High-dose insulin

Why this therapy is interesting

Hyperinsulinaemic euglycaemic therapy asks a clinician to give a shocked patient an insulin dose an order of magnitude above anything they will have prescribed before — an hourly rate at or above the patient's entire normal daily requirement — and to be confident enough about it to keep going for hours. It is one of the few treatments in emergency medicine where the intervention itself looks, on its face, like the error.

The reasoning is metabolic rather than receptor-based, which is what makes it unusual. A myocardium under stress switches its preferred substrate from free fatty acids to carbohydrate. Calcium channel blocker poisoning both impairs insulin release from pancreatic β-cells and induces insulin resistance, so exactly when the heart needs to burn glucose it becomes least able to take it up. High-dose insulin corrects that — it does not antagonise the poison at all.

Mechanism

The metabolic switch

The proposal is that insulin "allow[s] the switch of the cell metabolism from fatty acids to carbohydrates that is required in stress conditions, especially in the myocardium and vascular smooth muscle, resulting in an improvement in cardiac contractility and restored peripheral resistances."¹ Three components are usually described:

Substrate switch

Provision of glucose to a myocardium that can no longer efficiently oxidise fatty acids, improving contractility

Direct inotropy

Insulin increases intracellular calcium transport in myocytes, independently of the L-type channel the poison is blocking

Vasodilatation

Insulin increases endothelial nitric oxide synthase activity via the PI3-kinase pathway, raising nitric oxide and improving microvascular perfusion²

Why it is slow

The haemodynamic benefit depends on a metabolic change rather than a receptor occupancy, so it arrives late — "frequently occurring within 30 to 45 minutes of starting HIET."¹ This single fact governs how the therapy should be used: it must be started before the patient is peri-arrest, because if you wait until it is obviously needed you will spend the next half hour without it.

Indications

Where it sits in calcium channel blocker poisoning

The 2017 expert consensus recommendations place intravenous calcium, high-dose insulin, and noradrenaline and/or adrenaline in the first-line tier, "prioritized based on desired effect", alongside suggestions of dobutamine or adrenaline in cardiogenic shock and atropine for symptomatic bradycardia or conduction disturbance. For patients refractory to first-line treatment with myocardial dysfunction, and for those in refractory shock or peri-arrest, the recommendation is incremental doses of high-dose insulin.³ Lipid emulsion and pacing sit further down, in the refractory tier.

Beta-blocker poisoning

The evidence here is thinner and largely extrapolated, but the same metabolic argument applies and high-dose insulin is widely used, generally after or alongside glucagon.⁴ Toxicity from beta-blockers with membrane-stabilising activity (propranolol) or potassium channel blockade (sotalol) has additional problems that insulin does not address — see sodium bicarbonate and magnesium sulphate respectively.

The regimen

Insulin

• Bolus: 1 unit/kg of soluble human insulin (Actrapid, Humulin S) IV
• Infusion: 0.5–1 unit/kg/hour, titrated upward against haemodynamic response
• Escalation to considerably higher rates is described in refractory cases; the ceiling varies substantially between protocols, and is a decision to take with NPIS rather than from a monograph
• Continue until haemodynamics improve and the poison is clearing — this is typically hours, not minutes

Glucose

• Give glucose with the bolus unless the patient is already significantly hyperglycaemic. A bolus of 25–50 g (50–100 mL of 50%, or 250–500 mL of 10%) is conventional — note the cited source describes the continuous infusion rather than a bolus, so treat the bolus figure as local convention
• Then a continuous glucose infusion, conventionally 15–30 g/hour in an adult, using 10% or a more concentrated solution¹
• Titrate to euglycaemia — the source literature targets roughly 6–8 mmol/L
• Concentrated glucose is a vesicant. Anything above 10% wants central access, which is another reason to start planning the line early

Potassium

• Insulin drives potassium intracellularly; hypokalaemia is expected
• Most patients do not become severely hypokalaemic, and supplementation is generally aimed at preventing severe deficiency rather than normalising the number¹
• A moderately low serum potassium here reflects a shift, not a deficit — over-replacing it produces rebound hyperkalaemia when the insulin stops

Monitoring

• Blood glucose — the cited source specifies at least every 30 minutes; many protocols check every 15 minutes initially, then hourly once stable
• Potassium — commonly 1–2 hourly; the interval is protocol-dependent rather than derived from trial data
• Continue glucose monitoring for hours after the insulin stops. The hypoglycaemia risk outlasts the infusion, and this is where the harm from HIET actually occurs
• Haemodynamic monitoring adequate to detect the fall in systemic vascular resistance — this therapy warrants critical care involvement from the outset

Adverse effects and hazards

• Hypoglycaemia — the principal risk, and it can be delayed and prolonged well beyond the infusion
• Hypokalaemia — from intracellular shift; usually modest
• Hypomagnesaemia and hypophosphataemia — from the same shift
• Volume overload — the glucose infusion is a substantial fluid load in a failing heart, particularly if dilute solutions are used
• Vasodilatation and worsening hypotension if given without vasopressor support
• Rebound hyperkalaemia if potassium has been aggressively replaced during the infusion

Practical use in the ED

1. Recognise early and commit early. In a significant calcium channel blocker ingestion — particularly verapamil or diltiazem, and particularly modified-release — decide about high-dose insulin while the patient still looks reasonable. The therapeutic lag makes late starting a poor plan.
2. Set up the vasopressor at the same time. Noradrenaline alongside insulin, not after it fails.
3. Get the glucose infusion running before or with the insulin bolus, and get the central line planned.
4. Call NPIS. Ceiling doses, duration and escalation are exactly the decisions the service exists to support, and the case will be long.
5. Escalate to critical care. This is not a therapy to run in a busy ED resus bay for six hours, and the monitoring burden alone justifies the referral.

Critical appraisal

1. There are no randomised controlled trials in humans. The evidence is animal models, individual case reports and case series, and expert consensus. Published series report high success rates, but a case series of a poisoning treatment is close to a definition of publication bias — a patient who received high-dose insulin and died is markedly less likely to be written up than one who recovered, so any pooled figure overstates the effect by an unknown amount.
2. Consensus is not evidence, but it is the honest ceiling here. The 2017 expert recommendations are the best available synthesis for calcium channel blocker poisoning,³ and they are explicitly a Delphi consensus rather than a systematic review of trials. Treating them as though they carried trial-level authority is a common misreading.
3. The vasopressor question is a live disagreement, not settled doctrine. Rietjens and colleagues published a specific argument in 2023 that high-dose insulin should not be used without vasopressors,² which implies that the practice of using it alone was, and possibly still is, common enough to need arguing against.
4. Dosing has drifted upward over time without a trial to justify the drift. Early descriptions used 0.5 unit/kg/h; ceilings of several units per kilogram per hour now appear in protocols. That escalation is driven by case reports of success at higher doses — a mechanism that reliably ratchets doses up and never brings them down.
5. The harm is real and it is late. Prolonged hypoglycaemia after the infusion stops is the documented injury from this therapy, and it happens after the patient has left the resuscitation room and the attention has moved on.

References

1. 1
   Lheureux PER, Zahir S, Gris M, Derrey A-S, Penaloza A. Bench-to-bedside review: hyperinsulinaemia/euglycaemia therapy in the management of overdose of calcium-channel blockers. Crit Care 2006;10(3):212. PMC Source of the mechanism, dosing, glucose and potassium statements above.
2. 2
   Rietjens SJ, van Riemsdijk TE, Sikma MA, de Lange DW. High-dose insulin should NOT be used without vasopressors in calcium channel blocker toxicity. Br J Clin Pharmacol 2023;89(4):1275–8. Wiley
3. 3
   St-Onge M, Anseeuw K, Cantrell FL, et al. Experts consensus recommendations for the management of calcium channel blocker poisoning in adults. Crit Care Med 2017;45(3):e306–15. PubMed
4. 4
   Engebretsen KM, Kaczmarek KM, Morgan J, Holger JS. High-dose insulin therapy in beta-blocker and calcium channel-blocker poisoning. Clin Toxicol 2011;49(4):277–83. PubMed
5. 5
   Krenz JR, Kaakeh Y. An overview of hyperinsulinemic-euglycemic therapy in calcium channel blocker and β-blocker overdose. Pharmacotherapy 2018;38(11):1130–42. PubMed
6. 6
   Actrapid 100 international units/ml solution for injection — Summary of Product Characteristics, Novo Nordisk. electronic Medicines Compendium. Licensed indication is diabetes mellitus only; section 4.2 covers intravenous administration by healthcare professionals. Verified 23 Aug 2026.
7. 7
   TOXBASE — calcium channel blockers; beta-blockers; high-dose insulin. National Poisons Information Service. toxbase.org (NHS login required. NPIS: 0344 892 0111.)