Assessment of lesions detected at mammographic screening: performance at first or repeat screening in the Florence...

To evaluate the assessment criteria and the results achieved in the detection of breast lesions at mammographic screening.

Review of cases assessed in the last screening round of Florence city (FC--first screening round: 29,522 subjects) and Florence district (FD --repeat screening round: 13,268 subjects) programmes.

Referral and biopsy rates, predictive values, and prevalence of cancers detected by screening were determined, as well as the frequency of the diagnostic procedures used at assessment, and their contribution to the final diagnosis according to the mammographic appearance of the suspected lesion. Assessment costs were estimated.

Referral rate (FC 4.2%; FD 1.8%), referral positive predictive value (FC 18.7%; FD 28.3%), surgical biopsy rate (FC 0.96%; FD 0.6%), benign/malignant biopsy ratio (FC 0.20; FD 0.13), and prevalence of cancers detected by screening (FC 0.78%; FD 0.5%) were all within the European Community (EC) recommended standards for screening performance. The benign biopsy rate was considerably lower than that of recommended standards. The cost for each subject assessed was 179,000 Italian lire at the first and 116,000 lire at repeat screening. The cost for each subject screened that was attributable to assessment was 7600 lire at the first or 2100 lire at repeat screening.

Limited referral rates and costs were achieved and the proportion of cancers detected by screening was high. The number of referrals was further reduced at repeat screening, and assessment had a limited impact on total screening costs. Detail or magnification mammography, palpation, sonography, and fine needle aspiration cytology all contributed to the final diagnosis and should be immediately available at the assessment clinic. The observed benign biopsy rate was particularly low and suggests that EC recommended standards should be modified.