Sodium, potassium and chloride are some of the most requested tests in any lab. They’re also some of the most consequential. A potassium result can send a patient to resus, and a sodium result can change how fast a clinician corrects fluids. Nearly all of them are measured with ion-selective electrodes (ISEs). Here’s how they work, and how they can trip you up.
What an ISE actually does
An ISE has a membrane that responds selectively to one ion. When the sample touches the membrane, a tiny voltage develops across it. That voltage is compared with a stable reference electrode. Roche’s method sheet explains that the voltage follows the Nernst equation: it changes with the logarithm of the ion’s activity. The electrode’s “slope” (in millivolts per tenfold change in concentration) tells you how responsive it is.
Different ions need different membranes. In one widely used system, for example, the sodium and potassium electrodes use neutral-carrier membranes and the chloride electrode uses an ion exchanger. The analyser calibrates against standards of known concentration to turn voltage into mmol/L.
Na, K and Cl in a sentence each
- Sodium (Na⁺): the main extracellular cation, and central to water balance. The Australasian harmonised adult reference interval is 135–145 mmol/L.
- Potassium (K⁺): mostly inside cells, so small changes in the plasma level matter a lot for heart and muscle function. The adult reference interval is 3.5–5.2 mmol/L.
- Chloride (Cl⁻): the main extracellular anion, and used in acid-base assessment, such as the anion gap.
Direct vs indirect ISE
This is the bit that confuses everyone, so let’s slow down.
Indirect ISE dilutes the sample before measurement. Most big lab chemistry analysers use it. The system mentioned above, for example, mixes 15 µL of serum or plasma with 450 µL of diluent. Dimeski and colleagues describe typical dilution ratios of 1:20 to 1:34. Diluting the sample lets the analyser use small sample volumes and widens the measuring range.
Direct ISE measures the sample undiluted. Blood gas analysers and many point-of-care devices use it.
Why does this matter? An ISE responds to ions dissolved in plasma water. But plasma isn’t all water. Normally about 7% of its volume is protein and lipid, the “non-aqueous” or “solid” phase. Indirect ISE results are standardised on the assumption that plasma contains 0.93 kg/L of water. When that assumption is wrong, the result is wrong.
The electrolyte exclusion effect
When protein or lipid takes up more of the plasma volume than usual, there’s less water per litre of plasma. The indirect method dilutes the whole sample, then calculates back as if the water content were normal, so it underestimates sodium. This is pseudohyponatraemia. It’s classically seen with very high triglycerides or paraproteins. Direct ISE avoids the problem because there’s no dilution step.
It works in reverse too. When protein is low, there’s more water than assumed, and indirect ISE overestimates sodium. This is pseudohypernatraemia, or a falsely “normal” result in someone who is actually hyponatraemic.
That second scenario is more common than people think. Using data from a Brisbane tertiary hospital, the same research group predicted important disagreement (4 mmol/L or more) between indirect and direct sodium for 25% of ICU samples and 8% of samples hospital-wide. Nearly all the ICU discrepancies (97%) were in hypoproteinaemic samples. The authors recommended hospital labs consider standardising on direct ISE for sodium.
The manufacturer’s method sheet acknowledges both directions. Elevated protein gives pseudohyponatraemia and decreased protein gives pseudohypernatraemia. It also recommends clearing grossly lipaemic specimens by ultracentrifugation.
Other error sources
Electrode ageing. Electrodes wear out. On the system described above, the sodium, potassium and chloride electrodes are due for replacement after 2 months or 9000 tests, and the reference electrode after 6 months. There’s also an acceptable slope range (50–68 mV/decade for sodium and potassium). A falling slope is a classic sign of a tired electrode.
Protein build-up. Serum and plasma are protein-rich, and protein can deposit on membranes and in the flow path over time. That can slow the response and make results drift. This is why manufacturers specify daily cleaning, and deproteinising solutions for the ISE flow path. The same system, for example, recalibrates every 24 hours and after cleaning or maintenance.
Sample problems. Fibrin microclots and separator gel can block the ISE system. Haemolysis falsely raises potassium. Red cells contain about 25 times more potassium than plasma, so the method sheet advises against using haemolysed samples for potassium. Serum potassium also reads higher than plasma potassium, because platelets release potassium during clotting.
Chemical interferents. Selective doesn’t mean perfect. Perchlorate treatment, for example, can cause falsely high chloride. The Dimeski, Badrick and St John review catalogues drug and matrix interferences for each electrode type.
Practical tips
- If a sodium doesn’t fit the clinical picture, look at the total protein, albumin and lipaemia index, and consider comparing it with a direct ISE result, such as a blood gas.
- Know which method each of your analysers uses. Blood gas and main-lab sodium may legitimately differ in patients with abnormal protein levels.
- Treat electrode slope, calibration and cleaning logs as early-warning systems, not just paperwork.
Key takeaways
- ISEs measure ion activity in plasma water, via a voltage that follows the Nernst equation.
- Indirect ISE dilutes the sample and assumes normal water content. Direct ISE doesn’t dilute.
- High protein or lipid gives falsely low sodium on indirect ISE. Low protein gives falsely high sodium.
- Electrodes age, collect protein and meet interferents, so maintenance and QC matter.