Friday, June 14, 2013

House of God's Law #3 (Revised)

At a cardiac arrest, the first procedure is to take your own pulse...and nobody else's once CPR is begun. 
(adapted from The House of God by Samuel Shem)

The Gist:  Worldwide resuscitation guidelines de-emphasize pulse checks using palpation due to poor sensitivity and increased no-flow time (1).  Yet, many in-hospital resuscitations still pause compressions for pulse checks, which is associated with poorer survival outcomes (2,3).  New* opportunities exist, taken within the context of the patient, to gauge return of spontaneous circulation (ROSC) including end-tidal CO2 (etCO2) and ultrasound that may lessen interruption in chest compressions.

The Case:  During one of my first official shifts in an ED as a medical student, I participated in a well executed code at Janus General.  Each individual had a clear role and my task was to feel for a pulse.  I kept my hand glued to the patient throughout the code, frightened that I might erroneously call something that was or wasn't actually present.  Eventually, after buckets of fluid, some epinephrine, and sodium bicarbonate (suspected DKA with profound hyperkalemia, which turned out to be the case),  I felt a bounding pulse in our patient and spoke up, still worried that it was the reverberation of my own heartbeat.

This case brought forth three questions, one of which was easily answered:
1.  What about the significance of the time off the chest?
  • I've seen a great deal of interruption in chest compressions throughout codes, mostly for transfers or provider switches, pulse checks, and defibrillation.  I began noticing these interruptions more frequently after I discovered that there's a 50% decrease in chance of ROSC for every 10 seconds of hands-off time (4).  I addressed defibrillation here, but pulse checks seem like a non-controversial, more intuitive, and easier means for intervention.
2.  Why aren't we following the guidelines?
3.  Was I crazy to be so worried that I would mis-judge the pulse? We're physicians, surely we can feel a pulse, right?

The FOAM (Free Open Access Medical Education):  On a recent Broome Docs podcast, Dr. Matt Dawson provided reassurance that my fear was not unprovoked.  He recounts a pediatric case in which the discrepancy between cardiac echo and palpation for pulse were not aligned and how that changed the resuscitation.

Do people really still stop for pulse checks?  Yes, and more than we think.
  • National guidelines and popular emphasis on uninterrupted CPR may be executed well in the FOAM world, simulation, and many centers; however, chest compressions are still interrupted more than we think. 
    •  A recent article by Souchtchenko et al in the Journal of Emergency Medicine articulates this point and provides some points for intervention.  
  • We often underestimate interruptions in chest compressions.  This is likely multifactorial, due to recall bias, overconfidence, and cognitive load.  It's easier to gauge CPR quality on the periphery of a code, where our minds have a lessened cognitive load.  Furthermore, it's difficult to accurately gauge elapsed time in such situations, particularly when even a 5 second pause has deleterious consequences on cerebral perfusion pressure (3).  Also, we're more likely to look kindly upon our own skills and actions when we are closer in proximity to the action.   
    • A study (n=40 resuscitations) by McInnis et al showed a 30% non-compliance with AHA guidelines regarding compressions and also demonstrated that providers often failed to recognize interruptions in chest compressions.  In a survey, the code leaders explained interruptions in compressions mostly as a result of pulse checks (37%) and defibrillation (24%).  
  • Knowledge translation likely plays a role in the discrepancy between what the guidelines recommend and what we practice.  Some community institutions may be more susceptible to slow uptake and implementation of new/changing practices.  Also, CPR is often one of those times in which rescuers run off of their ingrained algorithms, which can be difficult to change in the heat of practice.  
  • Equipment/Training.  We may very well know what we're supposed to do, but this may not translate into execution.  Capnography and ultrasound are not available in all resuscitation areas, despite what the FOAM world may have one believe.  
Pulse Palpation Literature: During cardiopulmonary resuscitation (CPR), pulse palpation lacks sensitivity and specificity and typically interrupt chest compressions thus, the guidelines advise against it (1).  According to some studies, only 55% of rescuers can identify a pulse within 60 seconds (6).  Furthermore, most equipment in the hospital allows for rhythm checks during compressions thereby obviating the need for that pause (during which some providers historically integrated a pulse check).  So, even if we're not pausing compressions solely for a pulse check, checking for a pulse may not be that helpful (except to occupy an eager medical student).

Tibballs et al 2010 (full text) was similar to a studies in adults by Dick et al and Eberle et al.
  • 209 health care professional "rescuers" were asked whether a pulse was palpable or not in a cohort of pediatric patients receiving forms of extracorporeal life support where some forms generated a palpable pulse while others did not.  Information regarding pressures and pulse forms was obscured
  • Pulse palpation: Correct determination 78% of the time.  Rescuer sensitivity 86%, specificity 64%.  Often took up to 30 seconds.
  • Limitations: pediatric population, gold standard, site of pulse check, on ECLS (supposedly blinded)
What now?: Ultrasound and capnography (ETCO2) offer ways to predict ROSC with limited to no interruption of chest compressions.  Dr. Reuben Strayer has a detailed cardiac arrest vector demonstrating resuscitation flow with no pulse checks and an Academic Life in Emergency Medicine post offers insight into achieving successful cardiac arrest outcomes.

Ultrasound: This is well-covered by FOAM sources, especially after an article by Blyth et al.  Highlights include the following posts: BestBets (see for landmark articles), EM Lit of Note, ScanCrit.
If you're an EMRAP subscriber, check out the Jan 2013 discussion.
  • Lack of cardiac activity on point of care limited echo is associated with poor likelihood of ROSC with a pooled negative LR of 0.18 (95% CI = 0.10 to 0.31), and a positive likelihood ratio of 4.26 (95% CI = 2.63 to 6.92).
  • Use the echo in the context of the individual patient/resuscitation. Experts such as Dr. Matthew Dawson suggest that echo be used in two situations: 
    • (1) Rule out reversible causes such as tamponade or massive PE.  
    • (2) After rounds of CPR as an excuse to cease CPR when resuscitation seems futile (3).
  • Limitations:  typically involves interruption in chest compressions, very rare occasions exist when a patient may not have cardiac activity on ultrasound and go on to walk out of the hospital.
  • On the horizon? Transesophageal echo during CPR, which limits interruptions but takes far more skill (Blaivas et al).  
Capnography.  There are excellent posts by St. Emyln's, ScanCrit, and a short PK talk by Jon Schonert of EMChatter on this topic so I won't rehash these.
  • Given consistent ventilation, the partial pressure etCO2 (PetCO2) correlates well with cardiac output during CPR.  During CPR, the carbon dioxide in the lungs is primary delivered by cardiac output.  
    • A rise in PetCO2 to ~35-40 mmHg or a sudden rise of 10mmHg over baseline during CPR has found to predict ROSC (Pokorna et al).  
    • Similarly, PetCO2 levels <10mmHg have been associated with unlikely ROSC or need to improve CPR quality. 
  • There are limitations to PetCO2 in the context of CPR.  
    • Cause of cardiac arrest may alter the PetCO2.  A respiratory cause may increase PetCO2 whereas large pulmonary embolism may cause very low PetCO2 although this is not currently an acceptable diagnostic modality by itself (see this meta-analysis).
    • Administration of sodium bicarbonate may increase PetCO2.
*Use of "new" is relative, as literature such as Salen et al is over a decade old.
1.  Berg RA, Hemphill  R, Abella  B, et al.  2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science. Circulation. 2010; 122: S685-S705.
2. Cunningham LM, Mattu A, O'Connor RE, Brady WJ.  Cardiopulmonary resuscitation for cardiac arrest: the importance of uninterrupted chest compressions in cardiac arrest resuscitation.  Am J Emerg Med. 2012 Oct;30(8):1630-8. 
3.  Mattu A, Bond M, Tewelde S, Brady W.  The Cardiac Literature 2010.  Am J Emerg Med.  2012 May;30(4):615-25.  (full text)
4.  Eftestøl T, Sunde K, Steen PA.  Effects of Interrupting Precordial Compressions on the Calculated Probability of Defibrillation Success During Out-of-Hospital Cardiac ArrestCirculation.2002; 105: 2270-2273
5.  Dawson, M.  Comments on Notes from the Community: Cardiac Ultrasound. .Jan 24, 2013.  Available at Accessed June 13, 2013. 
6.Dick WF, Eberle B, Wisser G, Schneider T. The carotid pulse check revisited: what if there is no pulse?Crit Care Med. 2000 Nov;28(11 Suppl):N183-5


  1. Great post. Pulse checks are sooo last year ;)

    In our service area we no longer check for a pulse without an EtCO2 waveform and value which prompt us to consider checking the rhythm (a jump to >30, or >40 without a jump). Only when those jive will we check for a pulse.

    With really good CPR we've seen EtCO2's in normal ranges with VF, and I've not witnessed a ROSC that did not have a sudden change in EtCO2.

    I think the caveats for EtCO2 in cardiac arrest are overblown, and it works exceedingly well as a decision to continue compressions or to stop for a pulse.

    My last working arrest was a GI bleed and had an EtCO2 in the 20s during the arrest, with a sudden jump to 55 mmHg at first ROSC. He lost palpable pulses on other occasions but maintained spontaneous ventilations and an EtCO2 >30, indicating to us a volume state problem. With the presence of a narrow complex sinus tach, we knew fluids--and blood, 14 units in the ED--not CPR were going to regain pulses (his heart was pumping, CPR would not improve upon that).

    Lastly, EtCO2 is measured well via mask, SGA, and ETT.

    1. Awesome, thanks for sharing your experience with EtCO2 in arrests. In several EDs I worked in, the EMS folks had EtCO2 but the ED did not...we would borrow theirs if the resuscitation was in progress. Baffling, but probably more common than many of us would like to think.

  2. Hi, Lauren! Great post. I agree that we now have the technology to avoid needless interruptions in chest compressions for things such as pulse checks and defibrillation. I also agree with your comment above about it being more common than we think that more EMS systems have EtCO2 than EDs. "When will we start to follow the guidelines" is the ultimate question. One day, I hope. One day.

  3. Same here in Qatar! EMS got capnography before ED. And they let us keep it while we ran the code.

  4. Great post Lauren, thanks for revising!

    Even though I am a huge ultrasound belieber, my concern is that I often see that the time needed to obtain that proper view to look for cardiac activity leads to an even increased loss of compression time compared to looking for a pulse. Maybe TEE will be the answer but I guess until that time we should warrant loosing compression time while looking for cardiac activity.

    Maybe have that eager medical student hold a stopwatch and count down the 10 seconds the bedside ultrasonographer has during compressor switch to obtain that view. After 10 seconds the new compressor should start compressing and the ultrasonographer has another shot during the next compressor switch. Or in the skinnier population obtain a subxiphoid view during compressions, hold that probe in place during compressor switch and look for cardiac activity while switching.