ABSTRACT
The quality of a breast-screening programme is measured not only by it's cancer detection rates (sensitivity), but also by it's recall rates and positive predictive values (specificity). False positive screens, which lead to assessments with benign outcomes, can cause psychological and physical morbidity in previously well women. In 1999 the Wiltshire breast-screening programme adopted a novel recall strategy whereby all films marked for recall are subject to consensus opinion--even women "recalled" by both original readers may be returned to routine screening. Since 1999 our recall rates have fallen from above to below the national averages (prevalent 7.2% and incident 3.0% in 2001/02). Specificity has increased over the same period such that approximately one in seven women assessed will be diagnosed with cancer. The consensus meetings provide a forum for teaching and are very popular with our film readers, who all report reduced levels of stress.