Reproducibility of pulmonary function tests under laboratory and field conditions.

The reproducibility of pulmonary function tests in the laboratory and in a mobile field survey vehicle has been studied. Groups of laboratory workers were studied at base and a random sample of 38 coalminers was examined in the mobile laboratory. The intra-subject variability of some newer tests of lung function, including closing volume and maximum flow at low lung volumes, has been compared with that of well-established tests, such as lung volumes and forced expiratory volume from two measurements made more than one day apart. Most measurements were slightly less reproducible in the study of coalminers than in the laboratory personnel. Conventional tests, such as forced expiratory volume in one second, lung volumes, single breath CO transfer factor, and exercise ventilation were very reproducible, the coefficients of variation (cov) being generally between 5% and 10%. The closing volume test, maximum expiratory flow at low lung volumes, and the single breath N2 index were less reproducible: cov between 15% and 39% in the miners. The forced expired time and volume of isoflow, measured only on laboratory workers, however, exhibited greater reproducibility than previously reported (cov = 10% and 15% respectively). It is suggested that, when assessing the repeatability of lung function tests, account should be taken of the circumstances in which the intra-subject variability was measured.