Breast cancer screening and overdiagnosis

Area of Science:

• Oncology
• Public Health
• Biostatistics

Background:

• Overdiagnosis, a consequence of cancer screening, is challenging to quantify accurately.
• Estimating overdiagnosis requires long-term comparative data from screened and unscreened cohorts, which is often unavailable.

Purpose of the Study:

• To quantify the variation in estimated breast cancer overdiagnosis in organized screening programs.
• To evaluate observed and simulated data, and demonstrate how modeling can provide insights into overdiagnosis.

Main Methods:

• Collaborative analysis of observed and simulated data from the International Cancer Screening Network.
• Evaluation of observational data, considering methodological pitfalls and study designs.
• Application of different modeling approaches to estimate overdiagnosis.

Main Results:

• Reliable data and modeling indicate overdiagnosis accounts for less than 10% of invasive breast cancer cases in women aged 50-69.
• Estimates exceeding 10% may stem from study design inaccuracies.
• Cohort study designs with extended follow-up can reduce discrepancies in observational estimates.

Conclusions:

• Overdiagnosis in breast cancer screening programs is estimated to be below 10% for invasive cases in women aged 50-69.
• Improved study designs, particularly cohort studies, are crucial for accurate overdiagnosis estimation.
• Modeling serves as a valuable tool for estimating overdiagnosis when direct comparative data is limited.