Since the publication of the systematic review by Wang and colleagues in 2017 (BMJ 2017;356:j138, doi:10.1136/bmj.j138, published 31 January 2017), the authors now have substantial concerns about three included articles, including one confirmed to be a non-randomised study and another two with integrity concerns (both retracted). The details of these three articles are described in two earlier Rapid Responses (https://www.bmj.com/content/356/bmj.j138/rapid-responses). After consulting The BMJ’s editorial team, the authors have reanalysed the data by excluding the three articles. As a result, 54 trials reporting on 7173 women were included in the revised version. The reanalyses include four outcomes (pregnancy, ovulation, multiple pregnancy and miscarriage). Across all comparisons and outcomes, the odds ratios after reanalyses were similar overall, and the confidence intervals were similar or slightly wider. However, the difference between letrozole and clomiphene on multiple pregnancy and the difference between the combination of clomiphene and metformin and clomiphene alone on ovulation…

Leave a Reply

Your email address will not be published. Required fields are marked *