Comparison of four methods for calculating diffusing capacity by the single breath method.

In 283 patients referred for testing in an outpatient pulmonary function laboratory, we studied the single-breath diffusing capacity of the lungs for carbon monoxide (Dco) using the Ogilvie (Og), Jones-Meade (JM), Epidemiological Standardization Project (ESP), and three-phase iterative methods (3PIT, similar to the three equation method). The Dco maneuvers were performed using automated equipment and American Thoracic Society (ATS) recommended procedures. There were small but significant differences in mean Dco, the ESP method yielding the largest, followed in order by JM, 3PIT, and Og methods. The 3PIT and JM Dcos were in close agreement in all degrees and patterns of pulmonary function abnormality. The Og Dco method was 6 percent less than JM in patients with normal pulmonary function, although the difference was less in patients with expiratory flow limitation, restriction, or reduced Dco. There were no differences in the reproducibility of Dco measurements among the methods. Based on these results and a review of the literature, we conclude the following: (1) when measuring single-breath Dco using automated equipment that follows ATS recommended procedures for collecting a single expired gas sample of 500-ml volume, calculated Dco is largest using ESP method, following by JM, 3PIT, and Og methods; (2) in patients with reduced Dco associated with obstructive or restrictive abnormalities, the Og, 3PIT, and JM timing methods are nearly equivalent; and (3) reproducibility of Dco is the same by all methods.