Skip to main content

Table 5 Summary of ROC curve parameters for predicting ICU admission

From: Characteristics of lymphocyte subsets and cytokine profiles of patients with COVID-19

Predictor

AUC

P value

95% CI

Cutoff value

Jordan index

Sensitivity (%)

Specificity

Age

0.766

 < 0.001

0.689–0.844

 > 53.50

0.437

69.8

73.9

Diabetes mellitus

0.626

0.006

0.532–0.720

 > 0.50

0.252

30.2

95.0

Lymphocyte

0.780

 < 0.001

0.709–0.851

 < 840.00

0.456

73.6

72.0

CD3+

0.806

 < 0.001

0.737–0.874

 < 510.50

0.512

71.7

79.5

CD4+

0.776

 < 0.001

0.705–0.847

 < 357.50

0.426

77.4

65.2

CD8+

0.796

 < 0.001

0.721–0.872

 < 173.50

0.505

67.9

82.6

CD19+

0.616

0.012

0.533–0.699

 < 148.00

0.196

81.1

38.5

NK

0.639

0.002

0.553–0.725

 < 134.50

0.270

71.7

55.3

IL-4

0.526

0.573

0.433–0.618

 < 1.27

0.096

26.4

83.2

IL-6

0.785

 < 0.001

0.705–0.864

 > 6.58

0.519

77.4

74.5

IL-10

0.760

 < 0.001

0.681–0.838

 > 2.95

0.494

77.4

72.0

IL-17

0.539

0.393

0.451–0.627

 < 1.27

0.096

69.8

39.8

TNF-α

0.539

0.390

0.452–0.626

 < 8.40

0.099

96.2

13.7

IFN-γ

0.548

0.295

0.460–0.636

 > 7.73

0.116

37.7

73.9

Combined predictor

0.887

 < 0.001

0.837–0.937

 > 0.309

0.631

77.4

85.7

  1. ROC receiver operating characteristic, ICU intensive care unit, NK natural killer, IL interleukin, TNF tumor necrosis factor, IFN interferon