Fig. 3.

The mean number of new Gd⁺ lesions between weeks 0 and 52 and upper 95 % CIs for patients with highly active RRMS or less active RRMS who were treated with DAC HYP or placebo. Highly active RRMS was defined as at least two relapses in the year prior to randomization and at least one Gd⁺ lesion at baseline, all other patients comprised the less active RRMS subgroup. p-values and percentage reduction were estimated from an ordinal logistic regression model adjusted for baseline lesion count in each disease activity subgroup. CI confidence interval, DAC HYP daclizumab high-yield process, Gd ⁺ gadolinium-enhancing, RRMS relapsing-remitting multiple sclerosis. ^a Percentage reductions represent the reduction over placebo in the risk of having greater Gd⁺ lesion activity