iPhoneIcon_BigA recent study published in New England Journal of Medicine has created a splash in EM social media. While it acknowledges that temperature control is important, it also suggests therapeutic cooling of cardiac arrest survivors does not work.

International guidelines recommend therapeutic hypothermia after resuscitation from cardiac arrest. The recommendation is   to cool the patient to 32°C to 34°C  for 12 to 24 hours after return of spontaneous circulation. This is based on animal studies and some randomised trials. Still, it is a lot more controversial than we tend to think. Recent systematic reviews doubt the efficacy of therapeutic hypothermia and the grade level of evidence is low.

Many of the studies supporting therapeutic hypothermia suffer from being small, evaluating select populations and coming with a high risk of bias. One specific problem some of these studies had was how fever in the control groups wasn’t treated. Fever is known to be associated with a poor outcome in ROSC patients. The study in NEJM tries to fix that with a more robust study, with wide inclusion criteria and where post cooling feber is treated in both controls as well as in the intervention group.

The study
Multi-centre randomised prospective trial. The study included 950 adult patients who were admitted unconscious with ROSC after prehospital cardiac arrest, irrespective of the presenting rythm. Unwitnessed arrests with asystolia as presenting rythm were excluded. As were patients with body temperatures below 30°C and those with suspected or known stroke.

These 950 were then 1:1 randomised into receiving therapeutic hypothermia with target temperatures of either 33°C or 36°C. The cooling period was 28 hours were the various sites used intravascular cooling catheters, cold fluids, ice-packs, surface cooling etc in order to reach the target temperatures. This was followed by 8 hours of rewarming for a total intervention period of 36 hours. Then for another 36 hours the only intervention was fever control with target body temperatures of below 37,5 °C.

Primary outcome was mortality up until 180 days. Secondary outcomes were neurological function at 180 days.

Primary outcome was deaths at the end of the trial. In the 33°C group the mortality rate was  50% while the 36°C group had a mortality of 48%. So, no benefit from using 33°C as a temperature goal over 36°C.

The secondary outcome was neurologic disability as defined by the modified Rankin scale and the cerebral performance category score. Again there was no significant difference.

Take-home message
This study could not demonstrate any benefit in mortality or neurologic outcome from cooling to 33°C as compared to 36°C. It suggests therapeutic hypothermia doesn’t work as well as we would like to think. It also suggests treating hyperthermia in the post-ROSC is important. Temperature control is important but hypothermia might not work.

cIt is likely to be controversial and will be scrutinised in detail.

I don’t have time to do that, but one thing that gets me at a glance is how the 33°C group were more likely to have a cardiac history.

But still…..ouch?

Study lives here:
The New England Journal of Medicine Targeted Temperature Management at 33°C versus 36°C after Cardiac Arrest. November 17, 2013 | N. Nielsen and Others

Comment from Thomas D:

Pre-hospital cooling
I’d also like to append the recent JAMA study on pre-hospital cooling to this post. If cooling works, cooling as early as possible should be better. That was the concept og the following study. Around 1300 OHCA patients were randomised to cooling prehospitally, or standard treatment. Nearly all patients in both groups were cooled/continued cooled at hospital arrival. The intervention group reached target temperature around 1 hour earlier than the standard group. There was no difference in mortality or neurological status at hospital discharge.

Cooling works – pre-arrest
This was another small hit for cool and the gang. We know from cold drownings or people who freeze to cardiac arrest that cooling is beneficial – but it seems that to rescue the brain, it needs to be cooled pre-arrest. Hard to do in most cardiac arrest patients…

Temperature management seems to avoid further injury, so the TTM concept is as important as ever – don’t let your post cardiac arrest patients slip into hyperthermia.

Effect of Prehospital Induction of Mild Hypothermia on Survival and Neurological Status Among Adults With Cardiac Arrest, JAMA, 2013

This entry was posted in Cardiology, Emergency Medicine, Intensive Care. Bookmark the permalink.


  1. Pingback: The LITFL Review 116 - LITFL

  2. Good point on the cardiac history.

    But certainly agree – avoiding hyperthermia is important post-ROSC.

  3. Pingback: All in a lather over TTM - LITFL

  4. Dr Peter Isherwood says:

    Interesting study, though not appropriate to label it with “it also suggests therapeutic cooling of cardiac arrest survivors does not work” in the tag line.

    This study compared different target temperatures not the concept.


  6. Pingback: BEST OF 2013 |

  7. Jeff Laabs RCP says:

    I would have to agree with Dr Isherwood. It is wrong to state TH doesn’t work. This study shows no improvement in outcomes at 36C vs. 33C and no improvement in early cooling (prehospital) vs late cooling. Where in the studies, does it say TH doesn’t work? There is an issue with this study. The control group was not allowed to be febrile compared to the original study that allowed patients to be febrile. Obviously this could have a major affect on outcomes.

    The other trial comparing early (prehospital) vs. late (in hospital) showed no benefit to early cooling. Ouch? Nope. Only for those that are emotionally invested in one technique over the other.

  8. Pingback: NEW CPR GUIDELINES |

  9. Pingback: NEW CPR GUIDELINES - Medicine Handbook

  10. Pingback: TIME TO BE COOL |

Leave a Reply

Your email address will not be published. Required fields are marked *