TABLE 4.
Predicted % probabilities and odds ratios (ORs, relative benefits) of relapse in 2 years, using baseline risk scores developed with the LASSO (least absolute shrinkage and selection operator) and prespecified models
| Benefits | Model | Treatment | All patients | Baseline risk <30% Low‐risk patients | Baseline risk >50% High‐risk patients |
|---|---|---|---|---|---|
| Absolute benefits (%) | LASSO | Dimethyl Fumarate | 62% | 18% | 93% |
| Glatiramer Acetate | 64% | 23% | 93% | ||
| Natalizumab | 54% | 20% | 82% | ||
| Prespecified | Dimethyl Fumarate | 53% | 20% | 84% | |
| Glatiramer Aceta | 56% | 23% | 86% | ||
| te Natalizumab | 46% | 23% | 69% | ||
| Relative benefits (OR) | LASSO | Dimethyl Fumarate vs placebo | 0.52 | 0.25 | 0.81 |
| Glatiramer Acetate vs placebo | 0.57 | 0.35 | 0.81 | ||
| Natalizumab vs placebo | 0.29 | 0.29 | 0.28 | ||
| Prespecified | Dimethyl Fumarate vs placebo | 0.42 | 0.31 | 0.53 | |
| Glatiramer Acetate vs placebo | 0.50 | 0.38 | 0.63 | ||
| Natalizumab vs placebo | 0.31 | 0.40 | 0.23 |
Note: Results are shown for all patients, for low‐risk patients (baseline risk <30%) and for high‐risk patients (baseline risk >50%) in the observed range of baseline risk. The cut‐offs have been chosen arbitrarily for illustrative purposes.