Peeraer 2015.

Methods	Single‐centre, 2‐arm parallel, open‐label RCT
Participants	472 women undergoing FET Inclusion criteria: women undergoing FET were eligible for the study when they had a regular cycle (between 21 and 35 days) and were aged 21‐45 years Exclusion criterion: FET after PGD Participants were similar in demographic and infertility characteristics at baseline
Interventions	Natural cycle (235 women) Women underwent a first pelvic US and blood analysis around day 10‐day 12 of the menstrual cycle. HCG administered when the leading follicle had a mean diameter of ≥ 17 mm and endometrial thickness ≥ 7 mm with serum estradiol levels preferably 150 ng/L. HMG cycle (237 women) Women started SC injections of gonadotrophins (follitropin plus LH) on day 2 of the menstrual cycle. Starting dose of gonadotrophins (37 or 75 IU) determined by the treating clinician, based on woman's age, BMI, basal serum FSH (days 2‐5) and (if applicable) response to previous ovarian stimulation. On day 6 or 7 of the menstrual cycle, a first US and serum hormonal analysis (E2, progesterone, LH, FSH) performed. Based on these results, dose of gonadotrophins could be adjusted if needed. In both natural cycle FET and HMG FET cycle groups, the follicular response was monitored by regular vaginal US and serum hormonal analysis. ET was performed the same way in both groups.
Outcomes	Live birth per ET Ongoing pregnancy per ET Clinical pregnancy per ET Endometrial thickness
Notes	Outcome data reported as per 'embryo transfer cycle' (dichotomous data) or not clearly stated and there were multiple transfers per woman
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Process of randomization not sufficiently explained.
Allocation concealment (selection bias)	Low risk	Allocations concealed in opaque sealed envelope.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Described as an open‐label trial, thus both participants and personnel were aware of the treatment protocols. However, non‐blinding of outcome assessors may not have affected some of the outcome measures as they were objectively assessed.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Analysis was based on per cycle transferred.
Selective reporting (reporting bias)	Low risk	Outcome measures were prespecified in the methods section.
Other bias	Low risk	Participants similar in demographic and infertility characteristics at baseline.