Dr Constantinos Kanaris, PICU and retrieval consultant; @DrKanaris

This blog is written with an aim to dispel the myth that 0.9% saline is superior to Hartmann’s solution (or indeed any other balanced solution). A common justification to this is that saline has no potassium yet balanced solutions have some potassium in them and therefore are “better for the patient”. This justification does not hold any water, either medically or physiologically. This is why:

Q1. How much K+ do Hartmann’s and 0.9% Saline solutions contain respectively? (click arrow to right to reveal answer)
a) 3mmol/L  & 1mmol /L
b) 3mmol/L & zero
c) 5mmol/L & zero
d) 5mmol/L & 1mmol/L  

Correct answer: c) 5mmol/L & zero

So, Hartmann’s solution has 5 mmol/L of K+ whereas 0.9% saline has no K+ content. With that in mind:

Q2. Which fluid do you think is safer to use in a patient with hyperkalaemia (lets assume that’s someone with a K+ more ore equal to 5.5mmol/L)?
a) Hartmann’s is safer
b) 0.9% saline is safer
c) Show me the answer!

Correct answer: a) Hartmann’s is safer

If you went for 0.9% saline, don’t feel too bad – you were not alone! This is a very common misconception, as this recent opinion poll on twitter demonstrates (https://twitter.com/DrKanaris/status/1267132909561004032?s=20) where, out of 2226 votes, 39.5% of healthcare providers answered that Hartmann’s was safer but 38.8% said that 0.9% saline would be safer.  Paediatric critical care teams are challenged on this very concept by the referring hospital team almost on a weekly basis.  

So, what’s the theory

There is a grand total of zero evidence that balanced solutions exacerbates hyperkalaemia. So why does this firmly held belief persist and why is it wrong?

There are 3 reasons to this, both in theory and in practice:

Q3. How much of total body potassium is intracellular and what is its concentration (roughly)?
a) 90% and 4.5mmol/L
b) 76% and 100 mmol/L
c) 95% and 120 mmol/L
d) 98% and 140mmol/ L

Correct answer: d) 98% and 140mml/L

bananas - PartyJollof

Reason 1:  Acidaemia can induce hyperkalaemia.

The bulk of total body potassium is intracellular (about 98%) with a concentration of roughly 140mmol/L. Even a minute shift from within the cell to the extracellular environment has potential to case a dramatic rise on extracellular K+ concentration, way more than any fluid we give a patient.

(Ab)normal saline is chloride rich and a sudden influx of chloride causes hyperchloraemia with resulting non-anion gap metabolic acidosis. If you recall, sodium bicarbonate (a base) is used in many hyperkalaemia protocols in order to alkalinise the extracellular environment and drive K+ to the intracellular compartment. The reverse happens when we acidify our patients by giving chloride rich solutions like 0.9% saline.  The resulting acidosis causes K+ to shift from inside the cell to the extracellular compartment in exchange for hydrogen ions, causing hyperkalaemia. Acidosis also causes decreased K+ secretion and increased reabsorption in the collecting duct, further contributing to hyperkalaemia.

Hartmann’s on the other hand is “effectively” a weak base as it contains 29mmol/L of lactate which gets converted to an equivalent amount of HCO3. Therefore, physiologically it is more likely to cause a K+ shift into the intracellular compartment, thereby reducing hyperkalaemia (1)

Multiple ion transport pathways directly or indirectly affect net K+ flux in skeletal muscle cells (Aronson et al)

Hang on; ‘lactate gets converted to bicarbonate’? What? Surely, that’s a typo?

Fortunately, it is not a typo. In fact, misunderstanding of the metabolism of lactate, used as a buffer in balanced crystalloid solutions such as Hartmann’s, is part of the reason why it is still common practice for ‘abnormal saline’ to be used instead of balanced fluids in patients with hyperkalaemia. This is how it works (disclaimer: physiology rabbit hole warning!):

Curiouser and Curiouser”: my journey down the UX rabbit hole

Hartmann’s solution has a pH of 6, making it acidic. However, it has an alkalinising effect when given IV because of the way lactate is metabolised to bicarbonate and through the consumption of H+ ions by ‘acid anions’. Let us explain further.

Firstly, just as lactate dehydrogenase (LDH) can catalyse the conversion of pyruvate to lactate in anaerobic respiration, the same reaction can go in reverse in what is known as the Cori cycle in the liver.

So we give lactate in Hartmann’s, and now we have pyruvate.

Pyruvate can then undergo oxidation (aerobic cellular respiration), and produce CO2 and H2O. This, with carbonic anhydrase as the catalyst, then becomes carbonic acid, which dissociates to HCO3- and H+.

Blood Carbonic Anhydrase Activity and Zinc Concentration in ...

So from giving lactate in Hartmann’s, we now have bicarbonate.

But we also have H+ ions, cancelling out the alkalinising effect. So what about them?

Enter the concept of ‘acid anions’. An acid anion is an anion capable of accepting an H+ ion. This essentially means it operates as a base, because it mops up protons. They are also known as the ‘conjugate base’ in conjugate acid-base pairs (eg. the chloride ion when hydrochloric acid dissociates to H+ and Cl-). So when these acid anions (in Hartmann’s solution, this is Cl-, which was bound to Na+ rather than H+, so no extra protons) are metabolised in the liver, they will consume H+.

This leads to a net production of bicarbonate and explains the alkalinising effect of Hartmann’s.

(image from derangedphysiology.com)

Reason 2: Relative concentrations

Q4. If, at standard conditions, you have two solutions of equal volume (let us assume 500ml of each), one having a K+ concentration of 10 mmol/L and one with a concentration of 4mmol/L, what will be the net K+ concentration if you mix the two together to make a litre of solution?
a) 14mmol/L
b) 12mmol/L
c) 7mmol/L
d) 5mmol/L

Correct answer: c) 7mmol/L

If we are having the conversation whether Hartmann’s will exacerbate your patient’s hyperkalaemia or not, it is safe to assume that your patient has a high enough K+ level for you to worry about them (lets say 5.5mmol/L or above). As we established earlier, Hartmann’s has 5mmol/L of K+ in, so significantly lower than our patient’s potassium concentration.

Mixing a more dilute solution with a more concentrated solution can never make the net solution more concentrated; it can only make it less concentrated. 

Reason 3: Relative volumes

The distribution volume of K+ is far greater than just the extracellular fluid volume. Any infusion with a physiologically appropriate K+ concentration will have minimal influence on serum potassium levels. Let us explain.

Here’s a theoretical patient for you: A 50kg teenager with hyperkalaemic. Her serum K+ is 7mmol/L .

Hyperkalemia – Core EM

Q5. What is the estimated total extracellular fluid volume of our patient (ECF is plasma plus interstitial fluid plus transcellular fluid)?
a) 10 litres
b) 20 litres
c) 7.5 litres
d) 2.5 litres

Correct answer: a) 10 litres

Remember this with the 60-40-20 rule. Total body water (TBW) is around 60% of body weight. Two thirds of TBW is intracellular, and one third is extracellular. Therefore ECF is around 20% of body weight. So in a 50kg teenager, that’s 10L. Anyway, we digress.

Now imagine we give her a litre of a theoretical solution containing 9 mmo/L of K+ (remember – this is a theoretical experiment and not to be tried in real life!!).

Q6. Post infusion, the K+ should be a weighted average of the two solutions mixed together. What is the net concentration after the infusion?
a) 8mmol/L
b) 7.18mmol/L
c) 7.68mmol/L
d) 8.28 mmol/L

Correct answer: b) 7.18 mmol/L
How did we get that?: (10 litres of 7mmol/L) + (1 litre of 9mmol/L) = 79mmol of potassium in 11 litres fluid. That’s 7.18mmol/L.

So, a measurable difference of 0.18mmol/L from our patient’s baseline, but a very small one. And remember, potassium balances continuously between the intracellular and extracellular space, meaning the volume of distribution is much higher than just the ECF volume. Thus our net K+ increase will be much lower than 0.18mmo/L.  Now remember that in our theoretical experiment we used a solution with nearly double the K+ concentration of Hartmann’s . The outcome of that experiment was that in theory, we would increase the serum K+ by less than 0.18mmol/L . We would therefore need huge volumes for such a fluid to cause a significant potassium rise, let alone a fluid with physiological K+ concentrations.

So what’s the evidence?

If you’re yet to be convinced by the theory, we have further evidence to take into account.

There are at least three prospective double-blind randomized controlled trials relevant to our subject.

These compare the effect of Hartmann’s solution VS (ab)normal saline on the potassium levels in patients with renal failure.

Study 1

The first study was published in 2005 by O’Malley (2). This is a prospective double blind RCT whereby the two fluids were used intraoperatively in 51 adult patients undergoing renal transplants. Primary outcome was the effect of fluid choice in post-operative creatinine levels. The study was stopped due to safety concerns as interim analysis showed that in the 0.9% Saline group 19% developed K+> 6 and 31% developed metabolic acidosis requiring bicarbonate correction.  None of the patients receiving Hartmann’s developed hyperkalaemia or acidosis requiring correction (p = 0.05 and 0.004 respectively). The authors concluded that Hartmann’s is safe and may be superior to saline in patients undergoing renal transplant.

Study 2

Khajavi and colleagues (3) repeated the study in 2008; the study was almost identical with 52 patients also undergoing renal transplant.  The mean change in serum potassium during the procedure was +0.5 mmol/L in the (Ab)normal saline group compared to -0.5 mmol/L in the Hartmann’s group (p < 0.001). The former also had more acidosis when compared to the latter.

Study 3

Finally the largest study with 74 patients was by Modi in 2012 (4) (bizarrely only published as a letter to editor).  Again, this was almost identical to the two studies above. K+ increased amongst the saline arm group by 0.37 mmoL/L (p < 0.05). There was no significant change in K+ in the Hartmann’s arm  

Like a bulk of evidence applied to paediatrics, all studies were performed on the adult population. There are other study limitations to consider too. The papers do not factor in intraoperative K+ surge due to tissue damage or apoptosis and the volumes given in the studies are very liberal. So liberal in fact that it would be reasonable to suggest that in all likelihood, a smaller infused volume would cause less dramatic swings in K+ and pH. 

In summary:

“We’ve always done it that way” is one of the most dangerous phrases you will hear in daily healthcare practice.

Long-standing protocols are there to be challenged. Nothing in medicine is carved in stone. Balanced solutions are not only not contraindicated in children with hyperkalaemia but they are a considerably safer option than (ab)normal saline .  The time to start prescribing Hartmann’s, Ringer’s lactate or even Plasmalyte, and challenge dogma is now.

1: If you are really interested in the intricacies of K+ homeostasis at a cellular level in response to pH I would highly recommend this paper, complete with easy to follow diagrams of the various co-transporter channels on the cell wall Aronson, P. S., & Giebisch, G. (2011). Effects of pH on potassium: new explanations for old observationsJournal of the American Society of Nephrology22(11), 1981-1989.
2: O’Malley, Catherine MN, Robert J. Frumento, Mark A. Hardy, Alan I. Benvenisty, Tricia E. Brentjens, John S. Mercer, and Elliott Bennett-Guerrero. “A randomized, double-blind comparison of lactated Ringer’s solution and 0.9% NaCl during renal transplantation.” Anesthesia & Analgesia 100, no. 5 (2005): 1518-1524.
3: Khajavi, M.R., Etezadi, F., Moharari, R.S., Imani, F., Meysamie, A.P., Khashayar, P. and Najafi, A., 2008. Effects of normal saline vs. lactated ringer’s during renal transplantationRenal failure30(5), pp.535-539
4: Modi, M.P., Vora, K.S., Parikh, G.P. and Shah, V.R., 2012. A comparative study of impact of infusion of Ringer’s Lactate solution versus normal saline on acid-base balance and serum electrolytes during live related renal transplantationSaudi Journal of Kidney Diseases and Transplantation23(1), p.135.

1 thought on “Hartmann’s in Hyperkalaemia: is that o(k)+?”

  1. This is one of the most enlightening pieces I have read in a while. It will certainly change my practice. Thank you

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.