No CPR in Trauma Arrest?

In October 2013, Scotty Weingart was on the show and suggested that closed chest CPR has no role in traumatic cardiac arrest. It made sense because, after all, the things that cause a traumatic arrest won’t be helped by closed chest compressions. Tension pneumothorax, pericardial tamponade, hypovolemia from exsanguination – pushing on the chest isn’t going to reverse any of those. But where is the evidence to support that claim? Don’t we always compress the chest when the heart has stopped?


There is, unfortunately, a dearth of data on this topic. Swami and I scoured the known literature and here’s what we found…

Lockey DJ et al. Development of a simple algorithm to guide the effective management of traumatic cardiac arrest. Resuscitation 2013; 84: 738-42.

The authors’ algorithm states that in all traumatic arrests, you should look for a penetrating injury to the chest or epigastrium. If you find one, crack the chest ASAP.

If you don’t see a penetrating injury, consider a medical cause for the arrest.

There are a few key procedures in trauma arrest and the study recommends using the HOT mnemonic.

  • H – Hypovolemia (control external hemorrhage, pelvic binding, long bone splinting and give blood)
  • O – Oxygenation (Airway management)
  • T – Tension PTX (Bilateral decompression and chest tubes)
  • Swami and I add in a second T for tamponade, making it the HOTT mnemonic.

If these three things are addressed and there’s no ROSC, you should consider stopping resuscitation. This simple list emphasizes a systematic approach so that you open the chest when it’s indicated without thinking about it too much.

Leis CC et al. Traumatic Cardiac Arrest: Should Advanced Life Support Be Initiated? Acute Care Surgery 2013; 74: 634-8.

Advanced life support in this article means advanced procedures, not ACLS and CPR. The authors argue that we should resuscitate patients in traumatic arrest stating that 49.1% attain ROSC and 6.6% have a good neurologic outcome, including 23.1% of children.

ALS here included IV fluids, intubation, chest tube insertion, pericardiocentesis and FAST in the field. There was no mention of chest compressions.

Bottom Line: The question as to whether or not closed chest CPR in traumatic cardiac arrest has not been well studied. In the literature we found,  chest compressions were rarely mentioned (if at all). Logistically, compressions can get in the way of life saving procedures and haven’t been shown to help (nor have they been shown not to help). Considering the pathophysiology of traumatic arrest, compressions don’t make a lot of sense in the emergency department, since pumping on a closed chest in a patient with no volume (hemorrhagic shock), a hole (pericardial effusion), or obstructed outflow (tension PTX) isn’t going to help.

Links mentioned at the end of the show

Find out more and register for ATLANTIS CME

Essentials of Emergency Medicine

SMACC 2015


  1. SAM GHALI (@EM_ResUS)

    Hi Rob,

    Great cast with Swami on a fantastic topic!

    I think one of the main issues here revolves around the following question: What the hell defines a “traumatic arrest” in the first place? I don’t think it’s very well-realized, but in most of these circumstances where the decision has been made to aggressively push forward with resuscitation of a “traumatic arrest”, we are in essence talking about “PEA”. Asystole is the common final pathway, and V-Tac/V-Fib happens, (especially in the “medical traumas” as you guys discussed) *but by and large these patients have what could/should be a perfusing rhythm, yet .. wait for it… someone can’t quite feel a pulse. This is a subset of “PEA”. The “Traumatic PEA” if you will (only with a much more narrow differential- generally, HOTT) Are we really going to continue to call these “arrests” just as we damagingly do in medical “PEA”?

    In my opinion, the term “PEA Arrest” is one of the most illogical and dangerous terms in all of medicine.

    I had a recent case at Resus General: Young lady brought in from scene s/p GSW to epigastrum. She was in shock with a pulse of 135; Non-invasive BP of 70/40. Tube placement confirmed via great End-Tidal Waveform, and b/l breath sounds. There was excellent sliding bilaterally (no pneumos), and no pleural fluid(no hemos). Cardiac view showed no effusion (no tamponade) & furthermore a hyperdynamic left ventricle. The heart was going CRAZY trying to pump out what little blood there was. While products were being hung, one of the team was feeling for a pulse and couldn’t quite feel one and announced this to the room. Another member of the team instinctively jumped on the chest and started closed chest CPR as if they’d done this drill time and time again before. I took over by stopping this right away. I was adamant. There was no way I was gonna allow this to happen. It turned out to be one of the greatest “ah,ha” teaching/learning moments for all. I put the probe back on the chest and showed them the walls of the left ventricle slamming together and pumping with the most vigor you’ll ever see a heart pumping-- I said “of course you don’t feel a pulse! This patient is in shock….hemorrhagic shock. She’s bleeding into her belly” I think at that moment, with my explaining as we watched the Echo, everyone in the room (*including the surgery team who completely bought into it as well) truly could see & grasp that the only potential effect chest compressions could have was to impair the filling of a heart that in no way needed help beating-- it needed products. It was like lightbulbs went on in everyone’s head. As products poured in her pulses quickly became much more palpable. A quick chest x-ray confirmed my ultrasound findings. Now she needed immediate hemorrhage control. She went straight up to OR for ex-lap w/ongoing resus and actually did great.

    This is how the “traumatic arrest” should go down. If she had pneumo(s)/hemo(s) we would’ve entered the chest. Tamponade? Would’ve cracked the chest. For the reasons you guys discussed I don’t think we will ever see “quality evidence” for this, but not sure we need any to know that the last thing this lady needed was someone pumping on her chest & pumping her full of epinephrine just bc someone at some moment couldn’t quite feel a pulse.. She didn’t even need her chest cracked.

    We will never see RCT’s on this. We must attain a better understanding of the pathophysiology taking place in the crashing patient in front of us, use logic & ultrasound to guide specific goal-oriented resuscitation, and realize the craziness & danger of using someone’s subjective finger sense to determine what the most (in)appropriate next course of action is.

    Rob- sorry for rant. This is one topic in Emergency Medicine I am most passionate about. Hope you’re doing well my friend.


  2. mwells786

    Hi Rob. On a similar theme to Sam, I was interested that in your podcast (and also in most of the other work I have heard / read on trauma arrests) you don’t comment on arrest rhythms. After all in a medical arrest pretty much the first thing we are interested in is the underlying rhythm as that defines our management.

    It seems to me that there is an underlying assumption that in most trauma arrest cases we are dealing with a very low output state (which we’d typically call PEA) -- and that it’s therefore entirely logical to immediately treat this in any way we can via the HOTT mnemonic. I think I would apply this to in a case with aystole but recent documented signs of life (bearing in mind that signs of life can be defined as electrical cardiac activity -- what happens in terms of time based indications / CIs to thoracotomy if a ‘PEA’ traumatic arrest deteriorates to asytole?). But what if I encountered a traumatic arrest with a VF rhythm? I’d find it hard not to defibrillate before moving to tackle HOTT. Or would this likely suggest a medical arrest which had then gone on to cause trauma?

    Very interested to hear other people’s views -- not least because I am in the middle of writing a traumatic cardiac arrest management protocol for my department, and I’ve got stuck at stage 1 -- rhythm recognition!


  3. Brian

    I am struggling a bit with the airway management portion of this, especially when in a very slow PEA or Asystole*.

    What is the benefit of managing an airway in a extremely low to no flow circulatory state? If there is an obstruction to blood flow or no blood to flow due to exsanguination what benefits are achieved?

    *I do see a point to oxygenation / ventilation in a rhythm that looks like it could be perfusing, but we just can’t detect it. If we believe a hypoxic event to have precipitated the arrest and there is a chance to circulate the oxygenated blood I get it, but this would require some sort of rhythm.

Awesome article, I know - please share your erudite thoughts...