Table 4.

Bivariate and multivariate analysis of primary endpoint (recurrent ipsilateral ischemic stroke or retinal artery occlusion) with Cox regression

	Bivariate analysis		Multivariate analysis
	HR (95% CI)	p value	HR (95% CI)	p value
Conventional ≥ 50% stenosis	1.0	Ref	1.0	Ref
Near-occlusion without full collapse	1.5 (0.7–3.4)	0.28	2.0 (0.9–4.5)	0.09
Near-occlusion with full collapse	2.7 (1.4–5.4)	0.005	2.6 (1.3–5.3)	0.006
Age (10-year increment)	1.1 (0.8–1.6)	0.49	1.2 (0.9–1.8)	0.26
Male sex	0.9 (0.5–1.6)	0.66	Not used	–
Previous myocardial infarction	1.1 (0.5–2.2)	0.78	Not used	–
Current angina	1.4 (0.7–2.8)	0.41	Not used	–
Heart failure	0.9 (0.3–2.8)	0.81	Not used	–
Current intermittent claudication	1.2 (0.4–3.4)	0.71	Not used	–
Previous arterial revascularization	1.4 (0.7–2.8)	0.28	Not used	–
Atrial fibrillation	1.8 (0.8–3.8)	0.14	Not used	–
Current smoking	0.8 (0.4–1.8)	0.64	Not used	–
Diabetes	0.7 (0.4–1.5)	0.41	Not used	–
Hypertension	2.6 (0.6–10.8)	0.18	2.6 (0.6–11.0)	0.19
Previous stroke	1.0 (0.5–2.3)	0.98	Not used	–
Ipsilateral event < 14 days before presenting event	0.7 (0.3–1.6)	0.41	0.8 (0.3–2.0)	0.69
Cerebrala presenting event	7.8 (1.1–56.4)	0.04	7.0 (0.96–51.5)	0.06
Sought health care at other hospital	0.9 (0.5–1.8)	0.85	Not used	–

All co-variates with p < 0.1 in bivariate analysis or in baseline analysis used in the multivariate model

HR hazard ratio

^aStroke and TIA merged to “cerebral” event, compared to amaurosis fugax and retinal artery occlusion merged to “retinal” event: stroke and TIA had similar high risk of primary endpoint (p = 0.82), whereas only one case with retinal artery occlusion and none with amaurosis fugax reached the primary endpoint