Table 9:
Sensitivity and Specificity of Mammography and Adjunct Ultrasound Compared With Mammography Alone in Women at High Risk for Breast Cancer
| Author, Year | Round | Screens, N | Sensitivity, % (95% CI)a | Specificity, % (95% CI)a | ||||
|---|---|---|---|---|---|---|---|---|
| Mammography | Mammography + Ultrasound | Absolute Change in Sensitivity | Mammography | Mammography + Ultrasound | Absolute Change in Specificity | |||
| Positive Test: BI-RADS 4, 5 | ||||||||
| Riedl et al, 201541 | All | 1,365 | 37.5 (24.2, 53.0) | 50.0 (35.2, 64.8) | 12.5 (NR) | 97.1 (96.1, 97.9) | 95.7 (94.5, 96.7) | −1.4 (NR) |
| Sardanelli et al, 201142 | All | M: 1,095 M+US: 1,047 | 50.0 (35.5, 64.5) | 62.5 (47.3, 76.0) | 12.5 (NR) | 99.0 (98.2, 99.5) | 97.6 (96.4, 98.4) | −1.45 (NR) NSb |
| Kuhl et al, 201038 | All | 1,679 | 33.3 (17.2, 53.9) | 48.1 (29.1, 67.6) | 14.8 (NR); NSc | 99.1 (98.5, 99.5) | 98.4 (97.5, 98.8) | −0.79 (NR) |
| Kuhl et al, 200539 | All | 1,452 | 32.6 (19.0, 48.5)a | 48.8 (33.3, 64.5)a | 16.28 (NR) | 96.8 (95.7, 97.7)a | 89.0 (87.2, 90.6)a | −7.81 (NR) |
| Positive Test: BI-RADS 3, 4, 5 | ||||||||
| Berg et al, 201236 | 1 | 2,659 | 55.6 (38.1, 72.1) | 94.4 (81.3, 99.3) | 38.9 (20.2, 57.5); P < .001 | 89.1 (87.8, 90.3) | 74.3 (72.6, 76.0) | −14.8 (−16.3, −13.2); P < .001 |
| 2,3 | 4,814 | 52.0 (40.1, 63.7) | 76.0 (64.7, 85.1) | 24.0 (14.7, 33.3); P < .001 | 91.3 (90.4, 92.0) | 84.1 (83.1, 85.2) | −7.1 (−8.0, −6.3); P < .001 | |
Abbreviations: BI-RADS, Breast Imaging Reporting and Data System; CI, confidence interval; M, mammography; N, number; NR, not reported; NS, not significant; US, ultrasound.
Calculated based on data provided in study.
Authors stated that the specificity ranged from 96% to 99% across imaging modalities and combinations without significant differences; however it is unclear if this applies to comparison between mammography alone and mammography with adjunct ultrasound.
Authors stated that the combination of ultrasound and mammography was not statistically significantly higher (P < .12) than mammography alone or ultrasound alone. It is unclear if the P value represents the comparison to mammography alone, ultrasound alone, or both.