Direct ISE versus indirect ISE plasma sodium measurement in the critically ill

Plasma sodium and chloride concentration are routinely measured using either a direct or indirect ion-specific electrode (ISE). Direct ISE is the technique employed in blood gas and most other point-of-care analyzers whereas indirect ISE is used in the majority of biochemistry profiling analyzers sited in central laboratories. 

In critically ill patients there is often a difference in the plasma sodium and/or chloride concentration when measured by the two techniques. These differences are the focus of this Australian study. Blood was sampled from 300 critically ill patients for measurement of plasma sodium and chloride by both direct and indirect ISE. 

The mean plasma sodium by indirect ISE (laboratory method) was 141 mmol/L compared with 139 mmol/L by direct ISE (blood gas analyzer). This mean difference of 2.1 mmol/L is statistically significant. The limits of agreement ranged from –1.6 mmol/L to +6.2 mmol/L. This exceeds 4 mmol/L, which is regarded as the upper limit of acceptable difference for sodium. 

For 39 (13 %) patients, direct ISE indicated hyponatremia, whereas indirect ISE results indicated normonatremia. In a further 20 (7 %) cases, direct ISE indicated normonatremia but laboratory-based indirect ISE indicated hypernatremia. The difference between direct and indirect ISE results was found to correlate with serum albumin concentration. 

The lower the plasma albumin concentration, the greater was the difference in plasma sodium by the two methods. In an extensive discussion of this observed relationship the authors provide an argument for the use of direct rather than indirect ISE for measurement of plasma sodium in patients, like the critically ill, who frequently have a reduced serum albumin concentration.

 
As far as plasma chloride is concerned, the overall mean of 300 results by indirect ISE was just 1 mmol/L less than the mean of results by direct ISE, but the limits of agreement ranged from –6.6 to +4.5 mmol/L. There was no correlation between serum albumin concentration and the difference between chloride results by the two methods.