Hypothermia Harmful for In-Hospital Cardiac Arrest?

October 06, 2016

The largest study to date evaluating therapeutic hypothermia for patients with in-hospital cardiac arrest has suggested that the treatment is not associated with higher rates of survival to discharge or favorable neurologic survival and may be associated with potential harm.

"We were surprised by the results. We maybe would not have been surprised by a neutral result, but we certainly didn't expect hypothermia to be harmful. There may have been an issue with overcooling in this study, though, as 1 in 5 patients who received hypothermia achieved a temperature below the recommended 32℃C, so this may explain the results," lead author, Paul S. Chan, MD, Saint Luke's Mid America Heart Institute, Kansas City, Missouri, told Medscape Medical News.

The observational study of data from a large US national registry of in-hospital cardiac arrest patients is published in the October 4 issue of JAMA.

Need for a Randomized Trial

Dr Chan explained that hypothermia has been shown to be beneficial in patients experiencing out-of-hospital cardiac arrest.

"But patients having in-hospital cardiac arrest are different. They are normally sicker and their heart rhythms are different. Also CPR [cardiopulmonary resuscitation] is generally started much sooner in in-hospital cardiac arrest patients — often within a minute of cardiac arrest, so the likelihood of getting some blood flow to the brain is higher. This may potentially limit the theorized benefit of therapeutic hypothermia to reduce free radical–mediated reperfusion injury from anoxic brain injury," he said.

"There is a critical need to conduct a randomized controlled trial of hypothermia in in-hospital cardiac arrest and not just assume because it is beneficial for some out-of-hospital cardiac patients then we can extrapolate to other groups. In the meantime I would not recommend it be used for in-hospital cardiac arrest patients," Dr Chan added.

The current study analyzed data from the Get With the Guidelines–Resuscitation (GWTG-R) registry, which includes 26,183 patients successfully resuscitated from an in-hospital cardiac arrest between 2002 and 2014 at 355 US hospitals.

Of these, 1568 patients (6.0%) were treated with therapeutic hypothermia, and 1524 of these patients were matched by propensity score to 3714 patients who did not receive hypothermia.

Results showed that after adjustment, therapeutic hypothermia was associated with lower in-hospital survival (27.4% vs 29.2%; relative risk, 0.88, P = .01), and this association was similar for nonshockable cardiac arrest rhythms and shockable cardiac arrest rhythms.

Therapeutic hypothermia was also associated with lower rates of favorable neurologic survival for the overall cohort (17.0% vs 20.5%; RR, 0.79; P < .001) and for both rhythm types.

When follow-up was extended to 1 year, there remained no survival advantage with therapeutic hypothermia treatment.

To investigate whether the worse outcomes in the hypothermia group may have been reflecting some unmeasured bias in the study — possibly that cooling was used in sicker patients — the researchers conducted a sensitivity analysis looking at outcomes in the first 24 hours.

"If hypothermia patients were actually sicker than those who did not receive cooling, then we would expect worse outcomes in the first 24 hours but we actually saw the opposite: Hypothermia patients had a slightly better outcome in the first day. This suggests that they were not sicker or alternatively that cooling may be protective but just for the very short term," Dr Chan reported.

He noted that international and American Heart Association guidelines recommend hypothermia for out-of-hospital cardiac arrest and suggest it should be considered for in-hospital cardiac arrest.

"The in-hospital recommendation is based on an extrapolation from out-of-hospital data. Two studies in out-of-hospital cardiac arrest have shown a benefit but only for one heart rhythm — patients in VF [ventricular fibrillation]. Patients with in-hospital cardiac arrest are less likely to have VF than out-of-hospital patients — they are more likely to be in asystole or to have pulseless electrical activity. Patients with these two rhythms in the out-of-hospital studies were less likely to respond to hypothermia," he said.

He also pointed out that studies in other populations, such as patients with meningitis or traumatic brain injury, have also suggested a harmful effect of hypothermia.

"So just because it works in one situation does not mean it works in others."

Dr Chan said the current study was the largest so far to look at the use of hypothermia among patients having in-hospital cardiac arrest, and it had the most rigorous methods.

"Two other studies have looked at hypothermia for in-hospital cardiac arrest. One found no effect, but that study did not control for the site at which care was given — they included hospitals without hypothermia capability.

"Another study, which suggested a benefit, only involved 42 in-hospital cardiac arrest patients, and although the study did use propensity score matching, it mainly included out-of-hospital patients, so the propensity scores were based on factors that were not really appropriate for in-hospital cardiac arrest patients.

"I would say our trial gives the best evidence so far on hypothermia for in-hospital cardiac arrest, and it suggests the procedure is not beneficial and could be harmful. But out trial still had an observational design. We need a randomized controlled trial to really know for sure."

"Excellent" Research

Commenting for Medscape Medical News, Sarah M. Perman, MD, University of Colorado School of Medicine, Aurora, who has also conducted research in this field, described the current study as "excellent."

"Given the findings in this paper — a pragmatic trial to address the many questions that still persist regarding the utility of therapeutic hypothermia (33℃C) or even targeted temperature management (36℃C) in this cohort might just be warranted," she added.

Dr Perman pointed out that because the cohort studied represented only 6% of the registry population, with such a small proportion unmeasured bias is likely. 

"Without coma as a variable in the GWTG-R, it is difficult to know the level of neurologic impairment patients had after return of spontaneous circulation, and the authors do state that little is known about the quality of postarrest care that was provided to patients who received therapeutic hypothermia," she said. 

"Nonetheless, the study does show potential poor outcomes for those treated with hypothermia in this in-hospital cardiac arrest cohort.  Just as the authors suggest, it is challenging to translate findings from the out-of-hospital cardiac arrest arena to in-hospital cardiac arrest."

The GWTG-R registry is sponsored by the American Heart Association. Dr Chan reported serving as a consultant for the American Heart Association, Optum Rx, and Johnson & Johnson.

JAMA. 2016;316:1375-1382. Abstract

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.