Abstract

Background: Polycystic ovary syndrome (PCOS) has a prevalence of 8–13%. Given the prevalence, diverse health impacts and variation in care, rigorous evidence‐based guidelines are needed in PCOS management. This systematic review with meta‐analyses aimed to investigate the effect of the combined oral contraceptive pill (COCP) and/or metformin in the management of hormonal and clinical features of PCOS, to inform international guidelines.

Methods: Electronic databases were searched systematically from inception until 11 January 2017 to inform the guideline process. Eligible studies were randomized controlled trials which investigated the effect of COCPs and/or metformin alone or combined on hormonal and clinical features in women with PCOS. Outcomes were prioritized as critical for informing a decision about an intervention or important or not important, according to GRADE. Articles were assessed by one author against selection criteria, in consultation with a second author. Data were double extracted independently by four authors, and data quality appraisal was completed. Meta‐analyses were conducted, where appropriate.

Results: Fifty‐six studies were eligible for inclusion. Outcomes prioritized by women and health professionals included the following: irregular cycles, insulin resistance, weight, BMI, thromboembolic events and gastrointestinal effects. In low‐quality evidence in adolescents, meta‐analyses demonstrated that metformin was better than COCP for BMI (mean difference [MD] −4.02 [−5.23, −2.81], P < 0.001); COCP was better than metformin for menstrual regulation (MD −0.19 [−0.25, −0.13], P < 0.00001). In low‐quality evidence in adults, meta‐analyses demonstrated that metformin was better than placebo for BMI (MD −0.48 [−0.94, −0.02], P = 0.04); metformin was better than COCP for fasting insulin (MD 4.00 [2.59, 5.41], P = 0.00001), whereas COCP was better than metformin for irregular cycles (MD 12.49 [1.34, 116.62], P = 0.03). Combined oral contraceptive pill alone was better than the combination with an anti‐androgen for BMI (MD −3.04 [−5.45, −0.64], P = 0.01). Metformin was associated with generally mild gastrointestinal adverse events. Differences in statistical significance were observed when outcomes were subgrouped by BMI.

Conclusions: This review identified that COCP therapy has benefits for management of hyperandrogenism and menstrual regulation. Metformin combined with the COCP may be useful for management of metabolic features. There is minimal evidence of benefits of adding an anti‐androgen to COCP therapy. Metformin alone has benefits for adult women for management of weight, hormonal and metabolic outcomes, especially for women with BMI ≥ 25 kg/m². There is inadequate evidence to suggest the optimal COCP formulation, or dosing regimen and formulation of metformin.