Hypoxemia during altitude exposure. A meta-analysis of chronic obstructive pulmonary disease

A previous study identified spirometric testing as a useful adjunct for estimating PaO2 during altitude exposure in patients with chronic obstructive pulmonary disease (COPD). We sought to examine the validity of this finding by quantitative analysis of recent published reports. We analyzed acute hypoxic exposures from five prior studies involving 71 patients. Across all studies, the change in arterial oxygen tension per unit change in inspired oxygen partial pressure (linear slope, dPaO2/dP1O2) correlated with the preexposure forced expiratory volume in 1 s (FEV1, p < 0.01). The correlation with FEV1 held for values weighted or unweighted by sample size, with rotating deletion of each study from analysis one at a time, and with semilog slope as the dependent variable. A formula derived from the semilog slope relationship with FEV1 gave accurate description of the mean hypoxic response in each prior study and individual responses from one study (n = 18): ln (PaO2alt/PaO2g) = Kn.(PIO2alt-PIO2g). We found that FEV1 modulated the values of kn in this study. We conclude, based on analysis of prior studies, that preexposure arterial oxygen tension and FEV1 both influence the prediction of PaO2 during hypoxic exposures in patients with COPD.