Automated breast ultrasound: fewer recalls, similar cancer detection

Design

Observational registry analysis, Breast Cancer Surveillance Consortium

Population

25,328 supplemental ultrasound examinations in women with dense breasts

Primary outcome

Abnormal interpretation, follow-up and cancer detection rates

Effect

Short-interval follow-up 3.0% vs 14.6%; cancer detection similar

This observational analysis used Breast Cancer Surveillance Consortium data from 27 facilities between 2017 and 2020: 21,110 handheld and 4,218 automated supplemental ultrasound examinations in women with dense breasts. It was published in July.

Automated ultrasound was associated with a lower abnormal interpretation rate (14.9% vs 19.2%), far fewer short-interval follow-up recommendations (3.0% vs 14.6%) and fewer false-positive initial assessments. Biopsy recommendation rates, positive predictive values and cancer detection did not differ.

The drop in short-interval follow-up is the clinically meaningful part: it is the recommendation that most burdens women without finding cancer. Because women were not randomised, differences in who received each test and how readers worked may explain some of this. For Indian practice, where handheld ultrasound is widely used for dense breasts, automated systems remain a capital and workflow question rather than a clear upgrade.

• Consider automated breast ultrasound for supplemental screening in dense breasts where it is available.
• Audit your own short-interval follow-up rate for handheld screening ultrasound.
• Expect similar cancer detection with either method in this setting.
• Do not assume automated ultrasound suits every woman; selection was not randomised.

Why it matters

It suggests supplemental screening can be made less burdensome without losing cancers.

Don't overread it

This was observational; women and readers were not randomised between methods.

The statistics, in plain English

Short-interval follow-up fell from about 15 to 3 in every 100 examinations. Cancer detection being similar means the drop in recalls did not come at the cost of missed cancers that could be measured here. Because this is observational, the groups may have differed in ways not adjusted for.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.