Differentiating interstitial lung diseases from other respiratory diseases using electronic nose technology

Respiratory Research

Table 3 Results of breath analysis in subgroups

Group 1		Group2	n =	Dataset	AUC	95% CI	Specificity	Sensitivity	Accuracy	NPV	PPV
ILD (ever smoking)	75	Asthma–COPD–Lung cancer (ever smoking)	72	Training	0.99	0.99–1.00	0.99	0.96	0.97	0.96	0.99
	37		36	Test	0.94	0.89–0.99	0.86	0.95	0.90	0.94	0.88
Never smoker	96	Ever smoker	223		0.66	0.60–0.73
Current smoker	30	Former smoker	193		0.80	0.73–0.87
Male sex	168	Female sex	154		0.67	0.61–0.73
Hospital EMC	254	Hospital FGV	73		0.64	0.53–0.74

Results of the cohort that is split in a training and test set are based on 2 principal components; results of the unsplit cohort are based on 1 principal component. ^Includes asthma and chronic obstructive pulmonary disease patients only, as the number of patients with interstitial lung disease and lung cancer were too small in medical center FGV. AUC: area under the curve; CI: confidence interval; COPD: chronic obstructive pulmonary disease; EMC: Erasmus Medical Center; FGV: Franciscus Gasthuis & Vlietland; ILD: interstitial lung disease; NPV: negative predictive value; PPV: positive predictive value

ISSN: 1465-993X