TABLE 3.

Predictive values of the different joint models for the prediction of CVA positive diagnosis.

Characteristics and variables	AUC	95% CI (AUC)	PCC	Sensitivity %	Specificity %	PPV	NPV	P-value
FENO + FEF25–75%pred	0.762	[0.677, 0.847]	67.14	78.57	62.24	47.14	87.14	0.003
FENO + ΔFEV1%	0.847	[0.778, 0.916]	77.86	88.10	73.47	58.73	93.51	0.076
FENO + ΔFEF25–75%	0.727	[0.631, 0.823]	77.86	50.00	89.80	67.74	80.73	0.160
FEF25–75%pred + ΔFEV1%	0.848	[0.779, 0.916]	76.43	90.48	70.41	56.72	94.52	0.069
ΔFEV1 + ΔFEV1%	0.837	[0.767, 0.907]	74.29	92.86	66.33	54.17	95.59	0.207
ΔFEV1% + ΔFEF25–75%	0.823	[0.751, 0.896]	72.86	90.48	65.31	52.78	94.12	0.575

The abbreviations are as defined for Tables 1, 2. Statistical significance is shown by bold font.