Table 3.
Comparisons of lifetime risk classification between ML-Adapt Boosting (ML-ADA) algorithm and the BOADICEA model (reference standard) for the breast cancer-free cohort.
| Risk age | Near-population risk BOADICEA risk < 17%, N = 21,283 | Moderate risk 17% ≤ BOADICEA risk < 30%, N = 11,685 | High-risk BOADICEA ≥ 30%, N = 3178 | ||||||
|---|---|---|---|---|---|---|---|---|---|
| ML-ADA < 17% | 17%≤ ML-ADA < 30% | ML-ADA ≥ 30% | ML-ADA < 17% | 17%≤ ML-ADA < 30% | ML-ADA ≥ 30% | ML-ADA < 17% | 17%≤ ML-ADA < 30% | ML-ADA ≥ 30% | |
| 20–29 (n = 4 959) | 2181 | 430 | 215 | 372 | 1050 | 233 | 17 | 41 | 420 |
| 30–39 (n = 5 277) | 2069 | 645 | 430 | 407 | 989 | 256 | 18 | 34 | 429 |
| 40–49 (n = 6 410) | 2466 | 832 | 625 | 442 | 1191 | 326 | 20 | 44 | 464 |
| 50–59 (n = 7 025) | 2681 | 899 | 751 | 535 | 1243 | 337 | 25 | 49 | 505 |
| 60–69 (n = 6 436) | 2037 | 745 | 849 | 570 | 1326 | 349 | 23 | 43 | 494 |
| 70–80 (n = 6 039) | 2116 | 871 | 441 | 465 | 1233 | 361 | 21 | 48 | 483 |
| Total | 13,550 | 4422 | 3311 | 2791 | 7032 | 1862 | 124 | 259 | 2795 |
| Concordance | 63.67% | – | – | – | 60.18% | – | – | – | 87.95% |
| Reclassification | – | 20.78% | 15.56% | 23.89% | – | 15.93% | 3.90% | 8.15% | – |
– Does not apply.
N = 36,146. ML machine learning, BOADICEA Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm, ADA adaptive boosting.