NEW YORK (Reuters Health) – Use of mechanical cardiopulmonary resuscitation devices (mCPR) for out-of-hospital cardiac arrest was associated with less favorable neurologic outcomes than manual CPR in a recent analysis.
“The use of mCPR to deliver CPR has become more widespread, although a survival advantage has not been demonstrated in randomized, controlled trials. Little is known about real-world use of mCPR or the association with outcomes,” write Dr. Joseph Rossano of Children’s Hospital of Philadelphia, Pennsylvania and colleagues in Circulation, online December 19.
To investigate, the team analyzed registry data from 2013 to 2015 on 80,861 individuals (median age, 62) who experienced nontraumatic out-of-hospital cardiac arrest, 35.1% of whom received bystander CPR. Researchers compared outcomes for those treated with mCPR (17,625) and those receiving manual CPR only (63,056).
Although time of cardiac arrest, time of first CPR, and timing of the interventions were not reliably reported for all participants, information was available for return of spontaneous circulation – whether it occurred before or after advanced life support measures were initiated.
Compared with patients receiving manual CPR, those receiving mCPR were more likely to have had an unwitnessed cardiac arrest (57.3% versus 55.7%), placement of an automated external defibrillator (33.3% versus 28.3%), placement of an advanced airway (87.4% versus 79.0%), use of an impedance threshold device (41.8% versus 13.4%) and prehospital targeted-temperature management (16.6% versus 12.2%) (P<0.05 for all).
The team also looked at EMS agency use of mCPR during the study period. Overall, use of mCPR increased from 20.6% to 23.4% (P<0.0001), and mCPR was used at least once by 41.9% (244 of 582) of agencies included in the study. Median mCPR use was 43.9% for agencies that used the devices. However, agency use varied greatly, with 21.7% of agencies using mCPR in more than 75% of cardiac arrests and 37.7% using mCPR in fewer than 25% of arrests. Survival to hospital discharge was greater in patients not receiving mCPR (11.3% versus 7.0% for overall survival, P<0.0001), as was neurologically favorable survival (9.5% versus 5.6%, P<0.0001).
A subgroup analysis that assessed neurologic outcome at hospital discharge excluded patients with a return of spontaneous circulation before advanced life support, because those patients were less likely to receive mCPR. After that exclusion, the likelihood of a neurologically favorable outcome was also greater in those not receiving mCPR (5.9% versus 4.6%; P<0.0001).
Overall survival and neurologically favorable survival were greatest in EMS agencies that never used mCPR compared with those that used mCPR either in fewer than 25% of cases or in 25% to 50% of cases. No significant difference was found in agency survival percentages for those that just did manual CPR and for those that used mCPR in 50% to 75% of cases and more than 75% of cases.
The key finding, Dr. Rossano told Reuters Health, is that “only a minority of (EMS agencies) used these devices and that the use of the devices was associated with worse overall survival and (worse) neurologically favorable survival at hospital discharge.” Asked when mCPR might be clinically useful, Dr. Rossano declined to comment, stating “that would be beyond the scope of our study.”
Dr. Christopher Granger of Duke University in Durham, North Carolina, coauthor of a Perspective article on improving out-of-hospital cardiac arrest care in the same issue of Circulation, said, “This study shows that use of a device that is widely used across the U.S. to provide mechanical CPR to patients in cardiac arrest was associated with worse outcomes, but it may have been because those patients were sicker.”
“In the context of more definitive randomized studies showing no benefit from mechanical CPR, ” he told Reuters Health by email, “this study suggests we should not be enthusiastic about investing in these devices, unless additional studies show some benefit.”
Dr. Ashish Panchal, director of the Center for EMS at The Ohio State University Wexner Medical Center in Columbus, observed, “In the prehospital setting, EMS professionals see cardiac arrests infrequently and may not use these devices regularly. Furthermore, with the distractions and challenges inherent in this environment, efficiency of device use and delivery of optimized resuscitation care is difficult.”
The study findings are “concerning when we consider that the favorable neurological outcome decreased from 9.5% to 5.6%,” he told Reuters Health by email. “However, this of course is in the (context) of settings where median device use was 44%, with large variations in agency use.”
“Interestingly, when mechanical CPR devices were used in more than 50% of cases, survival was identical to no mechanical CPR,” he observed. “This suggests that some of these findings are (really) related to implementation of devices in the prehospital setting.”
Dr. Panchal, who was not involved in the study, concluded, “This study highlights that real-world implementation is different from randomized, controlled trials. We should take caution that (this) device does not necessarily improve outcomes. But the use of these tools in conjunction with excellent training, optimized CPR performance, and continuous process improvement can improve outcomes from cardiac arrest.”
One coauthor has received fees from the Medtronic Foundation and other medical device companies. Dr. Granger also has received support from the Medtronic Foundation.
Pages: 1 2 | Multi-Page
9 Responses to “Mechanical CPR Devices Tied to Worse Outcomes”
January 18, 2017
Andy NisbetAll in the reporting ?. Most ROSC’s are probably achieved in the first 5 minutes of BLS with Defib being utilized and a good chain of survival preceding this. Certain cohorts of patients (hypothermia, OD, PE) will require prolonged CPR and this is where mechanical CPR is most beneficial. I doubt mechanical CPR is responsible for poorer outcomes but agree that by the time it is used as part of a resuscitation attempt (many systems will have this kit only as part of a secondary response to a cardiac arrest) the prognosis is probably poor, ps the same rationale can be applied to adrenaline and intubation. Most ROSC is achieved as a result of a tight chain of survival, early recognition, early CPR, early shock. But don’t become a mechanical CPR hater unless you have tried doing CPR in the back of an ambulance, or for over an hour because the patient condition or your system requires you to do so
January 18, 2017
Ken, RNFirst paragraph, the word is “than” and not “then.” Jeesh!
January 19, 2017
Dawn Antoline-WangCorrected-thank you!
January 19, 2017
Mark SmithThis is bad research. I’m curious what the researchers ties are to companies that compete with the manufacturers of these devices. Take a second and interpolate the information. Mechanical CPR devices are tied to poor neurological outcomes. Now how can they compare the etiologies of the different cardiac arrests, the co-morbidities of the patients, the downtime and response times, location and access to the patient, whether or not an ALS or BLS crew responded, time to first drug, first shock, etc… I can go on for days with the unknown variables. I’ll inject some common sense into this study just like I did into another study that was trying to link epinephrine to poor neurological outcomes also: These devices are EXTREMELY effective. They allow for continuous consistent compressions without fatigue. They allow compressions to continue when traditionally they were interrupted to move the patient, load the patient, or when rescuers became exhausted. The “poor neurological outcomes” are popping up because these devices (just like epinephrine…) are working and actually keeping these patients alive. More humans are surviving the cardiac arrest event. Naturally more are surviving neurologically deficient. The point is that they’re surviving which is what the equipment is intended to do. We have no way to know a patient’s neurological prognosis before we start a code, either manually or with a machine. The fact that this study even exists proves that those mechanical CPR devices are working as they should.
January 19, 2017
TomI would like to know when the last time one of these researchers did CPR in the back of an ambulance for 20 minutes.
Cardiac arrest is cardiac arrest, there is nothing you can do to make the patient worse since they are already dead.
Our area CPR protocols have the paramedic working a code on scene for 20-30 minutes before a transport decision is made. On top of that, transport time is 15-20 minutes to nearest facility. These machines have been a life saver for a volunteer service with limited manpower and a high median membership age. During the day we have a driver, a tech and an intercept paramedic. If the medic is doing the ALS thing, driver is driving, you have one person to do cpr for a 15 minute transport on top of 20 minutes on scene.
I have worked an equal number of cardiac arrests with a machine and without. Free om what I have observed, manual CPR resulted in about 50% transport 50% being called on scene after 20-30 minutes of CPR. With the machine, I have seen 100% transport rate, including twice now where the paramedic was planning on doing an on scene pronouncement after 20 minutes of CPR. Both patients were 80+ years old with pre existing heart problems. Both cases CPR was stopped, patient was reassessed and it was found that a pulse had been reestablished.
Must be nice to look at numbers from a desk, come work in the real world.
January 19, 2017
Jensen Heckler“Researchers compared outcomes for those treated with mCPR (17,625) and those receiving manual CPR only (63,056).”
How can you possibly publish this as a reliable study when your sample group of patients receiving manual CPR is nearly 50,000 patients larger than the sample group for those receiving mCPR?
January 19, 2017
JamesFTFA….
“No significant difference was found in agency survival percentages for those that just did manual CPR and for those that used mCPR in 50% to 75% of cases and more than 75% of cases.
The key finding, Dr. Rossano told Reuters Health, is that “only a minority of (EMS agencies) used these devices and that the use of the devices was associated with worse overall survival and (worse) neurologically favorable survival at hospital discharge.” Asked when mCPR might be clinically useful, Dr. Rossano declined to comment, stating “that would be beyond the scope of our study.”
How can you possibly make this conclusion in the paragraph following the results that directly counter it. Did Dr. Rossano even look at the actual results before making this statement? This is junk science and even worse reporting. Your headline is clickbait…
February 13, 2017
AshSeveral comments assail this study as “junk science” and “bad science”. It is not. What it is, is an association study. Association Studies cannot determine cause; they simply look for variables that are associated with improved or worsened outcomes related to the chosen item of interest — in this case mCPR.
The real problem is the widespread use of association studies to provide “evidence”. The media howls over these association studies and our culture sometimes changes behavior over them. Think about the number of headlines that say “meat is bad” followed by “meat is good” and so forth. These are all association studies. Misuse of these studies may, in part explain our country’s lack of trust in science.
The point of an association study is to develop ideas for further study — nothing more.
So, when you see tites and media headlines that use words like: “linked”, “tied to”, “associated with”, etc. Read them knowing what you are reading them for; possible new ideas for study.
The interesting thing about this study are the other variables associated with study endpoints. For instance, early AED, Impedance threshold devices, therapeutic hypothermia, advanced airway management and mCPR are therapies that in my mind suggest EMS agencies that are “cutting-edge”. Hummmmm
The strong association between “advanced-airway” and mCPR catches my eye. Hummmmm
It would be interesting if the authors used their data to look for an association between “advanced-airway” and outcomes. What is probably happening is that effective positive pressure ventilation via an ET tube reduces preload and thus cardiac output during CPR (lots of animal experiments) on this. Additionally, EMS provider knows that “airway-thrashes” are a tad more common than we like to think they are.
PS Experiments can help to establish causation — associations studies cannot.
June 22, 2017
Chip GetchellI believe we will ultimately find that mCPR produces benefit in certain circumstances, such as insufficient number of rescuers to do good CPR and for long transports to tertiary centers that provide PCI or ECMO for salvageable patients. Inclusion criteria and good training are essential. But for many OHCA, good manual CPR and full onscene ACLS may be best. And I have no Level 1 evidence to support what I just said, and I take money from no manufacturers or distributors.