Abstract

BACKGROUND AND OBJECTIVE: Although long-term oxygen therapy is one of few treatments known to improve survival in COPD, no patient subgroup likely to derive benefit has been identified in over 30 years. We assessed the prognostic value of point measurement of P(a) O(2) and proportion of ambulatory oximetry with saturations below 90% in this context. METHODS: Subjects were community-living patients with COPD and P(a) O(2) 56-70 mm Hg when stable. Baseline assessments included clinical, physiological and laboratory assessments, as well as 24-h ambulatory oximetry. Patients were followed to death from any cause, first exacerbation and first admission. RESULTS: Despite poor survival overall, there was no significant difference in prognosis between mildly hypoxaemic and moderately hypoxaemic patients. There were also no significant differences for secondary end-points (exacerbation and hospital admission). On multivariate analysis, trends were observed towards improved survival, with higher P(a) O(2) and greater proportion of ambulatory oximetry below 90%. CONCLUSIONS: Neither resting P(a) O(2) nor proportion of ambulatory oximetry below 90% saturation effectively predicted survival in COPD.