OBJECTIVE: To present an audit model for mammography results used on the examinations performed at "Centro de Diagnóstico Mamário da Santa Casa da Misericórdia do Rio de Janeiro", Rio de Janeiro, Brazil. MATERIALS AND METHODS: Between May 1997 and September 1998, 2,000 mammography examinations were performed, 1,570 in asymptomatic women, which results were assessed in this audit. RESULTS: The mammography examinations of the asymptomatic women were classified as follow: category 1 - 1,362 cases; category 2 - 9 cases; category 3 - 95 cases; category 4 - 24 cases; category 5 - 6 cases; category 0 - 74 cases. Out of 199 abnormal examinations (categories 0, 3, 4 and 5), 107 (54%) were followed-up and 92 (46%) were lost to follow-up. The 107 cases included in the study were classified as follows: category 3 - 48 cases; category 4 - 18 cases; category 5 - 4 cases; category 0 - 37 cases. No cases of cancer were detected among the 48 cases in category 3; two cases of cancer were seen among the 18 cases in category 4; all four cases in category 5 were cancer; one case of cancer was seen among the 37 cases in category 0. The audit values obtained were the following: positive predictive value (based on the exam with abnormal result) - 11%; positive predictive value (when there was indication for biopsy) - 31%. The rate of cancer detection in 1,570 women was of seven cases; the sensitivity was 100%; the specificity was 96%; and the recall rate was 4.7%. CONCLUSION: All these results reproduce data from the literature.
Audit; Mammography; BI-RADS
