Table 1.
Facility-reported practice procedures and performance measures (N = 44 facilities with 484 463 screening mammography examinations)*
| Facility characteristic | No. (%) | Specificity†, % (95% CI) | Sensitivity‡, % (95% CI) | PPV1§, % (95% CI) | PPV2‖, % (95% CI) |
| Facility structure and organization | |||||
| Is your mammography facility for-profit or not-for-profit? | |||||
| Not-for-profit | 20 (45.5) | 91.0 (89.6 to 92.4) | 82.2 (77.8 to 86.7) | 4.9(4.0 to 5.8) | 42.5 (34.5 to 50.6) |
| For-profit | 24 (54.6) | 89.5 (86.1 to 92.8) | 77.3 (67.9 to 86.7) | 3.5(2.7 to 4.3) | 35.5 (25.7 to 45.2) |
| Is this facility associated with an academic medical center? | |||||
| No | 34 (77.3) | 90.2 (87.9 to 92.6) | 78.8 (72.3 to 85.2) | 4.3 (3.6 to 5.0) | 41.4 (33.9 to 48.8) |
| Yes | 10 (22.7) | 89.9 (86.8 to 93.0) | 82.7 (74.9 to 90.6) | 3.6 (2.1 to 5.2) | 28.6 (22.2 to 35.0) |
| Facility volume (average no. of mammograms per year)¶ | |||||
| ≤1500 | 9 (20.5) | 93.4 (91.3 to 95.6) | 64.6 (38.2 to 91.0) | 4.1 (1.5 to 6.7) | 32.7 (19.6 to 45.7) |
| 1501–2500 | 12 (27.3) | 87.1 (80.4 to 93.8) | 87.5 (79.9 to 95.0) | 3.6 (2.3 to 4.9) | 24.9 (16.3 to 33.6) |
| 2501–6000 | 12 (27.3) | 89.7 (88.2 to 91.3) | 83.1 (78.8 to 87.4) | 4.1 (3.4 to 4.9) | 47.6 (36.6 to 58.6) |
| >6000 | 11 (25.0) | 91.2 (89.4 to 93.1) | 78.1 (73.1 to 83.0) | 4.7 (4.0 to 5.5) | 46.4 (31.3 to 61.6) |
| What percentage of your mammograms are screening mammograms? | |||||
| 1–74 | 15 (34.1) | 89.9 (88.1 to 91.7) | 82.2 (77.4 to 87.0) | 3.9 (3.0 to 4.7) | 47.9 (34.6 to 61.2) |
| 75–79 | 13 (29.5) | 88.0 (81.9 to 94.1) | 84.2 (77.7 to 90.8) | 4.3 (3.3 to 5.4) | 30.9 (21.6 to 40.1) |
| 80–100 | 16 (36.4) | 92.2 (90.5 to 93.8) | 73.6 (60.5 to 86.6) | 4.2 (2.8 to 5.6) | 36.9 (27.3 to 46.5) |
| What percentage of the screening mammograms done at your facility are interpreted at another facility?# | |||||
| 0 | 27 (61.4) | 90.5 (89.2 to 91.8) | 82.4 (78.7 to 86.1) | 4.6 (3.8 to 5.3) | 40.5 (34.1 to 46.9) |
| 80–100 | 17 (38.6) | 89.6 (84.9 to 94.4) | 74.9 (61.5 to 88.3) | 3.4 (2.4 to 4.5) | 35.6 (21.1 to 50.1) |
| What percentage of screening mammograms are interpreted in groups of 10 or more? | |||||
| 0–50 | 7 (17.5) | 88.1 (74.6 to 101.7) | 77.6 (57.8 to 97.4) | 4.6 (1.5 to 7.7) | 35.8 (19.4 to 52.3) |
| 51–100 | 33 (82.5) | 90.4 (89.3 to 91.4) | 80.0 (73.9 to 86.1) | 3.9 (3.3 to 4.5) | 40.7 (33.1 to 48.4) |
| Missing | 4 | ||||
| Does this facility offer diagnostic mammograms? | |||||
| No | 11 (25.0) | 92.0 (89.8 to 94.1) | 68.6 (47.9 to 89.4) | 3.2 (1.8 to 4.5) | 47.6 (25.2 to 70.1) |
| Yes | 33 (75.0) | 89.5 (87.1 to 92.0) | 82.9 (79.4 to 86.4) | 4.5 (3.8 to 5.1) | 36.6 (30.5 to 42.7) |
| Does this facility offer interventional services (FNA, core or vacuum-assisted biopsy, cyst aspirations, needle localization or other procedures)? | |||||
| No | 17 (39.5) | 89.2 (84.5 to 93.9) | 76.8 (65.0 to 88.7) | 3.5 (2.5 to 4.6) | 39.7 (25.4 to 54.0) |
| Yes | 26 (60.5) | 90.6 (89.3 to 91.9) | 83.1 (79.2 to 86.9) | 4.6 (3.9 to 5.4) | 38.3 (32.1 to 44.5) |
| Missing | 1 | ||||
| Interpretive and audit processes | |||||
| Are any screening mammograms performed at your facility interpreted by radiologist(s) who specialize in breast care? | |||||
| No | 27 (61.4) | 88.9 (86.0 to 91.9) | 84.2 (80.5 to 87.8) | 4.6 (3.8 to 5.3) | 34.5 (27.4 to 41.6) |
| Yes | 17 (38.6) | 92.1 (90.9 to 93.3) | 71.9 (59.1 to 84.7) | 3.5 (2.5 to 4.5) | 47.7 (35.6 to 59.8) |
| Are any screening mammograms from your facility interpreted by more than one radiologist? | |||||
| No | 21 (48.8) | 89.6 (85.8 to 93.3) | 84.2 (77.5 to 90.8) | 4.3 (3.2 to 5.3) | 33.4 (25.4 to 41.5) |
| Yes | 22 (51.2) | 90.8 (89.3 to 92.4) | 74.9 (66.4 to 83.5) | 4.0 (3.2 to 4.8) | 45.3 (36.0 to 54.6) |
| Missing | 1 | ||||
| How are decisions made for mammograms interpreted by more than one radiologist? | |||||
| Double readings not performed | 21 (48.8) | 89.6 (85.8 to 93.3) | 84.2 (77.5 to 90.8) | 4.3 (3.2 to 5.3) | 33.4 (25.4 to 41.5) |
| Independently | 19 (44.2) | 90.9 (89.2 to 92.6) | 74.0 (64.0 to 84.0) | 3.8 (2.9 to 4.7) | 46.3 (35.4 to 57.2) |
| By consensus | 3 (7.0) | 90.4 (80 to 100.8) | 80.6 (57.1 to 104.2) | 5.2 (2.9 to 7.5) | 39.8 (14.9 to 64.6) |
| Missing | 1 | ||||
| How often is individual radiologist-level performance data shared with radiologists? | |||||
| Once a year | 18 (40.9) | 89.9 (88.3 to 91.5) | 80.5 (69.3 to 91.7) | 3.7 (2.9 to 4.5) | 33.4 (25.3 to 41.5) |
| Two or more times per year | 19 (43.2) | 91.9 (90.5 to 93.4) | 78.6 (75.0 to 82.2) | 4.7 (3.6 to 5.8) | 39.2 (29.1 to 49.3) |
| Unknown | 7 (15.9) | 85.9 (73.3 to 98.6) | 79.9 (60.0 to 99.9) | 3.8 (1.9 to 5.6) | 52.1 (27.4 to 76.9) |
| How is this information reviewed? | |||||
| Reviewed together (with other radiologists) in a meeting | 21 (47.7) | 89.9 (88.5 to 91.2) | 81.8 (77.3 to 86.3) | 4.4 (3.8 to 5.0) | 36.3 (26.3 to 46.4) |
| Reviewed by facility or department manager or by lead radiologist alone | 8 (18.2) | 92.3 (90.9 to 93.8) | 77.2 (71.0 to 83.4) | 3.8 (2.2 to 5.4) | 41.9 (33.9 to 50.0) |
| Reviewed by each radiologist alone | 5 (11.4) | 93.6 (89.9 to 97.3) | 84.6 (71.4 to 97.8) | 5.9 (1.5 to 10.4) | 40.4 (26.2 to 54.6) |
| Unknown | 10 (22.7) | 87.3 (79.0 to 95.7) | 74.1 (51.4 to 96.7) | 3.0 (1.7 to 4.3) | 41.8 (21.9 to 61.6) |
Outcomes in bold have an overall P level of .1 or lower for a univariate association. CI = confidence interval; PPV1 = positive predictive value 1; PPV2 = positive predictive value 2; FNA = fine-needle aspiration; BI-RADS = American College of Radiology Breast Imaging Reporting and Data System.
Specificity was defined as the percentage of screening examinations that were given a negative BI-RADS assessment (BI-RADS 1, 2, or 3 with a recommendation of normal or short-interval follow-up) among those that did not have a breast cancer diagnosis during the follow-up period.
Sensitivity was defined as the percentage of screening mammograms that were given a positive BI-RADS assessment (BI-RADS 0, 4, 5, or 3 with a recommendation for immediate work-up) among those that had a diagnosis of invasive breast cancer or ductal carcinoma in situ during the follow-up period.
PPV1 was defined as the percentage of screens that were associated with a breast cancer diagnosis within 1 year of follow-up among those with a positive BI-RADS assessment.
PPV2 was defined as the percentage of screens that were associated with breast cancer diagnosis during follow-up among those with a BI-RADS assessment of 4 or 5 and a recommendation for biopsy, surgical consultation, or FNA.
Based on registry data.
No facilities reported that 1%–79% of the screening mammograms done at the facility were interpreted at another facility.