Figure 1From: Sample size calculations in pediatric clinical trials conducted in an ICU: a systematic review Impact of misspecification of control event rate on total sample size. The sample sizes required to detect RRs with 80% power at a 5% two-sided type I error level are shown. The dotted vertical line indicates the assumed CER. (See text for further explanation). CER, control group event rate; RR, risk ratio.Back to article page