Table 5. Influence of in- or exclusion of studies on estimated effects.

			RR (95% CrI)
	Number of studies	Number of patients	Ischaemic Stroke	ICH	Death	Ischaemic Stroke or TIA
Our meta-analysis, all included studies	37	1991	0.29 (0.08−0.77)	0.00 (0.00−0.05)	0.69 (0.23−1.99)	0.59 (0.19−1.45)
Engelter et al's meta-analysis, all included studies	36	1285	0.45 (0.15−1.41)	0.00 (0.00−0.36)	1.40 (0.46−3.79)	n/a
Menon et al's meta-analysis, all included studies	34	762	0.51 (0.13−1.62)	n/a	n/a	1.48 (0.41−4.33)
Engelter et al's versus our meta-analysis
Studies included in Engelter et al's and our meta-analysis	20	1109	0.64 (0.18−2.14)	0.00 (0.00−0.15)	1.80 (0.38−9.15)	n/a
Studies included in Engelter et al's but not in our meta-analysis	16	176	0.45 (0.15−1.41)	0.00 (0.00−26.68)	0.96 (0.13−4.32)	n/a
Studies included in our, but not in Engelter et al's meta-analysis	17	882	0.19 (0.02−0.97)	0.00 (0.001−0.003)	0.33 (0.03−1.80)	n/a
Menon et al's versus our meta-analysis
Studies included in Menon et al's and our meta-analysis	18	553	0.48 (0.10−1.76)	n/a	n/a	2.00 (0.44−7.52)
Studies included in Menon et al's but not in our meta-analysis	16	209	0.65 (0.02−13.72)	n/a	n/a	0.76 (0.10−4.05)
Studies included in our, but not in Menon et al's meta-analysis	19	1438	0.15 (0.02−0.70)	n/a	n/a	0.41 (0.14−1.03)

Outcomes included if reported in at least two out of the three available meta-analyses.

ICH = intracranial haemorrhage; TIA = transient ischaemic attack.