An ultrasound-based breast cancer screening programme in Shenzhen, China, has been used to analyse —across nearly 1.7 million examinations— which interval between tests and which management of each BI-RADS category best balance detecting cancers against avoiding unnecessary recalls. The authors conclude that a two-year interval is appropriate between the ages of 40 and 64.
What was known
Population screening aims to bring diagnosis forward, but every additional examination has a cost: anxiety, extra tests and biopsies, some of them avoidable. In programmes where ultrasound is the main test, large-scale data were lacking on how often it should be repeated and how to act at each level of suspicion on the BI-RADS scale.
What this work adds
The researchers analysed 1,694,838 ultrasound examinations reported with BI-RADS between 2018 and 2023. For BI-RADS 1 and 2 reports they assessed inter-screen intervals using four indicators: recall rate, cancer detection rate, positive predictive value and early-stage proportion. For BI-RADS 0, 3, 4 and 5 they examined the relationship between biopsy waiting time and those same indicators, This is a retrospective observational study of a single urban programme.
Main results
Among 337,497 intervals following a BI-RADS 1 or 2 report, longer intervals were associated with a higher detection rate (from 0.13 to 0.54 cases per 1,000 women) but also with more recalls (from 17.4% to 22.1%) and a lower positive predictive value (from 61.1% to 31.0%), with no change in the early-stage proportion. The interval that best balanced yield and burden was 13 to 24 months in women aged 40 or over, and 25 to 36 in younger women. Of the BI-RADS 4A lesions, 64.7% (10,909 of 16,871) were downgraded during imaging follow-up. In 487 biopsied BI-RADS 3 cases, delaying biopsy increased the positive predictive value without changing the early-stage proportion.
What it means
The authors conclude that a two-year interval is advisable between 40 and 64 years, extendable to three years below the age of 40, and that biopsy should not be delayed for BI-RADS 4B, 4C and 5. The high downgrade rate of BI-RADS 4A lesions supports image-based triage —still pending validation— before biopsy. Tailoring intervals and management pathways to the level of suspicion reduces the workload on services without losing detection capability.
A breast unit perspective
These figures come from an ultrasound screening programme in China, not from the Spanish screening programme, which relies on biennial mammography, so they are not directly transferable. The organisational idea that does matter: classifying women by level of suspicion and time to biopsy allows a unit to audit process quality with measurable indicators.
Reference: Sun W, Li H, Zhao X, Hu S, Qi H, Jin S, et al. Optimising screening intervals and BI-RADS management pathways in ultrasound-based breast cancer screening: real-world evidence from Shenzhen (2018-2023). Breast Cancer Research and Treatment, 2026. 10.1007/s10549-026-08045-w
