 35.9% of early-stage cases could represent potential overdiagnosis.
- Were the changes in early-stage incidence and mortality statistically significant?
- No. The 4.9% annual increase in early-stage incidence and the 6.0% annual decline in mortality were both statistically non-significant.
What this means
The observed trends are consistent with a maturing screening program, characterized by a significant reduction in late-stage disease alongside non-significant changes in early-stage detection and mortality. While early detection is a primary goal, the potential for overdiagnosis highlights the importance of evaluating screening's overall net impact on patient care and healthcare resources.
Limitations and uncertainties
- The findings are based on a retrospective analysis of a single regional program (KPCO), and estimates of potential overdiagnosis represent a range in which only a portion reflects true overdiagnosis.