Skip to main content

Table 5 Sensitivities, specificities, predictive values, and likelihood ratios for select base excess and hematocrit cutoff levels

From: Base excess and hematocrit predict response to indomethacin in very low birth weight infants with patent ductus arteriosus

Predicting variable

Cutoff level

Sensitivity (95% CI)

Specificity (95% CI)

Positive Predicting Value (95% CI)

Negative Predicting Value (95% CI)

Positive Likelihood Ratio (95% CI)

Negative Likelihood Ratio (95% CI)

Base excess

−6.56

100 (94,100)

44.83 (26,64)

79.49 (69,88)

100 (75,100)

1.81 (1.31,2.52)

0

−4.56a

96.77 (89,100)

79.31 (60,92)

90.91 (81,97)

92 (74,99)

4.68 (2.29,9.55)

0.0407 (.01,.16)

6.56

6.45 (2,16)

100 (88,100)

100 (40,100)

33.33 (24,44)

NA

0.0094 (.88,1.00)

Hematocrit

28

98.39 (91,100)

0 (0, 15)

67.78 (57,77)

0 (0,95)

0.98 (.95,1.02)

NA

40a

69.35 (56,80)

65.52 (46,82)

81.13 (68,91)

50.00 (33,67)

2.01 (1.19,3.41)

0.47 (0.31,0.70)

52

1.61 (0,.09)

96.55 (82,100)

50.00 (1,99)

31.46 (22,42)

0.47 (0.03,7.22)

1.02 (0.98,1.05)

  1. CI confidence interval
  2. aOptimal cutoff level; we have considered the value as the optimal cutoff for which the difference (sensitivity − specificity) is minimum