Evaluating breast cancer screening performance without registries using medico-administrative data.

Area of Science:

• Oncology
• Public Health
• Health Informatics

Background:

• The French Breast Cancer Screening Program (DOCS) aims for early breast cancer (BC) detection.
• Key performance indicators (KPIs) like sensitivity and cancer detection rate (CDR) are crucial for evaluating screening programs.
• Access to BC registries for linking screening data is limited in France, hindering performance metric calculation.

Purpose of the Study:

• To assess the feasibility and effectiveness of using medico-administrative data, specifically the French National Health Data System (SNDS), to calculate BC screening performance metrics.
• To establish an alternative methodology for evaluating screening programs in regions lacking accessible cancer registries.

Main Methods:

• Linked regional digital mammography screening data (2011-2020) with the French National Health Data System (SNDS).
• Followed women for 24 months post-screening to identify screen-detected and interval cancers.
• Calculated key performance indicators: sensitivity (SE), positive predictive value (PPV), cancer detection rate (CDR), and interval cancer rate (ICR).

Main Results:

• Analyzed 252,786 screening exams from women with a mean age of 61.
• Achieved a sensitivity (SE) of 77.9% and a positive predictive value (PPV) of 19.8%.
• Reported a cancer detection rate (CDR) of 10.9 per 1000 exams and an interval cancer rate (ICR) of 2.4 per 1000 exams.

Conclusions:

• Medico-administrative data (SNDS) provides a reproducible method for calculating breast cancer screening performance metrics in France.
• This approach serves as a valuable alternative in areas with limited or inaccessible cancer registries.
• Further validation in registry-covered regions is recommended to confirm the methodology's broader applicability.