Careers  |  Sign In  |  Register  |   Twitter

Hyperoxia following resuscitation predicts in-hospital mortality

Reuters Health • The Doctor's Channel Daily Newscast

NEW YORK (Reuters Health) – When spontaneous circulation is restored after cardiac arrest, patients with hyperoxia have higher rates of in-hospital mortality than those with hypoxia or normal oxygen levels, researchers reported on Tuesday.

The findings, published in the June 2nd issue of the Journal of the American Medical Association, add to an ongoing debate about the role of supplemental oxygen after resuscitation.

“It appears that oxygen can be a double-edged sword when delivered to the injured brain,” senior author Dr. Stephen Trzeciak, of Cooper University Hospital in Camden, New Jersey, told Reuters Health in an e-mail. “While it is intuitive that ongoing oxygen deprivation after resuscitation from cardiac arrest can worsen brain injury, studies in animal models have shown that excessive oxygen can also be harmful by increasing the production of oxygen free radicals.”

Guidelines from the American Heart Association currently recommend 100% oxygen during resuscitation. “But after the patient is resuscitated,” Dr. Trzeciak told Reuters Health, “the speed with which clinicians turn down the supplemental oxygen to the minimum amount that the patient requires is highly variable.”

“We believe that a treatment strategy of controlled reoxygenation after resuscitation from cardiac arrest could potentially improve patient outcomes,” he added, “but first we need to test this in a controlled clinical trial. That’s our next step.”

In the meantime, he said, “Our study indicates that exposure to very high oxygen levels after resuscitation from cardiac arrest is associated with worse clinical outcomes.”

Dr. Trzeciak and colleagues used data from the Project IMPACT database, which contains information from ICUs across the U.S. They included 6326 adults with non-traumatic cardiac arrest who had been resuscitated within 24 hours before ICU arrival, and whose arterial blood gases had been analyzed no later than 24 hours after arrival.

On arrival in the ICU, 18% had hyperoxia, defined as a partial pressure of arterial oxygen (PaO2) of 300 mm Hg or greater; 63% had hypoxia (PaO2 of less than 60 mm Hg) and 19% had normoxia.

Overall, 3561 of the patients died in the hospital (56%). Mortality was significantly higher in the hyperoxia group (63%) than in the hypoxia (57%, p<0.001) and normoxia (45%, p<0.001) groups. Kaplan-Meier analysis showed that survival diverged significantly over time for hyperoxia and normoxia (p<0.001).

Patients with hyperoxia were also significantly less likely than those with normoxia to be functionally independent when leaving the hospital.

Stepwise logistic regression analysis found several risk factors for in-hospital mortality, including age, nonindependent preadmission functional status, emergency department origin, active chemotherapy, and chronic renal failure. In addition, hypotension on ICU arrival, tachycardia and hypoxia all significantly predicted in-hospital death.

After adjusting for these factors, hyperoxia was still associated with increased mortality (odds ratio, 1.8).

But at this point, Dr. Trzeciak told Reuters Health, “the optimal oxygen level to strive for is not yet clear and will require additional study.”

“Given the rather conservative definition of hyperoxia (PaO2 at least 300 mm Hg), the true incidence of more moderate levels of hyperoxia is likely to be quite high,” Dr. Patrick M. Kochanek and Dr. Hulya Bayir, of the University of Pittsburgh School of Medicine in Pennsylvania, write in an editorial.

They add that “this finding underscores the possibility that further meaningful improvements in outcome might result from careful attention to appropriately titrating basic aspects of extracerebral physiology at the bedside, such as prevention of hyperoxia.”

Drs. Kochanek and Bayir note that the study could not address whether timing played a role. “It would have been informative,” they write, “to have provided an assessment of the temporal relationship of hyperoxia with outcomes because experimental work suggests the possibility that early hyperoxia rather than delayed postresuscitation hyperoxia is deleterious to the brain.”

JAMA 2010;303:2165-2171,2190-2191.