Limiting chronic obstructive pulmonary disease (COPD) screening to people with symptoms is not likely to advance general population screening for undiagnosed COPD, according to a study published in Respiratory Research.

Researchers used data collected from 1332 individuals older than 40 years of age who were part of a multilevel sampling from the general population and part of the Canadian Cohort of Obstructive Lung Disease Study (CanCOLD). Ethics approval for the CanCOLD study was granted from the review boards at each study site, and all participants gave consent.

Individuals from the CanCOLD study were monitored for up to 3 years and examined at 18-month intervals where diagnostic spirometry was used. Demographic data was collected, including current tobacco use, history of tobacco use, comorbidities, and respiratory issues. Information was also collected on severity and frequency of symptoms such as cough, phlegm, and wheezing. A logistical regression model was used with symptoms as independent variables, adjusted for demographic data. Receiver Operating Characteristic (ROC) curves determined the area under the curve (AUC) for each of the 3 symptoms.

Continue Reading

Individuals who had not reported previously diagnosed COPD, emphysema, or chronic bronchitis were thought to have COPD if they presented with persistent airflow restrictions. The incidence of undiagnosed COPD was 26%, and based on the GOLD spirometry grading system, 95% had mild to moderate disease.

Among the symptoms, wheeze performed the best according to the ROC curves (AUC=0.62), with cough and dyspnea following (AUC=0.57). The symptoms combined performed somewhat better than any symptom alone (AUC=0.64) and the AUC improved by approximately 0.06 when smoking history was combined with each of the symptoms (AUC=0.67 for a model that included pack-years and wheeze). The AUC was 0.71 using a model that included all of the individual symptoms and covariates.

Related Articles

The researchers concluded that “symptoms alone had poor performance in identifying patients with undiagnosed COPD.” Most symptoms had sensitivities and positive predictive values below 50%. They suggested using symptoms alongside characteristics such as smoking history.


Johnson KM, Tan WC, Bourbeau J, Sin DD, Sadatsafavi M; for the Canadian Cohort of Obstructive Lung Disease (CanCOLD) study and the Canadian Respiratory Research Network. The diagnostic performance of patient symptoms in screening for COPD. Respir Res. 2018;19(1):147.