Thread/

Do you look after patients with HHS?

UK guidelines use calculated osmolality = 2Na + Glucose + Urea

“What’s the problem?”

Including urea lowers the diagnostic threshold in patients with CKD
- a group at risk of treatment harm

But it gets much worse

#medtwitter

2/

You do the calculation & the recommendation is to use a cut-off of >320 mosmol/kg = hyperosmolar

But what if the patient has advanced CKD and their urea is always 30?

Well, there’s no mention of that in the 20+ page UK guidance....

http://www.diabetologists-abcd.org.uk/JBDS/JBDS_IP_HHS_Adults.pdf

3/

“So what happens?”

When these CKD patients come in with hyperglycaemia it is MUCH easier for them to meet the biochemical criteria for calculated hyperosmolality

= lots of IV fluids (+-insulin) if HHS protocol is started inappropriately
not a good look in late CKD

4/

“But it’s not the guidelines fault that urea is raising tonicity and risking fluids shifts in these CKD patients?”

Except it doesn’t
No, it really doesn’t in anybody

Let’s quickly detour via the concept of effective and ineffective osmoles

5/

Na = effective osmole = it contributes to fluid shifts
The sum has ‘2Na’ to simplify for equal number of effective anions (mostly Cl & HCO3)

Glucose can’t get into cells in absence of (enough) insulin, so also exerts pull on water in HHS

Plug: @TheSkeletonKG case8

6/

However, unlike glucose and sodium,

urea is an ineffective osmole

This means that urea:

moves quickly & freely across cell membranes
won’t contribute to raising extracellular tonicity
won’t cause the fluid shifts we fear in HHS (in CKD or pre-renal states)

7/

“So why include urea in the calculation?”

They justify inclusion by saying urea is an “indicator of severe dehydration”.

So they think it’s worth having a marker of disease severity in the formula, & in patients with normal baseline urea who look dry & sick, that’s fine.

8/

But though this ‘dehydration marker’ argument is fine for most patients, it falls apart in advanced CKD (a not uncommon condition for patients in and out with high blood sugars to have)

It also doesn’t work in other causes of urea that aren’t hypovolaemia

Plug: @BukuRenal https://twitter.com/bukurenal/status/1306500678790123520

9/

“So given urea is an ineffective osmole, why not just use effective osmolality = 2Na + Glucose instead?”

This is where they slip in a reason which is a slightly devastating critique of UK diabetologists within their own HHS guidelines (and I’m sure not true!)

10/

To be fair the guideline discourages using biochemistry alone & also stresses the hypovolaemic diagnostic criteria

BUT

A concern is that the clinical diagnosis of hypovolaemia has large inter-observer variability, & I’ve seen doctors focus effort on the calculation instead

11/

Take home points:

If you calculate osmolality in HHS including urea you’re using it as a ‘dehydration severity marker’

But when urea is simply at it’s high baseline this is illogical, & you decrease diagnostic specificity in a group at high risk of treatment harm

Fin/

Starting HHS protocols is risky given the large IV fluid volumes: in advanced CKD always discuss early with specialists

Guidelines should note calculation limitations & exclusions (& expect us to be able to learn the difference between effective & ineffective osmoles!)