*2.2. Statistical Analysis*

We compared the performance of DBT and ultrasound in women recalled for assessment after a positive screening mammogram to test the hypothesis that there is no difference between DBT and ultrasound in reducing unnecessary benign biopsies. The comparisons were performed according to breast density (BI-RADS A, B, C, and D). Relative risks at a 95% confidence interval were calculated to establish how DBT and ultrasound decreased the likelihood of an unnecessary biopsy following screening mammography. Using needle biopsy results as the reference standard, the number of cases requiring DBT and ultrasound assessment to prevent one unnecessary biopsy was estimated to determine the likelihood of benefit. The number needed to be assessed is inversely proportional to the risk reduction [1/(absolute risk reduction)]. The ideal screening number would be 1, in which all the women recalled for assessment with benign lesions have benefited. The association between the recalled mammographic abnormalities and breast density was analysed using the chi-square independence test ( χ2 continuity correction). The McNemar test ( χ2 continuity correction) was used to determine the statistical significance between DBT and ultrasound. A *p*-value ≤ 0.05 was considered statistically significant. These statistical analyses were conducted via the open-source Jamovi software (2.3.0).
