Evolution of breast cancer screening in countries with intermediate and increasing incidence of breast cancer

• Published in: Journal of medical screening 2006-01, Vol.13 Suppl 1 (2), p.S23-S27
• Main Authors: Wu, Grace Hui-Min, Chen, Li-Sheng, Chang, King-Jen, Hou, Ming-Feng, Chen, Shin-Chen, Liu, Tse-Jia, Huang, Chiun-Sheng, Hsu, Giu-Cheng, Yu, Chih-Cheng, Jeng, Li-Li, Chen, Shou-Tung, Chou, Yi-Hung, Wu, Chang-Ying, Shin-Lan, Koong, Chen, Tony Hsiu-Hsi
• Format: Article
• Language: English
• Subjects: Aged; Breast cancer; Breast Neoplasms - diagnosis; Breast Neoplasms - epidemiology; Humans; Incidence; Logistic Models; Mammography - methods; Mammography - statistics & numerical data; Mammography - trends; Mass Screening - methods; Mass Screening - statistics & numerical data; Mass Screening - trends; Middle Aged; Models, Statistical; Sound Spectrography - methods; Sound Spectrography - statistics & numerical data; Sound Spectrography - trends; Surveys and Questionnaires; Taiwan - epidemiology
• ISSN: 0969-1413; 1475-5793

Few studies have been published regarding the practice of breast cancer screening in Asian countries. The present study illustrates how the health policy for breast cancer screening has evolved in Taiwan from selective mammographic screening within a high-risk group, firstly to a programme of physical examination by public health nurses, and finally to a two-stage breast cancer screening programme, with a risk assessment followed by mammography for those at moderate to high risk. Breast cancer screening has evolved from 1995 to 2004 in Taiwan in three stages: (1) selective screening for breast cancer with mammography, ultrasound and physical examination only in first-degree relatives of breast cancer cases (1995-1998); (2) a programme of mass screening (1999-2001) with physical examination by public health nurses; and (3) two-stage breast cancer screening with a risk factor questionnaire and mammography for those deemed at moderate-to-high risk (2002-2004). The questionnaire was based on significant risk factors in a previous epidemiological study, in conjunction with the physical examination programme, a risk score was constructed from the logistic regression coefficients from the previous study, and women with a score above the median in the previous epidemiological study were assigned to mammography. Two-stage mammography screening had the most favourable results compared with the two previous screening regimes. It had a positive predictive value of recall after mammography of 14%, compared with 8% for selective screening and 2% for physical examination. Of screen-detected cancers in the two-stage programme, 71% were either ductal carcinoma in situ or stage T1, compared with 61% for selective screening and 60% for physical examination. The area under the receiver operating characteristic curve was 71% for the two-stage programme. For a low- to medium-risk country such as Taiwan, two-stage screening has acceptable parameters of recall and cancer detection, and compares well with other screening strategies.