Rates of recall for thorough examination and surgical biopsy following mammographic screening. Experience with the screening program in Florence

The authors evaluate the recall and biopsy rates during three years of mammographic screening. Recalls (373/25970 = 1.4%) were more frequent at the first (2.3%) than at subsequent screening rounds (1.2%) whereas the predictive value for cancer increased at further screening (11.7 vs 21.7%). Recalls are less frequent in older women, particularly at further screening (40-49 vs 50-59 vs 60-70: 1.7 vs 1.1 vs 0.9%) whereas predictive value is higher (14.8 vs 16.5 vs 36.5%). The presence of mammographic abnormalities accounts for the majority of recalls (1.3%) whereas subjective symptoms other than pain are a less frequent cause of recall (0.2%) although as predictive (22.7 vs 20.4%). Calls for biopsy were limited to cases with suspicion for cancer and this allowed a low biopsy rate (75 = 0.3%) and a high predictive value (92%). The observed results have greatly improved with respect to the past years, due to a higher specificity of recall criteria and to the systematic use of all modern tests (high frequency US, fine-needle aspiration cytology, either manual or US-guided or stereotaxic) at the time of diagnostic assessment. All screening programs, particularly the new ones, should monitor recall and biopsy rates and predictive values as they are indicators of efficiency.