Keeping comatose cardiac arrest survivors cold for up to three days worked no better than keeping them cold for six hours, according to the ICECAP trial published August 5 in JAMA. Researchers at 71 US hospitals randomly assigned 1,158 patients to cooling durations between 6 and 72 hours and found no improvement in brain function at 90 days, no difference in deaths, and no difference in any secondary outcome.
The trial stopped early after a planned interim look at the data showed that longer cooling was not going to win.
Key takeaways
- Cooling for 72 hours was no better than cooling for 6 hours for brain recovery after out-of-hospital cardiac arrest.
- It randomized 1,158 comatose patients at 71 US hospitals between June 2020 and June 2025.
- A separate finding matters more for families: some patients in the trial woke up more than two weeks after the arrest.
What the study found
Every patient in the trial was cooled to 33 degrees Celsius. The only thing that changed was how long the cooling lasted: 6, 12, 18, 24, 30, 36, 42, 48, 60, or 72 hours.
To be eligible, adults had to remain unconscious after an out-of-hospital cardiac arrest, reach a temperature below 34 degrees Celsius within four hours of the arrest, and be on a real temperature control device. The first 200 patients were split evenly between the 12, 24, and 48 hour groups. After that, an adaptive method steered new patients toward whichever durations were looking best.
Of the 1,158 patients randomized, 883 had nonshockable rhythms and 275 had shockable rhythms. The median age was 61 and 39.6% were female.
The main measure was brain function at 90 days, scored on a weighted modified Rankin Scale, which rates how much disability a person is left with. In the nonshockable group, the analysis put the chance that 6 hours was the shortest duration achieving the best possible result at 51% (posterior probability 0.51). In plain terms, the data could not tell the durations apart, so the shortest one is as good a bet as any. Results in the shockable group looked the same.
One secondary result stood out. According to Medical Xpress, the survival rate among nonshockable patients was higher than in previous studies involving cooling. William Meurer, MD, professor of emergency medicine and neurology and one of the trial leaders, told Medical Xpress, “Our study showed that some patients undergoing therapeutic hypothermia can wake up even after two weeks.”
Dr. Kumar’s take
I understand why “cool the brain longer” was so appealing. Cooling slows metabolism and buys injured brain tissue time, and in the lab, longer cooling looked better. That logic is clean, and it is exactly the kind of reasoning that gets a therapy adopted before anyone has tested the dose.
This trial is a win, not a failure. ICECAP compared cooling durations from 6 hours out to 72, and the long end bought nothing: no better brain function at 90 days, no fewer deaths, no difference on any secondary outcome. That makes the shortest duration the defensible default, and it means no one has to hold a patient at 33 degrees for three days on the assumption that longer must be safer. Cardiovascular medicine is full of ideas that make sense on paper and then do not hold up, which is why I keep coming back to what the measurements actually show.
Two limits are worth naming. Every patient here was cooled, so this trial says nothing about whether cooling itself beats no cooling. And every patient got below 34 degrees within four hours, so the answer applies to people cooled fast.
The finding I would want a family to hear is the one buried in the secondary data. If patients can regain consciousness two weeks out, then a decision made on day three or day five is being made on incomplete information. Early prognostication after cardiac arrest can become self-fulfilling: a grim prediction leads to withdrawal of support, and the prediction comes true because of the decision, not because of the brain injury.
What it means for you
If someone you love is in an ICU after a cardiac arrest, do not push for longer cooling. This is the best evidence we have that it adds nothing, and shorter cooling means less sedation and fewer of the complications that come with it.
Do ask the team how long they plan to wait before predicting recovery, what that prediction rests on, and what would change their mind. A patient who has not woken up on day four has not finished the story.
And if the arrest was the kind that never got a shock, those patients did better in this trial than in earlier ones. That is not a guarantee, but it is a reason not to give up early.