Chen 2016.

Study characteristics
Methods	Multicentre RCT Conducted: in 14 reproductive medical centres throughout China Enrolment: June 2013‐May 2014 Power calculation: stated Randomisation: an online central randomisation system (www.medresman.org) was used Timing randomisation: at day of oocyte retrieval Nature of intervention: day‐3 embryo cryopreservation by means of vitrification. Local investigators had the option to transfer day‐2 embryos if there were < 3 embryos on day 2 Follow‐up: cumulative live birth (including all FETs performed within 12 months after the initial transfer)
Participants	1508 women (746 freeze‐all, 762 control) Inclusion criteria: PCOS, using modified Rotterdam criteria (which included menstrual abnormalities (irregular uterine bleeding, oligomenorrhoea, or amenorrhoea) combined with either hyperandrogenism or polycystic ovaries) first IVF cycle Exclusion criteria: history of unilateral oophorectomy, recurrent spontaneous abortion (defined as ≥ 3 previous spontaneous pregnancy losses), congenital or acquired uterine malformations, abnormal results on parental karyotyping, or medical conditions that contraindicated ART or pregnancy
Interventions	For women who were assigned to the fresh embryo group, on day 3, 2 high‐quality embryos were picked out for fresh transfer and supernumerary embryos were transferred by means of vitrification. For women who were assigned to the FET group, there was no fresh transfer as all day‐3 embryos were cryopreserved for later transfer. Local investigators had the option to transfer day‐2 embryos if there were < 3 embryos on day 2. In cycles following the menstrual cycle with ovum pick‐up, after artificial endometrial preparation, on day 4 of the progesterone regimen, 2 day‐3 frozen embryos were thawed and transferred.
Outcomes	Primary outcome was a live birth, defined as delivery of any viable infant at ≥ 28 weeks of gestation during the first ET. Prespecified secondary outcomes included biochemical pregnancy, clinical pregnancy, ongoing pregnancy, singleton LBR, cLBR (including subsequent FET), pregnancy loss, moderate or severe OHSS, ectopic pregnancy, pregnancy and neonatal complications, and congenital anomalies.
Notes	Funding: supported by a grant from the National Basic Research Program of China, by grants from the National Natural Science Foundation of China, and by grants from the Thousand Talents Program (to Drs. Legro and H. Zhang).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	An online central randomisation system (www.medresman.org) was used to automatically generate the assignment sequence
Allocation concealment (selection bias)	Low risk	Assignment sequence was unknown to the clinical investigators
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding of doctors and participants was not possible due to the nature of the intervention.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessor blinding was not reported, however primary outcome is not likely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Data were analysed for all randomised women.
Selective reporting (reporting bias)	High risk	Some prespecified outcomes (e.g. time to pregnancy) were missing from the report
Other bias	Unclear risk	Not reported on blinding of doctors to interim analyses of outcomes of the study. Blinding of investigators was not reported (which is relevant for determining end of study).