Frosty the snowmanCooling in cardiac arrest has had its ups and downs. First, it was cool with 32-34 degrees, then TTM said 36 degrees was fine. So, what’s the optimal brain temperature post arrest? One of the problems with the TTM study was the time from ROSC to target temperature: 8 hours after randomisation. You won’t get any difference between 33 and 36 degrees if you reach 33 degrees after 8 hours. Could there be a hidden benefit with lower temperatures in the long cooling times?

For real?
A new article in the popular and well known Therapeutic Hypothermia and Temperature Management journal tries to systematically review its way to an answer, and concludes with the title Time to Cooling Is Associated with Resuscitation Outcomes. With all the new fake journals popping up, we had to do some quick research on this. The THTM journal has big names like Stephen Bernard as editors, so we’re pretty happy with the journal. And, more importantly, the article itself makes sense and seems solid.

We also know cooling works – if done pre-arrest on animals in experiments. They preserve brainfunction much better. The problem is brain cells die before we can cool them. Post-arrest cooling might not work, but maybe it helps if we cool quicker?

The study
Looking into studies on temperature management in cardiac arrest, excluding the ones that used intravenous fluids to cool the patient, as this probably adds badness through volume overloading and blood dilution. They chose 34°C as their target temperature, and pooled the results into two groups with times from ROSC to target temperature achieved: quick cooling (<3hrs) and slow cooling (>3.5hrs). They also seperated VT/VF pts from other CA pts. The endpoint was favourable recovery, meaning survivors with good neuro outcome (CPC 1-2).

As target temperature varies between centres and studies, this review article looks at cooling rate (degrees C/hr), time to 34 degrees as well as time to target temperature.

Cooling rateThere seems to be a linear relation between time to target temperature of 34°C and good neuro outcome. In the quick cooling group, the average time to target temperature was 2.5 hours with a cooling rate of 4.4°C/hr, and the slow cooling group had an average time to target temperature of ~5 hours with a cooling rate of 0.6°C/hr. A cooling time of less than 3 hours to achieve 34°C seemed to improve outcomes.

None of these studies had quick and slow cooled pts in the same trial, so all pts pools are collected from different trials with different settings, and then compared. Also, very few studies with rapid cooling are avaialble, so most have small sample sizes, but all were included. Small sample sizes might also hint to less robust trials. The slow cooling arm had lots of trials to choose from, so only trials with >100 pts were included. Lots of limitations and possible biases right there in the selection process, and hard to make any good comparison/analysis on.

The quick cooling of some patients might be a marker for overall effective medical treatment. Short ROSC to target temperature times might also mean short transport distances, meaning shorter time to arrival at hospital and higher level of care, etc etc.

This is review is far from solid evidence, but a hint that the last word in cooling might not be TTM. This review is a hint to look at how you choose to cool your patients. Cool aggressively, either your target temperature is 36, 34, 33 or 32. The review has some severe limitations to it. Still, the argument is sound and the collected data seem to back it up: time to cooling matters.

Time to Cooling Is Associated with Resuscitation Outcomes, Therapeutic Hypothermia and Temperature Management, 2016, 6 (4): 208-217

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

3 Responses to TIME TO BE COOL

  1. Vidar Magnusson says:

    Hi Thomas

    I like the idea of cooling and early cooling seems to make more sense than late cooling, but the only problem is that studies starting cooling as early as possible (post arrest) i.e. in the prehospital phase haven’t managed to show benefit…



    • Thomas D says:

      I agree, but the prehospital trials I’ve seen have cooled with cold saline. This is probably too slow (even when started prehospitally), and also gives problems with volume overload, as cardiac arrest pts aren’t volume depleted.

      It’s still a question if cooling helps at all when it’s post-arrest. We know it works pre-arrest in experimental settings, but post-arrest, hmm… The study in this post shows there *might* be a benefit to cooling quick and early hidden in the data, but it might also turn out to be a wild goose chase.

  2. mario arturo carrera sanchez says:

    The question is then how to cool these patients fast and safely

Leave a Reply

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