Five RCTs were included in the review, but only 4 RCTs were included in the meta-analyses (273 women with a total of 921 cycles). The number of patients or cycles in the fifth RCT was not reported.
In terms of study quality, 3 studies described the method of treatment allocation, one described adequate methods of allocation concealment, and one performed a power calculation.
Ovulation induction (4 studies): there was no statistically significant difference in ovulation between tamoxifen and clomiphene (OR 0.7555, 95% CI: 0.513, 1.111). No statistically significant heterogeneity was detected (P=0.141).
Pregnancy rate (3 studies with 743 cycles including 504 anovulatory cycles): there was no statistically significant difference between tamoxifen and clomiphene, either in pregnancies per cycle (OR 1.056, 95% CI: 0.583, 1.912) or pregnancies per anovulatory cycle (OR 1.162, 95% CI: 0.632, 2.134). No statistically significant heterogeneity was detected (P=0.815).
Live-birth rate (1 study, n=40 women): there was no statistically significant difference between tamoxifen and clomiphene in live births per cycle (OR 0.261, 95% CI: 0.005, 2.711).