Mechanical CPR: Three CHEERS or a boo?

There has been a fair bit about mechanical CPR devices floating around the FOAMasphere lately, so I thought I should probably do a post.

These devices are not exactly new (check out the Thumper, in use in Victoria in the 70s) However, there seems to be a surge in interest in these devices, and I must say there seems to me to have been a largely positive buzz about them in spite of the evidence for their effectiveness being somewhat lacking to say the least.

The LINC trial was published earlier this year, comparing normal resuscitation with that using the LUCAS mechanical CPR device.   No difference in survival to hospital, or discharge from hospital was found.  This backs up the data from the earlier CIRC trial that found the same results with a different device, this study that showed no benefit (despite what the conclusion says), and this study that showed harm (worse neurological outcomes) using mechanical devices.  Ongoing are the MECCA trial in Singapore (discussed here, although I cannot find it on PubMed) and the PaRAMeDIC trial from Europe which I believe will be published shortly.  It seems likely to me that both the MECCA and PaRAMeDIC trial will follow the same lines as the previously published research that shows that mechanical CPR devices are no better than people powered CPR.

Also published last month in Resuscitation was this small observational paper (subscription required) which noted a significantly increased number of injuries in patients who received mechanical CPR as compared to manual CPR.  These injuries were more common in non-survivors, and it is not suggested that any of the injuries contributed to the death.  CPR is a brutal intervention and none of us are phased when we feel a few ribs go; it just appears that mechanical CPR is more brutal than traditional CPR.  It is not reported what, if any, effect injuries had on the recovery of survivors.

In spite of the universally negative trials so far, most of the discussion I seem to see revolves around how great these devices are and is invariably in favour of using them.  Here is an excerpt from an email that landed in my inbox from a well known EMS journal a few weeks back:

EMS systems that are using mechanical compression devices have seen dramatic results in the number of patients not just resuscitated, but also leaving the hospital neurologically intact.

Attend this important Webcast Thursday to hear *** ******. MD, discuss how convinced he and many other medical and administrative leaders are that mechanical CPR offers significant benefits over manual CPR…

This concerns me.  This publication, like it or not, is influential in the sphere of EMS and does a great deal to shape opinion.  When it starts putting out stuff like this that is clearly counter to the published evidence we have a problem.  If there are systems that have found such results, why have they not been published?  Surely if the above comments are true, this represents a major improvement in cardiac arrest care and should be shared…

So why are we so uncritical of these devices?  I’m sure that partly it is because we love our toys, and these are pretty nifty toys.  It must also be noted that in the published trials, the devices are no worse than people powered CPR.  So it would seem that there is potential benefit to rescuers in terms of health and safety.  Not doing CPR is far easier and less likely to cause injury than doing CPR…  We must be clear though, when we discuss these devices, who the benefit (real or perceived) is to.  As things stand, when using these devices as part of a standard resus, there is no benefit to the patient, but there may be to us.  That’s ok, but we need to be clear and honest about that.

There are two areas where the perceived benefit is the greatest: in the rural setting or other settings where humans are in limited supply, and in transporting the cardiac arrest patient to hospital with CPR ongoing.

I agree that there is potential for patient benefit in the resource-poor area as having CPR done for you frees up a great many hands to do other stuff.  However I’m still a bit skeptical, as my experience with these devices is that they are not easy to set up, and actually require a fair number of hands to the pumps to ensure that it is done with minimal interruption to CPR.  Nonetheless, I can see that they could potentially be useful.

Transporting cardiac arrest patients to hospital is not something that is done in the services I have worked for, outside of exceptional circumstances.  I appreciate that this may be different elsewhere.  If we are going to transport, I believe that we need to have a clear goal in mind and take the patient to a center that can actually provide further, useful intervention.  Throwing the body in the back and driving like the clappers in the hope that an ED might be able to do “something” is just crazy.  CPR whilst moving with a mechanical device may be safer for the paramedics, but it is not any less futile.

There may be a ray of hope here, with the recent publication of the CHEER trial (subscription required) out of Melbourne.  This is a small, non-blinded pilot trial to see if it is feasible to continue with a larger blinded trial.  What we did was to take the patient who really should survive (the young, “healthy” person with a presumed reversible cause of arrest, who has refractory VF/VT, good CPR, short down-time, all that sort of good stuff) and do all our normal stuff: good CPR, drugs, intubation and so on.  We then load them into a CPR device, cool them and head to the Alfred Hospital, where they are put on ECMO in the ED and taken for PCI.

This study had remarkable success with 54% survival to discharge, neurologically intact. (for more on the trial, check out this post over at ScanCrit, along with some links to other interesting ECMO stuff) Of course this is a small, non-randomised trial, so we can’t hang our hat on this result yet.  All it does is give us a good basis for further research, and some hope that maybe mechanical CPR devices might be good for something other than blowing budgets.

So why does this study give us hope for a new paradigm in (mechanical) resuscitation?  It is part of an intensive, multidisciplinary effort in a system that is well resourced and already has a high level of cardiac arrest care with good outcomes.  Arrests are attended by 2 Advanced Life Support and 2 Intensive Care Paramedics, as well as Firefighters as first responders/CPR devices.  So the scene is set with ample manpower and ample brainpower (unless I’m on scene) to be able to make decisions and carry out tasks efficiently.  This is then followed up by aggressive care delivered by highly experienced Doctors in a large, quaternary hospital.  There is a clear reason to attempt transport in a small subset of patients with a specific outcome in mind from a hospital that is able to provide such care as required.

Perhaps more importantly though, are the patients we in whom we apply these interventions.  These interventions were limited to a select group of patients as noted above.  This was not for the high care nursing home patient who has finally succumbed to one of 37 different chronic diseases.  The patient’s in the CHEER trial are the patients we were always meant to be resuscitating before things got out of hand.  Consider this from one of the earliest publications on ‘modern’ resuscitation: “Not all dying patients should have cardiopulmonary resuscitation attempted.  Some evaluation should be made before proceeding.  The cardiac arrest should be sudden and unexpected.  The patient should not be in the terminal stages of a malignant or other chronic disease, and there should be some possibility of a return to a functional existence” (Jude et al. JAMA 1961 – subscription required)

Does this sound like most of your arrests?  Didn’t think so.  But we need to consider who we can successfully resuscitate as opposed to who we can just attempt resuscitation on.  To use these devices in such a fashion represents an enormous outlay and mobilisation of resources.  Attempting such efforts in all comers would be a poor use of the limited healthcare dollar that we all must be stewards of and (it would appear) not improve patient oriented outcomes.

With all of this in mind I think that the use of mechanical CPR devices has a genuine future for some patients.  We just need to accept that that particular future is not here yet.


Since posting this there have been two more pertinent studies published.  The first is this study by Boyle et al in the Netherlands Heart Journal.  Survival from cardiac arrest in the population studied was a very impressive 43%
Great news for mechanical CPR devices?  No.  This spectacular result was achieved by widespread bystander CPR and AED use.  I would be fascinated to see a cost/benefit analysis of this approach compared to proprietary CPR devices.

The other trial was the PaRAMeDIC trial, published today in the Lancet.  This is the large, pragmatic trial of mechanical CPR devices that we have been waiting for.  The result?  I’lll quote from the article itself: “We noted no evidence of improvement in 30 day survival with LUCAS-2 compared with manual compressions. On the basis of ours and other recent randomised trials, widespread adoption of mechanical CPR devices for routine use does not improve survival”

We may hold on to some hope that in a setting such as the CHEER trial we may yet see some utility for these devices, but as it stands we need to accept the obvious: they don’t work.  The best bang for our resuscitation buck seems (unsurprisingly) to lie in CPR training for the general public and widespread availability of AEDs.  It’s not sexy, it doesn’t make anyone any money, but it works.  

This entry was posted in Resuscitation and tagged , . Bookmark the permalink.

7 Responses to Mechanical CPR: Three CHEERS or a boo?

  1. Hi Robbie

    I enjoyed this post and agree with what you are saying – thanks a lot for the ping back

    We have been trialling a Zoll-Autopulse – there have been some major other practical issues that we have experienced:
    1) Unable to Cath
    2) Head bobbing around (a LOT)
    3) Slow to put on

    Your discussion of increased injuries was interesting to me because I always thought this was one of the up sides of mechanical CPR.

    Regarding the upsides, I agree that there’s a role in rural and pre-hospital setting where where are lack of rescuers or potential danger from being unrestrained in an accident.



  2. Brian says:

    Did anyone else get the email from a certain manufacturer of this type of device with a bold claim of “Increase Neurologic Recovery with the LUCAS device” ?

  3. Jake says:

    “CPR whilst moving with a mechanical device may be safer for the paramedics, but it is not any less futile.”

    I have to wonder about that. One of the oft-cited theories behind the increased survival to discharge rates when arrests are worked on scene instead of transported is that it is not possible to do continuous and high quality chest compressions in a moving vehicle, or while “cot surfing” when moving the patient on the stretcher from the scene to the ambulance or from the ambulance to the ER. The (theoretical) benefit from these devices is that they do constant, continuous, and consistent quality compressions, regardless of motion. You get the same compressions in a moving ambulance as you get on the floor of the patient’s bedroom, and they don’t have to be interrupted to lift them onto the cot, or while loading/unloading into the ambulance.

    Of course, as always, the proof will be in the evidence. But, other than improved consistency of compressions, it’s hard to see what these devices do that a human does not.

  4. Christopher says:

    The CHEER trial came out at roughly the same time I’ve begun pitching M-CPR for the transport of on-going resuscitation in favorable patients.

    During the webcast you mentioned, I asked the question: if we know that humans can do extremely well, why pay money? (I wanted to ask, “are folks not willing to do better manual CPR just that lazy? Do they just not care?”)

    Anyways, I’m looking forward to getting more aggressive with refractory patients!

  5. Jon Kavanagh says:

    Are we lumping everything in together? Prehospital vs hospital (I realize this is targeted towards Pre); Lucas vs Autopulse; level of injury due to potentially poor compressions vs mechanical; generation I vs II devices (if they exist)…

    Cardiac arrest management is too complex to point at one item as a silver bullet. Logically, a mechanical compressor should improve outcomes because it is doing exactly what is preached is the most important part of an arrest, that is quality compressions. Thus, in the absence of improved outcomes, is it the device at fault or is something else? Maybe the quality of the management of the patient needs to be examined alongside the study of the device?

    I think it was Dr. Weingart that identified that these devices’ potential may not lie in improved outcomes but in the increased staff productivity (whether EMS or ER) by freeing up a line of compressors and the ability to move these patients (transport to ER or within the hospital) while continuing excellent compressions.

    A multifaceted problem for sure.

  6. Pingback: CME 01/09/16 - Mechanical CPR: evidence and issues - Charlie's ED

  7. James Oswald says:

    I think it is likely that part of the problem with these studies is that we may be comparing mCPR done poorly VS manual CPR done well. Its a new skill and there is plenty about it to balls up. But just like a manual approach, with rigorous quality control, mCPR may be truly non-inferior or perhaps a bit better. When you add that to other benefits, it starts to look better (1).

    1. Ong MEH, Anantharaman V. Out-of-hospital cardiac arrest: manual or mechanical CPR? The Lancet. 2015 Mar 14;385(9972):920–2.

    The safety benefits are important and I think it won’t be long before mCPR will be seamlessly integrated with see-thru CPR/SAED to DCCS/continuously compress, further reducing hands off chest time and cognitive load.

    It is an inescapable truth that any algorithmic process will be given over to machines in the not too distant future. SAED will take over manual DCCS at some stage. STEMI interpretation/criteria will become complex enough that computers do it better. Decision support algorithms are popping up everywhere. I think its a good thing. It leaves more time for us to do all thinking.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s