Left Ventricular Assist Devices (LVADS)

LVAD Emergencies

These patients are super-complicated, luckily I got Zack Shinar, MD from Sharp Memorial in San Diego to try to wade through the morass.

All Situations

  • Call the patient’s VAD coordinator ASAP
  • These patients may not have a palpable pulse. Listen over the heart to hear if the motor is working. Then use mental status, skin color/temp, and the machine flashing Low Flow as indicators that perfusion badness is occurring. Do a bedside echo. The MAP should be ~65 on manual doppler BP, Automated BP devices may give you a MAP as well. A-line MAP is the most accurate.
  • Try not to cut or yank out the drive-line, ’cause that is embarrassing.

Poor Perfusion

  • When in doubt, consider a fluid bolus. VADS love volume. If you need to improve hemodynamics with a working LVAD, consider preload augmentation and possibly afterload reduction (if MAP is high).
  • Consider inotropes–if you think it is right heart failure, give dobutamine. If you think the patient is septic and has markedly reduced afterload, consider norepinephrine.
  • On echo:
  • Big RV, small LV=pulm hypertension or right heart strain/stemi. Correct hypoxemia and acidosis, consider volume, screen for RV STEMI, consider inotropes.
  • Small RV-give volume
  • Big RV & LV-pump failure or pump thrombosis.

Consider pump thrombosis–Signs of pump thrombosis are LVAD is hot, working hard, with high RPM, low flow, dilated RV/LV, and low MAP. Zack would give a bolus of 5000 U of Heparin in the decompensating LVAD that he thought was secondary to thrombosis (or if he just couldn’t figure out what was wrong with a failing device). He would also consider tPA if he really thought it was pump thrombosis and the patient was decompensating and peri-code.

On ECHO, a dilated RV/LV could be from pump thrombosis or non-working pump (electrical issue for example).if you think that is the problem, heparinize.

Machine Not Running

Check batteries. Make sure all of the lines are connected.


These folks are prone to bleeding from the anticoag (and probably additional plt dysfunction from the device if I had to guess). So if they have altered mental status or neuro findings–consider hemorrhagic stroke.

Patient appears Infected

Drive-line infection-look at the site at entry to the skin. If the patient appears septic and you can’t find a source, consider it a device infection until proven otherwise. Don’t yank the device. Treat for health-care associated infection covering both hospital gram negatives and MRSA.

Patient is Coding

We need to AVOID CPR until the patient needs it and at point, what is the alternative? Can you rip out the device with CPR-yes! Many of the CT surgeons recommend not to do CPR, but you can’t get deader than dead (I was not a philosophy major, so I could be wrong). Avoid CPR if at all possible, some of the 1st gen devices had hand-pumps you could use–the current generation don’t. If you’re the point where there is NOTHING else to do except CPR you need to use your clinical judgment.

Here is Zack’s clinical judgment:

CPR is not recommended by the manufacturers secondary to potential cannula dislodgement.  I would not do CPR unless the pump was NOT working and the patient had lost their BP (MAP of 0).  This is the one scenario where you have to perfuse the brain no matter what the cost.  All other scenarios I would focus on how to get that pump operating better (at all).

Joe Bellezo then adds:

Just agreeing with Zack’s thoughts on this. My approach to this is ‘Look, Listen, and feel” – assuming a comatose LVAD patient.

Look: …at all the connections. Everything connected? Ok. Look at the controller. Green Light on? Yes..ok. No? Troubleshoot for a problem with the VAD and keep working until you get the green light on the controller.

Listen: to the hum. Assuming a green light on the controller…there should be an LVAD Hum. No hum? the pump isn’t functional (duh). Find out why. Again, check all the connections and then touch the control box and check RPM, flow, etc). Pump thrombosis is your reversible problem here.

Feel: hot control box is not good and usually means thrombosis or dislodgement of the outflow cannula to the aorta…or distal obstruction like a dissection.

Compressions: here is my thought: if you’ve gone through all of the above and there is nothing to fix…then you have an LVAD patient who does not have a functional LVAD. I would treat them just as if they came to the ED the day before they got their LVAD: a patient with end stage heart failure and no blood pressure. I would begin chest compressions if their MAP was below 60 because they aren’t perfusing their vital organs and will die. I know this goes against Zack’s recommendations but that shows you that nobody really knows the best answer here. This patient will die. I say start the chest compressions and get inotropes going. Dobutamine or milrinone stat in addition to levophed. In other words…pretend they don’t have a VAD and aggressively resuscitate them. Yes, dislodging the pump is possible…but these patients are going to die anyway.


Though if I had to guess Zack and Joe would only do CPR long enough to crash the patient on to ECMO.

Algo by Dr. Higgins of MMC (Click for Full Size)



Read this PDF Now

An insanely good field guide from mylvad.com.It has device-specific recommendations. Read it NOW!

The site also has some excellent additional resources.

Additional Resources

Another great review on LVADs from Fire Engineering

the images from this post are from mylvad

Put your questions in the comments and I’ll send them on to Zack

Now on to the Podcast…

You finished the 'cast,
Now get CME credit

Not a subcriber yet? Why the heck not?
By subscribing, you can...

  • Get CME hours
  • Support the show
  • Write it off on your taxes or get reimbursed by your department

Sign Up Today!


Subscribe Now

If you enjoyed this post, you will almost certainly enjoy our others. Subscribe to our email list to keep informed on all of the ED Critical Care goodness.

This Post was by , MD, published 3 years ago. We never spam; we hate spammers! Spammers probably work for the Joint Commission.


  1. vaishali patel says

    about a year ago, we had an LVAD pt who was brought in for being dizzy & weak – he was put in our resus area. we were getting him hooked up to a monitor & while others were flitting about plugging in the LVAD, getting a line, trying to get a bp (which we couldn’t get), etc – when i looked up & noted that he was in VT at 200. oh sh** – following ACLS – is he stable or not? dunno – can’t get a bp reading but he’s awake & talking to us. our next big question was can you shock a pt with an LVAD? per the cardiomyopathy attending we had on the phone, YES. it worked … but was really really odd to shock a pt that has a machine assisting his heart.

    • Sam says

      Remember that LVADs support only the LV and not the RV (BiVADS are used only in acute situations mostly post-operative situations). RV output is mandatory for the LV to fill and if a patient is in VT or VF, there is not RV output and so LV does not fill and pump cannot function. Thus, its very important to maintain a rhythm- whether native/paced for the RV to function (+/- inotropes) for a normal LVAD function. Hence you would need to cardiovert the patient.

  2. says

    I enjoy reading case reports of sustained VF in these patients; who needs pulsatile blood flow anyways! Although, this leads me to wonder if EMS is in a position to cardiovert or pace in the field, should we have similar coagulation concerns as with “don’t mess with A-fib >48 hours”?

  3. Brandon says

    With regards to pump thrombosis, you will see an increase in POWER not necessarily RPMs. Second and Third gen pumps are usually fixed RPM. (With certain pumps you can see an transient decrease in RPMs that can be related to volume status as the pump detects a low flow state and tries to give the LV time to fill.)
    I agree with the rest, look for a hot controller/hot skin over the pump, lack of hum and high Watts (power). Also consider trans-thoracic echo while looking for aortic valve opening. If your patient appears poorly perfused, is hypotensive and still has aortic valve opening it could be a thrombosed /defective pump. You have to be cautious with FLOW numbers in the presence of pump thrombosis since flow is a calculated value that increases in a linear fashion with pump power.(it may be falsely elevated)
    When it comes to arrythmias in these patients treat them as you would any other patient. There is no such thing as stable sustained VF in these patients they should be treated immediadiately with defibrillation. These devices are electronically isolated and can stay running during defibrillation. In VT you need to look at the patient. If they are A&O, and the pump is not alarming consider pharmacological treatment. If the patient is not mentating and/or the pump is alarming consider cardioversion.
    I applaud all of whom have assumed care of this special patient population without formal training or a level of comfort with the device. Hopefully through blogs such as this one, as well as some other resources available we can share our knowledge and experiences.

    • Joe Bellezzo says

      Great comments, Brandon. I agree with all of your suggestions. Remember that these are left ventricular ASSIST devices, not bypass devices. They augment the left ventricle – not bypass it. So, while you may have a patient in VFIB and awake (we had a patient in VFIB drive themselves to our ED!!!!) it isn’t ideal and not sustainable. Cardiovert.

  4. Jeff says

    Thank you for this excellent Podcast. It is also relevent for EMS responders who may have an LVAD patient in their response areas and don’t know a lot about trouble shooting the device. The EMCRIT blog and Podcasts have been excellent learning opportunities, Thanks Scott!

  5. Joe Bellezzo says

    ****UPDATE regarding chest compressions in LVAD patients: We just reviewed every in-hospital LVAD patient who received CPR at our facility. We found 10 patients who received chest compressions after coding because of a nonfunctional LVAD (ie thrombosis, motor dysfunction, etc). 5 patients survived to hospital discharge after the LVAD issue was resolved. 5 patients died after brain-death. No LVADs were dislodged. I will probably publish this data but here are the preliminary results. BOTTOM LINE: LVAD not functioning? MAP 0? Do chest compressions.

    • Brandon says

      Joe, your data is interesting. I would love to do the same for my facility. What types of devices do you primarily see? I have seen one pump dislodgement secondary to chest compressions with a Gen II axial flow device, however, the patient was rushed to O.R. where the pump was successfully re-seated and the patient lived. Pump dislodgment is much more common in the axial flow devices that sit in a pump pocket as opposed to adjacent to the ventricle in the pericardial space.
      There is a small amount of animal research to support the use of chest compression in Gen III centrifugal flow devices. Not only is pump dislodgement unlikely, there is also evidence of cerebral and other end organ perfusion with compressions. This is not always the case with some of the axial flow devices as you can essentially create a loop between the device and the ventricle. While the manufacturer of the current (recently) FDA approved bridge to transplant Gen III centrifugal flow device is not prepared to make recommendations for or against chest compressions all of their reps have assured me that they feel compressions are safe.
      What I would hate to see is a surge in pre-hospital and facilities unfamiliar with mechanical circulatory support jumping to chest compressions with lack of a “pulse” in an unstable patient. I feel that it should still be seen as a last resort method until a full assessment of the problem and its possible solutions is completed, along with documentation of LVAD appropriate vital signs. I think this is great dialogue. Keep it up guys!

  6. Zack says

    We see Heartmate II LVADs. We are starting to trial HVAD (Heartware), not sure start date. I am unsure about the likelihood of dislodgment by generation or placement of pump, but it’s a great topic. It makes perfect sense that the Heartware device would have a much lower possibility of dislodgment given the size. Certainly, Heartmate II devices would seem less likely to dislodge than the old pulsatile VADs. Joe mentioned our data and I think in the non-working LVAD without perfusion, chest compressions are reasonable. I also fear about medics/physicians doing compressions on patients with a working VAD and perfusion but no pulse. Therefore, I temper this message by saying we should be sure they are not perfusing if we start compressions.
    In the podcast, I did not go into use of bedside ED ECHO or some of the unique clues from the controller for assessment of LVAD problems. I do like to use my own echo to assess LV and RV size and decide whether a pump thrombosis/AI (big RV/LV) or suction event/hypovolemic event (small LV). This can be a bit complicated to look at, but can give some good info.
    Also, just feeling the controller can give you a sense of how much work the machine is doing. A pump thrombosis or recurrent suction event can make the controller quite hot to touch.

    • Brandon says

      I agree, what is amazing about this patient population is how much you can get from a hands on physical assessment and remembering the basic components of cardiac output. Joe, I love the idea of “look, listen, and feel”. Great podcast and comments. There are some excellent clinical pearls that people should take away across the spectrum, from the pre-hospital environment to the ICU.

    • says

      Zack, thanks for making LVADs less scary. Regarding compressions: the Emergency Physicians Monthly article currently in print, summarizing your EMRAP interview, mentions abdominal compressions being preferable to chest compressions in these patients. Thoughts?

      • Brandon says

        This has been visited a few times in the past with regards to cardiac arrest, post-cardiac surgery via median sternotomy. The key concept at play here is Coronary Perfusion Pressure, which we all know is the vital component to ROSC. Coronary Perfusion pressure is defined as the “aortic-to-right atrial pressure gradient during the relaxation phase of cardiopulmonary resuscitation” (Paradis et al., 1990). Paradis and gang also found that only those with a CPP of 15 mmHg or greater had ROSC. There were various studies over the decades since that article was written that showed in human, swine and computer models it was possible to obtain a CPP of > 15 mmHg with Abdominal only CPR (known in the literature as AO-CPR, or OAC-CPR).
        We also know, thanks to Rottenberg et al., that in at least 1 case of HM II LVAD going into asystole in the O.R. while undergoing re-sternotomy for pump exchange that abdominal compressions were an effective bridge to ECMO (if the primary end point is ROSC) as this patient did achieve ROSC. What we do not know, however, is whether ROSC was a result of the ECMO, the compressions or a combination of both.
        So, I believe it is a possible alternative to External Cardiac Massage in patients with Left Ventricular Devices where ECM is contraindicated or prohibited. What we lack is data to back it up. This also brings up the questions of how fast to perform AO-CPR? how deep?, and what about hand placement? What does everyone else think about the subject?

        P.S. Sorry for letting the discussion die down, I would love to see some more people chime in with their thoughts.

        • Zack Shinar says

          Brandon and Bill,
          I completely agree with the emphasis we need on CPP and the stuff Norm did elucidating this is fantastic. What we also know is that even good chest compressions sometimes does not get adequate CPP. It takes amazing chest compressions (and some favorable anatomy) to get the CPP up to the upper tier of favorable survival (>25 mm Hg). I do not know the AO-CPR data well enough to say for sure but my educated guess is there’s no way in vivo we will be able to obtain these pressures with AO-CPR. My suggestion is forget about AO-CPR and do quality chest compressions. It’s likely with the LVAD patient population and the fact that there’s a device in the peri-diaphragmatic space that CPP is going to be very difficult to even reach the >15 mm Hg mark. My 2 cents.
          What do you all think?

  7. says

    Brandon–You have an amazing range of LVAD knowledge. Can you tell us who you are and what you do. Thanks for your comments!

    • Brandon says

      I am a Critical Care Flight RN for a hospital affiliated rotorwing/CCT program. My background is in Cardiothoracic ICU and Cardiovascular Recovery where we would commonly see devices such as HMII, HW, Centrimag for temporary extracorporeal (R, L, And BiVAD) support and of course ECMO. Our flight team also staffs the Level I Trauma center at our flagship hospital, which happens to be the intake point for LVAD emergencies. In my spare time I work with our outstanding VAD team doing pre-discharge patient teaching, and I have begun to work with some of the community healthcare facilities on LVAD related emergencies. It is certainly amazing to engage in conversation with some of the best in the field. I hope to contribute further!

      • says

        Fantastic. It is bringing folks like you, Zack, and Joe together in the same forum that makes this blog so much fun for me.

  8. minh le cong says

    great overview and discussion. we do IABP transports and a paed ECMO service in the state is just starting. LVAD transports are going to get more common so thankyou for this

  9. Bryan says

    Sorry, I’m a little late to the ballgame! We’re looking at establishing guidelines and training our rapid response and step down nurses with respect to coding the LVAD patient. Our CTICU nurses are experienced with it, but less so on the floor/step down. We’re having trouble finding any actual research on the subject, just expert opinion. This discussion will add a lot to it. Does anyone know of any published research data on this subject? Particularly with regard to CPR. Dr. Bellazo, your info is great and you should publish it. I’ve not found anything so far like it.

  10. lynn says

    My husband had his LVAD since September 23 came home October 23,was doing pretty good but recently has been complaining that the site from his line hurts most of the time especially when he trys to get up from bed or a chair has had some bleeding from the site-very slight no oder I am worried about this but he just shakes it off–is any of this normal and if so how long will it take for this to get better?
    Thank you for any help you can give me.

    • says

      So sorry, we can’t answer patient (or patient’s family’s) questions here. I recommend contacting your husband’s physicians. Be well.

  11. Mike says

    Hi Scott, I’m a paramedic here in Ga, and we have been receiving notifications of LVAD patients in our catchment area. Most notably one of my old partners just receivied one, and as such, I am seeking to obtain as much information as available. With regard to the debate of CPR for unconscious patients upon whom blood flow can not be confirmed with a dopler, I am inclined to agree that compressions would be indicated. We currenly use the LUCAS compression devices on our ambulances and my personal experiences with them has been phenominal to say the least. I am wondering if they might be a preferable adjunct to manual compressions in the decomponsated LVAD patient, as they are remarkably stable, and have demonstrated excellent perfusion?

    I do appreciate the dynamic flow of information herein, and have shared it with my colleagues here in the GA / Al prehospital care community.

  12. jeanne hupp says

    I would like to know all about OAC CPR how to perform this how many times breaths hands placement and is approved by the AHA and if so why is it not taught to instructors of CPR?


Speak Your Mind (Along with your name, job, and affiliation)