Salt of the Earth (Part 1)

Labs and Lytes 014

Author: Adam Drenzla
Reviewer: Chris Nickson

“You are the salt of the earth. But if the salt loses its saltiness, how can it be made salty again?“
— Jesus, Matthew 5:13, New International Version

A 79 year-old lady presented to the ED with one week of lethargy, nausea and decreased food intake. She was started on trimethoprim for a urinary tract infection 3 days prior to presentation. Since then she has had ongoing nausea, worsening confusion and unsteadiness on her feet.  A near fall prompted a neighbour to bring her to hospital.

Past medical history is remarkable for:

  • previous thyroidectomy for autoimmune thyroid disease, for which she is on replacement thyroxine
  • hypertension treated with a b-blocker
  • mild chronic airways disease

On examination:

  • she is a small, thin lady
  • She appears euvolaemic and is haemodynamicaly unremarkable
  • She is disoriented but alert
  • Nothing else of note!

Relevant investigations:

  • Na 114 mmol/L
  • K 3.5 mmol/L
  • albumin 28g/L
  • glucose 5 mmol/L
  • a morning cortisol and TFTs pending
  • Renal function is normal
  • Serum osmolality is 240 mOsm/kg
  • Urine osmolality is 450 mOsm/kg
  • Urine Na 110 mmol/L
  • CTB is unremarkable
  • CXR shows mild interstitial changes

A 1L fluid restriction for possible SIADH was imposed and she was admitted to ICU/HDU for electrolyte monitoring. On arrival an arterial line was inserted and a blood gas sent but before the line could be secured she had a short self-limiting seizure during which her arterial line and peripheral IV cannula were dislodged.

  • The arterial blood gas shows a sodium of 103 mmol/L

A wide bore peripheral cannula was immediately inserted in her cubital fossa and an infusion of 3% NaCl was commenced promptly after a full set of bloods was sent to the lab and a VBG was performed.

  • The venous sodium reported by the lab is 118 mmol/L
  • The VBG sodium is 112 mmol/L

Q1. How does the method of laboratory measurement affect the measured value of plasma sodium?

Depending on the method used, and the clinical situation, the sodium may be artificially high or low!

  • Most hospitals now measure plasma sodium with ion-selective electrodes (ISE)
  • There are two main methods – indirect ISE (used by most central lab analyzers) and direct ISE (used by point of care blood gas analyzers)(Burtis et al, 2012)

Differences in reported plasma sodium values between the two methods arise from two sources:

  1. The best studied is the effect of  changes in the relative size of the solid  versus water phases of plasma. The solid phase is that part of plasma composed of protein and lipids and normally contributes 7% of total plasma volume, the water phase constituting the remaining 93%. Sodium concentration differences in the order of 5-10 mmol/L  due to common variations in the size of the solid phase have been documented (e.g. Dimeski et al, 2012). Though rarely quantified one report suggests 46% of variability between direct and indirect ISE methods is due to changes in total protein concentration (Jones and Twomey, 2008).
  2. There is an additional variability between the two methods which has been less frequently quantified (differences seem to average in the range of 2-5 mmol/L). The source of this difference is less clear.

To put these differences in perspective:

  • About 4 mmol/L is generally considered a clinically significant difference in plasma sodium concentration (Smellie and Heald, 2007)
  • The coefficient of biological variation for an individual is <1% (1-2 mmol/L) as is the coefficient of variation for each method assessed independently (Loughrey et al, 2006)

‘Pseudohypo’ effect

Some reviews suggest that the rare but well known phenomenon of pseudohyponatremia has largely been eliminated by the replacement of flame spectrophotometry with ion selective electrodes, however the indirect ISE is prone to similar errors (Spasovski et al, 2014).

  • this is due to the ‘electrolyte exclusion effect’ that describes the exclusion of electrolytes from the fraction of total plasma volume that is occupied by solids, normally 7%. i.e. The main electrolytes are confined to the water phase, normally 93% (Burtis et al, 2012)
  • Methods such as flame photometry and indirect ISE that require dilution of a small sample of plasma with fixed volume of diluent assume that the water phase is a fixed 93% and thereby the degree of dilution is constant.
  • As the water phase decreases (as occurs in conditions such as hyperlipidemia), the same amount of diluent causes a relatively greater degree of dilution.
  • The assumption that the water phase is fixed therefore leads to a relative underestimation of the true plasma sodium concentration.
  • This “pseudohypo” effect is the basis for pseudohyponatremia or pseudonormonatremia and can lead to an underestimation of any degree of hypernatremia.

‘Pseudohyper’ effect

Less well known is that an increase in the water phase can also occur, leading to a relatively smaller degree of dilution by the fixed volume of diluent.

  • This is more common clinically, particularly in the ICU, due to conditions such as hypoproteinemia and hypoalbuminemia (Story et al, 2007).
  • The assumption that the water phase is fixed in this circumstance leads to overestimation of the true plasma sodium.
  • This overestimation or “pseudohyper” effect can cause pseudohypernatremia or pseudonormonatremia and of course can underestimate any degree of hyponatremia (Lang et al, 2002).

Dilution may not be the solution…

These methods for measuring sodium levels which rely on a dilution step calculate the concentration in total volume of plasma not the more physiologically and physicochemically relevant plasma water volume.

  • Normal plasma water sodium concentration is in fact 140/0.93 or about 150 mmol/L  — which is why normal saline has the sodium concentration it does, it is in fact isotonic to normal plasma water.
  • This negative “error” in plasma electrolyte analysis has been recognised for years . However, it was assumed that the volume fraction of water in plasma was constant enough to allow electrolyte reference intervals to be reported in total plasma concentrations rather than the more physiologic plasma water volume values (Burtis et al, 2012).

To make direct ISE sodium concentration comparable to these historic reference intervals most blood gas machines operate in “flame mode”.

  • This multiplies the directly measured concentration, which is proportional to plasma water phase ion activity in direct ISE, by 0.93. Therefore, although the plasma water volume fraction may vary widely from 93%, as long as the ion is not bound by protein the value is independent of the relative proportions of water and solid.
  • While reducing variability due to fluctuations of water phase size it does mean that as plasma water proportion increases in low protein states the value derived by flame mode becomes further removed from the true value in plasma water. Although this absolute value can easily be determined by dividing the concentration given in flame mode by 0.93 it is rarely done in clinical practice (and its clinical utility is not established) (Story et al, 2007).

Q2. How does the site of blood sampling affect the measured value of plasma sodium?

Sampling site can be important! From the little literature available it appears there are at least four ways different sampling sites can have clinically significant effects on the measurement of sodium concentration


  1. Arterial samples are more often measured by direct ISE in a blood gas machine (see above)
  2. Arterial blood can have a sodium concentration around 5 mmol/L lower than venous blood after an acute water load from the gastrointestinal tract (GIT)(Shafiee et al, 2005).
  3. Short periods of elevated sodium levels (up to 10-15 mmol/L) after convulsive seizures have been documented. The proposed mechanism is that high rates of muscle metabolism lead to cellular hyperosmolality which leads to water uptake.  Until a steady state is established this plasma sodium concentrations will be higher in venous blood draining major muscle groups than in the arterial blood (Welt et al, 1950).
  4. In paediatrics, capillary blood is commonly sampled and its sodium concentration has been shown to be on average 1-2 mmol/L lower than venous blood with a limit of agreement of around 8 mmol/L. Mechanisms postulated include a greater degree of haemolysis with capillary blood sampling leading to extrusion of sodium poor intracellular fluid into the plasma (Loughrey et al, 2006).

Unfortunately there has been little well conducted systematic research into these effects, particularly in clinically relevant settings.The extent and operation of these effects in combination and the possibility of as yet undiscovered effects and their influence is still unclear. For the severely hyponatremic patient in particular there is even less data.

  • There is only one report (Jain et al, 2009) that presents data stratified for serum sodium levelbutit takes no account of the known causes of measurement method or sampling site disagreement.
  • In a sample of 200 ICU patients, they compared measurements of venous blood assessed in the central lab via indirect ISE methods and arterial blood assessed by direct ISE in blood gas machines.
  • With hyponatremia <120 mmol/L the mean difference was 12.8 mmol/L with a maximum difference of 30 mmol/L.

Q3. What implications do these measurement differences have for patient management?

The discrepancies in serum sodium values between direct and indirect ISE are large enough to be clinically meaningful.

  • This will affect therapies and targets for correction of hyponatremia and hypernatremia (Spasovski et al, 2014).
  • It will affect calculation of osmolar and anion gaps and strong ion differences which could impact on need for further investigations and management (Morimatsu et al, 2003).

These effects will be particularly important in conditions of high and (more commonly) low solid phase states, and perhaps more so with  greater derangements in sodium concentration. Laboratories have different protocols for confirming indirect ISE determined sodium levels with direct ISE methods. These can vary depending on lipid, protein and sodium level and it might be worth knowing these triggers in your lab.

It is prudent to monitor sodium levels and their response to therapy with one method, and many experts agree that direct ISE is the method of choice.

There are also circumstances in which differences between arterial and venous sodium levels might be clinically significant (see Q4 and Q5). Awareness of these may at the least avoid management decisions based on misleading values.

Q4. Is oral fluid intake immediately prior to sodium measurement clinically important?

Yes, it can be…

  • As stated in Q2, acute water load results in a lower sodium concentration in arterial than venous blood.
  • It is this arterial sodium concentration which the brain “sees” and is physiologically important as it correlates with plasma vasopressin levels.

This “input physiology” is frequently neglected but can have potentially important clinical implications.

  • If a patient ingests a significant amount of electrolyte free water before arrival in the ED this can become a reservoir of free water input after initial blood tests and commencement of therapy. For example, in the early hours of admission water may continue to be absorbed despite commencement of fluid restriction. This can lead to further fall in sodium with worsening of hyponatremia and its clinical effects and can potentially be missed without arterial sampling as the large mass of the skeletal muscles takes up this extra water limiting the fall in venous sodium (Halperin et al, 2010).

If significant hypotonic fluid intake into the GIT cannot be excluded before hospital arrival it is prudent to assume that such intake has occurred and be aware that hypotonia may worsen substantially before it gets better. Fluid restriction cannot account or ameliorate for fluid that has already been ingested but not yet absorbed!

Q5. How should venous plasma sodium concentrations measured immediately after a seizure be interpreted?

Blood tests, particularly venous blood samples, taken shortly after a convulsive seizure could be misleading (as described in Q2).

  • Consider repeating venous plasma sodium measurements  5-10 mins after seizure termination, to allow time for the water absorbed by skeletal muscle to reinfuse into the circulation as cellular metabolism normalises.

References and Links

  • Budak YU, Huysal K, Polat M. Use of a blood gas analyzer and a laboratory autoanalyzer in routine practice to measure electrolytes in intensive care unit patients. BMC Anesthesiol. 2012 Aug 3;12:17. doi: 10.1186/1471-2253-12-17. PubMed PMID: 22862792; PubMed Central PMCID: PMC3431979.
  • Burtis CA, Ashwood ER, Bruns DE. Chapter 28.Electrolytes and Blood Gases, in, Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. 5th ed 2012. Elsevier
  • Chacko B, Peter JV, Patole S, Fleming JJ, Selvakumar R. Electrolytes assessed by point-of-care testing – Are the values comparable with results obtained from the central laboratory? Indian J Crit Care Med. 2011 Jan;15(1):24-9. doi: 10.4103/0972-5229.78219. PubMed PMID: 21633542; PubMed Central PMCID: PMC3097538.
  • Chow E, Fox N, Gama R. Effect of low serum total protein on sodium and potassium measurement by ion-selective electrodes in critically ill patients. Br J Biomed Sci. 2008;65(3):128-31. PubMed PMID: 18986099.
  • Dimeski G, Barnett RJ. Effects of total plasma protein concentration on plasma sodium, potassium and chloride measurements by an indirect ion selective electrode measuring system. Crit Care Resusc. 2005 Mar;7(1):12-5. PubMed PMID: 16548813
  • Dimeski G, Morgan TJ, Presneill JJ, Venkatesh B. Disagreement between ion selective electrode direct and indirect sodium measurements: estimation of the problem in a tertiary referral hospital. J Crit Care. 2012 Jun;27(3):326.e9-16. doi: 10.1016/j.jcrc.2011.11.003. Epub 2012 Jan 9. PubMed PMID: 22227082.
  • Halperin ML, Kamel KS, Goldstein MB. Fluid, Electrolyte and Acid-Base physiology. A problem-based approach. 4th ed. 2010. Elsevier.
  • Jain A, Subhan I, Joshi M. Comparison of the point-of-care blood gas analyzer versus the laboratory auto-analyzer for the measurement of electrolytes. Int J Emerg Med. 2009 Feb 24;2(2):117-20. doi: 10.1007/s12245-009-0091-1. PubMed PMID: 20157454; PubMed Central PMCID: PMC2700230.
  • Jones BJ, Twomey PJ. Relationship of the absolute difference between direct and indirect ion selective electrode measurement of serum sodium and the total protein concentration. J Clin Pathol. 2008 May;61(5):645-7. doi:10.1136/jcp.2007.050872. PubMed PMID: 18441158.
  • Lang T, Prinsloo P, Broughton AF, Lawson N, Marenah CB. Effect of low protein concentration on serum sodium measurement: pseudohypernatraemia and pseudonormonatraemia! Ann Clin Biochem. 2002 Jan;39(Pt 1):66-7. PubMed PMID: 11853193.
  • Loughrey CM, Hanna EV, McDonnell M, Archbold GP. Sodium measurement: effects of differing sampling and analytical methods. Ann Clin Biochem. 2006 Nov;43(Pt6):488-93. PubMed PMID: 17132280.
  • Macewen C, Watkinson P. Pitfalls in the Management of Severe Hyponatraemia. Nephron Clin Pract. 2012 Oct 5;120(4):c223-c227. [Epub ahead of print] PubMed PMID: 23051651.
  • Morimatsu H, Rocktäschel J, Bellomo R, Uchino S, Goldsmith D, Gutteridge G.Comparison of point-of-care versus central laboratory measurement of electrolyte concentrations on calculations of the anion gap and the strong ion difference. Anesthesiology. 2003 May;98(5):1077-84. PubMed PMID: 12717128.
  • Shafiee MA, Charest AF, Cheema-Dhadli S, Glick DN, Napolova O, Roozbeh J, Semenova E, Sharman A, Halperin ML. Defining conditions that lead to the retention of water: the importance of the arterial sodium concentration. Kidney Int. 2005 Feb;67(2):613-21. PubMed PMID: 15673308.
  • Smellie WS, Heald A. Hyponatraemia and hypernatraemia: pitfalls in testing. BMJ. 2007 Mar 3;334(7591):473-6. PubMed PMID: 17332587; PubMed Central PMCID: PMC1808124.
  • Spasovski G, Vanholder R, Allolio B, Annane D, Ball S, Bichet D, Decaux G, Fenske W, Hoorn E, 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. doi: 10.1530/EJE-13-1020. Print 2014 Mar.PubMed PMID: 24569125.
  • Story DA, Morimatsu H, Egi M, Bellomo R. The effect of albumin concentration on plasma sodium and chloride measurements in critically ill patients. Anesth Analg. 2007 Apr;104(4):893-7. PubMed PMID: 17377102.
  • Turchin A, Seifter JL, Seely EW. Clinical problem-solving. Mind the gap. N Engl J Med. 2003 Oct 9;349(15):1465-9. Erratum in: N Engl J Med. 2004 Feb 5;350(6):629. PubMed PMID: 14534340.
  • Welt LG, Orloff J, Kydd DM, Oltman JE. An example of cellular hyperosmolarity. J Clin Invest. 1950 Jul;29(7):935-9. PubMed PMID: 15436862; PubMed Central PMCID: PMC436130

One comment to “Salt of the Earth (Part 1)”
  1. Pingback: The LITFL Review 143 - LITFL

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.