Dysregulation of iron homeostasis in airways associated with persistent preschool wheezing

Respiratory Research

Table 2 ROC analysis of different models for the prediction of persistence of wheezing

Model	AUC (%) (CI)	P value	Sensitivity (CI)	Specificity (CI)	PPV (CI)	NPV (CI)
Iron + clinical predictors	0.84 (0.76–0.92)	< 0.001	0.75 (0.63–0.87)	0.81 (0.68–0.94)	0.84 (0.73–0.95)	0.71 (0.57–0.85)
Iron predictors + mAPI	0.82 (0.72–0.91)	< 0.001	0.98 (0.94–1.00)	0.56 (0.39–0.72)	0.75 (0.64–0.85)	0.95 (0.86–1.04)
Iron predictors	0.75 (0.65–0.86)	< 0.001	0.88 (0.78–0.97)	0.53 (0.37–0.69)	0.71 (0.60–0.83)	0.76 (0.59–0.93)
Clinical predictors	0.72 (0.61–0.84)	< 0.001	0.81 (0.70,0.92)	0.61 (0.45–0.77)	0.74 (0.62–0.86)	0.71 (0.55–0.87)
mAPI	0.67 (0.58–0.78)	0.001	0.54 (0.40–0.68)	0.81 (0.68–0.94)	0.79 (0.65–0.93)	0.57 (0.43–0.71)

AUC area under the curve, CI 95% confidence interval, mAPI modified Asthma Predictive Index, NPV negative predictive value, PPV positive predictive value, ROC receiver operating characteristic

ISSN: 1465-993X