A modified random-effect procedure for combining risk difference in sets of 2×2 tables from clinical trials · Study