Finding the optimal recall rate in breast cancer screening: results from the ROCS study

Area of Science:

• Radiology
• Medical Imaging
• Public Health

Background:

• Optimizing breast cancer screening involves balancing cancer detection rates with false-positive recalls.
• The Dutch Digital Mammography Screening Program's recall rate and its impact on detection require evaluation.

Purpose of the Study:

• To explore the relationship between recall rates and cancer detection in the Dutch Digital Mammography Screening Program.
• To determine the optimal recall rate for breast cancer screening.

Main Methods:

• 21 radiologists provided continuous Probability-of-Malignancy (PoM) scores.
• Receiver operating characteristic (ROC) curves were used to analyze recall-detection performance.
• Hierarchical bootstrapping was employed for error estimation (1000 resamples).

Main Results:

• The study analyzed 40,829 screening records, identifying 315 cancer cases.
• The overall recall rate was 2.33%, with a cancer detection rate of 6.4 per 1000 screens.
• Radiologist performance was high, with an Area Under the Curve (AUC) of 0.91.

Conclusions:

• Dutch screening radiologists exhibit high accuracy and operate at an optimal point for the recall-detection trade-off.
• The study highlights the effectiveness of ROC analysis in evaluating screening performance and setting targets.
• Continuous evaluation is crucial for maintaining a favorable benefit-to-harm ratio in screening programs.