Let's get salty part 2

More salt! Wait, less salt. Or at least the management of less salt. Let's review the evidence supporting the management of hyponatremia (or not)...

Aegisdottir H, Cooray C, Wirdefeldt K, Piehl F, Sveinsson O. Incidence of osmotic demyelination syndrome in Sweden: A nationwide study. Acta Neurol Scand. 2019 Nov;140(5):342-349.

Ayus JC, Moritz ML, Fuentes NA, Mejia JR, Alfonso JM, Shin S, Fralick M, Ciapponi A. Correction Rates and Clinical Outcomes in Hospitalized Adults With Severe Hyponatremia: A Systematic Review and Meta-Analysis. JAMA Intern Med. 2025 Jan 1;185(1):38-51. 

Chander S, Kumari R, Lohana AC, Rahaman Z, Parkash O, Shiwlani S, Mohammed YN, Wang HY, Chi H, Tan W, Kumar SK, Sindhu F. Urea to Treat Hyponatremia Due to Syndrome of Inappropriate Antidiuretic Hormone Secretion: A Systematic Review and Meta-Analysis. Am J Kidney Dis. 2025 Mar;85(3):303-319. 

Falhammar H, Andersson Franko M, Lindh JD, Skov J, Mannheimer B. Neurologic and psychiatric disorders following correction of profound hyponatremia - A cohort study. Eur J Intern Med. 2025 Oct;140:106394.

Krisanapan P, Vongsanim S, Pin-On P, Ruengorn C, Noppakun K. Efficacy of Furosemide, Oral Sodium Chloride, and Fluid Restriction for Treatment of Syndrome of Inappropriate Antidiuresis (SIAD): An Open-label Randomized Controlled Study (The EFFUSE-FLUID Trial). Am J Kidney Dis. 2020 Aug;76(2):203-212. 

Refardt J, Imber C, Sailer CO, Jeanloz N, Potasso L, Kutz A, Widmer A, Urwyler SA, Ebrahimi F, Vogt DR, Winzeler B, Christ-Crain M. A Randomized Trial of Empagliflozin to Increase Plasma Sodium Levels in Patients with the Syndrome of Inappropriate Antidiuresis. J Am Soc Nephrol. 2020 Mar;31(3):615-624. 

Refardt J, Imber C, Nobbenhuis R, Sailer CO, Haslbauer A, Monnerat S, Bathelt C, Vogt DR, Berres M, Winzeler B, Bridenbaugh SA, Christ-Crain M. Treatment Effect of the SGLT2 Inhibitor Empagliflozin on Chronic Syndrome of Inappropriate Antidiuresis: Results of a Randomized, Double-Blind, Placebo-Controlled, Crossover Trial. J Am Soc Nephrol. 2023 Feb 1;34(2):322-332. 

Rondon-Berrios H. Diagnostic and Therapeutic Strategies to Severe Hyponatremia in the Intensive Care Unit. J Intensive Care Med. 2024 Nov;39(11):1039-1054.

Seethapathy H, Zhao S, Ouyang T, Passos C, Sarang A, Cheung PW, Waikar SS, Steele DJR, Kalim S, Allegretti AS, Ayus JC, Nigwekar SU. Severe Hyponatremia Correction, Mortality, and Central Pontine Myelinolysis. NEJM Evid. 2023 Oct;2(10):EVIDoa2300107.

Schrier RW, Gross P, Gheorghiade M, Berl T, Verbalis JG, Czerwiec FS, Orlandi C; SALT Investigators. Tolvaptan, a selective oral vasopressin V2-receptor antagonist, for hyponatremia. N Engl J Med. 2006 Nov 16;355(20):2099-112. 

Spasovski G, Vanholder R, Allolio B, Annane D, Ball S, Bichet D, Decaux G, Fenske W, Hoorn EJ, Ichai C, Joannidis M, Soupart A, Zietse R, Haller M, van der Veer S, Van Biesen W, Nagler E; Hyponatraemia Guideline Development Group. Clinical practice guideline on diagnosis and treatment of hyponatraemia. Eur J Endocrinol. 2014 Feb 25;170(3):G1-47. 

Verbalis JG, Goldsmith SR, Greenberg A, Korzelius C, Schrier RW, Sterns RH, Thompson CJ. Diagnosis, evaluation, and treatment of hyponatremia: expert panel recommendations. Am J Med. 2013 Oct;126(10 Suppl 1):S1-42. 

Transcript

SPEAKER_00 0:16

Salut salut! This is the last episode I'm recording before going to France. I will not be working on podcasts while I'm there, so possibly the next episode will be a bit delayed. Not sorry. Because this episode has so many papers and so much core content, I will not be telling any stories. Many of you are probably thrilled about that. No more fast-forwarding, just straight to the meat. For those of you missing my fascinating banter, I'm hoping to post pictures of my trip on Instagram at BrightRedblood1. I will probably also have plenty of stories to share on a future podcast, hopefully not about navigating the French healthcare system with my son. Today we are getting back to our unfinished business with hyponatremia. Last episode, we exhaustively explored the pathophysiology, scrambling our brains a bit in the process. Hopefully, you've all had time to get yourself straight in the intervening two weeks, and now we're ready to talk about treatment and the evidence supporting treatment. Or not, as the case may be. Full disclosure up front, there is not a lot of great literature about treatment. Many of the papers I'm covering today are to highlight what we don't know. There are a few hyponetremia guidelines out that provide an excellent jumping-off point, though. One by a conglomerate of European associations for critical care, nephrology, and endocrinology, another by a panel of very knowledgeable, hopefully, experts in the US. They aren't that dissimilar. They seem to be unbiased. The research they cite is a little thin, but they readily acknowledge this fact. So refreshing. Caveat that these guidelines are over 10 years old now, and some studies have definitely been published since. Probably the biggest new revelations are about correction rate, but nothing definitive either. Let's just go ahead and delve into the management, and in the process, we will hopefully also shed light on exactly what the evidence does and does not tell

Acute vs chronic hyponatremia

SPEAKER_00 1:59

us. First, let's discuss this acute versus chronic hyponautremia business. I'm sure you have heard this dichotomized before, and when we learned about hyponatremia in med school, the difference was sold as very important, because if it's chronic and we correct too quickly, this is supposedly very, very bad and shameful. How do we determine if we're dealing with acute or chronic? Well, the cutoff is the 48-hour time point, roughly. In other words, chronic hyponatremia is still pretty acute. So why do we use 48 hours? Because that seems to be how long it takes for the brain to adapt to its new watery surroundings. It is during this acute time that brain edema can develop as water diffuses along osmotic gradients and into the more concentrated cells. The resultant increase in brain volume happens within a stubbornly rigid skull, such that an increase in size of 10% or more leads to herniation, blown pupils, and then a gruesome death. Good times. After about 48 hours, you are pretty much outside the window for brain swelling. Good, right? Well, then you have other problems. During the first 48 hours, because the brain is in flux or not fully adapted, you can conceivably correct the sodium pretty rapidly. You are just heading things off at the pass. After 48 hours, when the brain cells have already settled in, any correction forces them to readjust again, with the water then exiting back out of the cells, just kidding, potentially leading to the dreaded breakdown of myelin known as osmotic demyelation syndrome or ODS. Ah, so yes, the differentiation is important then, right? Well, in the real world, it can be hard to tell if we're seeing acute hyponeia much of the time. And even when we think it's acute, we have to be extra super spectacularly sure and then probably just play it safe anyway and treat it as chronic, which means mining the correction rate. Because if you're wrong about chronicity, possibly you have ruined some poor bastard's life. Although later we will discuss how the whole ODS thing is a bit overblown anyway. So, when might we see actual acute hyponatremia that we can be pretty sure of? It can happen with extreme exercise, like after you've just run an ultramarathon, for example, as one does. Okay, ultramarathoning is rare. My cousin has run the idider rod a few times, but he's completely insane. Actually, surprisingly very normal. But even then, apparently developing hyponatremia while ultramarathoning is even more rare, but maybe ludicrously common, depending on the study you read, it's not been well studied or studied well, I guess. It can also happen after some sort of acute catastrophic intracranial process in your local neurocritical care unit, where most of us medicine doctors don't hang out a lot unless you are in palliative care. Even in these cases, you likely won't trust the chronicity unless you see normal labs one minute and then they go haywire the next. Okay, but actually the designation might not be that important because true dangerous acute hyponatremia presents with severe symptoms. And guess what? Severe chronic hyponatremia, or severe supposedly chronic hyponatremia, also presents with severe symptoms. And you treat them both the same. Another full disclosure that I read a lot of chatter about how the severe symptoms and chronic hyponatremia are also caused by brain swelling. But then if it's past 48 hours, hasn't the brain adapted? Or maybe there's an acute on chronic situation happening, no exceptionally learned person would put their money down on the mechanism, so this less exceptionally learned person, that is me, also is not going to put her, I mean, my money down. But I guess it really doesn't matter. I just don't like things I can't explain. Moving on. What constitutes severe chronic hyponatremia in terms of the numbers? Usually a sodium of less than 120. Some literature references a sodium less than 125, though I'm not usually getting worked up unless it's less than 115 or I see severe symptoms. And usually severe symptoms don't present until you are down closer to around 115. What constitutes severe symptoms? Well, there is a consensus that seizure and coma are severe. That seems like desperately low-hanging fruit. Possibly vomiting and somnolence fall into this category as well, but more controversial. Beyond that, it's another direction the wind happens to be blowing on a given day situation.

Treatment of severe hyponatremia

SPEAKER_00 6:04

For practical purposes, if you have a patient very confused or somnolent or worse, whether acute or chronic, the patient is in extremis and needs immediate decisive intervention. Both the European and US guidelines recommend the same management for this type of patient, with the only difference being that the US guidelines support throwing caution to the wind with the correction rate in an acute setting, while the Europeans are more careful and indicate you shouldn't increase by more than 10 mil equivalents in the first 24 hours, regardless. But that's just because they assume everyone's stupid and may have mistaken chronic for acute, and to be honest, that's fair. But on to nuts and bolts. For all the ideologies of acute and chronic hyponatremia with severe or moderately severe symptoms, both of the guidelines recommend infusing hypertonic saline directly into the bloodstream. They also agree that in the very severe cases you should increase the serum sodium by about 5 ml equivalents in the first hour. Quite a bit. This is done using bolus doses of hypertonic saline, two boluses of 150 cc's, or up to three boluses of 100 cc's. Why hypertonic saline? Because that was what was done first, and there were a number of case series showing that it worked. We don't have head-to-head studies with alternatives. Certainly isn't going to be tested against placebo because killing people with medical experiments has been frowned upon since at least the late 1940s. And no one wants to test it versus other options when we know hypertinox saline is effective and the stakes are so high. Why bolus doses? Well, there are a number of trials looking at bolus versus continuous IV infusion, with the question being which strategy corrects just the right amount with just the right speed, without overcorrecting, and with a sustained outcome. Study results are mixed on this. In other words, it probably doesn't matter. We'll have a bigger discussion about correction rates toward the end. Most of the hyponatremia we're going to see, though, is going to be chronic and more moderate, meaning that you don't need to jump straight to IV therapy. This is also where treatment becomes etiology-specific and where there is a bit more research to support what we're doing. A bit, not a

Treatment of SIADH

SPEAKER_00 8:04

ton. Let's first discuss the management of syndrome of inappropriate antidiuretic hormone, or SiADH. Regardless of the trigger, all SiADH is basically treated the same. Foremost, as with anything, remove or treat the thing that is causing the problem in the first place. Get rid of that antipsychotic, SSRI, antiepileptic, if you can. Take out that brain tumor or lung tumor or pancreatic tumor, actually just call palliative care. But in any case, sometimes you can't get rid of the trigger or can't figure out the trigger, and that's where the fun starts. Remember how in SIADH the patient is holding on to too much free water? Well, the first thing you should do is the opposite. Restrict free water intake. There are some older studies looking at this, but they aren't great. They come to the conclusion that we should be restricting down to one liter of water per day or less. Some of them do a calculation based on urinosomes to determine just how little water the patient should be allowed, but that, frankly, is cruel. People get thirst angry? I don't often put folks on a less than 1.5 liter free water restriction because patients hate it and they aren't going to sustain it when they go home. Why set the patient up to fail at discharge? This brings up another question I can't answer for sure. Should we be restricting just free water or all fluids in general? Well, I couldn't find any papers on this after an exhaustive search and asking every AI chatbot I could get my hands on. After which I felt dirty and ashamed because I am completely anti-AI when it comes to clinical decision making. But maybe to find a reference is okay so I can then go to said primary literature myself. I don't know. I think AI is a harbinger of doom in both medicine and the arts, but at least I could then share my feelings of self-loathing with the AI chatbot who encouraged me to engage in some positive self-affirmations. But I digress, where was I? Oh yes, restriction of free water versus all fluids and lack of clarity. Some medical centers tell their patients to restrict only free water. That is, fluids very low in salt like water, coffee, juice, tea, and soda. Fluids higher in salt like broth, tomato juices, or milk products may not need to be restricted. Time for that V8 with soy sauce. But seriously, if anyone knows of any papers addressing this question, please get in touch. Unfortunately, free water restriction doesn't work in up to half of folks. I'm not convinced it's because it actually doesn't work or they just can't tolerate it. So what next? I add salt tabs. Yes, salt is more about volume regulation than blood sodium regulation, smardy pants, but if you ingest enough of it, it will raise your sodium. You can also start LASIKs, which gets rid of hypotonic fluid, meaning it has more water than solute, so that should increase your sodium concentration and also get rid of any excess volume you are holding on to because of the salt tabs. There was a relevant study published in 2020, the Effuse Fluid Trial, that looked at fluid restriction versus fluid restriction plus LASIKs versus fluid restriction plus LASIKs plus salt. I always expect a study with a clever acronym to be large and profound. I was also excited for a grand study on this topic. Well, this was not such a study, but the only one with head-to-head comparisons, and we should understand what we still don't know. It included 92 hospitalized patients with SIADH, mean sodium-125. Their fluid restriction was either 500cc, no crap, or 1 liter, about half and half. I'll point out that less than half of the 500cc group were adherent, so not feasible in the real world, or actually even in a hospital in a study setting. They did a power analysis but ultimately suffered from being grossly underpowered. There was a lot of variability in response to treatment among participants, such that the confidence intervals were wide and so difficult to pin down exactly what was happening. Despite the patients being in the hospital, they looked at outcomes over a long time course. Four, seven, fourteen, and twenty-eight days out from start of treatment, so probably applies more to outpatients without severe hyponatremia. Alright, why am I telling you about this study? What is the takeaway? Well, it did look like those on fluid restriction plus salt plus LASIKs may have had a slightly better response early on, but the wide confidence intervals destroyed any chance of finding something significant. But really, my point is that if someone cites this study as definitive evidence of anything, they don't know what they're talking about, so ignore them. We still just have no idea what we should be doing. Alright, what other options do we have? Well, you might use a Vapdan, like toll Vapdan. This is a vasopressin receptor antagonist that works by blocking the receptor that binds ADH in the late distal tubule and collecting duct, directly blocking the action of that pesky and appropriately secreted ADH. The first major trials reporting the use of Vapdans and hyponatremia were the SALT trials, published in the early 2000s. These were two multi-center randomized double-blind placebo-controlled trials at almost 100 sites, domestic and international, comparing toll Vapdan versus placebo to treat mild to moderate hyponatremia in the outpatient setting. About 40% of patients had SIADH, and the rest were pretty evenly divided between heart failure and cirrhosis. There were over 200 participants in each trial, meeting their recruitment goals. The average starting sodium was 125. It increased by about 4 mL equivalents by day 4 and about 6 m equivalents by day 30 in the toll Vapdan group compared to no change in the placebo group. This came at the expense of some increased thirst and urination in the treatment group. Not shocked. One interesting thing about the trials was that few patients were on free water restriction. Only 17% in the placebo group and 9% in the intervention group. So this study didn't tell us how Vaptans perform head-to-head against free water restriction. But there are some other studies out there indicating that tolvaptin does correct sodium more quickly into a greater magnitude. And everyone wants to solve their problems with a pill, right? Forget lifestyle modification, just open up, down the hatch, problem solved. Well, there are a few problems. First, overcorrection is a risk in over 30% of patients in one study. The other is cost. Tolvaptan is super expensive. But the takeaway is that it works in SIADH. What other cheaper options are available for SIADH? How about urea? This was something I hadn't heard about until a few years ago. I thought urea was only for dry skin and in pee. I thought it sounded compelling though because salt tabs are disgusting, and as I will reiterate, free water restriction is a mild form of torture only used by the CIA in the aftermath of 9-11. Or maybe that was waterboarding. Same difference, right? Well, the Europeans love urea because they are mostly public health systems and it's cheap and easily available. Why not jump on board with our more civilized friends across the Atlantic? There was a relevant article published in 2024, Urea to Treat Hyponatremia Due to Syndrome of Inappropriate Antidiuretic Hormone Secretion, a systematic review and meta-analysis by Chander et al. This wasn't the best paper ever, mostly because the studies it included weren't the best. They attempted a high-quality product though. Prisma guidelines were used. They included studies looking at patients with a diagnosis of SIADH receiving urea orally or by NG tube and reporting at least one outcome. They said they included only those that compared urea to a different treatment or placebo, but that wasn't actually true, which was weird to me. As I've mentioned before, the thing with meta-analyses is you need your studies to be pretty similar. Because if you are trying to group results, you want to be grouping apples with apples and not apples with chickens. Variability can come from many things. Inherent differences in the study population, like if one is looking at children and the other at adults. Different types of intervention, such as IV versus oral formulations. The study design can have a big impact too. Maybe the outcomes are measured on different days, like one is measured on day one and three and the other on days two and seven. Similarly, time of entry into the study can be a major factor. In one study, participants may enter the study at time of admission and another after a delay, so they're all at different places in their hospital course. And then differences in statistical tests and analytical approaches really matter, like maybe one measures RBC folate and the other serum folate, in which case both of those studies suck, so don't use them. Ultimately, they found 16 studies, no randomized control trials, all observational. One had a crossover design, which was probably the best. There was tremendous heterogeneity across the studies, so needless to say, you couldn't compare most of them to each other. Only six had a comparator group at all. I'm not even going to mention those without a comparator because for all we know, those folks would have just improved with time and the urea was irrelevant. For the six studies with a comparator group, one was rated as good quality, two moderate, three poor. Sample size was 13 to 129, but one study was that large number and most of the rest were under 30. Average starting sodium was about 125, so not severe. And what did they show? Okay, why am I even presenting this? Well, because there is interest in using urea, and we should at least understand what the lack of data tells us. Again. Anyway, sodium did increase generally. Three studies compared urea to fluid restriction. They worked about the same. Two studies compared urea to tolvaptin. Surprisingly, they worked about the same. Two compared to no intervention. It definitely worked better than nothing. No, there were not seven studies for those good at math, it's just that one study compared to both fluid restriction and nothing. So was urea helpful? Possibly maybe, which is a better song than adjudication of utility for a medical intervention, it at least doesn't appear to hurt patients medically. One patient across all the studies had reported overcorrection and no subsequent ODS. However, this is the kicker. Up to 50% of patients stop urea due to the taste. There's no sugar coating it, or maybe it should be sugar coated, it tastes like piss. What unanswered questions do I have about using urea? Well, if it's not a more pleasant intervention than salt tabs and free water restriction, both of which are cheap, then when should I be using it? I would want to know how it performs in patients in whom free water restriction or salt tabs are not effective, even when completely adherent. One study did purport to show improvement in those resistant to water restriction, but that was only five patients. I would also like to know if it provides any benefit when added to free water restriction and/or salt tabs. Hopefully there's more to come. In the meantime, I will probably try using it occasionally in patients without taste buds. Another intervention for SIADH you might see gaining traction in coming years is SGLT2 inhibitors. Theoretically, they work by increasing glucosuria, which increases electrolyte free water excretion. There have been two small trials looking at this, both completed by the same research group. One inpatient over the course of four days and another outpatient over the course of 30 days. In both studies, empaglyflosin increased serum sodium by an additional 3 to 4 ml equivalents versus placebo when combined with free water restriction. There was definitely a small risk of dehydration, even though they very specifically excluded folks with CKD3 or worse. But I do think this could be a potentially promising tool in our SIADH treatment toolbox, especially for our patients who already have other indications for SGLT2 inhibitors, which is like everyone, unless you're peeing on yourself. Okay, quick recapping management of SIADH. If it's severe, that hypertonic saline thing might be needed. If it's not severe, free water restriction is first line. Then it's sort of dealer's choice with salt tabs, LASIKs, and urea. Vaptans are expensive and appear to have a higher risk of overcorrection, but might be good in patients for whom other things aren't working. Consider an SGLT2 inhibitors. You will have to rely more on clinical gestalt than clinical

Too much water or too little solute

SPEAKER_00 19:36

data. Before we move on to treating appropriate ADH secretion with reduced effective circulating volume, all mentioned psychogenic polydipsia and low-solute diet as a brief palate cleanser. Well, don't let them drink. And get them to eat. Sodium usually gradually corrects in this setting as you let the kidneys do their thing without introducing further insult. If it's severe, yes, you still need to do that hypertonic saline thing initially, but if not, they will mostly sort themselves out. The end.

Reduced effective circulating volume: hypervolemia

SPEAKER_00 20:04

Okay, now reduced effective circulating volume. Remember how the body prioritizes volume over osms in desperate times? Sometimes this occurs when the body is total volume up. The fluid is there, it's just third spacing. Usually in heart failure, cirrhosis, or nephrotic syndrome. How do we treat this? Mostly we give loop diuretics. It's also a good idea to elevate the legs and apply compression to mobilize fluid, although this isn't strongly evidence-based. Loop diuretics work because they get rid of hypotonic fluid, and by drawing fluid out of the periphery and into the vessels, the problem should be mitigated. Although these folks typically remain borderline hyponeic at baseline. Vabdans have been investigated in heart failure and cirrhosis. Some of these folks were in the salt trials, remember, although most of the studies are looking at its role in diuresis and not improvement of hyponeatremia specifically. Patients do just about as well with tollvapdan as with a diuretic, or with a Vapdan as an adjunct to a diuretic, but as I've said before, Vapdans are super expensive and there is no reduction in mortality or rehospitalization compared to conventional diuretics, so probably not worth

Reduced effective circulating volume: hypovolemia and the DDAVP clamp

SPEAKER_00 21:08

using. Reduced effective circulating volume with low total body volume is a tricky situation. Easy enough to address, they need fluids or blood products now, but you can get yourself into a pickle pretty quickly. Remember how ADH is just a pinch hitter? Well, once the blood volume normalizes and the asthmoreceptor cells plump back up, ADH will be turned off, way off, because all of a sudden the hypothalamus is having a panic attack about this low sodium thing that is going on. And what happens when ADH is turned off completely? Well, much, much, much less water is reabsorbed in the collecting duct. The body is dumping free water as fast as it can to normalize the blood sodium. In this setting, the sodium can increase by two or more m equivalents per hour. And what is the thing we fear most in hyponatremia? Rapid correction. So Internalize this. Once volume is repleted through IV fluids or blood products, the body will start massively dumping free water in an attempt to rapidly correct the sodium. No! We need to keep a careful eye on this and mitigate accordingly. There are two options. One very non-evidence-based option is to place a folie and then match the patient's urine output with D5W flowing in by IV. I have done this and it worked fine. What I like about this method is that the patients are volume down to begin with, so you are also continuing to fix their fluid status in the process. The more sophisticated and possibly better method is to add back ADH as a medication, the so-called DDAVP clamp. This puts the brakes on sodium correction. There are no randomized control trials to support DDAVP clamping. The observational studies are typically looking at its use following or concurrently with IV hypertonic saline. Several strategies have been proposed. The proactive method, where you give DDAVP within two hours of starting hypertonic saline. The theoretical concern here is that you might worsen or prevent the correction of hyponatremia in someone with severe symptoms, but that doesn't seem to play out in practice. There is the reactive approach where you give it as the sodium starts to correct before it is out of control. And the third approach is called rescue, whereby you have already screwed up and are trying to save your ass. There is no consensus about which is best, although the save your ass method seems stupid on its face. Bottom line, there is utility in using DDAVP, we just don't have a lot to go on. There are some specifics about suggested dosing for this in the paper by Rondon Berio SciSight in the show notes,

History of correction rates and osmotic demyelination syndrome

SPEAKER_00 23:38

so take a look. Yeah, so now we've come to correction rates and the dreaded overcorrection. The horrible, no good, very bad complication we want to avoid is osmotic demyelination syndrome, or ODS. When the brain has adapted to hyponatremia and then suddenly it is exposed to a comparatively hypertonic environment, water diffuses out of the cells, brain volume shrinks, and leads to demyelination. I recall learning in med school about the so-called locked-in syndrome that can occur whereby you can only move your eyes, and otherwise you're like a brain in a jar for the rest of your life until someone mercifully puts you out of your misery. But there are actually a whole host of neurologic sequella that can occur. Bulbar symptoms like dysarthria and dysphagia are most common, paralysis of varying degrees, behavioral disturbances, movement disorders, seizures, lethargy, confusion, disorientation, abtendation, coma, so many things. Some of them which are also severe symptoms of hyponatremia. Confusing. And I feel like we're often looking at the sodium changing on labs and then immediately running over to the patient and poking them to make sure they're still moving, and then checking labs two hours later, and then immediately poking at the patient again. But the clinical manifestations of ODS start about two to six days after the overcorrection event. There are some definite predisposing factors that are good to be aware of. Think about alcohol use disorder, severe liver disease, malnutrition, low FOS, and low potassium. Also, you are much more likely to develop ODS if your initial sodium is really low, like 100 to 110. And the vast majority of patients who develop ODS have at least one risk factor. But even with risk factors, the risk overall is very low. And while we used to think that ODS was universally fatal, we now know that is very much not the case. So why did we think it was fatal? Basically because we only identified it on autopsy. So selection bias. We weren't looking at the living. Until the late 1970s, rapid correction of hyponatremia was thought to be necessary for urgent clinical improvement. ODS was a known entity, but its connection to hyponatremia and correction rate was not understood at all. Then in 1977, there was a report of a series of cases of hyponatremic patients who developed ODS, with the size of the demyelinating lesion roughly correlating with the sodium correction rate. Another group came to similar conclusions around that same time, prompting bench science experiments and cute fuzzy animal models everywhere. The evidence from cute fuzzy animal models did show that rapid correction of hyponatremia led to very much not cute demyelation. Although there continued to be disagreement in the medical community about what constituted a safe correction rate, or if it was other factors entirely in the setting of hyponeetremia that led to ODS, like the severity of hyponetremia or other electrolyte derangements. Current guidelines recommend correcting no more than 8 to 10 mil equivalents in the first 24 hours. This is based on a number of case reports and case series of ODS indicating that very few patients develop this complication at rates less than 8, although not none.

A few good Swedish studies on ODS

SPEAKER_00 26:38

There was an interesting qualitative study of ODS published in 2019 that looked at all cases of ODS in Sweden between 1997 and 2011 as listed in their large national health database. Only 83 such patients were identified out of a population of 10 million. 11 patients did not have hyponatremia, or it was unknown. We don't know why it happened in them. 72 of the 83 patients did have hyponatremia, and they had the risk factors. The mean serum sodium was 104, although a quarter did have an initial sodium more than 110. 70% had alcohol use disorder, and about the same number had hypokalemia. Notably, six cases occurred with a correction rate of 8 m equivalents per 24 hours or less, so the slower correction rate did not guarantee safety. The median correction rate of the remaining folks was 17. Yeah, that's kind of high. What symptoms were most common? Well, more than 80% reported dysarthria and dysphagia. 70% developed lymparesis. Only 12% were actually locked in. Fortunately, folks did recover well generally. 60% of patients recovered to functional independence. Six patients did die, but we all have to die sometime, right? A nice companion to this was another study done in Stockholm, Sweden, looking at development of ODS based on sodium correction rate and other risk factors. It included 7,623 patients with hyponatremia in which the 24-hour correction rate could be established. They stratified into four groups based on this rate. The average starting sodium was 122 across groups except those with a correction rate greater than 18, where it was 115. Because apparently it is really hard to control correction with a really slow starting sodium. This brings up an interesting question. Is it the correction rate that is the problem, or is it actually the initial sodium and the correction rate is just incidental but more likely to be high with a really low initial sodium? I don't know. Seven patients out of the more than 7,000 got ODS. Even with a correction rate of greater than 18, there were only two cases of ODS out of 560 patients. Another two cases occurred in the patients with a rate of 13 to 18, two cases with a rate of 9 to 12, and one with a correction rate of 8 or less. What about risk factors? Three of the ODS patients had an initial sodium of less than 105. Three patients had alcohol misuse, four had hypokolemia. They made the point that the initial serum sodium actually was more predictive of ODS than the correction rate. So ODS might be slightly more likely with higher correction rates than lower, but maybe more with lower initial sodium, but still super rare.

Hyponatremia and mortality

SPEAKER_00 29:22

The thing I think we forget when we talk about correction rate is that there is also plenty of better evidence that correcting too slowly increases mortality. More and more has been published on this recently, sort of circling back to the pre-1970s where correcting too slowly was the bigger concern. Let's take a look at two relatively recent studies investigating incidence of both ODS and mortality in patients with hyponatremia. The first is a systematic review and meta-analysis by AS et al. published in JAMIT Internal Medicine in 2025. They included studies looking at outcomes for hospitalized patients with severe hyponatremia, defined as less than 120 or less than 125 with severe symptoms, based on correction rate. The primary outcome was in hospital and 30-day mortality. The secondary outcome of interest was 90-day incidence of ODS. Results were stratified based on categories of correction rate ranging from very slow, or less than 4 to 6 mL equivalents per 24 hours, to very rapid, or more than 12 mile equivalents per 24 hours. 16 studies with 11,800 patients were included. They were all analyzed as observational. No randomization of correction rate, the chips just fell where they may. 11 of the studies had a high risk of bias in the confounding domain, meaning that something other than the correction rate might be responsible for the outcome. That is usually a problem with observational studies, and I was only interested in studies with outcomes adjusted for confounders. The population was by and large older, over 60 years old mostly. 13 of the 16 studies had a mean sodium of 115 or more. Only six of the studies reported on adjusted mortality rates, meaning other factors were controlled for, such as severity of initial illness and comorbidities, which is important in hospitalized patients because one patient's obsad admit for pneumonia is another man's ICU transition to hospice. In any case, relatively rapid correction of sodium, or more than the 8 to 10 recommended mill equivalents per 24 hours, resulted in 32 fewer in-hospital deaths per thousand treated patients compared to those slowly corrected, or less than 8. The number needed to treat to prevent one death was 31. Rapid correction resulted in 221 fewer deaths per thousand patients compared to very slow correction, or less than 4 to 6 mL equivalents per 24 hours, with a number needed to treat of 5. Rapid correction also was associated with significantly reduced long-term mortality, but results were unadjusted, so I'm not going to report them. Okay, so slow correction, bad, very slow, worse, but what about the dreaded ODS? In 14 studies with over 11,000 patients, only 33 cases were identified. 20 were in the rapid or very rapid correction group, and 11 in the slow or very slow group. The rate was higher in the rapid groups than the slow groups, 0.45% versus 0.135%, respectively. The number needed to harm for rapid or very rapid versus slow or very slow to cause one case of ODS? 317. So 317 patients would need to be corrected more slowly to prevent one case of ODS. Another way of looking at this is that for every one patient with ODS possibly prevented by keeping the correction rate around 8, 10 patients will die. To be fair, patients still might develop ODS with a lower correction rate too. And they still might die with a higher correction rate. But I think it's important to realize that the mortality risk is absolutely something that needs to be counterbalanced against the demyelination issue. The last study I want to mention was one of those included in the prior review, but I think it warrants discussing individually. This was a retrospective cohort study of 3,274 patients presenting to two different Boston area hospitals between 1993 and 2018 with severe hyponatremia defined as less than 120. Patients were divided into groups based on correction rate. Slow at less than 6 mL equivalents per 24 hours, fast at more than 10 mquivalents per 24 hours, and 6 to 10 was the benchmark. Outcomes were in hospital and 30-day mortality and incidence of ODS within 90 days of hospitalization. Length of stay was also assessed, but I don't actually care about this because I'm not trying to sell anything to the administrators today. They did a propensity score-weighted analysis when assessing mortality rates, controlling for a few things, including the admission sodium level and the Carlson Comorbidity Index. Overall, inpatient mortality was 9% and 30-day mortality was 14%. Mortality absolutely increased with declining correction rate. 5% inpatient and 8% 30-day mortality for rapid correction with an odds ratio of 0.76 in the weighted analysis, meaning that you were 24% less likely to die with rapid correction compared to the industry standard, so to speak. With slow correction, there was 13% inpatient mortality and 21% 30-day mortality, with an odds ratio of 1.54 in the weighted analysis, meaning you were 54% more likely to die with slow correction compared to the industry standard. Yikes. There were seven patients with ODS, a rate of 0.2% overall. Five had a correction rate of eight or less. Yes. Most were eight or less. All but one had risk factors like alcohol use disorder or other electrolyte disturbances. Now I'm not saying that we should run out and give hypertonic saline to everyone with abandon, but I am suggesting that we question whether our priorities are skewed. I think we need to make sure we are considering all potential benefits and risks when proceeding with correction. I will mention that there is one caveat when looking at the studies addressing rates of mortality versus ODS. Few patients had an initial starting sodium of less than 110. With lower sodium, there may be a higher risk of both mortality and ODS, but it is largely an evidence-free zone. Ugh. Again. Okay, so that was a lot. What are the takeaways? Well, first, we don't have a ton of great evidence about management of hyponatremia. I'm not sure we can definitively say any one approach is more right or more wrong due to lack of evidence. Probably fancy, expensive medications aren't necessary. Demeclocycline and lithium are not expensive, but also not a good idea, which is why I didn't mention them at all. The utility may lie in understanding how to diagnose the underlying cause and then treating it with old cheap standbys with careful monitoring. Hypertonic saline for anyone with severe symptoms, possibly with a DDAVP chaser, which should be given before the catastrophe and not after. As for correction rates, I think 10 to 12 might be ideal for patients with lower risk of ODS, that is, sodium-115 or above, otherwise normal electrolytes and no booze, maybe closer to 8 to 10 for those who are very high risk, initial sodium-105-ish, alcohol use disorder, hypocalemia, low FOSS. But overall, a little faster might be better than a little slower if your priority is saving lives. Now that we have beat hyponitremia to death, we can hopefully feel good about our bulging brains. From knowledge, not fluid imbalance, and you can spend the weekend dazzling your friends with your new hyponetremia management prowess. See you next time. Thank you.