HIGH-K trial sets up a direct test of glucose-only therapy for hyperkalaemia
Hyperkalaemia remains one of the most urgent electrolyte emergencies in emergency medicine, and insulin-dextrose therapy has long sat at the center of rapid potassium-lowering care. The treatment works, but its benefit comes with a familiar cost: iatrogenic hypoglycaemia, especially in patients without diabetes who do not have the same metabolic reserve to buffer a sharp insulin surge. The HIGH-K trial is designed to examine whether a simpler alternative — glucose-only therapy — can preserve potassium control while lowering the risk of treatment-induced low blood sugar.
The question is clinically sharp because the physiology is straightforward. Intravenous insulin drives potassium into cells, which helps lower serum potassium quickly, yet that same insulin can overshoot glucose homeostasis and leave patients vulnerable to dangerous neuroglycopenia. In non-diabetic adults, endogenous insulin secretion may respond to a glucose load, potentially providing enough intracellular potassium shift to help without the same degree of exogenous insulin exposure. If that mechanism holds in the emergency department, the tradeoff between potassium correction and glycaemic harm could look very different.
Emergency department hyperkalaemia care faces a safety tradeoff
The trial is a single-center, double-blind, randomized controlled study in adult non-diabetic patients presenting to an Australian emergency department with hyperkalaemia, defined here as a potassium concentration above 5.5 mmol/L. Ninety-five participants are randomized in a 1:1 ratio. One group receives glucose-only therapy consisting of a 100 mL bolus of 50% dextrose followed by a 250 mL infusion of 10% dextrose over two hours. The comparator group receives standard insulin-dextrose therapy, specifically 10 units of intravenous insulin with 25 g dextrose followed by a 250 mL saline infusion.
The study design reflects a practical bedside dilemma rather than an abstract pharmacology exercise. Emergency clinicians need an intervention that works quickly, can be delivered reliably, and does not create a second emergency while treating the first. Hypoglycaemia after insulin therapy is not a minor adverse effect in this setting; it can force extra glucose checks, rescue dextrose administration, prolonged observation, and sometimes escalation of care. A therapy that reduces potassium without exposing non-diabetic patients to that cascade would have immediate operational value.
Glucose-only treatment also tests a broader physiological idea: whether the body’s own insulin response, triggered by a concentrated dextrose load, can provide enough transcellular potassium movement to compete with exogenous insulin. The answer matters because the mechanism is plausible, but plausibility alone is not enough in emergency care. The margin between biochemical improvement and clinical harm is narrow, and the trial is built to measure that margin directly.
Hypoglycaemia is the primary safety endpoint
The primary safety outcome is the incidence of hypoglycaemia. That choice places the focus exactly where the controversy lives. Insulin-dextrose therapy is effective, yet the same intervention can push patients into a low-glucose state that is both preventable and clinically consequential. By centering hypoglycaemia, the trial does not treat safety as a secondary footnote; it makes glycaemic harm the main outcome against which any potassium-lowering strategy must be judged.
A one-sentence reality check belongs here. In non-diabetic patients, treatment for one electrolyte emergency can create another.
Beyond the safety endpoint, the study is also positioned to inform whether glucose-only therapy can stand as a credible emergency alternative rather than a compromise born of convenience. If the glucose arm achieves acceptable biochemical control with fewer low-glucose events, the case for changing front-line practice becomes stronger. If potassium lowering proves inadequate, or if glucose-only treatment still produces troublesome glycaemic instability, the emergency department will have a clearer boundary for where endogenous insulin release can and cannot be trusted.
Trial structure aims to isolate the treatment effect
Several features of the design strengthen the internal logic of the comparison. Randomization minimizes imbalance between groups at baseline, while double blinding reduces the chance that knowledge of treatment assignment influences monitoring or rescue decisions. In a condition where clinicians frequently intensify observation based on perceived risk, blinding matters because it helps separate the pharmacology from the behavior surrounding it.
The protocol also targets a clearly defined population: adults without diabetes. That restriction is important because the glucose-insulin relationship differs substantially in people with chronic dysglycaemia. Non-diabetic patients may be more vulnerable to exogenous insulin-induced hypoglycaemia, but they may also respond differently to a high-glucose load. Restricting the cohort makes the biological question cleaner, even if it narrows the immediate generalizability of the findings.
Single-center enrollment introduces another practical constraint. Emergency department practice patterns, monitoring intensity, and local protocols can shape both efficacy and safety outcomes. A single site can deliver consistency, but it also means the results will need cautious interpretation before they are applied broadly across hospitals with different patient mixes and response workflows.
Why this comparison matters at the bedside
The existing approach to hyperkalaemia is effective enough to remain standard, yet it is rarely elegant. Insulin-dextrose therapy buys time by shifting potassium intracellularly, but it does not remove potassium from the body, and it forces clinicians to watch closely for delayed hypoglycaemia after the initial correction phase. In busy emergency settings, that monitoring burden has a cost of its own. A glucose-only strategy, if effective, could simplify care for a subset of patients by using a physiologic pathway that depends less on exogenous insulin exposure.
The protocol’s two-hour dextrose infusion in both arms also shows how much of the comparison is about modulating glucose kinetics, not just selecting one drug over another. Continuous dextrose support may blunt the glycaemic fall that follows insulin administration, while the glucose-only arm aims to stimulate endogenous insulin release without administering insulin directly. Small differences in that balance may shape both the depth and duration of glucose changes, which is exactly why the trial’s safety focus is so relevant.
No efficacy or safety results are available yet, and the protocol does not answer whether glucose-only therapy will be enough to replace insulin-dextrose in routine care. What it does provide is a rigorous framework for testing a question emergency clinicians have been asking at the bedside for years: whether hyperkalaemia can be treated effectively without paying for it later in hypoglycaemia. Until those data arrive, insulin-dextrose remains the comparator that defines standard practice, and the HIGH-K trial will determine whether that standard can be made safer for non-diabetic adults.