A Nomogram for Predicting Cardiovascular Diseases in Chronic Obstructive Pulmonary Disease Patients

Table 6

Subgroup analysis of the predictive ability of the prediction model.

Subgroup	Characteristic	Testing set (95%CI)
COPD diagnosed by spirometry
	AUC	0.69 (0.53–0.85)
	Accuracy	0.83 (0.75–0.88)
	Sensitivity	0.38 (0.14–0.61)
	Specificity	0.88 (0.82–0.93)
	NPV	0.92 (0.87–0.97)
	PPV	0.27 (0.09–0.46)
COPD diagnosed from the questionnaire
	AUC	0.72 (0.64–0.80)
	Accuracy	0.63 (0.55–0.71)
	Sensitivity	0.70 (0.60–0.81)
	Specificity	0.61 (0.51–0.72)
	NPV	0.70 (0.60–0.81)
	PPV	0.57 (0.45–0.68)
Smoking (yes)
	AUC	0.79 (0.73–0.85)
	Accuracy	0.73 (0.67–0.79)
	Sensitivity	0.65 (0.54–0.76)
	Specificity	0.76 (0.70–0.83)
	NPV	0.84 (0.78–0.90)
	PPV	0.54 (0.44–0.64)
Smoking (no)
	AUC	0.64 (0.45–0.83)
	Accuracy	0.72 (0.59–0.83)
	Sensitivity	0.22 (0.01–0.49)
	Specificity	0.82 (0.71–0.92)
	NPV	0.85 (0.75–0.95)
	PPV	0.18 (0.01–0.41)

AUC: the area under the curve, NPV: negative predictive value, PPV: positive predictive value.