Mokhtari 2019.

Study characteristics
Methods	Double‐blinded randomised clinical trial ‐ parallel
Participants	Women with PCOS attending an infertility clinic for IUI (N = 198) Inclusion criteria: Women with PCOS aged between 20 to 40 years, having husbands with normal spermograms, having normal hysterosalpingography, having the Rotterdam diagnostic criteria for PCOS, no underlying endocrine diseases, and using no hormonal drugs within the past 3 months Exclusion criteria: Being deficient in an adequate ovarian response, suffering from ovarian hyperstimulation syndrome, and no history of treatment for infertility
Interventions	1. Melatonin tablet 3 mg (Nature Made, USA) Mean age 28.4 ± 5.5 (N = 98) 2. Placebo (made by Faculty of Pharmacy, Tehran University of Medical Sciences); mean age 29.3 ± 5.6 (N = 100) Treatment from the 3rd day of menstruation until the day of hCG administration
Outcomes	Chemical pregnancies Endometrial thickness (ET) on the day of IUI Clinical pregnancy
Notes	Location: P.O.Box: 1598718311, Department of Obstetrics and Gynecology, IVF Unit, Yas Hospital, Tehran University of Medical Sciences, Tehran, Iran Timeframe: from March 2017 to September 2017 Trial registration number: IRCT2017021132489N1 Informed consent: yes Ethical approval: yes Sample size power calculation: yes Funding:There is no financial support Conflict of interests:There is no conflict of interests ITT: no Author email: azmoudeh@sina.tums.ac.ir
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "the participants were divided into an intervention group receiving melatonin and a control group receiving placebo using a balanced block randomization technique. Considering blocks of 4 in this study, the Stata software was used to generate random‐number sequences from 1 to 6 until the desired sample size was achieved. Since the total number of modes to set two people in the blocks of 4 was 6 modes, if the generated number exceeded 6, the next number was regenerated regardless of the previous number".
Allocation concealment (selection bias)	Low risk	Quote: "Preparing the random allocation sequences of the participants, putting them in sealed airtight envelopes, and numbering them with a five‐digit serial number were all performed by a third person who was not involved in the study design. All the envelopes (n=188) having a random 5‐digit serial number were opened immediately after the completion of basic information and examination of the participants. Then, the participants were assigned to the intervention or control groups".
Blinding (performance bias and detection bias) All outcomes	Low risk	Quote: "double‐blinded randomized" placebo controlled trial
Incomplete outcome data (attrition bias) All outcomes	Low risk	Dropouts and reasons explained
Selective reporting (reporting bias)	Low risk	Trial registration number provided and clinical pregnancy reported
Other bias	Low risk	No other bias found