Hi, so my good friend Jeff Scott, ED/ICU doc and serious POCUSologist, asked me to summarize our current approach to fluid management, which is an amalgam of literature, physiology and bedside medicine-based evidence.
A few points to emphasize:
does my patient need fluid/ will he/she benefit from fluid.
is my patient fluid tolerant
is my patient fluid responsive – yes, it’s the last and least important
I figure we may follow this up with a discussion – that’s often the best way to get to the real clinical decision points, and it’s always interesting to hear the questions and ideas that come up, so looking forward to it!
I figured might as well make a mini podcast of it, so here it is:
So the discussions go on about volume status and POCUS, and recently one in particular made me realize that it is important to reframe the way we think about “volume status.” As Segun Olusanya (better known as @iceman_ex) said, “the IVC is not a fuel tank indicator,” and indeed it is not. But even if it was, would that be useful? If somehow, an 18 mm IVC (short axis circular or average of course!) corresponded exactly to a 0.70 ml/kg blood volume, would that be of any use?
No. Of course not.
I get asked this question in consult a lot. So I could be a stickler on principle and answer, whether verbally or in a consult, that the volume status cannot be precisely ascertained using POCUS, and keep walking down the hospital hallway.
But let’s instead reconsider the true clinical question for a moment. What does “volume status” mean when requested by a colleague. The truth is that he or she is likely asking you whether there is a need to give fluid, remove fluid, or stay the course.
Ahhhh. Now POCUS, the IVC and its friends can help. A lot. A lot more that most clinical examinations and chart reviews of weights or ins and outs can. Way more. Why? Because if you are cool, with a normodynamic heart and a small IVC, you are on the low side of volume. Now you may also have a lot of B lines from your pneumonia and the lack of volume tolerance will give your answer to be very careful with fluids. If you are warm and hyperdynamic with a small IVC and totally clear lungs, no elevated ICP and a soft belly, you may just be vasodilated but, if your BP is on the low side, some fluid is a fair go, so long as you follow closely thereafter for fluid stop points. If you have a low urine output, a big IVC, a pulsatile PV and a poor LV, you probably need lasix, no matter how clear your lungs are and even if your creatinine is rising, in fact, especially since it is rising.
The permutations are myriad. But that’s why we have MDs and are supposed to be able to integrate bits and pieces of physiological data to come up with an understanding of our patients. And POCUS gives us an unprecedented bedside view into this physiology.
So if you do have legit POCUS skills, and are able to do a bit more than a long axis M mode of the IVC, then try this instead:
“Sure thing, now tell me a bit about your patient – I imagine you’re debating whether to give some fluids or diurese?”
Forget about volume status in terms of absolutes. Just think of what the clinical question is, and give your colleagues the answers they need.
I think the patients will do a lot better that way.
So last week sometime we had an interesting twitter exchange which made me realize it is important to explain how some of us are using venous POCUS in different clinical scenarios, which is key, because the development of monosynaptic clinical reflexes with POCUS findings is a rabbit hole we should try not to go down. Instead, POCUS should be about asking the right question and taking that answer as a piece of the pathophysiologic puzzle facing us, which may mean intervening sometimes, and sometimes not, for the same given finding, but with different surroundings.
Thanks to those involved in that discussion – it is how we grow!
And here are some thoughts:
For those not up to speed on venous congestion POCUS I put up the chapter that Korbin Haycock, Rory Spiegel and I worked on in this earlier post.
Here are Korbin’s thoughts on this:
I’m very glad Dr. Eduardo Argaiz pointed this case out, as it brings up considerations apropos both chronic venous congestive cases as well as management of acute illness, particularly in sepsis, where we would expect patients to most likely be fluid responsive, but fluid tolerance is largely overlooked with current management strategies by the majority of clinicians.
Phil’s above audio commentary points out the difference is these two broad categories very nicely. If you didn’t listen to it–you should.
With respect to chronic venous congestive conditions, the knowledge and application of Doppler assessment to therapy will hopefully be the next advance in management at large. Already, I think there is more than adequate research available to show the value of Doppler POCUS (D’POCUS, D/POCUS, or DPOCUS?) in managing these patients. It’s only a matter of clinicians willing to commit to learning and integrate this technology into their skill set.
With respect to resuscitation of the acutely ill patient, there is by far less data, and we are probably into the realm of N=1 here, in terms of how to manage these patients. But, I personally believe–and I understand this is my opinion–that current trends in resuscitation (especially sepsis resuscitation), largely ignores the effect of over volume resuscitation and the potential downstream damage inflicted on our patients.
This theoretical damage of over aggressive fluid resuscitation is multifactorial, including glycocalyx shedding issues/endothelial dysfunction, positive fluid balance and EVLW causing increased mortality (which there is ample evidence for, I think), venous congestion leading to perfusion injuries to encapsulated organs, such as the kidney (AKI) and brain (congestive encephalopathy), and end organ edema leading to the perpetuation of a malignant inflammatory syndrome (portal HTN and gut edema).
In the case called out by Dr. Argaiz, (which can be reviewed by the previous post on this website) my patient had an IVC that whilst not plethoric, was not an IVC that one would expect to find in a patient with a typical distributive shock pattern (i.e. increased cardiac output, decreased SVR, and decreased RAP). Firstly, the complicating factor of atrial fibrillation with RVR was central to the patient’s shock state, however this was quickly addressed with rate control. However, in addition, this particular patient did exhibit additional signs of venous congestion. The portal vein was pulsatile and the intrarenal Doppler pattern was interrupted/bi-phasic in nature. Granted, a pulsatile PV Doppler could be interpreted as related to the hyper dynamic nature of septic shock (as the esteemed Dr. Denault correctly cautioned in his comments on the original post), however a less than flat IVC and the intrarenal findings gave weight to a venous congestive hypothesis as a cause the PV findings as well as a possible cause for his AKI evident on his initial labs.
With this particular case, given my personal global POCUS/FOCUS assessment of his increased LAP (high E/e’), RV dysfunction, RAP, PV, and intrarenal Doppler venous pattern, AND that fact that the RRI was insanely high with an AKI, I elected to treat my hypothetical construct of his renosarca with furosamide and his RRI with vasopressin (as the NE infusion did increase his MAP, BUT NOT decrease his RRI–which the vasopressin infusion did decrease, or so I presume as no other therapeutic interventions were given with respect to the time frame the RRI decreased).
In the end his kidneys had recovered by the next morning, which I’m sure that any intensivist will admit is the opposite of the norm, as the kidneys usually get, at least transiently worse initially-being the delicate sissies/whimps that they are. Whether this was because of the diuretic or the vasopressin, or something else, is debatable for sure, but it sure didn’t get better by 30 cc/kg of crystalloid mandated by CMS, because he got not a drop more than what was needed to push the diltiazem, the lasix, the antibiotics, and the vasopressors.
So to summarize, in the case of chronic cardiogenic venous congestion, clinician realization and adoption of Doppler assessment of this entity will likely be the next leap in improvement in the management of these patients. In the case of acute resuscitation, venous congestion may be a bit more nuanced, and a more comprehensive evaluation is in order in a case by case fashion. However, I think recognition of the issues of over aggressive volume administration will probably be the next frontier in sepsis resuscitation.
So here’s another interesting question as a follow up to the previous discussions:
Most people would probably agree that florid congestive signs on POCUS means the RV is unable to pass any more extra volume to the left heart; whilst the absence of those signs mean that the patient may be able to cope with some additional volume without immediately engorging the vital organs.
And my question is: what about those in between? i.e. the patients who start to develop some mild congestive features on POCUS.
For those who are on the verge of congestion, diuresis would push the RV to the left (i.e. steep part) of Starling curve resulting in significant CO drop; conversely, extra volume pushes the RV to the right (i.e. flat part) leading to congestion or even D-shape LV, directly hindering CO as well. This margin becomes even smaller in patients whose RV starts to fail (i.e. entire Starling curve shifted downwards)
Great, great question. The crux of this, I think, is deciding which is the greater issue, congestion or poor perfusion. Obviously they are intertwined, so the decision will be on a case by case basis. Jonathan alludes here to a narrow “balance point” between congestion and preload dependancy. My feeling – and we’ll see if we can get some consensus – is that this indeed narrow in patients with marked pulmonary hypertension. When patients have pure pump failure congestion, my clinical experience is that you can decongest plenty without drop in systemic CO, in fact it often improves, likely related to ventricular interdependance. So let’s go on…
I’ll illustrate my point with the following scenario:
for previously healthy middle aged patients intubated and admitted to the ICU for ARDS from severe pneumonia, they quite often develop some acute cor pulmonale after mechanically ventilated for several days even if the PEEP/driving pressure isn’t exceptionally high; and they usually have resp failure and shock to start with.
Given that they don’t have pre-existing heart disease, the only signs suggesting the emergence of cor pulmonale could be subtle, without structural changes like dilated RV (RVEDD at most at upper normal range) nor abnormal septal movements. You may see TAPSE dropping to marginal level and portal vein PW signal may become a bit more pulsatile. IVC looks full and RVSP usually rises but not skyrocket. The MV inflow pattern & E/E’ suggest rather normal LA filling pressure, not surprising from a previously healthy heart.
In this case, it isn’t the LV diastolic dysfunction that overly afterloads the RV; and it isn’t the RV dilation that impairs the (D-shape) LV from ventricular interdependence. Therefore I’d consider the right heart circulation & left heart circulation running purely in series, whereby limiting the RV preload could reduce the LV CO.
Now, if this patient goes into shock, would you consider fluid challenge or diuretics? Everyone probably would also get other therapies on board, e.g pressor, inotrope, source control etc. But when the patient’s BP is 80/40mmHg, I am more prone to giving some fluid as I believe that reducing preload in a septic patient can precipitate arrest; and that RV only directly impairs LV CO once the IVS starts to shift, which should take more time and thereby easier to monitor.
Interesting case that happens commonly – if you do POCUS and look for it rather than blind-ish management. Here, you have congestion, likely due to pulmonary disease, fluids, on a normal-ish RV (which also means it is unable to mount a huge PAP).
So personally – and will full disclosure that this is not evidence-based (as if there was any evidence in our resuscitative practices!), I would consider this a relative contraindication to fluids, given the non-volume-tolerant state (ALI/pneumonia/ARDS and portal pulsatility) of the patient. With pulsatility and signs of organ dysfunction I would be diuresing or pulling fluid off. We’ll see if we can get Rory to comment, as he has been doing a fair bit of this.
So in this patient it would be either no fluids, or diurese.
I don’t think one should have a general conception that reducing preload in a septic patient category is an issue. That may be so if you do not have the capability to look, and hence feel you should behave more cautiously. A septic patient with a tiny IVC may indeed be tipped over into low CO by removing fluids, but another with a full tank post resuscitation may benefit. So with the ability to assess hemodynamics, individualized approaches trump general ides and protocols. Much more to come on this in the next weeks as we break down a lot of interesting concepts in regards to vascular tone assessment and cardiac efficiency.
I fully appreciate how ambiguous this situation is and that in reality the only way to find out the treatment that works is often by trial and error. Serial assessment by POCUS is definitely needed and one may even put the entire fluid thing aside and focus on other treatments. But just want to know your take and the reasons behind.
Thanks again for all your work and these thought provoking posts; and my apologies for the supposedly quick question ending up being not so quick. It took me some effort to clearly delineate my question in mind.
Anyone interested in these topics should keep an eye out for the H&R2019 Tracks. A bunch of us are getting together before and during the conference and will be recording discussions on all these little cases and angles around hemodynamics and other fun resuscitationist topics.
So given the importance of these topics, the number of questions and discussions we’ve had on the twitterverse, and most importantly in the spirit of #FOAMed, here is the chapter from the POCUS book which was co-authored by Rory Spiegel (@EMnerd), Korbin Haycock (@korbinhaycockmd) and myself.
So a couple weeks ago I had the chance to sit in sunny Florida with Jeff Scott (@jsemccm), an ED-intensivist who runs the ED at Jackson South in Miami as well as rounding in the ICU at Jackson Memorial.
His group recently published an awesome article on TCD that pretty much made me realize I have to up my TCD game to the next level.
So there clearly is more to be done with TCD than I have been doing, and maybe it really has a place in the ED. I don’t work first line in the ED so initial stroke patients I only see if they deteriorate, but the idea of visualizing perfusion – or reperfusion – is really interesting.
So if you want to meet Jeff and have him teach you some POCUS TCD, don’t miss H&R2019 which is just around the corner. There aren’t many spots left! Jeff will be running a TCD workshop along with Rob Chen (@ottawaheartrob) which I’m really looking forward to!
Love to hear from anyone pushing the envelope of TCD (or any POCUS application). I believe we are only scratching the surface of what we can do with POCUS, and much study, based on front-line clinicians taking bold strides ahead to see what can be done.