Table 2. Risk of bias of included studies.

		Selection bias		Performance and detection bias		Attrition bias		Reporting bias		Other potential bias
	study	risk	explanation	risk	explanation	risk	explanation	risk	explanation	risk	explanation
Oocyte-based review	Nakahara 2010	U	Method of random allocation was not described.	H	Laboratory technicians were not blinded.	L	No loss of follow-up.	L	All investigated outcomes were published.	L	None.
	Cruz 2011	U	Method of random allocation was not described.	U	Blinding unclear.	L	No loss of follow-up.	L	All investigated outcomes were published.	L	None.
	Kirkegaard 2012	L	Block randomization using random number from sealed envelopes.	L	Laboratory technicians were blinded.	H	Large proportion of oocytes dropped out after randomization.	L	All investigated outcomes were published.	L	None.
	Wu 2016 (Part B)	U	Method of random allocation was not described.	H	Laboratory technicians were not blinded.	L	No loss of follow-up.	H	Did nor report pregnancy or implantation.	L	None.
Woman-based review	Park 2015	L	A web based randomization programme was used.	H	Laboratory technicians were not blinded.	L	Very small proportion of women dropped out.	L	All investigated outcomes were published.	L	None.
	Wu 2016 (Part A)	L	Computer randomization was used.	H	Laboratory technicians were not blinded.	H	Large proportion of women dropped out after randomization.	L	All investigated outcomes were published.	L	None.
	Kahraman 2013	L	A computer generated list was used.	H	Laboratory technicians were not blinded.	L	Small proportion of women dropped out and reasons were clearly stated.	L	All investigated outcomes were published.	L	None.
	Insua 2015	H	Patient were able to request the intervention.	H	Patients were not blinded.	H	Loss of follow up was not balanced.	H	The authors did not report clinical pregnancy and the primary outcome (ongoing pregnancy) was not assessed as reported in the trial register	H	IVI was a part owner of and had a long-standing financial interest with Fertilitech which only recently was sold. Fertilitech is the firm that manufactures EmbryoScope
	Matyas 2015	H	Paired randomization by two envelopes, researchers were awared of allocation before enrollment.	H	Laboratory technicians were not blinded.	H	Large proportion of women (20/140) dropped out.	L	All investigated outcomes were published.	L	None.
	Goodman 2016	L	A computer-generated random number sequence was used.	H	Laboratory technicians were not blinded.	H	Large proportion of women (52/287) dropped out.	L	All investigated outcomes were published.	L	None.

L, low risk of bias; H, high risk of bias; U, unclear risk of bias.