Breast ultrasound in Chinese hospitals: A cross-sectional study of the current status and influencing factors of BI-RADS utilization and diagnostic accuracy

Summary: Background: With the growing demand for breast screening in public health services and clinical care, ultrasound departments in China are facing tremendous challenges. Methods: A cross-sectional nationwide survey was conducted in 5,460 departments providing ultrasound diagnoses in mainland China from 2020 to 2021. The survey included general information about the ultrasound department, the characteristics of sonologists, the use of Breast Imaging Reporting and Data System (BI-RADS) templates, and the diagnostic accuracy rate of breast cancer ultrasound. Findings: There were on average 2.25 sonologists per 10,000 patients in mainland China per year. The average utilization rate of BI-RADS in Chinese hospitals was 87.02%. The GDP per capita of the province (P = 0.008), whether the hospital was specialized (P = 0.002) or a Tier 3 facility (P < 0.001), the percentage of doctors with master's and doctoral degrees (P < 0.001) and doctors ≤35 years (P = 0.005) were significantly and independently associated with the utilization rate of BI-RADS. The average diagnostic accuracy rate of breast cancer ultrasound in Chinese hospitals was 73.64%, and we observed significant positive associations between GDP per capita (P = 0.02), BI-RADS utilization rate (P = 0.019), and breast cancer ultrasound diagnostic accuracy. Interpretation: The utilization of BI-RADS templates effectively improved the diagnostic accuracy of ultrasound. Moreover, the survey summarized the current situation of departments and sonologists providing breast ultrasound diagnosis in mainland China, which helped monitor the development of the discipline and provide information for administrators to meet the growing demand. Funding: This work was supported by Natural Science Foundation of Beijing (7202156) and Foundation of ihecc (2019-C-0646-2).

Abstract licence: CC BY-NC

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.