Usefulness of Noninvasive Predictors of Oesophageal Varices in Black African Cirrhotic Patients in Côte d'Ivoire (West Africa)

Table 2

Logistic regression of factors predicting the presence of oesophageal varices or large oesophageal varices.

	beta (SE)	Adjusted OR (95% CI)	𝑃
Factors predicting OV
Gender	−2.5 (0.7)		0.003
Male		1
Female		0.1 (0.02–0.3)
Platelet count (cell/mm^3)	2.5 (0.7)		0.0003
≥93000		1
<93000		12.4 (3.2–47.7)
Spleen size (mm)	0.04 (0.01)	1.04 (1.01–1.07)	0.002
Factors predicting large OV
Gender	−2.1 (0.6)		0.001
Male		1
Female		0.1 (0.03–0.4)
Platelet count (cell/mm^3)	1.4 (0.5)		0.01
≥93000		1
<93000		4.0 (1.4–11.72)
Spleen size (mm)	0.04 (0.01)	1.05 (1.02–1.1)	<0.0001
Child-Pugh score			<0.0001
Class A		1
Class B	−1.6 (0.8)	0.2 (0.04–0.9)
Class C	−0.7 (0.8)	0.5 (0.1–2.3)

Intercept of model predicting OV = −4.2, C-index: 0.878, 𝑅² = 0.31, 𝑃<0.0001.
Intercept of model predicting LOV = −4.2, C-index: 0.850, 𝑅² = 0.30, 𝑃<0.0001.
Beta: coefficient estimates, OR: odds ratio, SE: standard error, CI: confidence interval.
Platelet count dichotomized by the median value because of large value of coefficient estimates of platelet count (−10^-5) when used as continuous variable not corrected by the exact method, Logistic regression analysis was computed on 110 patients due to one missing value in those with no OV.