David A Kaminsky1, Todd Whitman, Peter W Callas. 1. Pulmonary Disease and Critical Care Medicine, University of Vermont College of Medicine, Given D-302, 89 Beaumont Avenue, Burlington, VT 05405, USA. david.kaminsky@uvm.edu
Abstract
BACKGROUND: The diffusing capacity of the lung is usually reported as both the diffusing capacity (DLCO) and the diffusing capacity divided by the alveolar volume (DLCO/VA). However, it is unclear which measure to use when interpreting pulmonary gas exchange. We therefore conducted this study to determine whether the DLCO or the DLCO/VA is a better predictor of oxygen desaturation with exercise. METHODS: We retrospectively analyzed the pulmonary function records of all patients who had measurement of their diffusing capacity and 6-min walk oximetry in our university pulmonary function laboratory over a 2-year period. RESULTS: There were data available on 97 patients, most of who had interstitial lung disease and/or lung volume restriction. The median DLCO was 51% predicted and the median DLCO/VA was 64% predicted. The prevalence of exercise desaturation was 43%. The overall sensitivity and specificity as determined by the area under the receiver operator characteristic (ROC) curve was higher for DLCO than DLCO/VA, with an optimal cut-off of normal of 55% predicted. The positive predictive values were equally low for both measures, ranging from 50% to 70%. After adjustment for VA, there were no differences between the ROC curves or predictive values for DLCO and DLCO/VA. CONCLUSION: After adjusting for VA, neither the DLCO nor the DLCO/VA was better at predicting oxygen desaturation with exercise. The optimal cut-off of normal was 55% predicted.
BACKGROUND: The diffusing capacity of the lung is usually reported as both the diffusing capacity (DLCO) and the diffusing capacity divided by the alveolar volume (DLCO/VA). However, it is unclear which measure to use when interpreting pulmonary gas exchange. We therefore conducted this study to determine whether the DLCO or the DLCO/VA is a better predictor of oxygen desaturation with exercise. METHODS: We retrospectively analyzed the pulmonary function records of all patients who had measurement of their diffusing capacity and 6-min walk oximetry in our university pulmonary function laboratory over a 2-year period. RESULTS: There were data available on 97 patients, most of who had interstitial lung disease and/or lung volume restriction. The median DLCO was 51% predicted and the median DLCO/VA was 64% predicted. The prevalence of exercise desaturation was 43%. The overall sensitivity and specificity as determined by the area under the receiver operator characteristic (ROC) curve was higher for DLCO than DLCO/VA, with an optimal cut-off of normal of 55% predicted. The positive predictive values were equally low for both measures, ranging from 50% to 70%. After adjustment for VA, there were no differences between the ROC curves or predictive values for DLCO and DLCO/VA. CONCLUSION: After adjusting for VA, neither the DLCO nor the DLCO/VA was better at predicting oxygen desaturation with exercise. The optimal cut-off of normal was 55% predicted.