The EMBER Project

Approaching Zero Interruption CPR

Building Better Cardiac Arrest Care – Part 4

By now it should be clear that many of the innovations in delivering cardiac arrest care revolve around doing the basics more effectively. The posts in this series have highlighted new tools to help us achieve a more physiology-driven approach to cardiac arrest care, but the strategy remain focused on maximizing coronary perfusion pressure (CPP) by delivering high quality chest compressions of appropriate rate, depth and recoil, and also minimizing interruptions in those chest compressions.

emr05122008_fig1The importance of this last requirement cannot be overstated. After the first few minutes of a VF arrest, myocardial ATP has diminished to a critical level. If coronary perfusion pressure (CPP) is not improved prior to defibrillation, the likelihood of getting return of spontaneous circulation (ROSC) drops precipitously

A core concept related to optimal CPP from chest compressions is that chest compressions are not a simple on/off switch when it comes to CPP. Interrupting them for even a short period of time causes a rapid drop in CPP. Once chest compressions are interrupted, it requires a significant amount of time to achieve pressures that are once again adequate enough for successful defibrillation to occur.


The reason compression only CPR works is based on this current understanding: that in most cardiac arrest cases coronary perfusion is more important than oxygenation and ventilation. Anything that interrupts compressions will have a large negative effect on CPP that extends well beyond the actual time of the interrupted chest compressions.

Reducing Rhythm Checks
After research into this critical understanding of CPP was published, by pioneers like Dr Gordon Ewy cardiac arrest algorithms slowly attempted to walk back their previous recommendations, and began recommending reduced interruptions for reasons such as breath delivery. But until recently, most cardiac arrest algorithms still require frequent rhythm and pulse check interruptions at arbitrary intervals that are unrelated to the physiologic condition of the patient. The problem is summarized and highlighted nicely here by R.E.B.E.L EM within the graphics below:

Two Pauses Before Using Precharging

Smart providers like @srrezaie of REBEL EM have realized that pre-charging the defibrillator prior to the rhythm check can streamline the first and second pauses into one event. Here is the new suggested sequence:


One Pause After Using Precharging

This ingenious and simple change in the algorithm reduces interruptions, but it still requires several seconds to check the rhythm at each two-minute checkpoint. If an organized rhythm is detected, then chest compressions may be interrupted for and even longer period of time while attempts are made to feel a pulse.

When possible, the physiologic needs of the patient should always be prioritized over the arbitrary steps of any algorithm.

This is arbitrary checkpoint, one created by the historic limitations of technology,  bears no relationship to the physiologic condition of your cardiac arrest patient and what their needs are at the moment of the rhythm or pulse check.

 Algorithmic care often exists when there is a dearth of data about the underlying problem. But “best guess” care is not ideal care. When possible, the physiologic needs of the patient should always take priority over the arbitrary steps of any standardized algorithm.


Filtering Software

Filtering Software

In cardiac arrest patients, knowing what rhythm the patient is in without  interruptions would improve our attempts to meet both the physiologic demands of the patient, and allow us to identify and treat a reversible cause of cardiac arrest without compromising either goal.

This is now possible: by using an accelerometer with integrated monitor/defibrillator pads, and combining it with filtering software, compression artifact can now be subtracted from an underlying cardiac rhythm. This allows for a shockable rhythm to be detected during active chest compressions without needing to pause for the usual rhythm checks.

The defibrillator may be charged as in the above approach, and then a quick hands-off period of under a second is all that is needed to deliver that shock before resuming compressions. If there is no shockable rhythm then there is no need to stop for rhythm checks at all, further reducing the arbitrary two-minute interruption cycle of the current approach. If the underlying rhythm is again a VF/VT then the process below can be repeated. (We will discuss the possible return of stacked shocks and how to address refractory VF in a later post).


With Filtering Software

Reducing pulse checks

A similar use of the filtering software combined with ETCO2 can be used for pulse checks. In the current approach, rhythm checks that show an organized rhythm will also involve a potentially extended pulse check, further creating and prolonging detrimental interruptions in chest compressions.

In the new approach, if you encounter what appears to be an organized rhythm, then the next step should not be to check for a pulse, but to check the ETCO2.  A significant rise in ETCO2 should encourage you to check a pulse.  Going forward, a zero interruption vision in chest compressions should be the goal.

The use of other markers beyond ETCO2 and pulse checks for determining a perfusing rhythm will be our next post.  Watch this space.

What Are They & How To Get Them

Building Better Cardiac Arrest Care – Part 3

High-quality CPR is the primary component influencing survival from cardiac arrest, but there is considerable variation in monitoring, implementation, and quality improvement. – The AHA


maxresdefaultThe phrase “high quality” chest compressions has taken center stage in cardiac arrest care. Research providing insights into coronary perfusion during chest compressions (and its central importance in achieving ROSC) makes them the foundation that all other cardiac arrest care is build upon. This post is all about what they are, and how to get them in the current real world of cardiac arrest care.

But what does “high quality” really mean? And how can we tell if we’re achieving our goal? Without being able to identify and correct the factors intrinsic to good chest compressions during an arrest, the phrase “high quality” remains merely a slogan, and leaves us with the traditional algorithmic model of: chest compressions, meds, pulse check, repeat. A combination of new tools, physiologic markers, and crisis resource management techniques are needed to improve the current delivery of this central piece of cardiac arrest care, and give the phrase “high quality” real meaning.

Identifying the Challenges

Chest compressions, more than any other procedure physicians are trained in, appear to be one of the easiest and most basic skills we learn.  It’s a common mistake made by providers early in their training to believe that being “certified” in the mechanics of chest compressions and the required ACLS algorithms, will equate with providing good cardiac arrest care (I know, I was one of them).  Nothing could be further from the truth, and here’s why.

Even If You’re Doing It Right, You Might Be Doing It Wrong

Anyone who participates in cardiac arrest care knows the innumerable barriers to good chest compressions: start with patient clothing, and body habitus, poor external landmarks, blood, sweat, other secretions all mixed in with a squirt of ultrasound gel to get things nice and slippery. Add space restrictions, poor working surfaces, bed height, stool placement, provider fatigue, and variability in chest compression skills along with a chaotic environment and competing distractions such as distraught family members, airway management, vascular access, medication dosing, and you have just some of the many issues that translate into generally ineffective chest compressions during an  arrest.

To make matters worse, even if we achieve “ideal” chest compressions based on current guidelines, we may still be doing it wrong. Studies using transesophageal echo (TEE) during cardiac arrest find that the area of maximal compression during standard CPR can often significantly obstruct forward flow, and that initial hand placement based only on standard external landmarks are frequently sub-optimal.

We’re Not That Good At Self-Correcting 

“Providers are notoriously poor at estimating whether their own compressions are too fast, too slow, too deep etc”  Excerpts From: “Cardiac Arrest.” Oren Friedman, 2016. iBooks. 

When it comes to providers assessing the quality of their own chest compressions, the truth is we’re bad at it. Combine this with the other barriers to effective CC’s mentioned above, and the resulting ubiquity of bad chest compressions (or at least inconsistently good ones) in cardiac arrest care is not such a mystery.

“The code leader needs to continuously monitor the overall quality of compressions, and ensure proper rotation of providers” Excerpt From: “Cardiac Arrest.” Oren Friedman, 2016. iBooks.

To overcome this problem, it has traditionally been the responsibility of the team leader to constantly monitor and direct the details of this essential element of cardiac arrest care, in addition to maintaining focused on the big picture: gathering information, managing the rest of the team, and all other aspects of resuscitation care. Even with task delegation, this approach to crisis resource management in a cardiac arrest can allow focus on chest compressions to dissipate, negatively impacting overall chest compression delivery.

We Can’t Fix it if We Don’t Know it’s Broken 

Finally, what we really want to know during CPR is how effective our efforts are in improving coronary perfusion pressure (CPP). Real-time, physiology-driven decisions during CPR makes sense, but as yet there are no simple methods of directly measuring CPP during most cardiac arrests. For now, how well we’re actually doing with CPP in our patients is anyone’s guess, but there are surrogate  physiologic markers of CPP that can give us some guidance. In addition, there are tools that can guide chest compressions and improve the mechanics of delivery.

So What Can You Do?

In part one of this series, I broke down cardiac arrest care into two core goals:

  1. Rapidly optimizing cardio-cerebral perfusion.
  2. Finding and treating reversible causes of cardiac arrest.

When the multiple demands of these goals compete or interfere with each other, even an experienced code leader can become distracted or overwhelmed: loss of focus on CPR, even for a short time, can rapidly lead to inadequate coronary perfusion and a diminishing likelihood of achieving ROSC. Complete focus lock on good CPR can delay potential life-saving treatment of reversible causes of cardiac arrest.


Even brief interruptions in chest compressions can negatively impact CPP

Managing these competing demands in cardiac arrest care is a crisis resource management issue, and should be no different from managing any other critical action in resuscitation care.

To me it makes sense (when possible) to divide these roles and assign a chest compression leader (CC leader) with the responsibility of optimizing and maintaining coronary perfusion. This CC leader should be told that their sole focus is orchestrating and maintaining high quality, uninterrupted chest compressions.

To be effective, this role should be clearly defined within your team training prior to a cardiac arrest, and a qualified provider should be identified early on within the resuscitation. Ideally this person should have a clear grasp of all the factors that determine good chest compressions:

  • Proper hand placement, rate, depth, and recoil.
  • Height of the stretcher & stool placement
  • Proper rigid board placement if needed
  • Monitoring for fatigue and CC team rotation
  • Placement and managing of mechanical devices
  • Understanding of chest compression fraction

The “CC leader” can then independently direct the “CC providers” to improve their performance and adjust their efforts appropriately based on the available real-time feedback tools.

If possible, identify and assign a chest compression quality leader early on in your resuscitation.

I’m unsure why, in my experience this is done so infrequently. Perhaps it’s an integral part of cardiac arrest care in other places. If so, please share your experience with me. My belief is that this is another unintended consequence of algorithmic care in cardiac arrest training: where most code leaders feel they need to be in charge of the resuscitation based on the requirements of ACLS, where “running the code” means CPR.

Whatever the historical reason for this, assigning someone to manage the procedure of chest compressions during a code should be no different from assigning a qualified provider to manage the airway, or place a central line. All successful resuscitation care starts with getting control of the room: by assigning critical roles to team members and communicating effectively the chaos is reduced and the cognitive noise is subdued.

This model also fits smartly with the concept of cognitive offloading (presented so well recently by Salim Rezaie MD @srrezaie) during cardiac arrest care. An idea that was also highlighted in the first post of this series. Assigning this first core goal of cardiac arrest care (the optimization and maintenance of cardio-cerebral perfusion) in the same manner that airway management and other critical roles are assigned, frees the team leader to focus on the second core goal: finding and treating reversible causes.

Do I Really Have “High Quality” Chest Compressions? Getting Real-Time Feedback

Without the arrest team being able to identify and correct the factors intrinsic to good chest compressions in real-time, most CPR delivery during a resuscitation will be bad – or at least inconsistently good. 

Assigning a CC leader early in the resuscitation addresses the crisis resource management issues of delivering high quality chest compression, but how can we know if our “high quality” work is translating into improved coronary perfusion?

While there’s no way to directly measure CPP, there are several tools that can be employed to assess the effectiveness of our efforts.  The ones listed below are easily integrated int current cardiac arrest care, are readily available and accessible for most providers, and may already be in your department.  Being able to use these tools effectively to assess “quality” and adjust your approach accordingly should be an important skill for the experienced CC leader, and become a larger focus of advanced cardiac arrest training in the future.

The CPR Dashboard (feedback for the CC provider)

The Zoll-R now has a dashboard that provides detailed real-time visual feedback on CPR quality. The rate, depth, and release of each compression is captured with a sensor which is built into the pads. The system then gives visual and audio prompts to guide chest compressions.

While this tool can be fooled by factors like soft hospital bed surfaces, it’s a welcome addition in giving the code leader, the cc leader and the active cc provider real-time feedback on how well they’re doing.

ETCO2 (feedback on perfusion)

coetco2The best tool we have currently for assessing coronary perfusion is the use of ETCO2 as a marker of blood flow through the lungs during CPR. This tool is now widely available and a realistic option for most providers. Along with the chest compression quality dashboard, the CC leader can use this information to guide the teams attempts at improving coronary perfusion.

In much the same way we’re trained to adjust something in our approach to a difficult intubation  after our first attempt fails, ETCO2 in cardiac arrest can give us real-time feedback on the quality of our CPR.  If ETCO2 is not rising above 15-20 mmHg or more, change what you’re doing! For example. larger patients may need deeper compressions; backboard placement may give added efficiency to your CPR and adjustment of hand placement may change the point of maximal compression to align better with the ventricles.

Ultrasound ( feedback on hand positioning)

TEE is not yet a realistic option for most emergency departments or in hospital cardiac arrests, but bedside TTE (beyond the extended FAST exam to search for reversible causes of PEA arrest) may be a useful adjunct to give the team feedback on the quality of chest compressions. (For more on TEE and its potential use in physiology-driven cardiac arrest care, I highly recommend viewing this wonderful presentation by Felipe Teran MD @FTeranmd)

TTE ultrasound for this reason during cardiac arrest should never take priority over uninterrupted chest compressions, but probe placement in the sub-xiphoid or apical view may help optimize hand placement by assessing the area of maximum compression. Compressions that clearly land on the aortic root or LVOT should be moved caudally over the ventricles. I add this to my cardiac arrest care when, despite optimizing chest compressions, the ETCO2 does not rise.

Chest Compression Fraction

Chest compression fraction is a relatively new term in the CA world to describe the proportion of time chest compressions are done by providers during a cardiac arrest. This is also integrated into the CPR dashboard, and may be the most important real-time feedback tool the new system offers, because it’s a measure of how well your team is doing in minimizing chest compression interruptions.

Increased chest compression fraction is independently predictive of better survival in patients suffering a prehospital ventricular fibrillation/tachycardia cardiac arrest. Circulation. 2009 Sep 29; 120(13): 1241–1247.

This is FOAMtastic

cover225x225Finally this is a great place for me to introduce you to another fantastic FOAM resource on cardiac arrest care by Dr. Oren Friedman @OrenFriedman. It’s geared toward the expert resuscitationist, and covers the latest thinking on everything from chest compressions to ECMO. Oren also addresses much of the teamwork and communication skills required to turn the ABCs of cardiac arrest care into high level resuscitation skills. It’s an incredible new resource and a great jumping off point for the many of the posts in this series, including this one. So go ahead, read it.  I’ll wait…

That’s it for now. Our next topic in this series will be on how to minimize interruptions in chest compressions during a cardiac arrest.

Watch this space

Building Better Cardiac Arrest Care – Part 2


For many years the approach to patients in cardiac arrest has been held hostage by algorithmic care that stifled innovation and did nothing to improve overall survival. Now a combination of new technology, and a realization that one size does not fit all, has led to innovative approaches to the two goals of cardiac arrest care:

  1. To rapidly optimize cardio-cerebral perfusion.
  2. To find & treat reversible causes of cardiac arrest.

Like the previous post on the Lucas device & Cognitive Offloading, this is part of the “Building Better Cardiac Arrest Care” series that will highlight both new tools and new concepts that move us away from algorithmic care in favor of approaches designed to fit the needs of local environments – where decisions are driven by real-time feedback on the quality of our CPR, and a more rapid assessment of the patient’s pathophysiology.

Today, we’re going to introduce the Zoll-R defibrillator by sharing an incredibly useful resource by my friend Dr Jim Horowitz. This iBook is filled with useful tips and videos on how to use the device (neither of us have a financial interest in Zoll), and over the course of the series Jim will join us to highlight effective strategies for integrating the new features of this tool into cardiac arrest care.

What’s New

If you’re already familiar with the new Zoll-R, you know it provides all the same features and has the same “knobology” as older models but with three big changes:

  • An accelerometer that is placed on the chest with the pads
  • Integration of the monitor leads with the defibrillation pads into one unit
  • Filtering software that allows you to see the patient’s underlying cardiac rhythm.


Combined with a new dashboard, integrated ETCO2, and the standard features of a monitor/defibrillator the new machine can give you detailed real-time feedback on the quality of your resuscitation in ways not previously available.

Over the next few posts we will discuss how these new tools come into play in cardiac arrest care, and how to use them to optimize cardio-cerebral perfusion by: assessing quality of chest compressions, minimizing interruptions, and streamlining rhythm and pulse checks.

— Check this space…

Zoll-R Quick Reference Guide:  4-1da-r-series-quick-guide

Building Better Cardiac Arrest Care – Part 1


For many years the approach to patients in cardiac arrest has been held hostage by algorithmic care that stifled innovation and did nothing to improve overall survival. Now a combination of new technology, and a realization that one size does not fit all, has led to innovative approaches in care.

The result is an opportunity to move away from the status quo in favor of approaches designed to fit the needs of local environments – where decisions are driven by real-time feedback on the quality of our CPR and the patient’s pathophysiology in order to achieve the two core goals of advanced cardiac arrest care:

  1. To rapidly optimize cardio-cerebral perfusion.
  2. To find and treat reversible causes of cardiac arrest.

The Building Better Cardiac Arrest Care series is about physiology-driven resuscitation in cardiac arrest care, highlighting new concepts and new tools to improve our approach to these patients.


If you asked me about mechanical CPR in the ED a year ago, I would have said, “why would I want another tool cluttering my resus bay that hasn’t been shown to improve outcomes? Well, we recently got a Lucas CPR compression system in our ED, and its arrival has coincided with a great post by Dr Salim Rezaie on cognitive offloading (using physical action to alter the information processing requirements of a task to reduce cognitive demand) during cardiac arrest. So I’ve decided to put a discussion of the two together, since I think there is no better way to frame the argument for using one than Dr Razaie’s post.

Within the choreography of a resuscitation, multiple critical actions need to occur, but which ones?  Each action we take is a calculated choice. With finite time and cognitive bandwidth, every action we say yes to is also concomitantly a no to others. Small changes in the choices we make to achieve our goals during a resuscitation have the potential to significantly impact the quality of our cardiac arrest care. 


Beyond ACLS: Cognitively Offloading During a Cardiac Arrest

beyond-acls-765x575That’s why I love posts like Salim’s on cognitive offloading during a cardiac arrest. He’s taken the time to deconstruct a standard ACLS approach with the goal of reducing our cognitive burden to give us a better chance at rapidly transitioning to the important task of defining the problem behind the cardiac arrest.

We all know the H’s & T’s and the importance of reviewing potentially reversible causes of cardiac arrest. It’s also no mystery that the faster you can get to thinking about them, the faster you can make lifesaving decisions about care.  But if the basic requirements of the ACLS algorithm keep you incessantly occupied by a multitude of details that demand your full attention (monitoring for quality CPR, issues with IV access, repeated medication dosing, time wasted on prolonged pulse checks) then how realistic is it in the real world of cardiac resuscitation to expect you are going to have enough time to find the cause and reverse the problem?

And what if you’re working in a resource poor environment with too few hands, or you have a patient with difficult access, or is 400 pounds and requires herculean strength to maintain high quality CPR?  Well then, you may never get there at all – or at the very least your arrival may be significantly delayed.

Finding a better pathway to that cognitive space is Salim’s goal. His solution? Leverage the concept of cognitive offloading to get you there faster by rethinking the basic tasks required for optimal perfusion during CPR so you have more time to think.  To me this makes a lot of sense.

A rapid sequence review of recommendations for cognitive offloading during a cardiac arrest

Now on to the Lucas – With Dr Jim Horowitz

Which brings us to mechanical CPR. It turns out that about the same time Salim posted we were getting familiar with our new Lucas device.  It’s benefit is not simply replacing the physical work of CPR with a machine, but reducing the cognitive work needed to ensure your team maintains high quality chest compressions throughout a prolonged resuscitation: watching for provider fatigue, calling for new CPR providers, ensuring the right depth, rate, and quality of compressions, and directing the CPR providers throughout a code are all tasks that distract a team leader.

For Jim (our VTE and ECMO expert and my favorite cardiologist to have at the bedside during an arrest) the benefits of offloading CPR are obvious: it means more time to initiate ECMO.  And as he mentions in the video, mechanical CPR tends to make codes quieter, and makes placing lines and intubation easier during active CPR. This is significant offloading in action and can reduce distractions or delays in getting to that all important cognitive space.

I wish I’d had the Lucas 2 in some of the rural hospitals I’ve worked in, where it was often me and one nurse on an overnight, and I had to grab the clerk to help with CPR. I was lucky if I could get a LMA and an IO in quickly enough to take my turn doing CPR.

Cognitive offloading is something most good Emergency Physicians do intuitively to get through their day, but the concept was never explicitly taught to me during my training. I vote that it should become a core content lecture for every residency program in the country.

More to come.

Thanks to Dr Jim Horowitz for coming and demonstrating the Lucas 2 device to our residents and faculty. You can also download his iBook manual for the Lucas 2  for free here.

(None of us have any conflicts of interest with this device).


%d bloggers like this: