Chest, Vol 82, 630-638, Copyright © 1982 by American College of Chest Physicians

ARTICLES

The UCLA population studies of chronic obstructive respiratory disease. 5. Agreement and disagreement of tests in identifying abnormal lung function

R Detels, DP Tashkin, MS Simmons, HE Carmichael Jr, JW Sayre, SN Rokaw and AH Coulson

The relative sensitivity and specificity of symptoms, spirometry, the flow-volume (FV) curve, the single-breath nitrogen test (SBNT), and specific airway conductance (SGaw) were estimated by cross comparison of these tests with the FEV1/FVC ratio and other tests in 1,201 residents of Los Angeles. Spirometry and the flow-volume curve were estimated to be the most sensitive, yet reasonably specific composite measures, and FEV1/FVC and V75 to be the most sensitive individual indices. Among current smokers, FEV1/FVC was more frequently abnormal (23 percent) than FEF25-75% (16 percent), showed higher concordance with an abnormal FEF25-75% (88 percent), and identified a greater percentage of individuals as abnormal when the FEF25-75% was normal (9 percent) than vice versa. SGaw and SBNT showed a poor concordance with FEV1/FVC and identified more individuals as abnormal who had a normal FEV1/FVC. Although the proportion of current smokers with productive cough was relatively high (26 percent), none of the function tests showed a reasonable concordance with this symptom. These findings suggest that: 1) spirometry, the flow-volume curve, V75 and FEV1/FVC are relatively sensitive and specific indicators of respiratory impairment; 2) spirometry and the FV curve provide similar but not completely overlapping information; 3) SGaw and SBNT yield different (or less specific) information than the other tests; and 4) effects of cigarette smoking identified by productive cough are different from those identified by other tests.