ABSTRACT
Background: To validate the BIRADS in mammography, the calculation of its predictive value in each center is required, as recommended by the American College of Radiology. Aim: To determine the predictive value of the BIRADS system in our center. Material and Methods: All ultrasound guided needle percutaneous biopsies, performed at our center between 2006 and 2010 were reviewed. Predictive value, sensitivity, specificity and diagnostic accuracy of BIRADS were calculated, with a confidence interval of 95%. Results: Of 1,313 biopsies available, 1,058 met the inclusion criteria. Fifty eight percent of biopsies were performed to women with mammographies classified as BIRADS 4 or 5. The presence of cancer in mammographies classified as BIRADS 0 was 4%. The prevalence of cancer for mammographies BIRADS 1, 2, 3, 4 and 5 were 0, 3, 2.7, 17.7 and 72.4% respectively. The positive and negative predictive values of BIRADS classification were 55 and 92 % respectively. Conclusions: In our institution BIRADS classification 4 and 5 has a high positive predictive value for detecting cancer as in developed countries.