Evidence Base: Pulmonary

Accuracy in Diagnosis of COPD

Question 1: How helpful are the history and examination in predicting the finding of obstructive airways disease (COPD) in a patient?

Study of 167 patients with COPD

History

History	LR+	LR-	Sensitivity	Specificity
Smoking (ever)	1.2 (1.1-1.4)	.23 (.08-.64)	.94	.24
Dyspnea	1.2 (1.02-1.5)	.55 (.31-.98)	0.82	0.33
Cough	1.8 (1.2-2.6)	.69 (.52-.90)	0.51	0.71
Wheezing	3.1 (1.9-5.2)	.58 (.45-.75)	0.51	0.84
Chronic Bronchitis	3.8 (2-7.2)	.66 (.53-.81)	0.42	0.89
Orthopnea	1.6 (.79-3.4)	.92 (.80-.1.05)	0.19	0.88

back to top

Examination

History	LR+	LR-	Sensitivity	Specificity
Decreased Breath Sounds	1.2 (1.1-1.4)	.23 (.08-.64)	.94	.24
Wheezing	1.2 (1.02-1.5)	.55 (.31-.98)	0.82	0.33
Rales	1.8 (1.2-2.6)	.69 (.52-.90)	0.51	0.71
Cough	3.1 (1.9-5.2)	.58 (.45-.75)	0.51	0.84
Subxyphoid Impulse	3.8 (2-7.2)	.66 (.53-.81)	0.42	0.89

back to top

Peak Flowmeter (nomogram)

Nomogram for predicting probability of airflow obstruction at the bedside

How To Use:

• In first column, select A scale for "no wheezing" or B scale for "wheezing"
• In last column, use either "no or yes" or best of 3 "PEF" rate trials scale
• Place a ruler on # of yrs smoking and selection in last column-- intersection in middle column will give appropriate probability.

• 0.95 to determine the odds of at least moderate disease
• 0.35 to determine the odds of at least severe disease

Conclusion
Peak expiratory flow is not necessary to making the Diagnosis of COPD at the bedside.

back to top

Question 2: Can the diagnosis of moderate COPD be made on clinical basis alone?

Clinical Component	LR+	Sensitivity	Specificity	Accuracy	ROC+
History*	Infinite	.40	1.00	90%	0.80
History plus Breathsounds**	38 (22.5-Infinite)	.67	.98z	93%	0.88
History, Breathsounds & Flowmeter < 200L/min	16.9 (13.8-22.1)	.77	.95	92%	0.90
*History of chronic bronchitis or emphysema, smoked 70 or more pack-years **FEV1, 60% predicted or FEV1/FVC <.6 +area under ROC curve

Adapted from Badget, RG, et al.

back to top