Table 2.

Univariate and multivariable logistic regression model for predictors of bleeding

Variable	Univariable logistic regression		Multivariable logistic regression
	OR (95% CI)	P	OR (95% CI)	P
sGPVI
<43 ng/mL	1 (Ref)	—	1 (Ref)	—
≥43 ng/mL	4.59 (2.65-7.96)	<.001	2.81 (1.51-5.23)	.001
ICU status*
No	1(Ref)	—	1 (Ref)	—
Yes	4.18 (2.37-7.36)	<.001	2.76 (1.46-5.21)	.002
Sepsis†
No	1 (Ref)	—	1 (Ref)	—
Yes	2.89 (1.64-5.10)	<.001	2.27 (1.13-4.56)	.021
Cardiopulmonary bypass surgery
No	1 (ref)	—	1 (Ref)	—
Yes	1.79 (1.04- 3.07)	.036	1.50 (0.76-2.93)	.238
Platelet count, ×109/L‡
≥151	1 (Ref)	—	1 (Ref)	—
150-100	1.44 (0.44-4.67)	.542	0.89 (0.25- 3.22)	.858
50-100	1.26 (0.42-3.74)	.680	0.86 (0.26- 2.81)	.801
≤49	3.58 (1.20-10.66)	.222	1.72 (0.52- 5.66)	.376

Ref, reference.

^*Hospitalized in a medical or surgical critical care unit at time of enrollment in study.

^†International Classification of Diseases, Ninth Revision, code for sepsis during hospitalization sGPVI level as drawn.

^‡Platelet count on day of sGPVI blood draw.