Skip to main content

Table 2 Classification results from plots of metrics for oxygenation, proximal pressure amplitude (ΔPproximal) and tidal volume (Vt) plotted against the continuous distending pressure (CDP) with corresponding sensitivity, specificity, positive and negative predictive value calculated using the respiratory inductance plethysmography–CDP plots as reference, and Fleiss’ Kappa

From: Physiologic responses to a staircase lung volume optimization maneuver in pediatric high-frequency oscillatory ventilation

Variable

Changes observed

LVOM outcome

Sensitivity

Specificity

Positive predictive value

Negative predictive value

ROC AUC (95% CI)

Fleiss

kappa

Responsive

Unresponsive

During lung volume optimization maneuvre

 Oxygenation–CDP

Increase, suggestive for lung volume recruitment

36/41

6/13

87.8

53.8

85.7%

58.3%

0.71 (0.53–0.89)

0.431

No change or decrease

5/41

7/13

 ΔPproximal–CDP

Decrease, suggestive for increased compliance

12/41

3/13

29.3

76.9

80.0%

25.6%

0.53 (0.35–0.71)

0.782

No change or increase

29/41

10/13

 Vt–CDP curve

Increase, suggestive for lung volume recruitment

31/35

10/11

88.6

9.1

75.6%

20.0%

0.49 (0.29–0.68)

0.866

 

No change or decrease

4/35

1/11

After lung volume optimization maneuvre

 SpO2/FiO2 ratio at the end of the maneuvre

Increase ≥ 20%

22/41

5/13

53.7

61.5

81.4%

29.6%

0.58 (0.40–0.75)

N/A

Decrease or increase < 20%

19/41

8/13

 SpO2/FiO2 ratio 1 h after the maneuvre

Increase ≥ 20%

38/41

3/13

92.7

76.9

92.6%

76.9%

0.85 (0.70–0.99)

N/A

Decrease or increase < 20%

3/41

10/13

  1. The lower panel summarizes changes in SpO2/FiO2 ratio at the last applied CDP during the decremental phase (T2) and one hour after (T3) after the lung volume optimization maneuver (LVOM). Data are depicted as median (25–75 interquartile range) or absolute number. ROC AUC receiver operating characteristic area under the curve; 95% confidence interval