Volume 21, Number 1—January 2015
Workshop on Use of Intravenous Immunoglobulin in Hand, Foot and Mouth Disease in Southeast Asia
|Absolute risk of progression†||Noninferiority margin Δ‡
*Sample sizes are total sample sizes (i.e., for both groups combined) but do not account for loss-to-follow-up or increases in sample size that would be required if formal interim analyses with stopping boundaries are planned. IVIg, intravenous immunoglobulin.
†Assumed to be identical for both groups.
‡The noninferiority margin Δ is the absolute risk increase in the placebo arm to be exclude with the trial. That is, if the assumptions are correct, the sample sizes above guarantees that we can reject the hypothesis that placebo increases the absolute risk by +Δ (or more) at the 1-sided 2.5% significance level with 90% power. (Or, equivalently, the sample sizes above guarantee a probability of 90% that the 2-sided 95% CI for the absolute risk difference excludes an excess risk of +Δ, at worst, in the placebo arm.)