Survival in COPD patients under long-term oxygen therapy

Abstract

Randomized clinical trials have provided clear evidence that long-term oxygen therapy (LTOT) increases life expectancy in severely hypoxemic COPD patients. However, in real-life settings, in global COPD cohorts, a paradoxical 2-3-fold increased risk of death has been reported in patients under LTOT. This discrepancy could be explained by a subgroup of patients under LTOT who do not meet the current guidelines and for whom LTOT would be associated with poor prognosis. This retrospective study of a global COPD cohort therefore sought to: (1) determine if a subgroup of patients under LTOT without severe hypoxemia could be distinguished, and (2) compare the mortality risk according to hypoxemia severity and LTOT prescription. The sample was taken from a database (NCT02055885) on 191 stable COPD patients (age: 65.1±9.8 years, 85 women) admitted for a pulmonary rehabilitation program between 2009 and 2012 and followed until 1 January 2018. Uni- and multivariate Cox proportional hazard ratio (HR) models were used to examine the associations between clinical characteristics (age, sex, blood gases, FEV1 …), LTOT according to PaO2 level, and mortality. Forty patients (21%) were under LTOT at PR entry despite not meeting the O2 prescription criteria. Patients under LTOT had a nearly 2-fold higher mortality risk adjusted by covariates (HRmultivariate=1.83; p=0.009). Furthermore, the higher mortality risk under LTOT was specific to the patients under LTOT without severe hypoxemia (HRmultivariate=2.39; p=0.04). The association between LTOT and mortality might be attributed to a subgroup under LTOT despite not/no longer meeting the LTOT criteria. Further studies are needed to identify the physiological basis of this phenomenon.