INCIDENT ROUND CANCERS - WHAT LESSONS CAN WE LEARN

At the time of implementation of the National Health Service Breast Sc reening Programme a range of targets were set, and have since been upd ated. This study looks at a subset of cancers detected by screening - the incident round cancers - which are not currently identified and ad dressed by the targets. In our first incident round we screened 24 838 women and detected 112 breast cancers in lll women. On the basis of a two-stage retrospective evaluation of the prevalent round mammograms in these 111 women, we classified the prevalent mammograms as false ne gative 21 (19%), minimal signs 32 (28%) and new tumours 59 (53%). Anal ysis of the radiographic features revealed a disproportionate number w ith parenchymal deformity (21%) and asymmetry (17%) among the missed t umours when compared to all tumours detected by our screening programm e (12% and <1%, respectively), Identification and quantification of th ese incidence detected missed cancers could be incorporated into the Q uality Assurance Data Set, and used as another performance indicator. Reviewing the mammograms proved a valuable teaching exercise and has a llowed us to develop a test set of challenging mammograms.