To test questions usable in an ambulatory clinic to identify persons likely to have chronic obstructive pulmonary disease (COPD).

Analyses were performed as part of a study to identify patients with likely COPD in the Glenfield UK primary care clinic. Patients age 40 and older were recruited based on one of the following criteria: (1) respiratory medications in previous 2 years; (2) history of smoking or (3) history of asthma with no current medications based on case notes. Consenting patients reported smoking history, symptoms, and personal and family history of respiratory conditions. Spirometry with reversibility was conducted to ATS standards. Analyses were performed on this database to test questions for identifying patients with COPD from a sample of patients with a positive smoking history. Multivariate logistic regression identified the question set that best discriminated COPD from other conditions using receiver operating characteristic curves. The usefulness of a simple scoring system was assessed.

The study sample included 369 current and former smokers. Patients were diagnosed as:   (16.8%);   (8.1%); or no obstructive lung   (75.1%). The best questions for discriminating between persons with and without COPD included items on age, dyspnoea on exertion, and wheeze. This set of questions identified COPD patients with a sensitivity of 77.4–87.1% and specificity of 71.3–76.2%.

A simple questionnaire can facilitate the diagnosis of COPD in a primary care setting.