Optimal combinations of lung function tests in the detection of various types of early lung disease.

The ability to detect early lung disease with different combinations of lung function tests was assessed by discriminant analysis. A number of lung function tests were performed in 224 never-smokers, 232 smokers, 111 pneumoconiotic subjects and 137 asthmatic patients. The discriminatory capacities of different combinations of test variables are presented. For detecting lung damage induced by tobacco smoke, a combination of the transfer factor and the slope of the alveolar plateau (phase III) increased the sensitivity from 18% to 32% at a specificity of 95%, compared with phase III alone. Dynamic spirometry did not add to the discriminatory capacity. Patients with asthma could be separated from reference subjects by airway resistance, Phase III or a combination of variables in dynamic spirometry. Pneumoconiotic subjects were best identified by a combination of the transfer factor, volumic compliance and phase III. Closing capacity divided by total lung capacity (TLC) and FEV1/TLC further improved the discrimination between different subgroups.