Can moderate chronic obstructive pulmonary disease be diagnosed by historical and physical findings alone?

The value of the history and physical examination in diagnosing chronic obstructive pulmonary disease (COPD) is uncertain. This study was undertaken to determine the best clinical predictors of COPD and to define the incremental changes in the ability to diagnose COPD that occur when the physical examination findings and then the peak flowmeter results are added to the pulmonary history. Ninety-two outpatients with a self-reported history of cigarette smoking or COPD completed a pulmonary history questionnaire and received peak flow and spirometric testing. The subjects were independently examined for 12 physical signs by 4 internists blinded to all other results. Multivariate analyses identified independent predictors of clinically significant, moderate COPD, defined as a forced expiratory volume in 1 second (FEV1) less than 60% of the predicted value or a FEV1/FVC (forced vital capacity) less than 60%. Fifteen subjects (16%) had moderate COPD. Two historical variables from the questionnaire--previous diagnosis of COPD and smoking (70 or more pack-years)--significantly entered a logistic regression model that diagnosed COPD with a sensitivity of 40% and a specificity of 100%. Only the physical sign of diminished breath sounds significantly added to the historical model to yield a mean sensitivity of 67% and a mean specificity of 98%. The peak flow result (best cutoff value was less than 200 L/min) significantly added to the models of only one of the four physicians for a mean final sensitivity of 77% and a specificity of 95%. Subjects with none of the three historical and physical variables had a 3% prevalence of COPD; this prevalence was unchanged by adding the peak flow results. Diminished breath sounds were the best predictor of moderate COPD. A sequential increase in sensitivity and a minimal decrease in specificity occurred when the quality of breath sounds was added first to the medical history, followed by the peak flow result. The chance of COPD was very unlikely with a normal history and physical examination.