Potassium Disorders Viva

Long Case Viva Defence

Candidate's opening statement (model answer)

"Mr Patel is a 72-year-old man with type 2 diabetes for 20 years, ischaemic heart disease with a prior myocardial infarction, heart failure with reduced ejection fraction of 30 percent, and stage 3b chronic kidney disease with a baseline eGFR of 32, who presents with two days of progressive weakness and one day of confusion, on a background of viral gastroenteritis five days ago with vomiting and poor oral intake. He is on a full cardiorenal regimen — bisoprolol, sacubitril/valsartan, spironolactone, frusemide, empagliflozin, atorvastatin, aspirin, and metformin. [1]

On examination he is drowsy but rousable, his blood pressure is 144 over 84, heart rate 56, the JVP is just visible, his chest is clear, and he has no oedema. His ECG shows wide QRS complexes at 150 milliseconds with loss of P waves and a sinusoidal morphology. A venous blood gas returns potassium 7.6, pH 7.28, bicarbonate 16, glucose 9.4, lactate 2.1, sodium 136, haemoglobin 102. [1]

His main problems are:

1. Severe hyperkalaemia — potassium 7.6 — with a pre-terminal ECG in sine-wave morphology, an imminent cardiac arrest risk
2. Metabolic acidosis — pH 7.28, bicarbonate 16 — contributing to the transcellular potassium shift and a feature of his CKD
3. Bradycardia from hyperkalaemic conduction toxicity
4. Stage 3b chronic kidney disease on a cardiorenal drug regimen — sacubitril/valsartan, spironolactone, and empagliflozin all raising potassium, with metformin now held for the acute kidney injury risk
5. Recent gastroenteritis with volume depletion — the precipitant, reducing distal sodium delivery and potassium excretion
6. Heart failure with reduced ejection fraction and ischaemic heart disease — the indication for his mortality-reducing drugs and the context for his cardiovascular risk
7. Encephalopathy — likely multifactorial from the acidosis, the electrolyte disturbance, and the underlying illness [1]

His dominant clinical threat is the arrhythmia from the hyperkalaemia — he can deteriorate to asystole in minutes. My immediate priorities are to stabilise his myocardial membrane with calcium, shift potassium into cells with insulin-dextrose, salbutamol, and bicarbonate given he is acidotic, remove potassium with frusemide and a binder, address the precipitant and review the medications, and call the nephrology team early for the possibility of dialysis. I will manage him in the high dependency unit with continuous cardiac and glucose monitoring, watching especially for rebound hyperkalaemia and for the hypoglycaemia that can complicate insulin-dextrose therapy." [1]

Examiner probing questions and model answers

Q1: "The ECG shows a sine wave. Walk me through exactly what you do in the first 15 minutes." [1]

"First I confirm the patient has a pulse and a perfusing rhythm — if he is in cardiac arrest, I follow the hyperkalaemic arrest protocol with calcium chloride 10 mL of 10% IV, then the standard approach. Assuming he has a pulse, I put him on continuous cardiac monitoring, secure two large-bore cannulae, give high-flow oxygen, and send a repeat venous blood gas immediately to confirm the potassium and exclude pseudohyperkalaemia — though the ECG changes make a true hyperkalaemia near-certain. I send blood for a group and hold, full blood count, urea and electrolytes including magnesium, liver function, troponin, and a venous lactate. Then I give calcium gluconate 10 mL of 10 percent — that is 2.25 millimoles of calcium — intravenously over 2 to 5 minutes, with continuous ECG monitoring, watching for the QRS to narrow. If the ECG has not improved after 5 minutes I repeat the dose once. Calcium does not lower the potassium but it stabilises the membrane and prevents arrhythmia while the other therapies work. Immediately after the calcium I give insulin 10 units of rapid-acting insulin with 25 grams of intravenous dextrose, salbutamol 10 to 20 milligrams nebulised, and — because he is acidotic with a bicarbonate of 16 — sodium bicarbonate 50 millimoles of 8.4 percent intravenously over 15 to 30 minutes. I check a blood glucose at baseline, at 30 minutes, at 1 hour, and then hourly for 4 to 6 hours, because hypoglycaemia is the commonest serious complication of insulin-dextrose and can be delayed. I give frusemide 80 milligrams intravenously to promote potassium excretion, start a potassium binder — sodium zirconium cyclosilicate 10 grams three times daily for the first 48 hours — and call the nephrology team to alert them to the possibility of dialysis. I reassess the potassium and the ECG at 1 to 2 hours." [1]

Q2: "He is on sacubitril/valsartan and spironolactone. What do you do with these drugs, and what is the long-term plan?" [1]

"In the acute setting I hold both the sacubitril/valsartan and the spironolactone immediately, along with the metformin — which I hold because of the acute kidney injury risk on top of CKD — and I review the empagliflozin, which I would also hold during the acute illness given the volume depletion and the euglycaemic ketoacidosis risk. The frusemide and bisoprolol I continue. The critical point, though, is that these drugs are the cornerstone of guideline-directed medical therapy for heart failure with reduced ejection fraction, and each of them — the ARNI, the MRA, and the SGLT2 inhibitor — has a proven mortality and hospitalisation benefit. Permanently ceasing them would worsen his outcome. So my long-term plan is to reintroduce them once the potassium is controlled and the acute illness has resolved — typically within days to a week — starting at a lower dose and titrating up. To enable this, I would maintain him on an oral potassium binder such as patiromer or sodium zirconium cyclosilicate, which have been shown in the OPAL-HK and HARMONIZE trials to control potassium and allow continuation of RAAS inhibitors in patients with CKD and heart failure. I would monitor his potassium and renal function one to two weeks after each dose change. I would also give him a sick-day plan — to hold the RAAS inhibitor, the SGLT2 inhibitor, and the diuretics during any future acute illness with volume depletion, which is exactly the scenario that precipitated this admission." [1]

Q3: "When would you dialyse him, and what are the cautions?" [1]

"The indications for urgent haemodialysis in hyperkalaemia are: refractory hyperkalaemia despite medical therapy, severe hyperkalaemia with ECG changes in a patient who is oligoanuric or has end-stage CKD, and ongoing potassium release as in rhabdomyolysis or tumour lysis. This patient presented with a pre-terminal ECG but has stage 3b CKD with expected preserved urine output, and I would expect him to respond to the medical protocol — calcium, insulin-dextrose, salbutamol, bicarbonate, frusemide, and a binder. I would reassess the potassium and the ECG at 1 to 2 hours. If the potassium fails to fall below 6.5, or the ECG fails to improve, or he becomes oliguric or haemodynamically unstable, I would proceed to haemodialysis via temporary vascular access. If he is haemodynamically unstable I would prefer continuous renal replacement therapy — CVVHDF — which removes potassium more slowly but sustainably. The cautions are three. First, rebound — the insulin-dextrose wears off at 4 to 6 hours, and if the underlying cause is ongoing, the potassium will rise again; ongoing removal with a binder and a diuretic, or repeat dialysis, may be needed. Second, the rate of potassium fall — a rapid fall can cause arrhythmia from the reverse shift, so I monitor the ECG. Third, the dialysis itself — vascular access, anticoagulation, and haemodynamic shifts all carry risk in a patient with heart failure and ischaemic heart disease. I would involve the nephrology team early, and the intensive care team if he is unstable." [1]

Q4: "Why does acidosis raise potassium, and why does bicarbonate have a role here but not in non-acidotic hyperkalaemia?" [1]

"Acidosis raises potassium by a transcellular shift. In metabolic acidosis, hydrogen ions move into the cell to be buffered, and to maintain electroneutrality potassium moves out into the extracellular fluid. The relationship is approximately a 0.6 millimole per litre rise in potassium for every 0.1 unit fall in pH, though it is loose and is greater with mineral — that is, hyperchloraemic — acidosis than with organic acidosis such as lactic or ketoacidosis, because organic anions like lactate and ketones move into the cell with the hydrogen, mitigating the potassium shift. Bicarbonate works by correcting the acidosis, which reverses the transcellular shift and moves potassium back into cells. In the acidotic patient, bicarbonate is an effective adjunct to insulin-dextrose and salbutamol. In the non-acidotic patient, bicarbonate is ineffective as monotherapy — it does not shift potassium by a direct mechanism independent of pH correction, and its modern role is confined to the acidotic or acidaemic patient. It also carries a sodium and volume load, which is problematic in CKD, heart failure, and hypertension. So I would not use bicarbonate as my primary potassium-lowering agent in a non-acidotic hyperkalaemia — insulin-dextrose and salbutamol are the first-line shift agents there." [1]

Q5: "Let's switch to hypokalaemia. A 28-year-old woman is admitted with potassium 2.3. She has a normal magnesium, metabolic alkalosis, hypertension, and a high aldosterone with a suppressed renin. What is the diagnosis, and how do you confirm it?" [1]

"The pattern is hypertension with hypokalaemia, metabolic alkalosis, a high aldosterone, and a suppressed renin — which is primary aldosteronism, Conn's syndrome. This is the most common cause of secondary hypertension, present in 5 to 10 percent of all hypertensives and up to 20 percent of those with resistant hypertension. The next step is confirmation with a suppression test — an oral salt loading test (urinary aldosterone after three days of high salt), or a saline infusion test (aldosterone after 2 litres of normal saline over 4 hours), or a fludrocortisone suppression test. Failure to suppress aldosterone confirms autonomous production. I would then localise with an adrenal CT, and — because imaging is imperfect at distinguishing adenoma from bilateral hyperplasia, and because the management differs — I would do adrenal venous sampling to confirm whether the aldosterone is unilateral (an adenoma, curable by adrenalectomy) or bilateral (hyperplasia, managed medically with spironolactone or eplerenone). Before any of this I would ensure the screening was done correctly — the aldosterone-to-renin ratio is affected by posture, time of day, potassium, and many drugs including diuretics, beta-blockers, ACE inhibitors, and spironolactone, and these need to be controlled or interpreted carefully. Importantly, I would correct the hypokalaemia first, because hypokalaemia suppresses aldosterone secretion and can cause a false-negative ARR." [1]

Q6: "What if she were normotensive with a high renin and high aldosterone?" [1]

"Then the picture changes. A normotensive patient with hypokalaemia, metabolic alkalosis, and high renin and aldosterone has the inherited tubulopathy pattern — Bartter or Gitelman syndrome. The discriminator is the urine calcium. Gitelman syndrome, which is a defect of the thiazide-sensitive sodium-chloride cotransporter in the distal convoluted tubule, causes hypocalciuria and hypomagnesaemia, and presents in adolescence or adulthood. Bartter syndrome, which is a defect of the Na-K-2Cl cotransporter in the thick ascending limb, causes hypercalciuria and presents in childhood with more severe disease, often with a history of polyhydramnios and growth failure. I would also do a urine diuretic screen to exclude surreptitious diuretic use, which can mimic either. The management is lifelong potassium and magnesium supplementation, often with a potassium-sparing diuretic such as amiloride, and a liberal salt intake. Gitelman has a better prognosis." [1]

Q7: "What is the role of the novel potassium binders, and how do they compare with sodium polystyrene sulfonate?" [1]

"The modern binders are patiromer and sodium zirconium cyclosilicate, or SZC. Patiromer is a non-absorbed polymer that binds potassium in the colon; in the OPAL-HK trial it produced a sustained 0.8 millimole per litre reduction and allowed continuation of RAAS inhibitors in CKD. The dose is 8.4 grams once daily, titrated. SZC is an inorganic crystal that exchanges sodium or hydrogen for potassium throughout the gastrointestinal tract; in the HARMONIZE trial it produced a rapid reduction within 4 hours and sustained control at 28 days. The dose is 10 grams three times daily for up to 48 hours for acute correction, then 5 grams once daily for maintenance. SZC has a faster onset and is approved for acute use in some regions; it carries a sodium load, so caution in heart failure. Patiromer can cause hypomagnesaemia and gastrointestinal effects. Both are well tolerated compared with sodium polystyrene sulfonate, or resonium, which is the older resin. Resonium is slower, less predictable, and has been associated with intestinal necrosis — especially when given as a retention enema in the post-operative or ileus patient — and the FDA issued a warning. Resonium has been largely superseded by patiromer and SZC in modern practice but is still used where the newer agents are unavailable or unaffordable. The key point is that binders are too slow for the acute severe hyperkalaemia — they are for subacute and chronic management, especially to enable continuation of RAAS inhibitors in CKD and heart failure." [1]

Q8: "What is pseudohyperkalaemia, and how do you avoid treating it inappropriately?" [1]

"Pseudohyperkalaemia is a falsely elevated serum potassium from the release of potassium from cells during or after venesection. The causes are a difficult draw with prolonged tourniquet time and fist clenching, haemolysed sample, delayed transport, severe leucocytosis above 100 or thrombocytosis above 700 from cell lysis in the tube, and ex vivo cell lysis in haematological malignancy. The first step before any treatment of a significant hyperkalaemia is to confirm the value with a point-of-care venous or arterial blood gas on a free-flowing sample run immediately — this avoids the cell-lysis and handling artefacts. A serum-to-plasma potassium difference greater than 0.4 millimoles per litre on paired samples confirms pseudohyperkalaemia. Treating a pseudohyperkalaemia with calcium and insulin-dextrose exposes the patient to real harm — hypoglycaemia, volume overload, line complications — for no benefit. The discipline is: any reported hyperkalaemia, especially if the ECG looks normal or the patient is well and the context suggests artefact, gets an immediate blood gas before treatment. If the patient has ECG changes consistent with hyperkalaemia and is unstable, I treat first and confirm after — but I am vigilant for the pseudohyperkalaemia pattern, particularly in haematology patients." [1]

Short Case Discussion

Scenario: "This patient has an ECG with peaked T waves and broad QRS. Discuss the ECG changes of hyperkalaemia and the immediate response."

Candidate presentation (model): [1]

"The ECG shows the classic progression of hyperkalaemic cardiotoxicity. The earliest change is the peaked T wave — tall, narrow-based, symmetrical, with a tented appearance, best seen in the precordial leads, and accompanied by a shortened QT interval. This reflects accelerated terminal repolarisation and typically appears at a potassium of 5.5 to 6.5. As the potassium rises further, into the 6.5 to 7.5 range, atrial and atrioventricular conduction slows — the PR interval prolongs, the P waves flatten and then disappear, and the QRS begins to widen. Above 7.5 to 8.5 the QRS widens further and begins to merge with the T wave, conduction blocks and escape rhythms appear, and ventricular ectopy becomes prominent. The pre-terminal pattern is the sine wave — the QRS and T wave merge into a continuous sinusoidal trace, which signals imminent ventricular fibrillation, asystole, or pulseless electrical activity. The progression is a useful teaching tool but individual patients do not always follow it — a near-normal ECG does not exclude dangerous hyperkalaemia, and I treat the patient and the trend, not the single ECG. [1]

My immediate response to ECG changes of hyperkalaemia is to give calcium gluconate 10 millilitres of 10 percent intravenously over 2 to 5 minutes — this stabilises the myocardial membrane by raising the threshold for excitation and reverses the conduction effects of hyperkalaemia within minutes. I repeat once at 5 minutes if the ECG changes persist. Calcium does not lower the potassium, so I follow immediately with shift therapy — insulin 10 units with 25 grams of intravenous dextrose, salbutamol 10 to 20 milligrams nebulised, and bicarbonate if the patient is acidotic — and removal therapy with frusemide, a binder, and dialysis as indicated. I place the patient on continuous cardiac monitoring and reassess the ECG and potassium frequently." [1]

Examiner: "Why can a patient with a severe hyperkalaemia have a normal-looking ECG?" [1]

"The ECG has imperfect sensitivity for hyperkalaemia. The relationship between the serum potassium and the ECG is variable — some patients maintain a relatively normal-looking ECG at potassium levels that would be expected to produce marked changes, and the first manifestation can be the arrhythmia or arrest itself. This is particularly described in patients with chronic hyperkalaemia, in dialysis patients who have adapted, and in patients whose other electrolytes — calcium, magnesium, sodium — modify the membrane effect. The practical implication is that I never dismiss a significant hyperkalaemia because the ECG looks normal; I treat the number and the patient, with the ECG as one input." [1]

Examiner: "Describe the ECG of hypokalaemia and the arrhythmia risk." [1]

"The hallmark is repolarisation abnormality. The earliest change is T wave flattening or inversion with minor ST depression, at a potassium of 3.0 to 3.5. As the potassium falls further, prominent U waves appear — small positive deflections after the T wave, best seen in V2 to V4 — and may merge with the T wave giving an apparent long QT which is actually a long QU. Below 2.5, the corrected QT is genuinely prolonged, premature atrial and ventricular complexes appear, and the patient is at risk of torsades de pointes — especially with concurrent QT-prolonging drugs such as macrolides, fluoroquinolones, antipsychotics, methadone, and ondansetron, and in the patient on digoxin, where hypokalaemia potentiates digoxin toxicity. The U wave taller than the preceding T wave is highly suggestive of significant hypokalaemia. My response is to replace potassium — oral if mild to moderate, intravenous if severe or symptomatic — and crucially to check and correct the magnesium first, because hypomagnesaemia causes refractory hypokalaemia through renal potassium wasting." [1]