This study gives high quality evidence that hyperoxia post-arrest is harmful. For each hour of hyperoxia (PaO2 >300), risk of poor neurological outcome rose by 3%. Number needed to harm was 8.
Why does this matter?
Once a patient has return of spontaneous circulation (ROSC) there are a few things we can do to aggravate the initial insult. One of them is to let the patient have hyperthermia, hence targeted temperature management (TTM). Hypotension and hypoxia are also harmful. But too much of a good thing, namely oxygen, is harmful too. Prior studies have been very small, retrospective, and conflicting. This study is much stronger.
Too much of a good thing
This was a multi-center prospective study of 280 post-arrest patients who were ventilated and had TTM. PaO2 was checked at 1 and 6 hours by protocol. Of the 280 patients, 105 had hyperoxia. Harm began when the PaO2 was >/=300. 77% of patients with hyperoxia vs 65% without had a poor neurological outcome (modified Rankin 3 or greater), NNH = 8, assuming the association is true. The relative risk of a poor neurological outcome was statistically adjusted to account for known confounders, such as prolonged CPR, initial rhythm VF/VT, hypotension, multiple comorbidities, etc. and was 1.23, meaning there was a 23% greater risk of a bad outcome for those with hyperoxia. Risk of an adverse neurological outcome increased 3% for each hour of hyperoxia. The strength of this study was its prospective design and protocolized measurement of PaO2 in all patients at hours 1 and 6. In practical terms, this means we cannot aim for a SpO2 of 100%. Rather, shoot for the mid-90s in post-arrest patients in the first 6 hours after arrest. Interestingly, they found a poor correlation between SaO2 and PaO2, meaning we should not rely too much on bedside SpO2. This is a time when an ABG is helpful. Rapidly titrating down the FiO2 is a simple, free way to give these patients the best possible chance for recovery.
Association Between Early Hyperoxia Exposure After Resuscitation from Cardiac Arrest and Neurological Disability: A Prospective Multi-Center Protocol-Directed Cohort Study. Circulation. 2018 Feb 1. pii: CIRCULATIONAHA.117.032054. doi: 10.1161/CIRCULATIONAHA.117.032054. [Epub ahead of print]
Peer reviewed by Thomas Davis, MD.