Research Article

A Nomogram to Predict Benign/Malignant Mediastinal Lymph Nodes Based on EBUS Sonographic Features

Table 7

Univariate and multivariate logistic regression analyses of the training set for the EBUS-TBNA diagnosed benign cohort.

CharacteristicsUnivariateMultivariate
OR (95% CI) valueOR (95% CI) value

Long-axis2.3 (1.55–3.41)<0.0012.05 (1.14–3.7)0.017
Short-axis
 ≤1Reference
 >12.66 (1.25–5.65)0.0111.2 (0.4–3.61)0.740
Long-axis/short-axis ratio
 <1.5Reference
 ≥1.50.83 (0.39–1.76)0.63
Number of passes per node1.39 (1.03–1.87)0.0321.0 (0.67–1.48)0.998
Echo intensity
 HypoechoicReference
 Hyperechoic1.16 (0.12–11.4)0.899
 Isoechoic2.32 (0.38–14.28)0.364
Echogenicity
 HomogeneousReference
 Heterogeneous6.07 (3.05–12.07)<0.0015.46 (2.33–12.78)<0.001
Margin
 IndistinctReference
 Distinct0.32 (0.11–0.96)0.0420.28 (0.08–1.04)0.057
Blood flow
 LackingReference
 Rich1.54 (0.69–3.43)0.286
Fusion
 AbsentReference
 Present3.56 (1.73–7.33)0.0011.07 (0.39–2.93)0.902
Central hilar structure
 PresentReference
 Absent14.84 (6.49–33.92)<0.00113.11 (4.74–36.27)<0.001