Cancer Research UK : Breast Screening - Performance of mammography

Breast Screening - Performance of mammography

This page contains information on the performance of mammography taken from the CancerStats report - Breast Screening - UK published in 2003. Sections include sensitivity by age, specificity, positive predictive value and interval cancers. The Issues around Breast Screening page has information on two view mammography.

Sensitivity by age

To calculate sensitivity the number of true cancers must be known. The number of true cancers is estimated from the number of screen-detected cancers plus the number of cancers that are detected in a suitable time interval, often 12 months. The sensitivity will be lowered if the time since mammography is increased as additional cancers will be detected. These additional cancers will consist of those that were missed by mammography and true interval cancers that have developed since the last screening.

The sensitivity of mammography in women aged over 50 has been estimated to range from 68% to over 90%, with most trials and programmes achieving sensitivities of around 85%. In women aged 40-49 the sensitivity has been reported to be lower, with estimates between 62% and 76%.1

Mammography is less effective in identifying cancers in women under 50 because breast tissue tends to be denser in pre-menopausal women. The sensitivity of mammography is much lower in women with dense breasts than those with predominantly fatty breasts.2 Furthermore, cancers found in younger women tend to be more aggressive and grow faster.3,4

Specificity

The specificity of breast screening by mammography ranges between 82% and 97%.1

Positive predictive value

Estimates of positive predictive value in the UK Breast Screening Programme range from 6% to 8% for prevalent screening, meaning that 6% to 8% of women recalled for further tests after their first screening have cancer. The positive predictive value is higher for incident screens and has been estimated as between 12% and 14%.5

Interval cancers

A necessary condition for effective screening is that the incidence of cancer occurring in the interval between screens (interval cancers) is kept relatively low. Interval cancers can occur through failure to detect an abnormality at the time of screening (false negative result), or as a new event after a negative screen (true interval cancer). The effectiveness of a screening programme depends on both the sensitivity of the screening test and the frequency of screening. Less frequent screening will lead to an increase in the number of true interval cancers. As the overall interval cancer incidence approaches that of the unscreened population, the benefits of screening disappear.

Measures of interval cancers can be expressed in two ways: as the rate of interval cancers per women years, or as a proportion of the underlying incidence of breast cancer in the absence of breast screening. It has been assumed that the background incidence of breast cancer, without breast screening, is 22 per 10,000 person years for women aged 50-64.6

Table 3.17-10 shows the standards set for the UK Breast Screening Programme and results from some individual programmes.

Table 3.1: Interval cancers occurring in the Breast Screening Programme, UK women aged 50-64

Download this table (15KB)

The rates of interval cancers have improved over time, a likely result of the introduction in 1995 of two view mammography in prevalent screens and improvements in techniques and reading standards within the National Screening Programme.

References for performance of mammography

IARC, Screening Techniques.IARC Handbooks of Cancer Prevention: Breast Cancer Screening, ed. B.F. Vainio H. Vol. 7. 2002, Lyon: IARC Press.
Mandelson, M.T., et al.,Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst, 2000. 92(13): p. 1081-7
Jmor, S., et al., Breast cancer in women aged 35 and under: prognosis and survival. J R Coll Surg Edinb, 2002. 47(5): p. 693-9
Kothari, A. and I.S. Fentiman, Diagnostic delays in breast cancer and impact on survival. Int J Clin Pract, 2003. 57(3): p. 200-3

Blanks, R.G., S.M. Moss, and J. Patnick, Results from the UK NHS breast screening programme 1994-1999. J Med Screen, 2000. 7(4): p. 195-8
IARC Handbooks of Cancer Prevention Effectiveness of Screening: Breast Cancer Screening. ed. B.F. Vainio H. Vol. 7. 2002, Lyon: IARC Press. p.119-156.
Moss, S. and R. Blanks, Calculating appropriate target cancer detection rates and expected interval cancer rates for the UK NHS Breast Screening Programme. Interval Cancer Working Group. J Epidemiol Community Health, 1998. 52(2): p. 111-5
Day, N., et al., Monitoring interval cancers in breast screening programmes: the east Anglian experience. Quality Assurance Management Group of the East Anglian Breast Screening Programme. J Med Screen, 1995. 2(4): p. 180-5
Everington, D., et al., The Scottish breast screening programme's experience of monitoring interval cancers. J Med Screen, 1999. 6(1): p. 21-7

Fielder, H., et al., Results from 10 years of breast screening in Wales. J Med Screen, 2001. 8(1): p. 21-3