Introduction

This chapter is concerned with what could be called “epidemiological” quality assurance (QA). In this aspect of QA we are mostly concerned with estimating the achievement of organized screening programs in relation to breast cancer mortality reduction and factors related to sensitivity and specificity. We are not directly interested in QA related to other factors, such as those of interest to the physicist or breast surgeon, except where they may impact on the test sensitivity and specificity. To begin with it is useful to discuss some of the fundamental measures used in assessing screening performance and why these are often difficult to measure in practice. This then leads to a discussion of randomized controlled trials (RCTs), and how they can be used to measure screening efficiency in organized national screening programs, and how improvements in performance can be made to such programs.

Fundamental measures of screening performance

The objective of breast cancer screening is to reduce deaths from breast cancer by detecting and treating the disease at an earlier stage than that at which the disease would have presented clinically. Following the screening test (mammography) women whose mammograms show possible signs of malignancy are referred (or “recalled”) for further diagnostic tests and those whose mammograms appear normal are returned to routine recall. Two important measures of screening performance are the sensitivity and specificity of the screening test. The test sensitivity is the proportion of women with breast cancer correctly referred for further assessment by the screening test.