InfusionFox
Electrolytes & Fluids · Clinical background

Insulin + dextrose (K-shifting)

Potassium-shifting therapy for life-threatening hyperkalemia. Regular insulin 0.25–0.5 U/kg IV bolus with concurrent dextrose 1–2 g per unit insulin, followed by dextrose CRI 2.5–5% for 4–6 hours to prevent rebound hypoglycemia. Onset 15–30 min, duration 4–6 hr, typical K reduction 0.5–1.2 mEq/L. Insulin activates Na/K-ATPase, shifting K intracellularly. It does NOT remove K from the body. Definitive treatment of the underlying cause is required for the K problem to fundamentally resolve.

Open calculator →

Clinical background

Insulin/dextrose is the workhorse of K-lowering therapy in life-threatening hyperkalemia. It works by shifting potassium from the extracellular space into cells, driving down serum K within 15–30 minutes and holding it down for 4–6 hours. The approach is borrowed directly from human ICU practice and is well-supported in both species. The key practical points are recognizing that this is a shifting therapy (total-body K is unchanged), preparing for rebound hypoglycemia (the dextrose bolus and follow-up CRI exist for this reason), and not relying on insulin alone, definitive treatment of the underlying cause is what fundamentally resolves the K problem.

Mechanism

Insulin activates Na/K-ATPase in skeletal muscle, the largest reservoir of K-shifting capacity in the body. The pump exchanges three intracellular Na⁺ for two extracellular K⁺, dropping serum K. Critically, this is independent of insulin’s hypoglycemic effect, the K-shifting is a direct receptor-mediated activation of the pump. But insulin also drives glucose into cells, and the resulting hypoglycemia (in a non-diabetic patient with normal pancreatic function) can be severe and prolonged. The concurrent dextrose bolus prevents this.

In humans, 10 U of regular insulin shifts approximately 30 mEq of K intracellularly. The veterinary literature is less precise on the magnitude, but typical K reduction in clinical practice is 0.5–1.2 mEq/L, enough to take a patient from a critical 8.5 down to a less-critical 7.3, but not enough to “fix” a severe hyperkalemia on its own. The drug is bridging therapy; the underlying cause must be addressed.

The effect onset is 15–30 minutes (slower than calcium gluconate’s 1–3 minutes, calcium is the immediate intervention if there are ECG changes). Duration is 4–6 hours, after which K can rebound if the underlying cause hasn’t resolved.

Indications

Same as calcium gluconate: life-threatening hyperkalemia in any species. The two prototypes are:

Other settings where insulin/dextrose contributes: oliguric AKI, tumor lysis syndrome, severe rhabdomyolysis. The protocol works in all of these because the mechanism (driving K intracellularly) is independent of cause.

The threshold for using insulin/dextrose, like calcium, is K plus ECG findings rather than K alone. Most clinicians give insulin/dextrose any time K > 7 mEq/L or with concerning ECG changes at any K. For mild hyperkalemia (5.5–6.5 mEq/L) without ECG changes, treating the underlying cause and supporting with fluid therapy is usually sufficient.

Dosing

Insulin: 0.25–0.5 U/kg regular crystalline insulin IV bolus.

Lower end (0.25) is the safer default for blocked cats and other small patients, the K reduction is similar at the lower dose, but the hypoglycemia risk is meaningfully reduced. Upper end (0.5) for severe persistent hyperkalemia or large patients where the lower dose’s effect has been inadequate. Higher than 0.5 U/kg does not reliably lower K further but increases hypoglycemia risk. The InfusionFox calculator clamps at 0.5 U/kg.

Dextrose: 1–2 g per unit of insulin given concurrently IV.

Default 2 g per unit (the higher, safer ratio for hypoglycemia prevention). Lower ratios may be acceptable in diabetic patients with established hyperglycemia (because they have higher baseline glucose to start with and are less likely to drop dangerously), but do not reduce the dextrose ratio for non-diabetic patients in hyperkalemic crisis, the consequences of unanticipated hypoglycemia are severe.

For a typical 5 kg blocked cat: 0.25 U/kg × 5 = 1.25 U regular insulin + 2.5 g dextrose (= 5 mL of D50, diluted before bolus).

For a 18 kg Addisonian Standard Poodle: 0.25 U/kg × 18 = 4.5 U regular insulin + 9 g dextrose (= 18 mL of D50, diluted).

Use regular crystalline insulin only

This protocol is built around regular (short-acting) crystalline insulin. Humulin R or Novolin R. Do not substitute other insulin types:

Use a U-100 insulin syringe for accuracy. Doses are small, a 5 kg cat gets 1.25 U at the standard 0.25 U/kg dose, which is 0.0125 mL of U-100. That volume is below the accuracy threshold of standard 1 mL or 3 mL syringes. The U-100 syringe is calibrated in units directly, so you draw to the unit mark and the math doesn’t matter.

Diluting the dextrose

D50 (50% dextrose, 0.5 g/mL) has an osmolarity of approximately 2500 mOsm/L, about 8× plasma osmolarity. Undiluted IV bolus causes phlebitis and probable endothelial injury, and is associated with thrombophlebitis at peripheral catheter sites.

Standard practice: dilute the D50 bolus before giving.

Two acceptable approaches:

  1. D25 dilution (quick option): mix the calculated D50 volume with an equal volume of 0.9% NaCl. The resulting D25 (~1250 mOsm/L) is still hyperosmolar but is acceptable for slow IV bolus over 1–2 minutes through a peripheral line.

  2. D5–D10 dilution (safer): dilute the D50 into 50–100 mL of saline (final concentration D5–D10, ≈250–500 mOsm/L) and give over 5–10 minutes. This is the safer approach for non-emergent dosing or when the patient has a small peripheral catheter.

In a critical patient with an obvious central or large-bore peripheral catheter, the D25 dilution is acceptable to save time. For the 5 kg cat with a 22-gauge cephalic catheter, the D5–D10 dilution is probably better practice. Both work clinically.

The post-bolus dextrose CRI

After the bolus, the patient enters a 4–6 hour window during which insulin’s K-shifting effect persists AND the patient is at risk of rebound hypoglycemia as the bolus dextrose is metabolized but insulin’s hypoglycemic effect continues.

Standard follow-up: 2.5–5% dextrose CRI added to the maintenance fluids for 4–6 hours.

Practical implementation: take the patient’s existing fluid bag and either: - Add concentrated dextrose to bring it to 2.5–5%, OR - Switch to a pre-made D5W or D5-half-saline bag at the same maintenance rate

The exact concentration depends on the patient’s baseline BG trajectory. A patient who’s drifting toward 80–100 mg/dL needs 5%; one staying steady at 150–200 may only need 2.5%.

Monitor BG every 1 hour × 6 hours. This is not optional. Symptomatic hypoglycemia in cats can be subtle, weakness, lethargy, decreased mentation, seizures, and easy to attribute to the underlying disease. The rule: a recheck BG drives every dextrose CRI adjustment.

If BG drops below 80 mg/dL despite the CRI, give a 1 mL/kg D50 bolus (diluted) and increase the CRI dextrose concentration. If BG drops below 50 mg/dL or the patient has neurologic signs, treat as a medical emergency with repeat dextrose bolus and continuous monitoring until stable.

Recheck K and decide next step

Recheck serum K at 30 minutes and 1 hour after the bolus. Expected reduction is 0.5–1.2 mEq/L. If K has fallen as expected and the underlying cause is being addressed, continue fluid therapy and the dextrose CRI; recheck K every 2–4 hours.

If K hasn’t fallen at 1 hour or has rebounded:

Insulin shifts K; it doesn’t remove K

This is the most important conceptual point. After insulin/dextrose, the patient’s serum K is lower because the K is now inside cells, not because it has been excreted. The total-body K excess is unchanged. When insulin’s effect wanes at 4–6 hours, the K can shift back out of cells if the underlying cause has not been corrected.

The K problem fundamentally resolves only when:

These definitive treatments must run in parallel with the bridging drug therapy. The drug bundle alone is not adequate for management of the K problem; it just keeps the patient alive long enough for the definitive intervention to work.

Beta-2 agonists as an adjunct

Inhaled albuterol (and IV/SQ terbutaline) shift K intracellularly via β₂-mediated Na/K-ATPase activation, a mechanism complementary to insulin but pharmacologically distinct. In human ICU practice, nebulized albuterol is a routine adjunct that adds another 0.5–1 mEq/L of K reduction beyond insulin alone.

In veterinary practice this approach is less established. There is some published support but it’s not part of the standard veterinary hyperkalemia bundle. Reasonable to consider as an adjunct when insulin/dextrose alone hasn’t reduced K adequately and the underlying cause hasn’t yet been corrected, but not first-line, not part of the typical workflow, and not a substitute for any of the standard interventions.

Adverse effects

The two clinically important effects:

Less common:

Sources