Comparison of model performance as a function of time
| Prediction time (months) | Image prediction (by algorithm) | Patient prediction (by algorithm) | Sensitivity | Specificity | Net benefit |
| Up to 15 | 0.73 (0.68 to 0.78) | 0.79 (0.74 to 0.84) | 0.45 (0.35 to 0.54) | 0.94 (0.92 to 0.95) | 0.02 (0.01 to 0.04) |
| 16–24 | 0.76 (0.71 to 0.80) | 0.80 (0.76 to 0.84) | 0.42 (0.34 to 0.51) | 0.93 (0.91 to 0.94) | 0.02 (0.01 to 0.03) |
| 25–36 | 0.69 (0.63 to 0.74) | 0.75 (0.68 to 0.80) | 0.30 (0.20 to 0.41) | 0.94 (0.91 to 0.96) | 0.02 (0.01 to 0.04) |
| 36+ | 0.82 (0.73 to 0.90) | 0.88 (0.67 to 0.96) | 0.67 (0.30 to 0.89) | 0.97 (0.91 to 1.00) | 0.07 (0.03 to 0.15) |
Sensitivity, specificity and net benefit are calculated at model score 0.5, for additional cut-off points see online supplemental appendix Q. In the parentheses, see 95% CI.