Assuming that UAI increases risk of HIV transmission by approximately 80% with a significance level of 0.05 and a power of 80%, the minimum necessary sample size was 150 men[13]. Rounding up, the planned sample size was 200 for each of the three sites for a total of 600 men.