CPR Induced Consciousness: 30 Second Video


An 82 year old healthy and high functioning gentleman had been feeling unwell for several days with progressive dyspnea. He walked into the emergency department with his son, sat down in a chair in ambulatory triage and promptly became unresponsive. He was wheeled into the resus bay where he was apneic and pulseless with a PEA rhythm. Cardiac arrest care included chest compressions, laryngeal mask ventilation, epinephrine, calcium, and bicarbonate, without ROSC. Ultrasound initially showed no cardiac activity which progressed to minimal cardiac activity without palpable pulses, ECG demonstrated ST elevations in V1 and V2. 50 mg alteplase was given without improvement of echo or ROSC; a second 50 mg dose was given 15 minutes later which also had no discernible effect.

Throughout the period of cardiac arrest care, including during periods of cardiac standstill, the patient became awake during chest compressions and opened his eyes, tracked staff, made purposeful movements toward the laryngeal tube and toward the person doing chest compressions as well as flailing, distressed lower extremity movements. 100 mg ketamine was given to good effect, repeated once. After 75 minutes, the decision was made with the patient’s son that further resuscitative efforts were unlikely to be successful. Chest compressions were halted, and the patient was pronounced deceased.


CPR-induced consciousness (CPRIC) was described in Lewinter’s 1989 case report out of Henry Ford Hospital; the authors were more concerned with competing models of chest compression physiology and venoarterial carbon dioxide gradients but report that “forward blood flow produced was sufficient to maintain consciousness.”  CPRIC was more explicitly addressed in 1994 by Quinn and his colleagues at the University of Ottawa who were studying an active compression-decompression device but conclude with a question:

“Is it inhumane to do CPR without sedation or analgesia? These agents were used in our case because of the need to restrain the patient and to facilitate intubation. Are there other patients who are aware of the resuscitation, but who cannot express their fears or discomfort?” 

One of the papers’ reviewers comments:

“The details of the patient’s resuscitation reflect several interventions that should be questioned. First, the large dose of midazolam administered may have reduced vasomotor tone and reduced coronary perfusion pressure.”

to which the authors respond:

“To paralyze and not sedate our patient would have been inhumane. Midazolam was used for sedation because it has less myocardial depressant effects than thiopental. Ketamine would have been the best choice in hindsight.”

We may never know how frequently cardiac arrest patients are aware during resuscitation, because few cardiac arrest patients live to tell the tale, however 2% of cardiac arrest survivors in the subtly-named AWARE study described visual awareness during cardiac arrest. In 2008, Bihari & Rajajee described CPRIC in the ICU and declare:

“While concerns about the adverse effects of sedative and analgesic medications in such a situation may have merit, it is worthwhile to consider that a patient in this situation is quite unlikely to survive despite maximal resuscitative efforts…Alleviating pain and suffering at this likely terminal stage should outweigh concerns about potential side effects.”

The incidence of CPR-induced consciousness seems to be increasing and at least one investigator is making it a clinical niche with three papers so far in press. The best of these is a registry study of 112 prehospital cases of CPRIC showing a 0.7% incidence, a weakly positive association between CPRIC and favorable patient outcomes, and a weakly negative association between the use of sedatives to treat CPRIC and favorable patient outcomes. A provider survey study described a prehospital scenario where “The patient was GCS 15 during CPR which was very unsettling for the crews involved. At one point the patient was begging us to stop as the CPR was hurting. He said ‘tell my family I love them and I’m ok with dying, please let me die.’”

Clinicians who care for cardiac arrest patients should be aware of–and have a plan for–CPRIC. Members of the resuscitation team may be alarmed by a patient who is conscious during chest compressions and should be reassured that this is a well-described phenomenon demonstrating the potential of high quality chest compressions and high quality cardiac arrest care to perfuse vital organs. As captured in the video, our patient was conscious during pulse checks (i.e. while chest compressions were paused), illustrating how perfusion pressures fall gradually following the cessation of cardiac output.

It seems inhumane to leave a patient awake while an artificial airway is in place and while the patient is receiving chest compressions, not to mention tube thoracostomy, pericardiocentesis, etc. Pound et al describe a patient so active during CPR that he was able to significantly interfere with resuscitative efforts. Most CPRIC authors recommend intra-arrest sedation with midazolam or ketamine; one group proposes a CPRIC sedation protocol.

After the decision was made with family to terminate resuscitation, our patient was allowed to emerge from ketamine dissociation to say goodbye to his loved ones while at least partially conscious. This emotionally charged circumstance has not been previously described but may merit further consideration as cardiac arrest care improves and CPR-induced consciousness becomes more common.

Thank you to the patient’s relatives and the clinicians who consented to the distribution of video footage for educational purposes.

The Causes of Cardiac Arrest

It has been repeatedly demonstrated that the use of drugs empirically does not work in cardiac arrest. Despite this, ACLS often feels like an algorithmic march down a menu of pharmaceuticals, one drug after the next, hold compressions, ok still asystole, have we tried bicarb? Until everyone has had enough and the patient is pronounced.

There are three things you can do to benefit your patient in cardiac arrest. The first is good chest compressions, which serves mostly to buy time. The second is shocking shockable rhythms, and the third is identifying and treating the cause of cardiac arrest. The use of a vasopressor in the right dose probably offers some benefit; unfortunately we can’t figure out what the right dose is and epinephrine 1 mg every few minutes is almost certainly way too much.

Which drug is next keeps you busy and is the focus of how ACLS is taught, which is why it’s so dangerous: it feels right but not only does not benefit the patient, which drug is next distracts you from the most important task in cardiac arrest, which is figuring out why the patient in front of you just died, so you can offer a treatment to address it.

Causes of CA

Dysrhythmia is a central cause of cardiac arrest; as soon as a cardiac monitor is available, check a rhythm and if shockable, shock immediately. If you’re not sure if the rhythm is shockable, shock. The harm caused by an unindicated shock is trivial compared to the harm of untreated pulseless vtach or vfib.

Acute coronary syndrome is a hugely important cause of cardiac arrest, but causes cardiac arrest predominantly by dysrhythmia, which you will fix by checking the rhythm early and often. ACS can also cause cardiac arrest by pump failure, these patients may benefit from emergent reperfusion and would ideally be in a cath lab, if we knew that acute coronary artery occlusion was the problem. If you think ACS is the cause of arrest, cath lab is not an option, and the patient seems salvageable, give a thrombolytic. The data is equivocal in this context but I think the majority of resuscitationists would offer it for hearts too good to die.

Any sufficiently severe insult to airway or breathing can cause cardiac arrest, fortunately they are all treated at once with 100% FiO2 through an endotracheal tube. If the cause of cardiac arrest is unlikely to be primarily A or B, intubation can be deferred in favor of a supraglottic device; however, if A or B is the problem, all efforts should be directed at establishing oxygenation, usually via endotracheal intubation. Diverting your finite cognitive powers to ACLS drugs in respiratory arrest is particularly poor care.

There are three common, crucially important obstructive causes of cardiac arrest. Pericardial tamponade and tension pneumothorax are easily excluded and should be explicitly excluded in every case of undifferentiated arrest. Massive pulmonary embolism is amenable to reperfusion therapy, usually with intravenous thrombolysis–consider risk factors, exam signs, and ultrasound the legs and RV when the context fits.

Hyperkalemia is common, immediately lethal, and readily treatable. Calcium should be given to most arrested patients with kidney disease, and hyperK is a great reason to push for point of care electrolytes testing in your shop. Hypokalemia is much less common but also easily treated and a particular consideration in malnourished patients. Likewise hypoglycemia–fingerstick testing is routine if POC blood gas not available, and empiric use of D50 is reasonable in diabetics or liver patients.

Hemorrhage is of course most relevant in trauma, which may be occult, but the peritoneum is easily sonographed and uncrossmatched blood remains one of our best therapies. Ahemorrhagic hypovolemia is even more treatable though an uncommon cause of arrest in developed countries.

There are several key antidotes to important toxins, intralipid is a rising arrest therapy to consider if a cardiotropic medication is suspected; bicarb, digibind, and hydroxocobalamin may also restore circulation in the poisoned patient.

Hyperthermia is hopefully easily identified; treatment can be tricky but should involve some form of icewater in most cases. Hypothermia is an important cause and effect of cardiac arrest and determining which can sometimes be a challenge.

Anaphylaxis as a cause of cardiac arrest will be treated by airway management and the epinephrine you’re giving anyway. Once the patient has arrested from sepsis or an aortic or intracranial catastrophe, it is unlikely that you will be able to meaningfully reverse the process.

Though excellent, step-by-step algorithms exist, it can be difficult to remember this differential while managing an arrested patient as well as your team of providers which, when large enough, invariably obstructs rather than augments care. However, once you’re used to taking care of pulseless patients, the differential can be streamlined.

Causes Highlighted

You’re not going to forget to look for and shock shockable rhythms or intubate. Hyperthermia and hypothermia announce themselves, anaphylaxis will be treated with your supportive therapies and the last three causes are more or less irreversible once the patient is dead. This just leaves eight causes which are treated with six therapies. Six is the perfect number of items for a mnemonic.


If you’ve ever been to Antwerp you know it is gloomy as hell, and you would never, ever carry bling through gloomy Antwerp. Needle, calcium, blood/fluid, thrombolytic, glucose and antidote are the six things to remind yourself to consider in the arrested patient. It’s the Hs and Ts, evolved for our post-ACLS age.

Lastly, if you are able to put arrested patients on bypass/ECMO, you are playing a different game than the rest of us and that game has a different set of rules. Come up with your own mnemonic that includes Antwerp.

Resuscitative Hysterotomy in the Emergency Department

Guest post by Greg Press


The second perimortem C-section of my career happened last week. The first was almost ten years ago. While I’m hardly an expert (takes three to be an expert), I have some pointers worth sharing.

The first case happened on the first day of my first job, in Houston, having just completed my ultrasound fellowship in New York. The night started with a guy shot in the right chest: intubation and chest tube. His girlfriend was brought in next, pregnant, shot in the head: perimortem c-section and neonatal intubation.

The second was just the other day. It was my first code at my new job at a small hospital in Wellington, New Zealand. A young girl hanged herself, was found by family who called EMS. They found in her VFib, defibrillated, started CPR, intubated and called to warn us of her arrival. They reported she was in asystole, a bad sign for mom. They reported she was 22 weeks pregnant, a bad sign for baby.

There have been a few PMCS reviews circulating recently in the EM world and I agree with most of their recommendations. But I have differing thoughts on a few points.

Do not memorize the number 24. Or 23. Or 25, or whatever the gestational age your neonatologists say they can currently save premies. You are unlikely to know the precise age–the mother is generally the most reliable source for this information, but generally not in this circumstance. We understood the hanged woman to be 22-weeks pregnant; post-mortem estimates placed the baby at 26 weeks. More importantly, you are not doing this procedure for the baby: PMCS is a resuscitative intervention for the mother. You’re trying to save mom, first and foremost, and then baby, maybe. So, whether the gestational age is 22, 24, or 26 weeks is irrelevant.

But you don’t want to do this procedure for a first-trimester pregnancy, so when do you do it?

  1. When you have a reported gestational estimate anywhere near viability. Or…

  2. When the tummy is big. If the fundal height is above the umbilicus, great. If this assessment escapes your mind in the chaos of the moment, just recognize that the tummy is big. Or…

  3. When the baby looks big on ultrasound. Ultrasound does not need to be performed routinely, particularly if the mother is bursting with baby. But if you’re not certain, put the probe on the belly, not to measure biparietal diameter, just to see is this a big baby? You will probably not be able to resist the urge to look for a fetal heartbeat, but do not delay the intervention of interest.

Do not memorize the number 4. This is the supposed number of minutes after maternal arrest when we see a precipitous decline in fetal neurologic outcome and survival. The number 4 is even more irrelevant than the number 24, because, again, the procedure is done primarily for the mother. And the procedure is done as soon as possible. As soon as possible is the only time consideration.

So what is most important to know? There are two things you must do and a few nuances to keep in mind.

Assuming you have some forewarning, the two things you must do are GET HELP and GET STUFF.

Call the obstetrician.  Call the NICU, or PICU, or whomever can help you take care of the baby.

Get stuff for mom. Get a kit. You may only have a thoracotomy kit or some other all purpose ED kit – get it. If you don’t have a kit, get a scalpel, real surgical scissors, towels, and clamps for the cord.

Get stuff for baby. Get a baby warmer. Get a neonatal BVM. Get a neonatal intubation kit, and an IO, and the dose of epinephrine.

Assign someone to manually displace the uterus to the left. Tell this person she must do this nonstop until the scalpel hits the skin. And then she must immediately shift focus to assure CPR continues until all efforts cease. There will be others responsible for CPR, ideally a line of people performing CPR, but it’s easy to neglect once scalpel hits skin, and so it is important to have someone assigned as insurance just at the moment when this oversight is most likely to occur.

Forget the mommy-tilting ramp. It will be hard to find the ramp, harder to jerry-rig one, and harder still to perform CPR and the procedure on a tilt. Assign a tummy-pusher.

The patient will arrive, and if she’s arrested you will have to secure the airway, perform CPR, obtain IO or IV access, give drugs, possibly defibrillate. This will seem simple because you do it all the time.

And then you will have to perform the perimortem C-section. Perhaps an obstetrician has by now arrived. If not, two thoughts might cross your mind: there’s very little to gain and I’m not going to be a hero. Erase both of those thoughts; there’s even less to lose and it’s time to be a hero.

Cut a vertical incision from the top of the belly’s curve to the pubis. With an earlier pregnancy and heftier mother, there may be fat to get through to the peritoneal wall – you can use your fingers as claws to bluntly dissect a parting to the peritoneum. Then cut through the peritoneum vertically, ideally with scissors (you can use the scalpel to initiate an opening inferiorly). Pull out the big uterus and cut it open vertically in the same manner until you get the baby out. Be careful not to cut the baby.

When the baby comes out it will be very alarming. If it is very premature, it will look like a purple alien. The first time I saw this it was disturbing. The second time I was quite aware how disturbing it was for everyone else in the room. The cord will have to be clamped twice and cut in between. Grab the baby, wrap it in a warm towel, bring it to the warmer you have prepared. Most likely you will immediately start bagging the neonate, unless it is near-term and crying. Look for signs of life such as attempts to breathe, cry or move. Both times for me there have been none. If there is no heartbeat or signs of life, start CPR. To determine if there is a heartbeat, listen with your stethoscope over the chest or feel for umbilical pulsations. If there is a heartbeat and it is slow (<60 bpm) just bag the baby, if it doesn’t improve quickly (30 seconds) start CPR. If the eyelids are fused it is probably too premature to be viable. If the baby is an early premie and bagging and CPR do not resuscitate the child I would stop efforts. If it is a full or near-term baby that cannot be resuscitated by bagging and compressions, intubate, obtain IO/UV access and start in with epinephrine and advanced neonatal life support.

Once more: the delivery is a resuscitative intervention for the mother first and foremost. Make sure maternal CPR and resuscitation continue throughout and after delivery. Don’t let the splayed open belly and uterus concern you, just pack them with towels, it will bleed less than you anticipate. Mom is likely young and healthy. After a few minutes our mother had a strong pulse and a good blood pressure. She is alive in the ICU as of this writing.

Despite having read this and feeling prepared, expect a shit storm. The unique scenario of your case will bring unanticipated events. Get help and get stuff (mommy and baby stuff) early. Assign the tummy-pusher-turned-CPR-advocate. Cut the tummy, cut the uterus, cut the cord. Expect to bag the baby and perform compressions unless near term and breathing. Continue efforts with mom until after the procedure.

Finally, I would like to comment on a dramatic difference between the Houston and New Zealand experiences. Following the case in New Zealand, there have so far been two debriefing sessions. Both have been held to address the clinical and emotional aspects of such a difficult case. In Houston, ten minutes after PMCS we moved on to the next case, another critically injured patient, and that was that. Everybody involved in the New Zealand case was invited to the second debriefing session, from the ambulance crew to the students to the doctors and nurses from the ED, obstetrics and the ICU. A psychologist ran the session, and though I was skeptical, she was quite good. Was very interesting to hear how students, pregnant nurses, old cranky doctors and everyone in between felt about the experience. It was just as important to air out clinical concerns and perspectives unique to each specialty, to avoid frustrations and move closer to optimal care for next time. I suggest you try to make this happen for the next once-in-a-lifetime case you experience.

Greg Press

A correspondence from New Zealand

From Rose 2015 as summarized by Cabrera

Approach to Undifferentiated Cardiac Arrest

The resuscitation of a patient who is in cardiac arrest of unknown etiology in PEA or asystole proceeds with the simultaneous management of the ABCs and addressing the likely elements in the differential. Think Hs and Ts.

– Hypoxia (Place LMA or ETT, provide 100% oxygen)
– Hypovolemia (Bedside ultrasound for free abdominal fluid/AAA, consider a NS bolus – if suspicion for hemorrhage is high, administer uncross-matched blood)
– Hypo/hyperkalemia (Consider calcium chloride, especially in the patient with suspected renal insufficiency)
– Hypoglycemia (Consider D50)
– Hypothermia (Warm the cold patient)
– Hydrogen ion/acidosis (Consider bicarb, especially if toxicology is suspected)
– Toxins (In addition to bicarb, consider empiric antidotes – cyanide kit, digibind, naloxone, intralipid)
– Tamponade (Perform bedside ultrasound)
– Tension pneumothorax (Perform bedside ultrasound / consider needle or tube thoracostomy)
– Thrombosis (Consider thrombolysis for AMI or, especially PE)
– Trauma (The entire cranium and posterior thorax should be inspected if occult trauma is entertained)

Although the administration epinephrine (and, formerly, atropine) is emphasized in codes, these agents do not benefit arrested patients. Your only chance to reanimate a patient in PEA or asystole is to reverse the underlying cause. Take your own pulse, take a step back, and think Hs and Ts.