Hyperkalemia in advanced CKD with a misleading ECG
A 74-year-old woman with diabetic kidney disease (baseline creatinine 2.6 mg/dL, eGFR 18 mL/min/1.73 m²) is admitted with 3 days of vomiting and reduced oral intake. She takes ramipril 10 mg daily, spironolactone 25 mg daily, and metformin 1 g twice daily. Vitals: BP 102/64, HR 52, afebrile. Exam shows dry mucous membranes and no peripheral edema. Labs: K 6.9 mmol/L, Na 131, bicarbonate 16 mmol/L, urea 98 mg/dL, creatinine 3.4 mg/dL, glucose 210 mg/dL, calcium 8.4 mg/dL. ECG shows peaked T waves in V2-V4 and a QRS of 108 ms. She is oliguric.
1. Prioritize your immediate management over the first 60 minutes. For each intervention, state the goal, expected onset, and the specific reason it is chosen in this patient.
Read the worked answer
1. Stabilize the myocardium - IV calcium gluconate 10% 10-30 mL (or calcium chloride if central access)
- Goal: prevent ventricular arrhythmia by stabilizing cardiac membranes.
- Onset: 1-3 minutes, duration 30-60 minutes.
- Rationale: peaked T waves plus QRS widening (108 ms) indicate cardiac membrane instability. This is the only intervention that treats the cardiac effect, not the potassium level. Repeat if ECG changes persist after 5-10 minutes.
2. Shift potassium intracellularly - IV regular insulin 10 units with 25 g dextrose
- Goal: lower serum K by ~0.5-1.0 mmol/L within 15-30 minutes.
- Rationale: fastest reliable potassium-lowering therapy. Monitor glucose hourly for at least 6 hours (risk of late hypoglycemia, especially with renal impairment).
- Salbutamol nebulized 10-20 mg can be added as adjunct; avoid if tachyarrhythmia.
- Sodium bicarbonate is reasonable given metabolic acidosis (bicarbonate 16), but is slower and should not replace insulin-dextrose.
3. Remove potassium - prepare for dialysis
- She is oliguric with eGFR 18 and refractory hyperkalemia; urgent nephrology consultation for hemodialysis is the definitive treatment.
- Potassium binders (e.g., sodium zirconium cyclosilicate or patiromer) can be started but are too slow for acute severe hyperkalemia.
- Loop diuretic (furosemide) may be considered if she has some residual urine output, but oliguria limits efficacy and worsens volume depletion.
4. Stop offending agents
- Hold ramipril and spironolactone immediately; hold metformin given AKI on CKD (risk of lactic acidosis).
5. Avoid calcium polystyrene sulfonate with sorbitol in this setting (bowel ischemia risk, slow onset).
2. The patient's ECG shows peaked T waves but the QRS is only mildly widened. A junior resident argues that because the ECG is 'not that bad', calcium can be deferred and insulin-dextrose alone is sufficient. Construct the counterargument using the physiology of hyperkalemic cardiotoxicity.
Read the worked answer
Counterargument: ECG severity underestimates cardiac risk, and calcium should not be deferred.
Points to make:
- ECG changes are a late and unreliable marker of hyperkalemic cardiotoxicity. The progression from peaked T waves to sine wave is not linear, and ventricular fibrillation or asystole can occur without warning even with modest ECG changes.
- Peaked T waves plus QRS widening already represent membrane instability; the QRS of 108 ms is abnormal and indicates slowed conduction.
- Insulin-dextrose does not stabilize the myocardium - it only shifts potassium, and the effect takes 15-30 minutes to peak. During that window the heart remains at risk.
- Calcium is fast, safe, and does not lower potassium - its sole purpose is membrane stabilization and it should be given whenever there are ECG changes or a high suspicion of cardiotoxicity.
- Caveats: use calcium gluconate rather than calcium chloride peripherally (less tissue injury); in patients on digoxin, give calcium slowly and cautiously because of the theoretical 'stone heart' concern, but do not withhold it in life-threatening hyperkalemia.
- Bottom line: treat the patient and the potassium, not just the ECG. Calcium is given concurrently with, not instead of, potassium-lowering therapy.
3. After acute stabilization, the team must decide on chronic therapy. Discuss how you would adjust her medication regimen and what monitoring and follow-up you would arrange, including the evidence for continuing versus stopping RAAS blockade in advanced CKD.
Read the worked answer
Medication review
- Stop spironolactone permanently - mineralocorticoid receptor antagonists are a major cause of hyperkalemia in advanced CKD and the risk-benefit is unfavorable here.
- Ramipril: The acute hyperkalemia was precipitated by volume depletion and AKI. After recovery, RAAS blockade can often be cautiously restarted at a lower dose because of its renal and cardiovascular benefits, but only with close potassium and creatinine monitoring (e.g., 1-2 weeks after initiation and after any dose change).
- Metformin: hold during AKI; restart only if eGFR remains above 30 and renal function stabilizes.
- Consider adding an SGLT2 inhibitor if eGFR permits (many guidelines allow initiation down to eGFR 20), as it slows CKD progression and may reduce hyperkalemia risk; however, this is a shared decision given her age and volume status.
- Dietary potassium counseling with a renal dietitian.
Monitoring
- Check potassium and creatinine at 1 week, then 2-4 weeks after any RAAS change.
- Educate on sick-day rules: hold ramipril, metformin, and diuretics during vomiting or poor intake.
Evidence base
- Observational and RCT data (e.g., subgroup analyses of RALES and EMPHASIS-HF, and CKD-specific studies) show that MRAs increase hyperkalemia risk substantially in eGFR <30 and require close monitoring; many clinicians avoid them in this range.
- For ACE inhibitors, discontinuation is associated with worse cardiovascular outcomes in observational studies, and guidelines generally recommend continuing with monitoring rather than routinely stopping, except in life-threatening hyperkalemia or dialysis-dependent patients.
- Nephrology referral is indicated given eGFR 18 and recurrent hyperkalemia; discuss renal replacement therapy planning.