1. Methodological quality of included studies.

Study ID	No. patients	Placebo RCT	Allocation blinded	Recipients blinded	Outcome blinded	Loss of follow‐up	Multi‐centre	Risk of bias
Appleton 1999	40	yes	yes	yes	yes	no	yes	low
Askalan 2003	9	no	not clear	no	some	no	no	high
Baram 1996	29	no	yes	no	yes	none	no	high
Chiron 1997	22	no	not clear	no	no	none	yes	high
Debus 2004	51	yes	yes	yes	no	no	yes	low
Dreifuss 1986	52	no	not clear	no	yes	yes	yes	high
Dyken 1985	17	yes	not clear	yes	yes	yes	no	high
Elterman 2010	221	no	not clear	no	no	yes	yes	high
Hrachovy 1983	24	no	not clear	yes	not clear	none	unclear	high
Hrachovy 1989	24	no	not clear	not clear	not clear	none	unclear	high
Hrachovy 1994	59	no	not clear	no	yes	yes	unclear	high
Lux 2004	107	no	yes	no	no except developmental outcome	yes	yes	high
Bitton 2012	68	yes	not clear	yes	yes	yes	yes	low
Shu 2009	30	no	not clear	not clear	not clear	not clear	not clear	high
Tsai 2009	56	yes	not clear	yes	yes	not clear	not clear	high
Vigevano 1997	42	no	not clear	no	not clear	none	no	high
Yanagaki 1999	26	no	not clear	not clear	not clear	yes	no	high
Zou 2010	38	no	not clear	no	no	no	no	unclear