Combined Electrolyte Panel Pathway

Evidence & Clinical Pearls

1. Why the Order of Correction Decides the Outcome

Two rules sit above everything else. An immediate cardiac or neurological threat is corrected before any metabolic sequencing: hyperkalaemia with ECG change, Torsades, symptomatic hypocalcaemia with tetany or laryngospasm, and hypotonic hyponatraemia with seizures. Then magnesium comes first among the rest, because both hypokalaemia and hypocalcaemia are refractory until it is replaced.

1. Magnesium. The gate on everything below it.
2. Potassium. Before calcium, because a potassium below 3.0 is itself arrhythmogenic.
3. Calcium, unless the phosphate is high, in which case the phosphate is lowered first.
4. Phosphate.
5. Sodium, last among the routine steps, because potassium replacement itself raises the serum sodium and that rise has to be counted before any further sodium is given.

Replacing potassium raises the serum sodium by roughly 1 mEq/L of sodium for each mEq/L equivalent of potassium given, because potassium entering cells displaces intracellular sodium outward. In a hyponatraemic patient this is a real route to overcorrection. Sixty mEq of potassium chloride given to a 60 kg patient moves the sodium by an amount that is not trivial against a 24 hour budget of 8 mEq/L. The Adrogué-Madias equation has a second form, with the infusate potassium in the numerator, for exactly this reason. This tool uses it.

2. The Magnesium Gate, and the Mechanism Behind It

Potassium leaves the distal nephron through the ROMK channel, and that channel is held partly shut by intracellular magnesium plugging its pore from the cytosolic side. Normal intracellular free magnesium, 0.5 to 1.0 mmol/L, sits on the steep part of that inhibition curve. Deplete the magnesium and the plug comes out: ROMK opens, and the kidney wastes potassium however much is given.

• Huang and Kuo estimate that more than half of clinically significant hypokalaemia carries a concurrent magnesium deficiency.
• Magnesium is also required for parathyroid hormone secretion and for its action at the receptor, which is why severe hypomagnesaemia produces a functional hypoparathyroidism with a low or inappropriately normal PTH.
• Any hypokalaemia or hypocalcaemia that will not correct is a magnesium level you have not sent.
• In Indian practice the recurring causes are diuretics, alcohol use, chronic diarrhoea, long term proton pump inhibitors, and amphotericin B or an aminoglycoside.

3. Correcting Sodium for Glucose, and Which Sodium to Act On

Three numbers are in circulation for the fall in measured sodium per 100 mg/dL of glucose, and the two current guidelines have not converged.

This tool prints both, as a range, and names the source of each. Hyperglycaemia produces a translocational hyponatraemia: glucose pulls water out of cells, the measured sodium falls, and the serum tonicity is normal or high. There is no cerebral oedema and no case for hypertonic saline. Hypotonic hyponatraemia is different, and it is confirmed by a measured serum osmolality below 275 mOsm/kg.

• Grade the severity and decide about hypertonic saline on the glucose-corrected sodium.
• Track the trajectory over 24 hours on the effective osmolality, which is twice the sodium plus the glucose divided by 18.
• The measured sodium will rise on its own as the glucose falls, and that rise is expected rather than a change in tonicity. The 2024 consensus is explicit that it is not a reason to give hypotonic fluid.

A previous version of this module used a piecewise rule, 1.6 up to a glucose of 400 and 2.4 above it, and attributed it to Hillier. That rule is not published anywhere, and it took Hillier's headline figure while discarding the slope he actually measured above his own threshold. It has been removed.

4. The Sample May Be Lying

A potassium that does not fit the ECG is the commonest laboratory artefact on a medical ward, and treating an artefact with calcium and insulin carries its own risk.

• Haemolysis at the needle, a prolonged tourniquet, fist clenching, a delayed or cooled sample and EDTA carryover all raise it. So do very high cell counts.
• Pseudohyperkalaemia was found in about 14 per cent of serum samples with a platelet count above 500, and about 25 per cent with a white cell count above 50. Potassium leaves platelets during clotting, and leaves leucocytes when the sample sits.
• Confirm with a whole blood potassium on a blood gas analyser, or with paired plasma and serum samples. The plasma is normal in thrombocytosis, because no clot forms.
• Two exceptions. In chronic lymphocytic leukaemia with a very high count the pattern reverses and the plasma is the falsely high one. In hyperleucocytosis with blasts the sample can read falsely low at room temperature, and the remedy is to ice it rather than warm it.

5. The Calcium-Phosphate Product Is No Longer a Target

KDOQI set a target of a calcium-phosphate product below 55 mg²/dL² in 2003. KDIGO recommended against it in 2009 and carried that recommendation unchanged into the 2017 update, in recommendation 3.1.5: individual calcium and phosphate values, read together, should guide practice rather than the mathematical construct of the product.

• KDIGO 2017 sets no numeric phosphate target at all, and says only to lower an elevated phosphate toward the normal range, because binder trials in non-dialysis CKD found that treating a normal or near-normal phosphate increased arterial calcification.
• The product survives in one narrower role. In tumour lysis and rhabdomyolysis it is invoked as an acute physicochemical precipitation risk in a patient whose phosphate is rising fast, which is why calcium is withheld for asymptomatic hypocalcaemia there. That is not a chronic treatment target. This tool computes the product only when the phosphate is high, labels it as a precipitation risk, and says plainly that it is not something to treat toward.
• The Indian commentary on KDIGO 2017 does suggest working ranges for Indian practice, 3.4 to 6.0 mg/dL in CKD 3 to 4 and 3.5 to 5.5 mg/dL in stage 5, on pragmatic and cost grounds. Those are an Indian adaptation, not a KDIGO target, and the tool says so where it uses them.

6. Adjusted Calcium, and When Not to Trust It

The Payne formula, adjusted calcium equals measured calcium plus 0.8 times (4.0 minus albumin in g/dL), was derived in 1973 from 200 patients at one laboratory with the assays of that era. It has since performed poorly against measured ionised calcium in the settings where it is used most.

• In one prospective ICU cohort it identified 15.2 per cent of truly hypocalcaemic patients against 62.8 per cent for the uncorrected total calcium, and albumin added nothing once total calcium, pH and sex were in the model.
• In patients with an albumin below 30 g/L, three quarters of those called normocalcaemic by the formula were in fact hypocalcaemic. Where the decision turns on the calcium, measure the ionised calcium.
• Ionised calcium moves with pH, by roughly 5 per cent, about 0.05 mmol/L, for every 0.1 unit: alkalosis lowers it, acidosis raises it, because calcium and hydrogen compete for the same histidine imidazole sites on albumin. That is why acute hyperventilation produces perioral tingling and carpopedal spasm at an unchanged total calcium.

7. Two Rules This Module Deliberately Does Not Teach

Two rules that are still widely taught are not used here.

• The transtubular potassium gradient has been withdrawn by the people who devised it. Kamel and Halperin showed in 2011 that its central assumption, that no appreciable osmoles are reabsorbed downstream of the cortical collecting duct, is invalid because a large quantity of urea is recycled in the inner medullary collecting duct. In 2021 they wrote plainly that they do not use the TTKG in assessing patients with potassium disorders. This tool uses the spot urine potassium to creatinine ratio instead.
• Potassium is not corrected for pH. The taught figure of 0.6 mEq/L per 0.1 unit rests on limited data, and the effect depends on which acid is present. Mineral acidosis, meaning hyperchloraemic acidosis, does raise serum potassium, because chloride cannot follow hydrogen into the cell and potassium must leave to preserve electroneutrality. Organic acidosis largely does not, because lactate and beta-hydroxybutyrate enter the cell with the hydrogen ion on monocarboxylate transporters. The hyperkalaemia of DKA comes from insulin deficiency and hypertonicity, not from the ketoacidosis. The practical corollary is the dangerous one: a potassium of 3.8 in DKA is not reassuring, it is alarming.

8. Recognisable Combined Patterns

9. Reference Ranges and Severity Bands

Conversions used by this tool: magnesium mg/dL divided by 2.43 gives mmol/L, and multiplied by 0.823 gives mEq/L. Calcium mg/dL multiplied by 0.25 gives mmol/L. Phosphate mg/dL divided by 3.1 gives mmol/L. Blood urea mg/dL divided by 6 gives mmol/L, and blood urea is 2.14 times BUN. Creatinine mg/dL multiplied by 88.4 gives µmol/L. Glucose mg/dL divided by 18 gives mmol/L. Urate mg/dL multiplied by 59.48 gives µmol/L.

The anion gap reference range fell from the taught 12 plus or minus 4 when laboratories moved to ion-selective electrodes, which read a higher chloride. Take the range from your own laboratory's report rather than from this table.

10. What This Tool Does Not Cover

The scope box on the assessment tab carries the boundaries: children, pregnancy and eclampsia, the cause as against the pattern, the limits of the acid-base screen, and the absence of drug prices. Two are worth stating here.

• There is no Indian national guideline on sodium disorders. API, ICMR and the Indian Society of Nephrology have published none, so the European 2014 and the American 2013 documents stand unopposed here. This tool says which of the two each number comes from, and where they disagree, as they do on the correction rate, it prints both rather than choosing for you.
• No drug prices. Costs vary between states and between hospitals, they date quickly, and a figure that cannot be traced to a National Pharmaceutical Pricing Authority notification does not belong on a page that carries a clinician's name. Where cost genuinely drives a choice, as it does between phosphate binders, the tool gives the clinical argument instead: the elemental calcium load, and the vascular calcification risk that follows it.