Skip to main navigation
Skip notifications
Access to POEMs and Essential Evidence Plus will no longer be included in CMA membership as of Dec. 1, 2023.
Search
Liberal oxygen therapy in hospitalized patients: too much of a good thing? (IOTA)
Clinical Question
Does liberal oxygen therapy worsen outcomes for hospitalized patients?
Bottom line
The use of liberal oxygen therapy results in increased short-term and long-term mortality in acutely ill hospitalized patients with a broad range of illnesses. A dose-response relationship is seen, with increasing risk of death associated with higher oxygen saturations. 1a
Reference
Study design: Meta-analysis (randomized controlled trials)
Funding: Self-funded or unfunded
Setting: Inpatient (any location)
Synopsis
These investigators searched the Cochrane Central Register of Controlled Trials, MEDLINE, Embase, and other databases to find randomized controlled trials that compared the effect of liberal and conservative oxygen therapies on acutely ill hospitalized adults. Outcomes of interest included mortality, disability, hospital-acquired infections, hospital-acquired pneumonia, and hospital length of stay. Two reviewers independently selected studies for inclusion, extracted data, and assessed study quality. Twenty-five randomized controlled trials were selected; 18 had low risk of bias. The 16,037 patients represented in these trials included patients with critical illness, trauma, sepsis, stroke, myocardial infarction, cardiac arrest, and those who had emergency surgery. As compared with a conservative strategy, a liberal oxygen strategy led to increased deaths in the hospital (relative risk [RR] 1.21; 95% CI 1.03 - 1.43), at 30 days (RR 1.14; 1.01 - 1.28), and at a median of 3 months of follow-up (RR 1.10; 1.00 - 1.20). Additionally, as oxygen saturation increased, mortality risk increased, both in the hospital and at the longest follow-up. Although the individual trials varied in their definitions of liberal and conservative oxygen therapy, the point estimates for mortality were consistent with no variation due to heterogeneity, suggesting a true effect. The authors report that you would need to treat 71 patients with liberal oxygen therapy to cause 1 death (number needed to treat to harm 71; 37 - 1000). No differences were seen overall in disability, hospital-acquired infections, hospital-acquired pneumonia, or hospital length of stay. Although a subgroup analysis showed fewer hospital-acquired infections with liberal oxygen therapy in patients who had emergency surgery, the quality of this evidence was low. This finding was not seen with medical patients.
Reviewer
Nita Shrikant Kulkarni, MD
Assistant Professor in Hospital Medicine
Northwestern University
Chicago, IL
Comments
We recently made the switch to using conservative oxygen therapy by all of our paramedics in Ontario. Great to see this review which fully supports our change.
Good poem
difficult to find " hard " evidence but many times the " nursing indications " to give / start O2 is not well documented other than " protocol ".
Patient with marginal saturation most likely live at home with same and do not need O2 once hospitalized.
Also effect of administrated O2 is in many cases not optimal.
Limiting Oxygen exposure in Hyperbaric Oxygen patients with CO inhalation to Hyperbaric Oxygen and no normobaric oxygen between treatments
How do we know that the patient to receive the liberal oxygen therapy weren’t sick right to begin with?