Metformin versus clomiphene citrate for infertility in non-obese women with polycystic ovary syndrome: a systematic review and meta-analysis.

UNLABELLED: BACKGROUND Recent studies suggest that metformin may be more effective in women with polycystic ovary syndrome (PCOS) who are non-obese. The objective here is to determine and compare the effectiveness of metformin and clomiphene citrate for improving fertility outcomes in women with PCOS and a BMI < 32 kg/m(2) (BMI 32 kg/m(2) was used to allow for international differences in BMI values which determine access to infertility therapy through the public health system). METHODS Databases were searched for English language articles until July 2011. INCLUSION CRITERIA: women of any age, ethnicity and weight with PCOS diagnosed by all current criteria, who are infertile; at least 1000 mg of any type of metformin at any frequency, including slow release and standard release, compared with any type, dose and frequency of clomiphene citrate. OUTCOMES: rates of ovulation, live birth, pregnancy, multiple pregnancies, miscarriage, adverse events, quality of life and cost effectiveness. Data were extracted and risk of bias assessed. A random effects model was used for meta-analyses of data, using risk ratios (relative risk). RESULTS The search returned 4981 articles, 580 articles addressed metformin or clomiphene citrate and four randomized controlled trials (RCTs) comparing metformin with clomiphene citrate were included. Upon meta-analysis of the four RCTs, we were unable to detect a statistically significant difference between the two interventions for any outcome in women with PCOS and a BMI < 32 kg/m(2), owing to significant heterogeneity across the RCTs. CONCLUSIONS Owing to conflicting findings and heterogeneity across the included RCTs, there is insufficient evidence to establish a difference between metformin and clomiphene citrate in terms of ovulation, pregnancy, live birth, miscarriage and multiple pregnancy rates in women with PCOS and a BMI < 32 kg/m(2). However, a lack of superiority of one treatment is not evidence for equivalence, and further methodologically rigorous trials are required to determine whether there is a difference in effectiveness between metformin and placebo (or no treatment) or between metformin and clomiphene citrate for ovulation induction in women with PCOS who are non-obese. Until then, caution should be exercised when prescribing metformin as first line pharmacological therapy in this group of women.