Table 2.
Non-parametric bootstrap analysis and parameter estimates of the final model for Gd+ lesion count (Model 5 mixture negative binomial model)
| Model parameter | Description | Point estimate | Non-parametric bootstrap (500 replicates) | |
|---|---|---|---|---|
| Median | 95 % CI | |||
| μ 1 | Baseline mean Gd+ lesion count for a typical subject in the lower baseline lesion activity sub-population | 0.48 | 0.474 | (0.382, 0.609) |
| R | Ratio of the mean Gd+ lesion count for a typical subject in higher baseline lesion activity group to lower baseline lesion activity group | 3.53 | 3.58 | (2.58, 4.73) |
| μ 2 | Baseline mean Gd+ lesion count for a typical subject in the higher baseline lesion activity sub-population | 1.69 | 1.70 | |
| r1 | Dispersion parameter for baseline λ in the lower baseline lesion activity sub-population | 44.8 | 44.7 | (39.6, 50.8) |
| r2 | Dispersion parameter for baseline λ in the higher baseline lesion activity sub-population | 0.499 | 0.496 | (0.398, 0.586) |
| θ | Proportion of subjects with lower baseline lesion activity | 0.602 | 0.602 | (0.561, 0.651) |
| β | Slope of AUC effect on log(λ) | −0.0256 | −0.0257 | (−0.0304, −0.0216) |
| t1/2 | Half-life of drug effect onset time (days) | 115 | 113.8 | (73.8, 179.6) |
| σ2 | Variance of random effect on baseline λ in log scale for the higher baseline lesion activity sub-population | 1.25 | 1.22 | (1.00, 1.46) |