Can the Inferior Vena Cava Ultrasound guide our fluid administration in the ED? Of course it can!
So I was getting on the plane to Las Vegas for Essentials 2012, on my iphone was the latest from Mike and Matt of the Ultrasound Podcast. Up pops Mike “the Rock” Stone interviewing my buddy, Haney Mallemat; these two ultrasound gurus discuss some ultrasound soundbites, but then… They both state that IVC ultrasound is useless for determining fluid responsiveness. It is worth taking a listen to that episode if you have a moment. So how can two brilliant guys get it so wrong? They just had their focus knob turned all the way to the right. Lets optimize their settings with an EMCrit Podcast.
Mechanically Vented Patients
Now, most of the podcast bashed IVC in spont breathing patients, but there was some overflow disparaging of IVC in mech vented patients, so let's get that out of the way first. There is plenty of literature for these patients. Put them on a temporary, high tidal volume (10 ml/kg). Get an IVC shot and if it increases in size by 15-18% (depending on the study), the patient is fluid responsive.
- Intensive Care Med. 2004 Sep;30(9):1740
- Intensive Care Med. 2004 Sep;30(9):1834
- Neurocrit Care. 2010;13:3
- J Trauma. 2007;63:495
- J Intensive Care Med. 2011 Mar-Apr;26(2):116
Spontaneously Breathing Patients
Now as the two ultrasound masters allude to, there have been a few studies showing IVC ultrasound assessment merely correlates with CVP (it actually correlates with respirophasic CVP) and then use that fact to write off the IVC. Now we have maligned CVP as a marker of fluid responsiveness so IVC is crap as well, right?
There is evidence for the use of IVC as a marker of fluid status. In patients with ultrafiltration for congestive heart failure (Intensive Care Med. 2010 Apr;36(4):692-6) as well as fluid removal during hemodialysis (Clin J Am Soc Nephrol 2006;1:749 and Nephrol Dial Trans 1989;4:563). There was also a trauma study showing that fluid resuscitated patients with IVC collapse were more likely than those without to have recurrent hypotension (J Trauma. 2007 Dec;63(6):1245).
There was also a study just published in the Aussie EM Journal. This study was severely limited by the fact that none of these patients had any significant IVC collapse and the criterion standard is not a test any of us consider useful for cardiac index measurements; further, looking at the tables, some of the responder group did not seem to have any sig. increase in their CI in response to fluid. (Emerg Med Aust 2012;24:534).
And a meta-analysis study showing IVC's relation to fluid status (AJEM 2012;30:1414).
Luckily, there was also a recently published study with the table below (Crit Care 2012;16:R188).
Here is figure 1 from the study
Here is the area under the cure (AUC); you notice there are points with much higher specificity.
Update:
Hot off the presses, this study is more reassuring: (Shock 2013;39(2):155)
It lends additional credence to the use of dynamic IVC for fluid responsiveness.
It's not Fluid Responsiveness, It's Fluid Tolerance!
This is the crux of the matter. In the ED, we want to give a bunch of fluid a modicum of fluid, but not if we are going to cause pulmonary edema. The term, fluid tolerance, is a perfect description of this idea. Responsiveness is great, but all we want to make sure of is that we are not going to do harm with additional fluid. This term was introduced to me by an amazing ED Intensivist named Chad Meyers. His lecture on this stuff from the ALLNYCEM Conference will go up very soon. IVC ultrasound is the perfect guide to fluid tolerance. CVP is fine as well, but why expose the patient to a central line.
More
Mike discussed a study he was part of that showed where you measure the IVC matters (Acad Emerg Med. 2010 Jan;17(1):96-9). Make sure you don't measure at the diaphragm; measure distal to the hepatic veins or at least 3 cm past the diaphragm.
The Predicting Fluid Responsiveness Chapter at Crashingpatient.com has a bunch more info on all of this
Here is the video for how to perform the IVC ultrasound:
Now Mike is an amazing doc–he offhandedly mentions throughout the podcast, none of this crap matters until you have already given 6 liters of fluid or so. But a lot of folks are too scared wisely will not be doing that, hence the need for IVC ultrasound. Now there are unfortunately also references to the fluid depleted patient showing signs like cracked lips and a generally sere appearance; yeah, not so much. If we could assess volume status on vasodilated septic patients, that would be swell, but all of sepsis literature stands against our ability to do so.
Want to know why these studies may show results all over the map? It's probably b/c the IVC is tough to measure unless you are good as demonstrated here.
Passive Leg Raise
Instead of IVC, Mike and Haney recommend using the passive leg raise with echo assessment of cardiac output before and after. You only need to do this once to realize its utility exists merely on paper. Much better is simply…
Giving a Fluid Challenge
Check cardiac index using echo, give a liter of fluid, and then recheck cardiac index.
Up until this week, I would have advocated considering using NICOM bioreactance, but a study just published makes me a bit leery until more data are in.
Most Recent Article
Additional New Information
More on EMCrit
Original RUSH Article(Opens in a new browser tab)
Additional Resources
You Need an EMCrit Membership to see this content. Login here if you already have one.
Professor
Nassau University Medical Center
No conflicts of interest (coi).
- EMCrit 385 – Eye Trauma I – Retrobulbar Hematoma, Orbital Compartment Syndrome, and Paracanthal Decompressions - October 5, 2024
- EMCrit Wee – Did this Really Just HAPPEN? – The HAPPEN Trial Hot Take – NIPPV for COPD - September 29, 2024
- EMCrit 384 – The Vascular Guy on Vascular Access - September 23, 2024
Great podcast Scott! You know how much I love this stuff. I got to read carefully read through some of the papers before I start accurately debating, but a couple of points for now: 1) IVC distensibility (and pulse pressure variation) are great and well validated tests in the mechanically ventilated patient. I don’t recall our discussion overflowing to bashing this, but for the record I’m down with IVC Distensibiltiy in the mechanical vent patients and it should be encouraged more. 2) The IVC would be a great tool at the bedside because it has a steep learning curve and… Read more »
Great points brother, here and on the hangout. To respond to your last point: If the patient is still showing collapse, I have no problem giving them 6-8 liters. At that point, adding on a low dose of norepi, ~2 mcg, will cause preload constriction without afterload constriction. Nice thing about the IVC is it will not be affected by the vasopressors. So then recheck your IVC.
As always a great podcast, Scott. I look forward to getting to talk about this ad nauseum once I get a real internet connection (thanks for nothing, Amtrak…) Until then, my thoughts: 1) Like Haney, I don’t remember going too far into ventilated patients and IVC measurements. It appears to work in experienced hands, but let’s face it, there’s a lot of ways to skin that cat with PPV, SVV, etc etc. Spontaneously breathing patients are the real issue. 2) I’ll admit (as I did in the podcast) that a flat or collapsible IVC likely predicts the need for volume.… Read more »
Stone– Love debating you, you are at least as aggressive a doc as I am which is a rarity. So most of the assessments even in intubated patients suck. PPV/SVV all require circumstances that rarely exist. There is little wrong with giving 5-6 liters empirically, but some of your patients will be in vasodilatory shock and actually would have benefited from just getting put on some norepi at the 2 liter mark. In our collaborative, if folks can’t/won’t monitor we ask for 4 liters empiric fluid bolus. All patients with something like sepsis should get a 2 liter bolus. At… Read more »
Hi Scott Sorry I missed the hangout. Haney and I had a quick chat and I completely agree with you. I work as one of those smaller ED docs and we are often flying blind in the early resuscitation of sepsis. IVC is a quick and easy test that works well over on the left side of the Frank-Starling curve. There is a quantum leap between IVC assessment and doing an accurate LVOT doppler (if you can even get the Apical 4 on the screen #%!$) – so this is a huge technological block to what they are suggesting. One… Read more »
Amen brother
G’day lads My comment takes a more philosophical perspective. Just because giving fluid increases cardiac index – just because a patient is fluid responsive – doesn’t mean fluid should be given. In many critical illness situations we don’t know what the appropriate cardiac index should be. The critical care pendulum – particularly from the ICU perspective – seems to be swinging back towards avoiding the administration of excessive fluid. A wet patient causes problems that become apparent beyound the resus room. I think we do need ways of assessing fluid status rather than empiric 4-6L fluid boluses – exactly how… Read more »
Thanks my friend; looking forward to our shift.
Wow Scott, it brings a tear to my eye every time I think of all the good stuff you’ve been giving us with your podcasts. I really like this concept of fluid tolerance – is there any literature supporting the safety of giving fluid until the ivc stops collapsing (i.e. not causing pulmonary edema or any long term problems with our ICU collegues downstream)? I wasn’t aware of this concept previously and have been using ivc up to a point where there was still some respiratory collapse by eyeballing it then switching to percent change in VTI in the LVOT… Read more »
Here’s what we know. IVC correlates with CVP (we at least all agree on that). Elevated CVP occurs before the onset of clinical pulmonary edema; you would have to go back a few decades to find that literature.
Love the post – and love the “fluid tolerance” wording – semantics, semantics… I bet the IVC is stoked for being such a topic of debate as it had been ignored for far too long while the aorta was being loved. Ill just say this, and, yes, I may sound like a radiologist here, but we need to take our ultrasound images as one data point when assessing the patient and “correlate clinically”…. while knowing the difference in IVC physiology with intubated versus non-intubated patients and various cardiac diagonoses – – -but that comes with clinical correlation. Im going to… Read more »
Stone is so mean to this poor tube, I thought it deserved some love : )
Absolutely agree about looking at the heart along with the ivc, it is only a few cm away!
We still need to set up our FAST Cast; can’t wait.
Hi Scott, Thank you for your fantastic podcasts, like having a bolus of good ideas. I followed the debate on IVC and was a bit confused. Excellent concept “fluid tolerance”. I doesn’t trust any single ultrasound measurement for fluid responsiveness or tolerance, what I do is a combination off just eye ball EF and myocardial contractility in standard projections, IVC and lung ultrasound. Then I might add VTI and LVOT before and after bolus. I think this is a useful approach particularly to CHF patients with sepsis or pneumonia. Not uncommonly these patients have an MI on top of everything… Read more »
Yep your practice sounds like mine and most of the ultrasound folks out there. Get a gestalt and go. We had to protocolize for the sepsis collab.
s
Hi Scott! I have to completely agree with you. That being said, there are a number of factors that explain why the “traditional” ultrasound assessment of the IVC may be less-than-ideal. The first and I think the most important is that the AP diameter in long-axis, 3cm below was (arbitrarily?) chosen as THE MEASURE. For anyone who has been doing bedside ultrasound for a while, you’ve certainly noticed that not all IVC’s are shaped the same. Some are really round, some quite elliptoid – and they can be so in either axis or diagonally. Furthermore, their collapse may be also… Read more »
excellent additions. Long axis is easier to teach, but I think in the hands of a skilled guru, both are better just as you say.
Nice work all round on ultrasound of IVC. However, Scott’s original post approvingly quotes that 2010 study by Wallace that came up with the gem ‘Clinicians should avoid measuring IVC collapsibility index at the junction of the right atrium and the IVC’, because IVC measurements here didn’t correlate with measurements further caudally. This was an amazing & gutsy conclusion to draw when you look at the gold standard Wallace used: there wasn’t one. So in fact for all that anyone knows [including Wallace], perhaps the IVC/RA junction was the correct site and all the others are crap. Furthermore: Yanagawa found… Read more »
Justin,
Agree on lungs. So, just fill until you start to se some B-lines? Sounds simple enough. Easy to learn an fast to perform. Maby a “harder” endpoint to stop filling at. Most of us do IVC and echo as well, the real question is probably what does IVC really add to lung ultrasound and viseverse. Did you come across any papers?
Cheers,
Pelle
Pelle,
What you describe there is essentially Daniel Lichtenstein’s FALLS protocol published in Chest in 2009. Certainly whe diffuse B lines start to appear its time to stop, but the real question is, especially when using crystalloids, should we not stop earlier? Fluid balance and morbidity studies certainly suggest it. I don’t believe an evidence-based answer to that exists yet!
Cheers
Phil
Daniel seems to like the lungs best, but I agree with Phillipe–seems we have gone too far if interstitial edema is our endpoint.
So many good points. I think we are in agreement on the following:
-Absolute measurements send us astray
-M-Mode is prob more likely to send us astray as the IVC moves with ventilation and your measurements may be very inaccurate as a result (see article just pub in Crit Ultrasound Journal)
-Eyeball of collapse is the way to go
-If the lung is dry on ultrasound, one more reason to push forward. If the lung shows signs of interstitial edema, this could be capillary leak syndrome and should not be used to withhold fluid.
Great podcast! Thanks
Can I make a comment about the last article you mention in the show notes- “Cardiac output measurements using the bioreactance technique in critically ill patients”. The study tries to correlate absolute values for CO between a bioreactance-based technique and thermodilution by PA catheter. The correlation was found to be poor (in 11 patients). The real utility of continuous non-invasive CO monitoring is that it can be used to measure the change in CO over time, and therefore the patients response to therapy. The absolute value of cardiac output is less important; will will never know if a certain CO… Read more »
thanks Scott et al for the debate! made me reflect on what I do in a prehospital setting as well as in ED. To be honest I cant say one approach is better than another and what you suggest sounds reasonable but to equate IVC USS directed sepsis fluid resus to what Rivers did etc, is a bit of a stretch in my view. logically it makes sense but like DSI I think you have to collect enough cases to prove the theory at least does not harm vs benefit. I am sure we all have seen overzealous fluid admin… Read more »
One thing all three of us agree on is the IVC equates to CVP, so you absolutely can make the jump to replacing the CVP element of EGDT with the existing evidence. We also are presenting our first 6000 patients from the NYC sepsis collaborative at SCCM in Jan. 2000 had IVC ultrasound as their fluid assessment.
that sounds great . look forward to reviewing the results!
Great podcast & discussion!
I won’t wade into whether or not passive leg raise actually gives you useful information, but you can do it without awkwardly lifting the legs.
start with:
-bed all the way up
-HOB 45 degrees
then:
-put the bed into extreme Trendelenberg
Now the torso should be flat with the legs up.
Now whether or not it gives you useful information is another question…
Nope, doesn’t quite make it. Have tried it with a few stretchers. If it works on yours, shoot a video.
Hi, Not so much about the USS technicalities but with regards to significant fluid resuscitation, if you look at the FEAST study (http://www.nejm.org/doi/full/10.1056/NEJMoa1101549) you have to wonder about the down side of high volume fluid resuscitation. The study has issues in terms of generalisability and is in a paeds largely sick malaria population BUT was initially planned to show a benefit to high volume fluid resus and instead showed higher mortality with this. It at least raises questions that we haven’t yet answered properly What this means IMHO is that we don’t necessarily understand the pathophys of severe sepsis as… Read more »
thanks Dave. Glad you cited FEAST as I was not sure if bringing up a paediatric paper here in an essentially adult critical care forum was appropriate or not. There are a few issues with FEAST but yes it does raise the question of harm with aggressive fluid resus in septic kids in Africa. Boyd et al did this retrospective review in the VASST trial http://xa.yimg.com/kq/groups/16749867/1999291950/name/Fluid%20resuscitation%20in%20septic%20shock.pdf It calls into question the CVP targets for sepsis resus, ergo even if IVC and CVP correlate, are we promoting the right thing here? NYC Sepsis collaborative data will be useful to review as… Read more »
ummmm not so much. The review of the VASST showed us that pt’s with high cvps (like the sepsis-induced cardiomyopathy and the sepsis-induced renal failure patients) had worse outcomes, not patients who were fluid resuscitated to a CVP goal did worse. As to what FEAST can tell us about the resuscitation of adult septic patients in the 1st world, I have absolutely no idea.
Hi Sorry for my non perfect english (I work in a medium hospital in Italy). I read with great interest your post. I would like to give my contribution: First problem: CVP must not be more than our benchmark! Second: the single measurement of the IVC is not a viable solution but only monitoring the collapse index gives us important information. Third: only ecographic study of heart, lung and IVC (together) provide the information we need. Fourth: the study of the lung must be our point of reference. The appearance of B lines is the benchmark: B lines = stop… Read more »
Gaetano-
thanks for commenting! We are in agreement, no point in measuring IVC in spont. breathing pt. Agree always echo. Sorry, can’t agree about the lungs. If they you have an A-line profile, then all is well; but interstitial syndrome (mult. b-lines) can be from many reasons in a severe septic pt like alveolar capillary leak.
I agree with you that the only problem is the patient with a B or B’ profile at the outset. In this case the lung is not useful and the continuous control of the IVC is our only weapon.
Thanks
Bye
Gaetano
Loved the podcast and I think the concept of fluid tolerance is a great one…. Im not sure how I could practice EM without echo now and realise that I was totally crap at assessing volume status before hand….! I think IVC assessment combined with functional echo assessment is a great way to characterise the nature of the hypoperfusion state, including sepsis, cardiac failure and the various cocktails of both that seems to present. Most of the patients I see are not in what you would call septic shock (obviously some are), but many (given the increasing proportion of failing… Read more »
James,
Amazing comments! Now that is truly interesting about aortic root diameter cancelling out in the equation. Is that definitely true? I must admit I’ve never done the actual calcs, just let the machine do the math. It seems intuitive that if it is a multiplicative value, then you are right, don’t need to calculate the aboslute just the relative change. v. cool.
re: ‘is that definitely true’.. mathematically, you can remove root diameter (and root area, the end destination for that variable) and still get a percentage change from before/after VTIs. ie, if you correctly assume that root diameter is not going to change, then SV and LVOT VTI will change proportionately. if the HR stays the same (probably only in paced patients, denervated/neuropathic pts, and fluid unresponsive patients, i would guess) then the CO would also change proportionately to VTI. you can calculate change in CO without root diameter, just like you can calculate change in SV without root diameter. (different… Read more »
This is fantastic! One less annoyance; all I care about is the delta not the absolute for this purpose.
Lets simplify it for the average ED doc in a non academic institution.On Initial Ed Eval. Pt septic. No cvp line. How much and how fast to give fluids in a healthy person.?. What about the chf or esrd patient?
The figure above… I had a hard time tracking down the original article with the citation provided ( Crit Care 2012;16:R188 ). I think that may be an error. Maybe the volume number and Author would help? Here’s a listing of the journal’s issues: http://www.ncbi.nlm.nih.gov/pmc/journals/9/
Found that link, and its a free full text article… The figure Scott uses is from Muller et al. Critical Care 2012, 16(5):R188
http://ccforum.com/content/16/5/R188
After SMACC and particularly the enlightening presentations by Bowra and Nalos on ultrasound, and Myburgh and Harris on shock and cardiac output measurement provided a useful overview of the literature in regard to both IVC U/S, fluid management and circulatory support in shock. It certainly reinforced the prevailing ICU view that excessive fluid administration is deleterious (particularly in sepsis) resulting in tissue oedema which is very difficult to mobilise later. Bowra’s literature search raised sufficient doubt over the precision of IVC ultrasound to accurately guide fluid replacement and certainly anything greater than a 0.9 cm IVC or less than 75%… Read more »
Not sure these are really the take home messages I would take from those two lectures. That approach is going to put every septic pt in an ICU or HDU and the evidence presented would lead to v. different conclusions except for #1.
A patient in septicaemic shock is likely going to need ICU anyway (or at pressors and more invasive monitoring) if they don’t respond to a couple litres of fluid. Pushing 6 litres into a septic patient purely based on IVC ultrasound be a way of keeping the patient in the ICU longer than you need. It’s not a powerful enough tool alone for monitoring haemodynamics..