Skip to main content
. Author manuscript; available in PMC: 2019 Aug 1.
Published in final edited form as: Cancer. 2018 Jun 22;124(16):3319–3328. doi: 10.1002/cncr.31528

Table 3.

Model Performance of the BCSC and BBD-BC Models

BCSC
Endpoint N Concordance Invasive BC (%) C-statistic (95% CI) Observed Events Calibration Predicted Events Pred./Obs. (95% CI)
Screening Cohort*
5 years 1089 20 (1.8%) 0.59 (0.46, 0.71) 20 26.12 1.31 (0.94, 2.25)
 10 years 1089 54 (5.0%) 0.63 (0.56, 0.71) 54 51.02 0.94 (0.85, 1.43)
BBD-BC
Endpoint N Concordance Invasive or DCIS BC (%) C-statistic (95% CI) Calibration Observed Events Predicted Events Pred./Obs. (95% CI)
Screening Cohort*
5 years 1089 27 (2.5%) 0.69 (0.58, 0.80) 27 29.93 1.11 (0.84, 1.77)
 10 years 1089 69 (6.3%) 0.66 (0.59, 0.73) 69 57.54 0.83 (0.77, 1.20)
Full BBD Cohort**
5 years 2142 56 (2.6%) 0.68 (0.60, 0.75) 56 49.70 0.89 (0.71, 1.21)
 10 years 2142 118 (5.5%) 0.66 (0.60, 0.71) 118 95.34 0.81 (0.70, 1.00)

BCSC= Breast Cancer Surveillance Consortium; BBD-BD= Benign Breast Disease to Breast Cancer; DCIS=ductal carcinoma in situ; BC=Breast Cancer; Pred. = Predicted; Obs. = Observed;

*

Screening cohort, n=1089 ages 35–74 and with mammogram;

**

Full cohort, n=2142, ages 18 and up