Getting Chilly Quickly 2…. Early hypothermia for ROSC?
By now, we all are thoroughly convinced as to the benefit of therapeutic hypothermia after cardiac arrest. (if you don’t do this in your hospital, you really should not be seeing OOHCA patients….). In addition, the consensus seems to be that time does matter and we should be doing this ASAP after ROSC. The evidence isn’t conclusive I’ll grant you (as Simon discussed in his post). Our beloved National Institute for Clinical Excellence says we should be cooling “as soon as possible after the cardiac arrest“.
Of course, we ED docs are an impatient bunch, and the idea of waiting around for something to happen is difficult for some of us. Knowing this, it is inevitable that someone was going to ask:
Why wait for ROSC?
Can’t we be cooling the patient down during CPR? This intuitively seems like a good idea. If we start during arrest, we get patients to target temperature faster. There is also a physiological rationale: one of the theoretical mechanisms for TH is the attenuation of reperfusion injury. Surely getting the patient cold before reperfusion occurs has to be good thing, right?
Alas, in a EBM world, “it seemed like a good idea” or “I was bored” are no longer acceptable justifications for treatment. We need to see some evidence here. So we ask ourselves:
Is intra-arrest hypothermia (IAH) better than traditional post-arrest therapeutic hypothermia (PATH). Does it increase our rates of ROSC, of survival and most importantly, the number of patients who can walk out of hospital and lead a normal life?
The evidence is scarce. There is one RCT in humans out there. These guys took about 200 OOHCA patient (witnessed, all rhythms). In addition to standard ALS, they began cooling down their treatment group using a intranasal device.
So did it help? Well, as a feasibility study it was successful. The patients got to target temp sooner, and the process appeared safe, However it did not detect any change in outcomes, as it was not powered to do so.
Interesting, but not really relevant to our question. I imagine most ED’s will be using ice and cold fluids (and maybe a cooling blanket) to start hypothermia, so studies looking at intranasal coolant administration (or cold liquid ventilation (!)) are not that relevant to us.
So that’s it for human RCTs. There are some animal studies which may be of interest to us. This study, found dramatically better outcomes when cooling (via cold fluid loading and then intravascular device) 10min post arrest than 20min post arrest. This one found IAH with an intravascular device improved outcomes in pigs. IAH with cold fluid loading did not affect any useful outcomes.
Fridge or cupboard?
So where to go now? Thinking through it, a thought occurs. Realistically, any IAH taking place in the ED will be via cold fluid loading, yes? In my practice, and I suspect in most of yours, patients are already getting 1-2 litres of crystalloid during an arrest anyway. Are we causing any harm by getting these fluids out of the fridge rather than the cupboard?
If we aren’t causing harm, then reducing the time to target temperature seems a good enough reason to start cooling intra arrest.
Garret et al performed a retrospective analysis comparing IAH with PATH. They found that 2000mls of cold saline started ASAP during CPR improved rates of prehospital ROSC. There was a trend towards better outcomes, but no significant improvement.
That seems to be it as far as evidence goes…. Can we draw any conclusions? Having sat and gone though what literature is out there I believe that:
So I’m going to fall off the fence, and say that I think we should be doing this. The evidence isn’t great I grant you, but I think there is enough to change practice.
Hopefully I am being controversial enough to provoke a debate. Anyone disagree and think I’m over reaching? Have I missed any important evidence? Comment below, abuse me on Twitter, join me in the FOAM!
P.S – no conflicts of interest to declare, (apart from being impatient). I mention the Rhinochill (TM) system, but have absolutely no involvement with the device or the manufacturer.