A Bioclinical Pattern for the Early Diagnosis of Cardioembolic Stroke

Table 3

Results of multiple logistic regression model for the likelihood of a stroke of cardioembolic origin among 114 patients. Odds ratios (OR) and 95% confidence intervals (95% CI) are estimated from multiple logistic regression models including and excluding atrial fibrillation (AF).

	Cardioembolic stroke		Model with AF			Model without AF
	No	Yes	OR*	95% CI	𝑃 value	OR*	95% CI	𝑃 value
Sex
Female	39	12	1.00	Reference		1.00	Reference
Male	41	22	1.31	0.38–4.54	0.67	1.36	0.47–3.93	0.57
Age (years)	80	34	0.96	0.90–1.02	0.16	1.00	0.95–1.05	0.92
Baseline NIHSS score	80	34	1.10	0.99–1.22	0.09	1.14	1.04–1.24	0.005
NT-proBNP (pg/mL)
<200	66	12	1.0	Reference		1.0	Reference
≥200	14	22	3.96	1.02–15.36	0.047	5.90	2.03–17.14	0.01
Globulin/albumin
<0.70	25	3	1.0	Reference		1.0	Reference
≥0.70	55	31	3.84	0.70–20.98	0.12	4.13	1.00–17.06	0.05
Atrial fibrillation
No	70	6	1.00	Reference
Yes	10	28	24.67	6.46–94.22	<0.001
Area under ROC curve			0.91			0.84

∗Each variable is adjusted for the others.