Table 2.
Screening Mammography Accuracy and Outcomes in Women With a Personal History of Breast Cancer (PHBC) and Matched Screening Mammograms in non-PHBC Women
| Measure of Accuracy or Outcome (Total Screens or Total Cancers, Both Groups) |
Screens With PHBC (95% CI)a |
Screens Without PHBC (95% CI)b |
P Value |
|---|---|---|---|
| No. of screening mammograms (117 660 screens) [117 740 screens]c | 58 830 [58 870]c | 58 830 [58 870]c | |
| No. of in situ or invasive breast cancers (956 cancers) [997 cancers]c | 615 [655]c | 341 [342]c | |
| Cancer rate/1000 mammogramsd (117 660 screens) [117 740 screens]c | 10.5 (9.7–11.3) [11.1 {10.3–12.0}]c |
5.8 (5.2–6.4) [5.8 {5.2–6.5}]c |
<.001 <.001 |
| No. of cancers detected on screening mammography | 402 | 261 | |
| CDR/1000 mammograms (117 660 screens) | 6.8 (6.2–7.5) | 4.4 (3.9–5.0) | <.001 |
| DCIS detection rate (117 660 screens) | 2.0 (1.7–2.4) | 0.8 (0.6–1.0) | <001 |
| Invasive cancer detection rate (117 660 screens) | 4.8 (4.3–5.4) | 3.7 (3.2–4.2) | .002 |
| CDR/1000 mammograms in PHBC of DCIS (12 135 screens)e | 11.5 (9.8–13.6) | ||
| CDR/1000 mammograms in PHBC of invasive cancer (46 695 screens)f | 5.6 (5.0–6.3) | ||
| No. of interval cancers | 213 | 80 | |
| ICR/1000 mammograms (117 660 screens) | 3.6 (3.2–4.1) | 1.4 (1.1–1.7) | <.001 |
| Interval cancers, % (956 cancers) | 34.6 (31.0–38.5) [213/615] |
23.5 (19.3–28.3) [80/341] |
<.001 |
| Mammograms with additional same-day imaging or recommendation, based on initial screening recommendation, % (117 422 screens) | 18.1 (17.6–18.6) [10 612/58 696] |
8.3 (8.1–8.5) [4862/58 726] |
<.001 |
| AIR, based on final assessment (117 660 screens)g | 2.3 (2.2–2.5) [1377/58 830] |
1.4 (1.3–1.5) [847/58 830] |
<.001 |
| Mammograms positive and recommended for fine-needle aspiration, biopsy, or surgical consultation, based on final recommendation, % (117 640 screens) | 2.2 (2.1–2.3) [1280/58 815] |
1.4 (1.3–1.5) [798/58 825] |
<.001 |
| Sensitivity, % (956 screens associated with cancer)g | 65.4 (61.5–69.0) [402/615] |
76.5 (71.7–80.7) [261/341] |
<.001 |
| Sensitivity for detection of invasive cancers, % (749 screens associated with invasive cancer)g | 61.1 (56.6–65.4) [284/465] |
75.7 (70.4–80.3) [215/284] |
<.001 |
| Sensitivity for detection of DCIS, % (207 screens associated with DCIS)g | 78.7 (71.4–84.5) [118/150] |
80.7 (68.4–89.0) [46/57] |
.74 |
| Specificity, % (116704 screens not associated with cancer)g | 98.3 (98.2–98.4) [57 240/58 215] |
99.0 (98.9–99.1) [57 903/58 489] |
<.001 |
| PPV, % (2224 positive screen results)h | 29.2 (26.8–31.7) [402/1377] |
30.8 (27.8–34.0) [261/847] |
.42 |
Abbreviations: AIR, abnormal interpretation rate; CDR, cancer detection rate; CI, confidence interval; DCIS, ductal carcinoma in situ; ICR, interval cancer rate; PPV, positive predictive value.
Forty PHBC screens from women with a history of mastectomy, associated with ipsilateral cancer recurrence during followup, were excluded from analysis of accuracy; all rates are per 1000 screens, and all percentages are shown with numerator and denominator.
Forty screens without PHBC (matched screening examinations) were excluded from analysis of accuracy; all rates are per 1000 screens, and all percentages are shown with numerator and denominator.
Numbers and rates in square brackets include 40 screens with ipsilateral cancer (recurrence on mastectomy side), or 40 matched screens in the cohort without PHBC.
Cancer rate refers to all cancers identified in screening participants (screen-detected and interval cancers).
CDR in women with PHBC of DCIS (12135 screens) included DCIS detection rate of 4.9 per 1000 mammograms (95% CI, 3.8–6.3) and invasive CDR of 6.7 per 1000 mammograms (95% CI, 5.4–8.3).
CDR in women with PHBC of invasive cancer (46695 screens) included DCIS detection rate of 1.3 per 1000 mammograms (95% CI, 1.0–1.6) and invasive CDR of 4.3 per 1000 mammograms (95% CI, 3.8–5.0).
Estimates based on positive mammogram results are defined as Breast Imaging Reporting and Data System (BI-RADS) assessments 4 or 5 or a BI-RADS assessment of 0 or 3 with a recommendation for biopsy, fine-needle aspiration, or surgical consultation. Negative mammogram results included BI-RADS assessments 1 or 2 or an assessment of 3 without recommendation for biopsy, fine-needle aspiration, or surgical consultation. Missing final result was BI-RADS assessment of 0 with recommendation for additional imaging or unspecified evaluation or missing recommendation. All remaining assessments of 0 were considered negative.
Percentage of positive mammogram results on final imaging assessment associated with a cancer diagnosis.