Monitoring the performance of breast screening programmes: Allowing for geographical variation in breast cancer incidence

Abstract

Among the myriad of problems associated with evaluating the performance of individual screening programmes in the National Health Service breast screening programme is that of correcting for background breast cancer incidence in the catchment areas of individual programmes. The background incidence will affect not only the expected cancer detection rate at screening, but also the interval cancer rates. This paper proposes a method that can be used to correct for background incidence. The method can be used either to correct the crude cancer detection rate or the age adjusted cancer detection rate using a measure such as the standardised detection ratio. Variation in background incidence and age distribution of screened women are the two major factors affecting the expected cancer detection rate for individual screening programmes. Control of these two variables should allow more effective evaluation of individual screening programme performance.