Tasty Morsels of Critical Care Andy Neill

Welcome back to the tasty morsels of critical care podcast.

Today we cover an incredibly common inpatient issue – hypnatraemia. We’ll often find 1 or 2 of these in our high dependency unit at any given time, mainly due to the requirement for frequent testing of Na levels that seems beyond the remit of normal ward level care. The approach I describe here is neither comprehensive or especially robust but it is how I approach it. Caveat emptor and all that.

The over bearing demyelinating elephant in the room in hyponatraemia is the risk of osmotic demyelinating syndrome (the pathology formerly known as central pontine myelinolysis). If we correct the Na too fast will our patients end up with a severe brain injury? This is rare but is a very real phenomenon.The brain is actually quite good at adapting to sodium levels that have lowered over a few days or weeks. Hence why the slow developing sodium of 120 often causes minimal or no symptoms. However once the patient is in this adapted state (as mentioned this probably is after a few days at a minimum) then a rapid return to baseline sodium can cause ODS. By contrast a rapid drop in sodium, eg over a few hours drinking litres of unnecessary water during a marathon, is poorly tolerated but the plus side is it can be corrected fairly rapidly without harm.

Most of the hyponatraemia we see admitted through the ED will be hypoosmotic hyponatraemia. The bucket here will include heart failure, cirrhosis, SIADH, tea and toast and beer potomania. I’m going to put these common ones to one side for a minute and look at some of the niche exam ones.

For example, i said hypoosmotic hyponnatraemia there, so presumably there could be an isotonic and a hypertonic verison. There is indeed. The isotonic hyponatraemias are usually from spurious results. For example, when you have high lipids (super high, like high enough to cause pancreatitis high) or high proteins (eg high paraproteins like myleoma) the measurement method can underestimate the sodium. You can work this out by always sending a serum osmolality. If this is normal but the Na is 125 and your calculated osmolality is low, then you have an isoosmotic hyponatraemia. You should then check the lipids and the protein. Hypertonic hyponatraemia is another strange beast. This time the tonicity is high from something else such as high glucose or mannitol drawing water from cells into plasma. Again a mix of clinical context and a serum osm will help you out here.

Let’s go back to the bread and butter (or should i say the “tea and toast”) hyponatraemia, the hypotonic or hypoosmotic hyponatraemia. Context as always will give you lots of clues, if the patient has consumed nothing but beer for weeks then the likely causes is beer potomania. If the patient has a new cancer then SIADH is high up your list.

I confess I lean heavily on the approach you can see on Deranged Physiology and have Alex Yartsev’s flow diagram saved on my phone and i look at it almost every time i’m trying to work this out. The first test (assuming you’ve confirmed this is hypotonic hyponatraemia) in this algorithm is urinary osm, the question you are asking here is whether the kidneys are doing what they’re meant to be doing in the face of a low sodium. A normal sane and functioning kidney will try and lose water to conentrate the plasma in order to bring the sodium back up to normal, in other words the kidney should be producing a dilute urine with a low osm. Next step is to check the concentration of sodium in this dilute urine. If the kidney is doing what it should be doing it should be holding onto to all the sodium it can and urine sodium should be low.

Welcome back to the tasty morsels of critical care podcast.

Today we’ll cover some key exam content, all be it not something you’re likely to run into in the ICU too often. The thyroid is a deceptive little organ, tucked in the neck, quietly secreting hormones and interfering in negative feedback loops. It usually restricts its mischief to outpatient clinics by running hot or cold on a chronic basis, occasionally hypertrophying and interfering with its more important neighbour the airway. But every now and then in a pique it decides it’s fed up of this low level mischief and uses its deeply embedded relationship with the rest of the body to wreak havoc.

We’ll split this into 2 parts, one when the thyroid goes on strike and is under active and the other when it goes bananas and secretes far too much hormone

Some basic physiology. Thyroid hormones are essential for all organ systems. The active forms are T3 and T4. T3 is generally the more active one. They are synthesised by incorporating iodine into tyrosine residues in thyroglobulin in the thyroid gland. Hence how iodine deficiency can cause a deficit in thyroid hromone. Their release into the circulation is stimulated by TSH. TSH causes endocytosis of this thyroglobulin into the follicular cells where they undergo hydrolysis into T3 and T4 which is released into the circulation. Both are highly protein bound to thyroid binding globulin.

Our first relevant condition is the wonderfully named thyroid storm. Most commonly you might see this as part of untreated Grave’s disease. It can be precipitated by the usual physiological stressors such as surgery or sepsis etc…

Expect to see (at least in an exam scenario)



* fever

* tachycardia or fast AF

* jaundice

* delirium

* heart failure

* eye signs or a goitre consistent with thyroid disease



For awareness there is a clinical prediction tool that rejoices in the name Burch-Wartofsky Point Scale. This includes most of the features listed above. It’s clear that the features listed above are fairly non specific and like always it’s likely just sepsis. But if something in the spidey sense tingles then finding undetectable TSH and high T3 or T4 should really get you going. In reality this is an incredibly rare diagnosis, one which in its fulminant form i have yet to see. Or perhaps more accurately one that i have failed to diagnose as yet. This is of course hardly surprising as it is hopefully clear by now on this podcast that I am not especially good at what i do and continue to put my appointment to my current job down as some kind of administrative error that is yet to be detected.

Once you’ve decided you’ve made the diagnosis then you’ll need a few basic principles of treatment. Firstly do a bit of resuscitation. There may well be some co existing sepsis so give some antibiotics. If they’re hypoxic give some oxygen. They may need some fluid or indeed they may be in congestive heart failure. The key is to do an assessment, this likely includes having a sneaky peak at the heart and the lungs with ultrasound. A commonly recommended treatment is propanolol to help with the tachycardia. Many patients will be hyperdynamic and tachycardic and giving a beta blocker may well be a good idea but giving a negative inotrope to someone who’s heart is a bit clapped out is generally considered bad form. The key message is to assess comprehensively and then decide.

For specific therapies, your list should include some steroids, this reduces the release of thyroid hormone from the gland. There is occasionally some coexisting adrenal insufficiency so you’ll treat that as well.

Welcome back to the tasty morsels of critical care podcast.

Today we’ll talk about one of the niche and shall I say “advanced” in inverted commas therapies in intensive care practice. ECMO. And to be precise we’ll be talking about VV ECMO. Indeed saying that you are “putting someone on ECMO” is a woefully incomplete sentence as the support and physiological difference between venovenous ECMO and venoarterial ECMO is really rather profound.

The post will be an intentionally broad description of the therapy and perhaps less on the nuances of managing a patient on VV ECMO, as at fellowship exam level I suspect you’d only be expected to have an overview of what it it is, what it can (and can’t do) and when to ask for it. I acknowledge the glaring gaps in the post and the likely criminal omission of the oxygen carrying capacity calculation. It would be fair to call this an idiot’s guide. And given that these posts are generated from my own notes then we all know who the idiot in that title it refers to is.

We’ll start at its simplest level, which is how i try to describe to friends and non medical people about how ECMO works. Blood is removed from the veins in one pipe and put through an artificial lung type device where CO2 is removed and Oxygen added, then blood is returned to the veins via a second pipe. If you’re lungs don’t work so well then the device can replace a lot of their function in the short term. Lay person explanation ends.

The degree to which we can replace lung function, primarily the degree to which we can oxygenate, is determined by the amount of the venous return coming back to the heart we can divert through the machine. Let’s say the cardiac output is a healthy 5L/min. That means that 5L/min is being ejected from the left ventricle and 5L/min is returning to the right ventricle. If the lungs aren’t working well then we need to capture at least 60% or so of this venous return and stick it through the oxygenator in order to maintain tolerable saturation of haemoglobin with oxygen. So in our example we’ll have to be siphoning off at least 3L/min from the venous return, putting it through the oxygenator and returning it back to the right side of the heart. With me so far?

It is at this stage that we immediately run into one of the physics challenges of VV ECMO. Pulling off 3L/min of blood requires pipes of substantial diameter. Typically these are in the 23 to 27Fr range. (ie 8-9mm internal diameter). You want to place this drainage pipe somewhere where there is a high flow of blood in a large vessel capable of accommodating it. Typically this will be in the SVC or the IVC, typically reached by an insertion point in the IJ or femoral vein respectively. It becomes really quite tricky to drain more than 3L/min of blood (or 60% of the venous return) with a single pipe as you can really only drain either the SVC (venous return from the upper body) or the IVC (venous return from the lower body) and as should be obvious the venous return from the body is split between these. In addition to the limitations of the physical size of the pipes you have to remember that the vessels within which these pipes are placed are not rigid fixed stented things, they dilate and contract in response to intravascular volume and intravascular tone. If you try to suck blood out of them with too much negative pressure the vessels will collapse around the pipe blocking all the holes and stopping all drainage.

All this to say that oxygenation is determined by the proportion of venous return we can divert through the ECMO machine. And capturing that venous return should be the priority when it comes to deciding on drainage pipe size and placement. once the blood is out of the body and through the oxygenator it turns out that it’s quite east t...

Welcome back to the tasty morsels of critical care podcast.

Way back in the way back in tasty morsel number 43 we discussed inotropes and vasopressors but there was a noticeable AHD analogue shaped hole in that post that i promised to discuss at a future stage. Well, that time has come and it’s time to run through vasopressin.

You probably first encourntered vasopressin when you heard about ADH in medical school. Anti diruetic hormone, named for what it stops Its discussion in medical school involved delving into the world of endocrinology and negative feedback loops. Something we will be studiously avoiding here. Vasopressin is an ADH analogue, very simillar in structure with very similar effects. As such vasopresin exhibits the same ADH effects but this maxes out at very low doses, much lower than what we use in sepsis. At the very high doses we use, much higher than the pituitary can secrete, it acts as a pure pressor without the inotropic effect we’re use to when using more familiar agents like noradrenaline or adrenaline.

How does it work? Well this is where the fun beings. We’re used to messing around with the adrenergic receptors but vasopressin opens up a whole new bunch of confusing letters that have a whole myriad of effects. Some of these receptors are even shared with other molecules like oxytocin. The main we’re interested in is the V1 receptor, this is found throughout vascular smooth muscle. Stimulating it causes calcium release from the sarcoplasmic reticulum leading to increased vascular tone. Note noradrenaline has the same mechanism (ca release) just through a different receptor. This vasoconstriction affects pretty much all the vasculature including things like the coronaries (not so good) but does seem to spare the pulmonary arteries meaning it may be good in those with pulmonary hypertension.

What other receptors is it worth knowing about? both for exams and the all important one-upmanship on the ward round. V2 receptors are mainly in the renal collecting ducts, this is where we get the ADH effect primarily be increasing the number and effect of something called aquaporin 2 channels. The V3 receptor causes increased ACTH, increasing cortisol secretion, and then there are the OTR and P2 receptors which my notes make no elaboration upon and i will make the dangerous assumption that they have no relevance to what we do in ICM.

Why pull out the vaso when we can get the same vasopressor effect from our beloved noradrenaline. In theory the vasopressin receptors should remain fully funcitonal in the depths of horrific metabolic acidosis that has led your patient into intensive care, the same acidosis in theory should be causing issues with the effectiveness of your catecholamines. It should cause less pulmonary arterial constriction than a catecholamine and should even have less tachyphylaxis. the above list of advantages seems to come straight from the manufacturers advert, so why doesn’t it come pre attached to every patient?

The issue gets a bit clouded due to the somewhat clouded evidence base. I’m going to run through a few of the bigger name trials that one may trot out in a viva type setting, and with all good controversial issues in ICM you could easily go the track of “on the one hand this and the other hand that” and come up with an answer with both buttocks firmly on the fence of the issue.

First up is the VASST trial, (Russel et al 2008 NEJM). Done in North America and Oz, they enrolled septic patients and randomised them to vasopressin vs a blinded infusion of 15mcg/min of norad. Once maxed out on the study drug, then open label additional norad could then be titrated to keep the MAP at target.

Welcome back to the tasty morsels of critical care podcast.

Today we’re going to verge into challenging territory for an audio podcast in that we’re going to the discuss the very visual topic of dynamic LV outflow tract obstruction. This is something fairly dependent on echocardiography for diagnosis which as you can imagine translates poorly to audio format.  This also means you’ll be denied my interpretative dance as i simulate the mitral valve leaflets being pulled over towards the septum via the Venturi effect. But alas i digress.

In essence dynamic LVOTO occurs when the closure point and tips of mitral valve tips are pulled into the left ventricular outflow tract during systole forming an anatomic obstruction to LV outflow thus reducing SV, CO, perfusion etc… This is reflected in poor blood pressure to which we respond by giving more catecholamines which makes this whole thing worse in a horrible cycle of nastiness.

Perhaps it’s best to start by identifying contexts where we should be on the look out for this. We’ll start with sepsis. Sepsis is a state of low systemic vascular resistance leading to reduced preload and afterload in the heart. The LV receives less than usual volume to stretch it and the low afterload makes it incredibly easy for the LV to empty itself of this load. This results in a small cavity LV where the LVOT and the mitral valve find themselves in much closer proximity than they are normally used to. If it gets out of hand bits of the mitral valve find themselves in the LVOT itself causing all kinds of bother.

The incidence of dynamic LVOTO in those with septic shock  is remarkably high and is reported to be  20% in one study from ICU echo guru Michel Slama. Even if it’s not that common it’s yet another reason why the super shocked patient should get a timely echo.

So let’s say we’re worried about our septic patient: within that cohort who is at risk? Classically it would be the older person with LVH or a thickened septal bulge, sometimes called a sigmoid septum. Going with that is a stiff ventricle that fills poorly and has diastolic dysfunction.

As noted at the beginning it’s clear that echo is a key part of the diagnosis here and if you do one you may see some of the baseline features just mentioned but with the addition of SAM or systolic anterior motion of the mitral valve. Most dynamic LVOTO has SAM but not all SAM has LVOTO. SAM can be quite a common out patient echo finding and so in addition to SAM you might want to look for flow acceleration. Just as a river approaching a narrow point accelerates and becomes turbulent so does blood flow in the LVOT approaching an unwelcome and intrusive mitral apparatus. This flow acceleration can be easily measured with doppler and produces characteristic patterns that get echo nerds like me all hot under the collar and is largely beyond the scope of the podcast.

Before we get onto management I want to mention another at risk cohort. These are usually easy to spot as they return from theatre with a big sternotomy following an AV replacement or mitral valve repair. To take the example of the aortic valve. Aortic stenosis leads to severe LVH as the LV has to generate an enormous pressure to get the crusty calcified stenotic valve to open. The heart slowly adapts and learns to live with this very high afterload. Then one day someone opens their chest and pops in a nice shiny new valve that opens like a dream. The LV is not used to this and continues to eject blood like pompeii on a bad day. This hyperdynamic contraction in an LV not used to it has a tendency to drag the mitral apparatus into the LVOT forming an obstruction. Patients following MV repair are also at risk as the change in shape of the annulus and final position of the coaptation point at end repair can also lead to the M...

Welcome back to the tasty morsels of critical care podcast.

Following hot on the heels of tasty morsel number 72 on cardio renal syndrome is its partner in nephron injury: hepatorenal syndrome. This gets covered in a sub section of Oh’s manual chapter 44 on liver issues but there are a variety of other sources mentioned at the end that are worth a read.

It can be a little tricky to pin down this diagnosis. A lot of that comes because it is a “syndrome”, ie a collection of clinical findings that someone has put into a big bucket and mixed around without paying too much attention to hard core diagnostic information like histology or a true pathological diagnosis.

To start with we need context. We should have an AKI in the setting of advanced chronic liver disease and portal hypertension ie cirrhosis. But of course there are multiple reasons for AKI in this context so we have to work through them a little before the label of hepatorenal gets attached. Our friends in the international club of ascites (yes that’s a thing, i didn’t make it up) suggest that you need an AKI with a failure to respond to simple things like withdrawal of nephrotoxic agents, treatment of infection and, importantly a decent trial of albumin. You also have to exclude intrinsic renal diseases that lose protein and blood but this is usually fairly straightforward to exclude. However you can quickly see that a lot of this is pretty nebulous and it can be hard to really draw a line under. As such it’s fair to say that your patient may have several causes for their AKI in cirrhosis and hepatorenal may only be part of the problem.

To take hepatorenal per se, what’s the purported pathogenesis? Well we think that increasing portal venous pressures and cirrhosis leads to splanchnic vascular vasodilation. The vessels in our gut lose tone and we develop this chronic high output, low SVR state. This state of reduced pressure leads to activation of the RAAS causing increased resistance in the renal arteries (in distinction to the very low resistance state of the splanchnic vasculature). As such, perfusing pressure to the glomerulus falls and GFR falls. This is reflected in oliguria and the usual renal response to a crisis of hanging onto Na for all its worth with a urine Na typically 10 if you do go looking for it. In the background you’ve got all this chronic ascites that is adding to the compartment pressure in the abdomen making things worse. That’s the basic bedtime story version of the pathophys that i received, I understand there’s a competing theory where the chronic bacterial translocation of a leaky liver leads to a chronic inflammatory process buggering the kidneys but i digress.

At this stage, it’s worth noting that just like our cardiorenal syndrome we can split hepatorenal into a couple of types. This is all about timing of onset. The in-patient with a rapidly rising creatinine in the hospital setting is more likely to have type I or acute HRS while the stable out patient cirrhotic with a gradually rising creatinine is going to have the type II or chronic HRS.

HRS itself can be precipitated by the usual chronic liver disease decompensations – ie , bleeding, infection, SBP and also large volume paracentesces without appropriate albumin replacement.

How should we treat. First off an important reminder that cirrhosis is not a reversible pathology and if you’re decompensating then the only real treatment to turn the whole thing around is a transplant. All the rest of it is a little bit like rearranging the deck chairs on the Titanic.