Heathcote 2013.

Methods	RCT Setting: Australia Method of randomisation: computer‐generated random number sequence performed by the hospital pharmacy department to ensure allocation concealment. Blinding: double blinding Number randomised: 27
Participants	Summary: metformin and CC vs CC and placebo Inclusion criteria: PCOS (Rotterdam criteria) Exclusion criteria: other causes of anovulation (hyperprolactinaemia, pituitary disease, hypothyroidism, congenital adrenal hyperplasia), diabetes mellitus, male factor subfertility Baseline characteristics of each group metformin and CC (n = 11) vs CC and placebo (n = 12): Mean age, years (SD) 29.3 (4.7) vs 28.7 (4.7) Mean BMI, kg/m2 (SD) 31.0 (7.1) vs 30.8 (6.1) CC‐resistant 4 (36.4%) vs 2 (16.7%) Dropouts: 4 (2 from metformin group and 2 from placebo group)
Interventions	Main intervention: Group 1: metformin 500 mg 3/d vs matched placebo Duration: 6 cycles Co‐interventions: CC 50 mg/d from day 3‐7. Women with BMI > 30 kg/m2 were referred to dietician
Outcomes	Primary: live birth rate Secondary: ovulation rate: serum progesterone > 10.6 nmol/L, miscarriage, gastrointestinal side effects
Notes	Unpublished paper with permission granted from the study authors (23 April 2019) to use the data for the review. The paper was supplied by the study authors.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer random‐number generator
Allocation concealment (selection bias)	Low risk	Pharmacy controlled allocation and dispensing
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and personnel were blinded
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Investigators were blinded
Incomplete outcome data (attrition bias) All outcomes	High risk	Dropouts stated with incomplete reasoning. 1 withdrew and 4 did not complete 6 cycles from the CC and placebo group, 2 withdrew and 4 did not complete 6 cycles from CC and metformin group
Selective reporting (reporting bias)	Low risk	The primary and secondary outcomes were clearly reported
Other bias	High risk	Not an ITT analysis; paper has not been peer reviewed