Search tips
Search criteria 


Logo of gutGutView this articleSubmit a manuscriptReceive email alertsContact usBMJ
Gut. 2007 December; 56(12): 1799.
PMCID: PMC2095685

Care is needed when using serum sodium levels to predict survival for patients with advanced liver disease

Londoño et al (Gut 2007;56:1283–90) report data from a retrospective cohort study and conclude that serum sodium is an independent predictor of survival for patients awaiting liver transplantation. We wish to highlight two issues that clinicians should be aware of before using serum sodium in this way.

First, two method types are commonly employed in modern hospital laboratories to measure serum sodium: indirect and direct. Unfortunately, the method used by the study authors is not described. Measuring sodium indirectly, most commonly using an indirect ion‐selective electrode or more rarely flame emission photometry, employs a sample dilution step. This makes such methods selective to changes in the volume of serum occupied by the solid phase, that is, the non‐water phase. Typically, this occurs with extreme protein and triglyceride concentrations. A direct ion‐selective electrode enables the sample to be measured without a dilution step and the problem above is avoided.

A common example of where the indirect method may give a false reading is the pseudohyponatraemia seen in multiple myeloma, where paraproteinaemia causes artefactually low sodium concentrations. It is also recognised that when serum protein levels are quite low, artefactually high sodium concentrations may result.1 At the extremes of serum protein or lipid levels in our laboratory, we have seen our indirect method underestimate the true serum sodium by up to 12 mmol/l and overestimate it by up to 8 mmol/l. Patients with advanced liver disease are at risk of hypoalbuminaemia; in addition, significant hypertriglyceridaemia can frequently accompany alcoholic liver disease.

Second, Londoño et al state that serum sodium meets most but not all of the ideal criteria for a variable used to estimate prognosis, as it may fluctuate following a number of therapeutic interventions. However, independently of interventions, sodium varies around its homeostatic set‐point, so called biological variation. This has led to a statistically significant change in serum sodium concentration being defined as 5 mmol/l.2 Such a change is likely to be greater in non‐healthy patients.

The accuracy and precision of the method used to measure serum sodium has clear implications when it is used to predict survival in patients with advanced liver disease, especially with a reported 12% decrease in survival per 1 mEq/l fall in serum sodium concentration within a certain range. Clinicians need to be confident that factors such as hypoalbuminaemia or hypertriglyceridaemia, which may lead to inaccurate concentrations by indirect methods, are not present, or that a direct electrode technique has been used. In addition, they should be aware of other sources of variation, both analytical and biological, before using small population‐based differences in serum sodium concentration to predict individual survival.


Competing interests: None declared.


1. Lang T, Prinsloo P, Broughton A F. et al Effect of low protein concentration on serum sodium measurement: pseudohypernatraemia and pseudonormonatraemia! Ann Clin Biochem 2002. 3966–67.67 [PubMed]
2. Smellie W S, Heald A. Hyponatraemia and hypernatraemia: pitfalls in testing. BMJ 2007. 334473–476.476 [PMC free article] [PubMed]

Articles from Gut are provided here courtesy of BMJ Publishing Group