Table 4.

Results of individual datasets

Article	DS	RR	95% CI	p-value
Welbury et al., 1991 [20]	01	0.64	0.27 - 1.51	0.31
Östlund et al., 1992 [21]	02	0.73	0.03 - 16.47	0.84
Taifour et al., 2002 [22]	03	0.57	0.24 - 1.36	0.21
	04	0.35	0.11 - 1.10	0.07
Mandari et al., 2003 [23]	05	0.18	0.05 - 0.59	0.005*
Frencken et al., 2007 [24]	06	0.52	0.25 - 1.08	0.08
Daou et al., 2009 [25]	07	4.36	0.51 - 37.09	0.18
	08	0.91	0.21 - 4.04	0.90
Mandari et al., 2001 [27]	09	0.56	0.25 - 1.24	0.15
Yu et al., 2004 [28]	10	Not estimable
	11	Not estimable
Svanberg, 1992 [29]	12	0.14	0.01 - 2.53	0.18
	13	0.33	0.12 - 0.91	0.03*
Qvist et al., 1997 [26]	14	0.59	0.28 - 1.25	0.17
	15	0.48	0.32 - 0.71	0.0003*
	16	0.57	0.38 - 0.87	0.0009*
	17	0.91	0.73 - 1.13	0.38

DS = Dataset number; RR = Relative risk; CI = Confidence interval; Not estimable = data from both treatment groups are essentially the same: p = 1.00.

* Statistically significant difference, in favour of GIC.