Sample Size
Based on an annual 10% recurrence rate of stroke
(including TIAs) in the usual care group,15 a 60% risk
reduction rate in the intervention group,20 and 2.5 years
of follow-up, the number of subjects required to detect
differences between the intervention group and the
usual care group was calculated to be 136 for each group
(2-tailed 5% significance level, power level of 80%).
Assuming that 10% of the subjects would be lost to
follow-up, the sample size was set at 152 for each group
and 304 for the 2 groups combined.