Episode 34 – The Death of Mechanical CPR (mCPR)?

Background: The two most important things that we can do in cardiac arrest to improve survival and neurologically intact outcomes is high quality CPR, with limited interruptions and early defibrillation. In the case of the former, the 2015 AHA/ACC CPR updates recommended a compression rate of 100 -120/min, a depth of 2 – 2.4in, allowing full recoil, and minimizing pauses. This is a lot to remember during a stressful code situation and one way many providers are offloading themselves cognitively is by the use of mechancical CPR (mCPR) devices.  In theory these devices compress at a fixed rate, and depth, with the added benefit that the machine simply does not tire out.  Additionally, use of this device allows another provider to be available for other procedures and interventions. A recent systematic review and meta-analysis in  looked at five randomized clinical trials with over 10,000 patients with out-of-hospital cardiac arrest (OHCA) (Gates 2015).  They concluded that there was no difference in ROSC, survival or survival with good neurological outcomes with the use of these devices compared to manual CPR. It is important to state that none of these studies showed increased harm either.  A new paper just published in Circulation however, argues that mCPR during OHCA was associated with lower neurologically intact survival.

Episode 34: The Death of Mechanical CPR (mCPR)?

Click here for Direct Download of Podcast


  • Buckler DG et al. Association of mechanical cardiopulmonary resuscitation device use with cardiac arrest outcomes: a population-based study using the CARES registry (Cardiac Arrest Registry to Enhance Survival). Circulation 2016; 134: 2131-33. PMID: 27994028

Clinical Question: How does mCPR effect outcomes in OHCA?

What They Did:

  • Cardiac Arrest Registry to Enhance Survival (CARES) is a US national registry of OHCA
  • This registry was analyzed for adults with non-traumatic OHCA
  • Compared patients treated with mCPR vs manual CPR


  • Primary: Neurologically favorable survival at hospital discharge (Defined as Cerebral Performance Category of 1 or 2)


  • 80,861 patients were included in the analysis
    • Median age was 62 years
    • 1% received bystander CPR


  • Used a multivariable logistic regression model to control for arrest characteristics (Including: age, arrest location, bystander CPR, AED use, witnessed arrest status, initial rhythm, post arrest targeted temperature management, successful advanced airway placement, and impedance threshold device use)
  • Large, real life look at the use of mCPR in OHCA as opposed to prior work that was done in research setting
  • Primary outcome is clinically important and patient centered


  • Retrospective secondary analysis
  • Time of arrest, time of first CPR, and timing of interventions were not reliably reported
  • Interventions aside from mCPR not detailed here (with the exception of advanced airway and ITD)
  • Large variations in use of mCPR use by various EMS services (Ranged from 21.7% to >75% of OHCA cases)

Author Conclusion: “In conclusion, the use of mCPR during out-of-hospital cardiac arrest was associated with lower neurologically favorable survival within emergency medical services agencies participating in CARES.  Although the use of mCPR devices increased during the study period, deployment rates remained highly variable, and the majority of agencies did not use them.  Further research is required to identify circumstances in which mCPR may benefit patients with out-of hospital cardiac arrest; however our data indicate that mCPR for routine cardiac arrest care was associated with worse outcomes.”

Special Guests:

Be sure to checkout Joe’s Rant on Mechanical CPR (mCPR) on the Podcast

Joe Bellezzo, MD
Chief, Department of Emergency Medicine
ED ECMO Director
Ultrasound Director
Sharp Memorial Hospital
San Diego, CA
Conference Director: Reanimate
Twitter: @edecmo

Scott Weingart Weighs In:

Audio Can Be Found HERE

Retrospective registry trials shouldn’t be published after RCT data is available because they don’t provide meaningful answers. The issue with retrospective registry trials is it can only demonstrate association not causation even with multivariate analysis. There are numerous RCTs now published evaluating mCPR vs manual CPR in the prehospital setting and the best evidence is there is no evidence of harm what so ever with the use of mCPR. Additionally, the trial at hand did not evaluate in-hospital use of mCPR, especially ED use. These patients have probably had many minutes of CPR already, so the question is what is the most effective way to continue CPR?

Scott Weingart, MD
Associate Professor of Emergency Medicine
Chief of Division of Emergency Critical Care
Stony Brook School of Medicine, Stony Brook, NY
Twitter: @emcrit
Blog: emcrit.org

Clinical Take Home Point: This is not the death of mCPR, and although the authors of this paper must be commended for taking confounding variables into account in this large trial, it is retrospective and observational and therefore conclusions must be taken with a grain of salt.

For More on This Topic Checkout:


  1. Buckler DG et al. Association of Mechanical Cardiopulmonary Resuscitation Device Use With Cardiac Arrest Outcomes: A Population-Based Study Using the CARES Registry (Cardiac Arrest Registry to Enhance Survival). Circulation 2016; 134: 2131 – 2133. PMID: 27994028
  2. Gates S et al. Mechanical Chest Compression for Out of Hospital Cardiac Arrest: Systematic Review and Meta-Analysis. Resuscitation 2015; 94: 91 – 7. PMID: 26190673

Post Peer Reviewed By: Anand Swaminathan (Twitter: @EMSwami)

Cite this article as: Salim Rezaie, "Episode 34 – The Death of Mechanical CPR (mCPR)?", REBEL EM blog, March 23, 2017. Available at: https://rebelem.com/episode-34-death-mechanical-cpr-mcpr/.
The following two tabs change content below.

Salim Rezaie

Emergency Physician at Greater San Antonio Emergency Physicians (GSEP)
Creator & Founder of REBEL EM

Like this article?

Share on facebook
Share on Facebook
Share on twitter
Share on Twitter
Share on linkedin
Share on Linkdin
Share on email
Share via Email

Want to support rebelem?

9 thoughts on “Episode 34 – The Death of Mechanical CPR (mCPR)?”

  1. I feel when you look at that many secondary outcomes and co-variates that the chance of statistical error skyrockets. I think a lot of the lack of success has to do with prehospital providers not being familiar with mCPR devices and just the lack of uniformity of level of care. Does getting a pair of volunteer EMT-Bs in rural wherever compare to EMT-P who run 20 of these calls a week? Just some random thoughts this morning

    • GM James,
      100% agree. One of the big things not discussed in this paper was time of application of device which potentially equals time without CPR. Appreciate you reading/listening and ty for your comment.


  2. One thing that’s worth mentioning about that study- It demonstrated that services that use it almost exclusively (I believe the actual number was over 80% of the time), do not have the disparate survival numbers that others have. This should suggest that it’s a lack of familiarity and training that lead to poorer outcomes.

    From my observation, there are several very correctable reasons why mCPR may not be as effective. One was mentioned above- people are taking too long to set it up. Without having a strategy that involves everyone understanding what they are doing, it can take anywhere from 30-seconds to a minute to set up mCPR. That is catastrophic downtime. I’ve seen videos demonstrating how the Lucas can be applied with no more than 5-6 seconds with hands off.

    Another reason is that not all the equipment is being used. Both Zoll and Physio should be faulted (To a very great degree) for allowing their individual straps for prevention of device-wandering to be able to be disregarded. I can’t count how many times I’ve seen a patient brought in with Lucas applied but without the neck strap, and the piston has wandered to the patient’s abdomen. Cardiac Arrest is probably the call that contributes to the greatest amount of scene-time anxiety in responders. If there are actions or interventions that can be disregarded or short-cutted, that is the call where they WILL be disregarded. Those devices should either not be allowed to operate without them, or should be made to be a nuisance (loud warning, etc) if the straps are not fastened.

    Lastly, they are being applied to early. In my area the first priority is getting the Lucas applied. Once it’s applied, responders intuitively transition to airway where they again stop compressions for an intubation attempt.

    In short, I don’t believe the devices themselves are the problem. The OHCA paradigm has too many fragmentation problems to allow them to fully succeed.

  3. Hi Sal, great post thanks.

    I found Joe’s comments on mCPR being ‘just better’ interesting. The real world he is talking about is one where mCPR use appears to be a (nearly) daily occurrence, so his shop’s experience, and therefore ability to apply a device is likely to be better than the real world of EMS, as James alludes to.

    I do have to disagree that good quality hCPR (human-CPR) simply does not happen in the real world. Zeiner et al (1) showed in a recent paper that even in “prolonged” arrests, well trained paramedics provide almost the same compression ratios as mechanical devices:

    “The compression ratios stayed high even in long (>20 min)resuscitation efforts. In the group with manual cc the compression ratio after 20 min was 83% (78–87) and the same as in the shorter resuscitations (82%(75–87)… and as high as in the group resuscitated with mechanical cc for more than 20 min with 82% (79–86)” pp.222

    Admittedly this is not for over 1 hour as Joe is talking about, however we should not dismiss the ability of humans (or even firefighters) to do good quality CPR. Indeed, Joe talks about patients regaining consciousness during CPR: in our service, which until recently ran exclusively hCPR, this was common enough that we published on it (2), and introduced guidelines to help manage it. I personally have had to sedate and paralyse patients in VF, as they interfere with resuscitation (and because I think it is a kindness that they not be aware, regardless of the outcome).

    I appreciate that this is a different scenario to being in the ER, waiting to bridge to ECMO, however in a real world cardiac arrest in the field, I would rather have good quality hCPR for my initial resuscitation, with a considered decision as to whether it is appropriate to attempt to get to ECMO, rather than a blanket approach of mCPR.

    It’s also worth noting that Zeiner also found that mCPR resulted in worse outcomes, despite the patients receiving mCPR being more likely to have favourable characteristics (younger, shorter downtime, bystander CPR, shockable rhythms). Zeiner’s study is significantly smaller than Bucklers, however it adds to the finding that mCPR is simply not the cardiac arrest saviour we seem to think it is.

    There are clearly patients in whom mCPR in the field, as a bridge to further care at an appropriately staffed and experienced facility, is a good idea. These patients are relatively few, and it is hard to justify the rampant use of mCPR in the greater population given the results we are discussing.
    Whilst it is good idea to have mCPR as an option, time and money would be well spent training EMS better, and more importantly, educating and empowering EMS to make more nuanced decisions around cardiac arrest pathways in individual patients.

    1. Zeiner, S., Sulzgruber, P., Datler, P., Keferböck, M., Poppe, M., Lobmeyr, E., … Sterz, F. (2015). Mechanical chest compression does not seem to improve outcome after out-of hospital cardiac arrest. A single center observational trial. Resuscitation, 96, 220–225. https://doi.org/10.1016/j.resuscitation.2015.07.051

    2. Olaussen, A., Nehme, Z., Shepherd, M., Jennings, P. A., Bernard, S., Mitra, B., & Smith, K. (2017). Consciousness induced during cardiopulmonary resuscitation: An observational study. Resuscitation, 113, 44–50. https://doi.org/10.1016/j.resuscitation.2017.01.018

    • Hi Rob,
      Appreciate you reading/listening and your thoughts. A few things from me:

      1. 100% agree that “Good” human CPR (hCPR) can happen, but I would have to say in my experience it is rare. Maybe, it is, we need to invest more time training.
      2. Cardiac Arrest never has a one size fits all, but it is important to have the right tools for the right situations. For example when doing an intubation, hopefully it is not happening exactly the same way every time. There should always be a Plan A, B, and C. I think of cardiac arrest the same way. There are situations where this device is absolutely indicated such as prolonged transport times with a two man/woman crew, or in the back of a helicopter (i.e. crew safety), or even to cognitively offload during a resuscitation when working in a small ED (i.e. not enough staff or being the lone physician).

      I think my conclusion from all this is mCPR is not dead, and there will be situations where it will be beneficial, but those situations should be taken on a case by case basis. Thanks again.


    • Wonderful Comments.

      FWIW: In our EMS system, when we implemented HP-CPR practices, we specifically trained on the LUCAS to get it down to under our 10 second “hard Limit” (3 second goal, 10 second hard limit on interruptions).

      What we found is that we could not consistently do it under 10 seconds EVERY TIME (even close) even in practice, much less in real performance.

      We therefore when to a TWO stage application. First the back-plate during one pre-planned 2 minute cycle “pause”, and the rest of the M-CPR during the next pre-planned 2 minute cycle “pause”.

      Worked much better that way.

      If other systems are taking 45 seconds to get the device on, and another 45 seconds to screw with it, then of course it would have worse outcomes.

  4. Next time, have the other opinion go against the other. Mechanical and conventional have their advantages. Machines have complications. Conventional has complications. Nothing worse then a one sided opinion on a great topic.

    • Hey there David,
      Agreed…not a black and white answer here. I have seen mechanical CPR devices take up to a minute to go on, or slip off during transport (i.e. pumping away on the stomach), they are expensive, and require training to know how to use them effectively. All that being said, the point of the blog/podcast was that this one study had some significant issues in methodology and to just abandon these devices from this one study would be wrong. There are scenarios in which each has its potential benefits and harms.


  5. What an excellant topic and bit of FOAMed. Thanks! I don’t agree with Weingart (gasp!) that Retrospective trials should not be published after RCT’s are conducted, but otherwise he brings up some good points too.

    At the end of the day the huge variability in EMS systems themselves may be the downfall of this study. As you know, if a system has poor resuscitation practices or insurmountable obstacles ( rural response times verses urban rescources, for example) then their data will bring down (or artificially inflate) the aggregate data of those systems who are well resourced or who have excellent resuscitation practices. Until we (as a profession) grade and rate and accredit the clinical performance of EMS systems the way we do trauma centers, Stroke Centers and PCI centers, and then publish that data in these studies, we will never know.

    Imagine the difference If we looked at this study and said: ” Of the study and control group, 1/3 of the participating EMS systems were level I EMS systems for SCA excellence. In those systems, Mechanical CPR was exactly equivalent to manual CPR”. Hypothetically that tells us a huge amount not only about M-CPR in both HP and non HP systems, but on the impact of high performing and low performing systems in general. The opposite is also true. A major intervention might have a huge impact in a one lower performing system where there is huge room for improvement, but almost none at all in a place like Seattle where they are already performing at the bleeding edge. Especially depending on which one has the larger run volume.

    With this study we simply do not know enough about the systems. We don’t know which ones are doing HP CPR and which ones are doing AHA CPR. We don’t know which ones are doing post arrest monitoring and are using feedback devices in the manual group, and which ones are not. Which ones applied the M-CPR only for transport late in the code, and which ones only applied it early in the code? What method did they use to apply it? Did they take 2 minutes to get it on, or did they take 10 seconds?

    In other-words, the study is so large than meaningful comparison is lost. What a problem to have in a study, huh?


Leave a Comment

Time limit is exhausted. Please reload CAPTCHA.