Diagnostic value of reversibility of chronic airway obstruction to separate asthma from chronic bronchitis: a statistical approach.

Our aim was to assess the usefulness of tests of reversibility of airways obstruction (AO) in the differential diagnosis of asthma (A) and chronic bronchitis (CB). We selected on strict clinical criteria 20 CB and 32 A patients with stable AO and measured maximal forced expiratory flows before and after increasing doses of inhaled salbutamol. The highest sensitivity (Se), specificity (Sp) and likelihood ratio (L) were obtained when the improvement in forced expiratory volume in one second (FEV1) was expressed as a percentage of predicted FEV1. Se, Sp and L were 0.97, 0.95 and 19.3 (highest possible L: 20) for an increase of 10% of predicted FEV1 on 0.2 mg of salbutamol, and 1.0, 1.0 and 20, respectively, for an increase of 15% on 1 mg of salbutamol. Conversely, L was only 2.8 when the widely used increase in FEV1 of 20% over baseline was tested. In a second group of 42 unselected patients, who were referred to the laboratory for routine lung function testing with a presenting diagnosis of A or CB, L was only 5.5 when the 10% increase in predicted FEV1 was used and even lower for the classical 20% increase over baseline FEV1. We submit that routine tests of reversibility of AO have little diagnostic usefulness in unselected patients with AO and that the commonly used criteria of 15 or 20% increase in FEV1 over baseline are of no value in separating A from CB.