When radiologists perform best: the learning curve in screening mammogram interpretation

Area of Science:

• Radiology
• Medical Imaging
• Breast Cancer Screening

Background:

• Screening mammography interpretation is a critical skill for early breast cancer detection.
• Radiologist experience significantly impacts diagnostic accuracy.
• Fellowship training in breast imaging may influence the learning curve for mammogram interpretation.

Purpose of the Study:

• To evaluate how clinical experience affects the interpretation accuracy of screening mammograms.
• To compare the learning curves of radiologists with and without specialized breast imaging fellowship training.

Main Methods:

• Analysis of 1,599,610 screening mammograms interpreted by 231 radiologists between 1996 and 2005.
• Radiologists' performance was assessed based on years of practice, categorized by fellowship training status.
• Statistical analysis accounted for case-mix and radiologist-level variations.

Main Results:

• Radiologists without fellowship training showed significant improvement in their first three years, with false-positive rates decreasing 11%-15% annually.
• The ratio of recalls to detected cancers improved with experience for non-fellowship-trained radiologists.
• Fellowship-trained radiologists demonstrated no significant learning curve, achieving high performance levels within their first year of practice.

Conclusions:

• Radiologists' screening mammogram interpretation skills enhance with early career experience.
• Targeted training and continuing medical education can optimize performance, reducing unnecessary work-ups while maintaining cancer detection rates.
• Breast imaging fellowship training facilitates rapid acquisition of expertise in mammogram interpretation.