Skip to main content
MedVellum
MCQsExamsAtlas
DashboardPricing
MBBS / Core medicine✳Dermatology✳ICU Fellowship (CICM)✳Anaesthesia✳Emergency Medicine✳Psychiatry Fellowship✳Paediatrics Fellowship✳Physician Medicine✳MCQs✳SAQs✳Vivas✳OSCE✳Evidence-first✳MBBS / Core medicine✳Dermatology✳ICU Fellowship (CICM)✳Anaesthesia✳Emergency Medicine✳Psychiatry Fellowship✳Paediatrics Fellowship✳Physician Medicine✳MCQs✳SAQs✳Vivas✳OSCE✳Evidence-first✳

MedVellum.

The folio

Exam-exhaustive medical education across every specialty — evidence-graded topics, engraved plates, and practice in every written and oral format. Educational content only — not medical advice.

llms.txt · psychiatry LLM catalog · sitemap

Atlas

  • Specialty atlas
  • MBBS / Core medicine
  • Dermatology
  • ICU Fellowship (CICM)
  • Anaesthesia
  • Emergency Medicine
  • Psychiatry Fellowship
  • Paediatrics Fellowship
  • Physician Medicine

Study & account

  • MCQ practice
  • Practice alias
  • Exam tools
  • Dashboard
  • Pricing
  • Sign in

© 2026 MedVellum. For education only — not a substitute for clinical judgement.

Folio edition · Set in Instrument Serif & Archivo

ICU TopicsPharmacology

ICU · Pharmacology

Insulin & Hypoglycaemics — Pharmacology

Also known as Insulin · Hypoglycaemics · Insulin types · Actrapid · Metformin · Sulfonylurea · GLP-1 · SGLT2 inhibitor · Variable-rate insulin infusion · Fixed-rate insulin infusion · NICE-SUGAR · Euglycaemic DKA · Metformin-associated lactic acidosis

ICU insulin and hypoglycaemics. INSULIN TYPES classified by onset/duration: RAPID analogues (lispro, aspart, glulisine — monomeric — onset 10-20 min, duration 3-5 h); SHORT/soluble (regular/Actrapid — forms hexamers — onset 30-60 min, duration 6-8 h — the IV preparation for ICU); INTERMEDIATE (isophane/NPH — protamine + zinc complex — cloudy — onset 1-2 h, duration 12-18 h); LONG (glargine — precipitates at neutral pH, flat peakless ~24 h; detemir — albumin-bound via myristic acid); ULTRA-LONG (degludec — multi-hexamers, 42+ h; icodec weekly). ICU INSULIN USE: DKA fixed-rate insulin infusion (FRII) 0.1 U/kg/h until ketones cleared; HHS lower dose 0.05 U/kg/h after aggressive fluid resuscitation; stress hyperglycaemia via variable-rate insulin infusion (VRII / sliding scale) targeting glucose 6-10 mmol/L per NICE-SUGAR (moderate control — tight control 4.4-6.1 increased mortality from hypoglycaemia, overturning the Leuven trial); insulin-dextrose for hyperkalaemia (intracellular K shift); GIK high-dose insulin for beta-blocker/calcium-channel-blocker toxicity. ORAL HYPOGLYCAEMICS: METFORMIN (biguanide — activates AMP kinase — suppresses hepatic gluconeogenesis — non-hypoglycaemic — UKPDS-34 mortality benefit in overweight T2DM — metformin-associated lactic acidosis (MALA) in renal failure/hypoxia/sepsis — removed by haemodialysis); SULFONYLUREAS (gliclazide, glibenclamide — close beta-cell K-ATP channel — stimulate insulin release — hypoglycaemia risk, especially glibenclamide/elderly/renal); DPP-4 inhibitors (sitagliptin — prolong incretins, glucose-dependent, weight-neutral, CV-safe per TECOS); SGLT2 inhibitors (dapagliflozin, empagliflozin — block Na-glucose cotransporter in proximal tubule — glycosuria — CV/renal benefit (EMPA-REG, CANVAS, CREDENCE) — euglycaemic DKA risk); GLP-1 agonists (exenatide, liraglutide, semaglutide — incretin mimetics — weight loss — CV benefit per LEADER/SUSTAIN-6). HYPOGLYCAEMIA MANAGEMENT: rule of 15 (15 g carbohydrate, recheck at 15 min); severe — 50% dextrose 25-50 mL IV (or 10% via central line) and glucagon 1 mg IM; octreotide 50-100 mcg SC for sulfonylurea-induced hypoglycaemia (inhibits insulin secretion).

high16 referencesUpdated 2 July 2026
On this page & tools

Your progress

Saved locally on this device.

Practise this topic

8 MCQs with explanations

Target exams

CICMFFICMEDIC

Red flags

NICE-SUGAR: TIGHT glucose control (4.4-6.1 mmol/L) INCREASED mortality vs moderate (6.1-8.3 mmol/L) — driven by severe hypoglycaemia (<2.2 mmol/L, RR ~1.5 for death). ICU target is 6-10 mmol/L; avoid aggressive titration.Sulfonylurea overdose (especially glibenclamide) causes PROLONGED refractory hypoglycaemia (24-48 h) — 50% dextrose triggers further insulin release and rebound — use OCTREOTIDE 50-100 mcg SC q6-8h to suppress endogenous insulin.SGLT2 inhibitors cause EUGLYCAEMIC DKA — glucose often <14 mmol/L but ketones high — precipitated by surgery, fasting, sepsis; check ketones/blood gas, not just glucose.Metformin-associated lactic acidosis (MALA): profound acidosis (pH <7.1, lactate >10), renal failure, hypoxia, sepsis — high mortality — treat with haemodialysis (clears metformin + lactate) and supportive care.DKA fixed-rate insulin infusion 0.1 U/kg/h requires aggressive POTASSIUM replacement (K+ falls 0.5-1 mmol/L per hour as acidosis corrects) — start K+ once serum K+ <5.5 and urine output established.

Your progress

Saved locally on this device.

Practise this topic

8 MCQs with explanations

Target exams

CICMFFICMEDIC

Red flags

NICE-SUGAR: TIGHT glucose control (4.4-6.1 mmol/L) INCREASED mortality vs moderate (6.1-8.3 mmol/L) — driven by severe hypoglycaemia (<2.2 mmol/L, RR ~1.5 for death). ICU target is 6-10 mmol/L; avoid aggressive titration.Sulfonylurea overdose (especially glibenclamide) causes PROLONGED refractory hypoglycaemia (24-48 h) — 50% dextrose triggers further insulin release and rebound — use OCTREOTIDE 50-100 mcg SC q6-8h to suppress endogenous insulin.SGLT2 inhibitors cause EUGLYCAEMIC DKA — glucose often <14 mmol/L but ketones high — precipitated by surgery, fasting, sepsis; check ketones/blood gas, not just glucose.Metformin-associated lactic acidosis (MALA): profound acidosis (pH <7.1, lactate >10), renal failure, hypoxia, sepsis — high mortality — treat with haemodialysis (clears metformin + lactate) and supportive care.DKA fixed-rate insulin infusion 0.1 U/kg/h requires aggressive POTASSIUM replacement (K+ falls 0.5-1 mmol/L per hour as acidosis corrects) — start K+ once serum K+ <5.5 and urine output established.
Cinematic ICU scene of a variable-rate insulin infusion pump beside a glucometer and a sliding-scale titration chart, clinical-blue lighting, medical educational, no faces, no text
FigureInsulin by the onset — the rapid monomeric analogues, the regular hexamer (the IV form), the basal. In the ICU the variable-rate intravenous infusion targets the glucose 6-10 mmol/L, titrated hourly; the harm is the hypoglycaemia, the under-treatment the hyperglycaemia that fuels the infection and the neuropathy.
[1]
Educational figure of insulin action and hypoglycaemia risk: IV regular insulin PK, NICE-SUGAR moderate glycaemic target, sulfonylurea prolonged hypoglycaemia
FigureInsulin drives glucose and potassium intracellularly — IV half-life is minutes, but sulfonylurea hypoglycaemia can last days and needs glucose infusion plus octreotide concepts.

Overview & definition

The one-paragraph exam answer

Insulin and hypoglycaemics are grouped by molecular type / mechanism for the oral agents and by onset and duration of action for insulin. INSULIN TYPES: rapid analogues (lispro, aspart, glulisine — engineered to be monomeric, onset 10-20 min, duration 3-5 h); short/soluble (regular — Actrapid — forms hexamers that dissociate slowly, onset 30-60 min, the IV preparation used throughout ICU); intermediate (isophane/NPH — protamine + zinc complex, cloudy, onset 1-2 h, 12-18 h); long (glargine — precipitates at neutral pH giving a flat, peakless ~24 h profile; detemir — albumin-bound via a myristic-acid side-chain); ultra-long (degludec — multi-hexamer formation, 42+ h). ICU INSULIN USE: DKA — fixed-rate insulin infusion (FRII) 0.1 U/kg/h continued until ketonaemia clears; HHS — lower dose 0.05 U/kg/h after aggressive fluid resuscitation (patients are profoundly dehydrated and insulin-sensitive); stress hyperglycaemia — variable-rate insulin infusion (VRII)/sliding scale targeting glucose 6-10 mmol/L per NICE-SUGAR (moderate control superior to the tight 4.4-6.1 mmol/L that NICE-SUGAR showed increased mortality through hypoglycaemia); insulin-dextrose for hyperkalaemia (10 U + 25-50 g dextrose shifts K+ intracellularly); GIK high-dose insulin for beta-blocker / calcium-channel-blocker toxicity (positive inotropy). ORAL HYPOGLYCAEMICS: metformin (biguanide — activates AMP kinase (AMPK) → ↓hepatic gluconeogenesis — non-hypoglycaemic — UKPDS-34 mortality benefit in overweight T2DM — metformin-associated lactic acidosis (MALA) in renal failure/hypoxia/sepsis, removed by haemodialysis); sulfonylureas (gliclazide, glibenclamide — close the β-cell K-ATP channel → insulin release — hypoglycaemia risk, especially glibenclamide/elderly/renal); DPP-4 inhibitors (sitagliptin — prolong incretins, glucose-dependent, weight-neutral, CV-safe per TECOS); SGLT2 inhibitors (dapagliflozin, empagliflozin — block Na-glucose cotransporter in the proximal tubule → glycosuria — CV/renal benefit per EMPA-REG, CANVAS, CREDENCE — euglycaemic DKA risk); GLP-1 agonists (exenatide, liraglutide, semaglutide — incretin mimetics — weight loss — CV benefit per LEADER/SUSTAIN-6). HYPOGLYCAEMIA MANAGEMENT: rule of 15 (15 g carbohydrate, recheck at 15 min); severe — 50% dextrose 25-50 mL IV (or 10% via central line) and glucagon 1 mg IM/SC (ineffective if glycogen-depleted — alcohol/starvation); octreotide 50-100 mcg SC q6-8h for sulfonylurea-induced hypoglycaemia (somatostatin analogue — inhibits endogenous insulin release).[1][1][3][10]

Insulin physiology and molecular pharmacology

Insulin synthesis, secretion and receptor signalling — the molecular basis of every insulin preparation

  1. SYNTHESIS in the pancreatic β-cell (proinsulin → insulin + C-peptide):

    • Translated on ribosomes of the rough ER as preproinsulin (110 aa), the signal peptide cleaved to proinsulin (86 aa).
    • In secretory granules, proinsulin is cleaved by prohormone convertases (PC1/3, PC2) at two basic-residue sites, releasing C-peptide (31 aa) and the mature insulin (51 aa = A-chain 21 aa + B-chain 30 aa, joined by two interchain disulphide bonds, with one intrachain disulphide on the A-chain).
    • Stored as a hexamer (six insulin molecules + two zinc ions) in the granule — the hexamer is the storage form; all subcutaneous preparations must dissociate to monomers before absorption, which is the rate-limiting step that distinguishes rapid from short-acting insulins.
    • Clinical correlate: C-peptide is co-secreted 1:1 with endogenous insulin and is measured to distinguish endogenous hyperinsulinaemia (e.g. insulinoma — high C-peptide) from exogenous insulin overdose (suppressed C-peptide). [1]
  2. SECRETION — glucose-stimulated insulin secretion (GSIS):

    • Glucose enters the β-cell via GLUT2 (low-affinity, high-capacity transporter that senses blood glucose).
    • Glucose is phosphorylated by glucokinase (the rate-limiting glucose sensor), metabolised → ↑ATP:ADP ratio.
    • ↑ATP closes the ATP-sensitive K+ channel (K-ATP) — composed of Kir6.2 pore + SUR1 regulatory subunit (the molecular target of sulfonylureas).
    • K-ATP closure → β-cell depolarisation → opens voltage-gated Ca2+ channels → Ca2+ influx → insulin granule exocytosis.
    • This K-ATP channel is exactly what sulfonylureas close directly (bypassing glucose) and what diazoxide opens (inhibiting insulin release — used for insulinoma and as an adjunct in sulfonylurea overdose). [1]
  3. INSULIN RECEPTOR — a receptor tyrosine kinase:

    • The insulin receptor is a heterotetramer (two extracellular α-subunits that bind insulin + two transmembrane β-subunits with intrinsic tyrosine-kinase domains).
    • Insulin binding → β-subunit autophosphorylation → phosphorylates insulin receptor substrate (IRS) proteins.
    • Two downstream cascades: (a) PI3K-Akt pathway → GLUT4 translocation to muscle/adipose membrane (glucose uptake), glycogen synthase activation (glycogen storage), mTOR (protein synthesis), anti-lipolysis — the metabolic arm; (b) Ras-MAPK pathway → gene transcription, cell growth, mitogenesis (the mechanism by which insulin is a growth factor).
    • Clinical correlate: insulin resistance (type 2 diabetes, sepsis, critical illness, steroids) is a post-receptor defect (impaired IRS-1/PI3K signalling); the MAPK arm is relatively preserved, contributing to atherosclerosis. [1]
  4. METABOLIC EFFECTS — net anabolic, storage-promoting:

    • Liver: ↓glycogenolysis, ↓gluconeogenesis, ↓ketogenesis, ↑glycogen synthesis.
    • Muscle: ↑glucose uptake (GLUT4), ↑glycogen synthesis, ↑protein anabolism (↓proteolysis).
    • Adipose: ↑lipogenesis, ↓lipolysis (↓free fatty acids, ↓ketogenesis — the reason insulin is the cornerstone of DKA treatment).
    • Electrolytes: drives K+, Mg2+, phosphate INTO cells (the basis of insulin-dextrose for hyperkalaemia, and the reason K+ falls during DKA treatment).
[1]

Insulin preparations — onset, peak, duration and formulation rationale

TypeOnset (SC)PeakDurationExamplesWhy it behaves this way
Rapid-acting analogue10-20 min1-2 h3-5 hLispro (B28 Pro/B29 Lys inverted), aspart (B28 Pro→Asp), glulisine (B3 Asn→Gly, B29 Lys→Glu)Single amino-acid substitutions prevent self-association → stays monomeric → absorbed immediately
Short-acting (soluble/regular)30-60 min2-4 h6-8 hRegular human insulin (Actrapid, Humulin R)Native insulin — self-associates into hexamers subcutaneously; must dissociate to dimers/monomers before absorption (rate-limiting). The only type given IV (no absorption step).
Intermediate1-2 h4-6 h12-18 hIsophane / NPH (Neutral Protamine Hagedorn)Complexed with protamine + zinc → cloudy suspension → slow dissolution/absorption. Resuspend before injecting (roll, don't shake).
Long-acting1-2 hPeakless (flat)20-24 hGlargine (A21 Asn→Gly + 2× Arg on B-chain → pH 4, precipitates at neutral SC pH); detemir (B30 Thr deleted + myristic-acid side-chain → binds albumin)Either precipitates (glargine) or binds albumin (detemir) → slow, flat, peakless absorption (low nocturnal hypoglycaemia)
Ultra-long1-2 hPeakless42+ hDegludec (multi-hexamer formation); icodec (once-weekly)Even slower release; degludec allows thrice-weekly dosing, icodec once weekly

Exam anchor points: rapid analogues ~15 min onset (lispro/aspart/glulisine); short/regular 30-60 min onset and the ONLY type that can be given IV; NPH isophane is the cloudy intermediate (must resuspend); glargine is the flat, peakless long-acting (precipitation mechanism).[1]

ICU insulin use

Glycaemic control targets in critical illness

The one-paragraph answer on ICU glucose targets — Leuven vs NICE-SUGAR

In 2001 the Leuven trial (van den Berghe) reported that tight glucose control (4.4-6.1 mmol/L, 80-110 mg/dL) with intensive IV insulin reduced mortality in a surgical ICU, sparking widespread adoption of 'tight glycaemic control'. The much larger multinational NICE-SUGAR trial (2009, n=6104) overturned this: patients randomised to the tight (4.5-6.0 mmol/L) target had HIGHER mortality (27.5%) than moderate control (target 6.1-8.3 mmol/L / ≤10 mmol/L; 24.9%), driven by severe hypoglycaemia (<2.2 mmol/L), which occurred in 6.8% of the tight arm vs 0.5% of the moderate arm and independently predicted death. Current ICU practice and the JBDS/ANZ guidance therefore target glucose 6-10 mmol/L, avoiding both hyperglycaemia (infection, oxidative stress, mortality) and the dangerous hypoglycaemia of aggressive control. The lesson: moderate, not tight, glucose control in the ICU.[1][2]

Approaches to glycaemic control in critical illness

StrategyGlucose targetEvidenceCurrent status
Tight glycaemic control (Leuven)4.4-6.1 mmol/L (80-110)van den Berghe 2001 — mortality benefit in surgical ICU[2]ABANDONED — superseded by NICE-SUGAR
Moderate control (NICE-SUGAR)6.1-8.3 mmol / 6-10 mmol/LNICE-SUGAR 2009 — lower mortality vs tight[1]STANDARD OF CARE in ICU
Conventional / permissive<10-12 mmol/LHistorical comparatorUsed if VRII resources limited
Sliding scale (historical)Reactive bolusesReactive, causes swingsLargely replaced by VRII

Variable-rate insulin infusion (VRII) and sliding scale

Setting up and titrating a variable-rate insulin infusion (VRII) for stress hyperglycaemia

  1. INDICATION: any critically ill patient with sustained glucose >10 mmol/L where the underlying illness (sepsis, steroids, pancreatitis, post-op, post-cardiac arrest) has caused insulin resistance and endogenous insulin cannot keep up. VRII provides smooth, titratable IV soluble insulin (Actrapid 50 U in 50 mL 0.9% saline = 1 U/mL). [1]

  2. CHOOSE THE RIGHT FLUID: run a substrate (glucose-containing) fluid alongside the insulin to prevent hypoglycaemia and ketogenesis — typically 10% glucose at 50-100 mL/h, or glucose-saline; do NOT run insulin without substrate unless treating DKA (where endogenous substrate is the ketones). Add K+ to the substrate bag. [1]

  3. USE A TRUSTED ALGORITHM: local sliding-scale tables prescribe the rate (U/h) by current glucose and rate of change. Most algorithms have three tiers — low-rate (insulin-naïve, low BMI, renal), standard, and high-rate (steroids, infection, high BMI). Start at the standard rate and re-check glucose hourly until stable. [1]

  4. MONITOR AND TITRATE: check glucose hourly until stable (within target for 4 h), then every 2 h. Aim to keep glucose 6-10 mmol/L, avoiding excursions >10 and never below 4. Adjust by one tier if persistently above or below target. [1]

  5. TRANSITION OFF safely: when the patient is eating / the acute insult has resolved and usual diabetes medication is appropriate, give the long-acting background insulin (e.g. glargine) and continue the VRII for at least 2-4 h overlap before stopping the IV insulin, to prevent rebound hyperglycaemia and ketogenesis. Never stop a VRII without background cover.

[1]

Diabetic ketoacidosis (DKA) — fixed-rate insulin infusion

The one-paragraph answer on insulin in DKA (fixed-rate insulin infusion)

In DKA the priority is to switch off ketogenesis, not merely to lower glucose — insulin is therefore given as a fixed-rate intravenous insulin infusion (FRII) at 0.1 U/kg/h (Actrapid 50 U in 50 mL saline), dosed on actual body weight, continuing at that fixed rate until the ketonaemia has cleared (blood ketones <0.6 mmol/L, venous bicarbonate ≥15 mmol/L, venous pH ≥7.3, anion gap closed). Glucose is expected to fall faster than ketones clear, so once blood glucose falls below ~14 mmol/L a 10% glucose substrate is started alongside (NOT reducing the insulin rate — the insulin rate stays at 0.1 U/kg/h until ketones clear). The expected glucose fall is 3 mmol/L/h; if it falls more slowly, increase the FRII to 0.15 U/kg/h. Critically, potassium falls steeply as acidosis corrects and insulin drives K+ intracellularly — replace K+ (typically 40-80 mmol per 24 h via substrate) once serum K+ <5.5 mmol/L. Continue the patient's usual long-acting background insulin throughout (or restart it) so that when the FRII stops there is basal cover and no rebound.[15][16]

DKA fixed-rate insulin infusion protocol — practical steps

  1. FLUIDS FIRST (0.9% saline, 1 L stat then over 1 h, then titrated) — restore circulating volume before/at the same time as insulin; the profoundly volume-depleted DKA patient is at risk of cardiovascular collapse. [1]

  2. START FIXED-RATE INSULIN 0.1 U/kg/h based on actual body weight (Actrapid 50 U in 50 mL 0.9% saline). This rate is NOT adjusted for glucose — the target is ketone suppression, which requires continuous tissue insulin levels. [1]

  3. POTASSIUM REPLACEMENT — measure K+ at baseline (often normal/high from acidosis despite total-body deficit):

    • If K+ >5.5 — omit K+, recheck in 2 h.
    • If K+ 3.5-5.5 — add 40 mmol/L to each substrate bag (the most common scenario).
    • If K+ <3.5 — defer insulin, give IV K+ (and cardiology input if <3.0) until K+ ≥3.5, then start insulin. [1]
  4. ADD GLUCOSE SUBSTRATE when blood glucose <14 mmol/L — switch to 10% glucose at 125 mL/h (with K+) alongside the saline, to allow continuation of the full FRII without hypoglycaemia. Do NOT reduce the insulin rate while ketones persist. [1]

  5. MONITOR — blood ketones (β-hydroxybutyrate), glucose, venous pH/bicarbonate and K+ every 1-2 h. Expected fall: glucose 3 mmol/L/h, ketones 0.5 mmol/L/h, bicarbonate ↑3 mmol/L/h. [1]

  6. STOP CRITERIA — ketones <0.6 mmol/L, bicarbonate ≥15, pH ≥7.3. Then convert to subcutaneous regimen: ensure long-acting background insulin given and a meal then stop the FRII 30-60 min later. [1]

  7. COMPLICATIONS — hypoglycaemia, hypokalaemia, cerebral oedema (mainly in children/young adults — avoid overly rapid correction; some units add phosphate only if <0.5), ARDS.

[1]

Hyperosmolar hyperglycaemic state (HHS)

The one-paragraph answer on insulin in HHS — lower dose, fluids first

HHS differs from DKA: minimal/no ketosis (residual insulin is enough to suppress lipolysis but not enough for normoglycaemia), extreme dehydration and hyperosmolality (osmolality >320, glucose often >30 mmol/L), and an older, comorbid population. Insulin plays a secondary role — glucose often falls dramatically with fluid resuscitation alone (volume depletion drives counter-regulatory hormones and concentrates glucose). The JBDS approach: aggressive fluid resuscitation (0.9% saline) first, then a LOWER insulin dose — 0.05 U/kg/h fixed-rate (half the DKA dose), begun only if glucose is not falling adequately with fluids (e.g. fall <5 mmol/L in the first 2-4 h). Excessive initial insulin risks osmotic shifts, hypovolaemia and thromboembolism, and precipitous falls in osmolality. Continue insulin until the patient is biochemically stable and re-hydrated, then transition to a subcutaneous regimen. HHS carries high mortality (~10-20%), driven by the precipitant (infection, MI) and thromboembolism — give prophylactic low-molecular-weight heparin.[14][16]

DKA vs HHS — insulin strategy contrast

FeatureDKAHHS
Insulin dose0.1 U/kg/h fixed-rate0.05 U/kg/h (lower; fluids first)
Primary target of insulinSwitch off ketogenesis (β-hydroxybutyrate <0.6)Lower glucose after rehydration
KetosisPresent (marked)Minimal/absent
Typical glucose15-30 mmol/LOften >30 mmol/L
Fluid deficit~6 L~9-10 L (more profound)
pH / bicarbonatepH <7.3, HCO3 <15pH >7.3, HCO3 >15
OsmolalityVariable>320 mOsm/kg (markedly raised)
PatientAny age, type 1 more commonOlder, type 2, comorbid
Key risk of over-treatment with insulinCerebral oedema (children), hypokalaemiaOsmotic shift, hypovolaemia, thromboembolism
Mortality<5% (adults)10-20%
[1]

Hyperkalaemia — insulin-dextrose

Insulin-dextrose for hyperkalaemia — mechanism and practical points

  1. MECHANISM: insulin activates the Na+/H+ exchanger → ↑intracellular Na+ → drives the Na+/K+-ATPase → shifts K+ out of serum INTO cells. Onset within 15 min, peak effect at 30-60 min, duration 4-6 h (the effect is transient — K+ will leak back out — so a definitive removal strategy, i.e. potassium-binding resin or dialysis, is needed if hyperkalaemia is ongoing). [1]

  2. REGIMEN: traditionally 10 units Actrapid + 25 g dextrose IV (25 mL of 50% or 125 mL of 20%); some units use 5 U + 25 g in dialysis-dependent renal patients to reduce hypoglycaemia. Give over 15-30 min. [1]

  3. HYPOGLYCAEMIA is the commonest complication (up to 20-75%), occurring 1-6 h later — especially in renal failure, anuric patients, and those without diabetes. Monitor glucose every 30-60 min for 4-6 h; consider a prophylactic 10% glucose infusion or smaller insulin dose in high-risk patients. [1]

  4. PRACTICAL: albuterol (salbutamol nebulised) has an additive K-lowering effect; combine for synergy. Insulin-dextrose does NOT remove K+ from the body — it only redistributes it. For total-body K+ removal use patiromer/sodium zirconium/potassium-binding resins or dialysis.

[1]

High-dose insulin / GIK for toxin-induced cardiogenic shock

High-dose insulin euglycaemia therapy (HIET) for calcium-channel-blocker / beta-blocker toxicity

In severe calcium-channel-blocker (CCB) or beta-blocker (BB) overdose, standard inotropes/vasopressors may fail because the toxin blocks the very receptors that inotropes act on. High-dose insulin euglycaemia therapy (HIET / GIK) — insulin 1 U/kg IV bolus then 0.5-1 U/kg/h (up to 10 U/kg/h), titrated with concomitant dextrose (and K+) to maintain euglycaemia — improves cardiac contractility and systemic vascular resistance by a mechanism independent of the blocked channels (shifts cardiac substrate utilisation toward carbohydrate, enhances intracellular Ca2+ handling). Give alongside fluids and vasopressors; monitor glucose (q10-20 min initially) and K+. It has become a first-line antidotal therapy for severe CCB/BB poisoning.[1]

Oral hypoglycaemics

The five classes of oral/injectable hypoglycaemic — mechanism, indication, signature toxicity

ClassAgent(s)MechanismHypoglycaemia riskSignature adverse effectOutcome evidence
BiguanideMetforminActivates AMP kinase (AMPK) → ↓hepatic gluconeogenesis, ↓insulin resistance, ↑peripheral glucose uptakeNone (does not stimulate insulin)Lactic acidosis (MALA) in renal failure/hypoxia/sepsis; GI upsetUKPDS-34 — ↓mortality, ↓MI in overweight T2DM[3]
SulfonylureaGliclazide, glibenclamide, glipizide, glimepirideClose β-cell K-ATP channel → depolarisation → Ca2+ influx → insulin release (glucose-independent)HIGH (especially glibenclamide, elderly, renal/hepatic)Weight gain; SIADH (chlorpropamide); prolonged hypoglycaemia in overdoseEffective glucose lowering; no CV benefit
DPP-4 inhibitor (gliptin)Sitagliptin, linagliptin, vildagliptin, saxagliptinInhibit DPP-4 → prolong endogenous GLP-1 / GIP incretins → glucose-dependent insulin releaseLow (glucose-dependent)Pancreatitis (rare); arthralgia; upper respiratoryTECOS — sitagliptin CV-neutral[9]
SGLT2 inhibitor (gliflozin)Dapagliflozin, empagliflozin, canagliflozin, ertugliflozinBlock Na-glucose cotransporter in proximal tubule → glycosuria (~60-80 g glucose/day), osmotic diuresisLow (insulin-independent) — but euglycaemic DKAGenital mycotic infections, volume depletion, Fournier's gangrene, ↑fractures (canagliflozin)EMPA-REG, CANVAS, CREDENCE — ↓CV death, ↓HF hospitalisation, ↓renal progression[4][5][6]
GLP-1 receptor agonistExenatide, liraglutide, semaglutide, dulaglutide, tirzepatide (injectable; oral semaglutide)Incretin mimetic → ↑glucose-dependent insulin release, ↓glucagon, ↓gastric emptying, ↑satiety → weight lossLow (glucose-dependent)Nausea/vomiting; pancreatitis (rare); early retinopathy worsening (semaglutide)LEADER / SUSTAIN-6 — ↓CV events, ↓mortality; ↓renal progression[7][8]

Metformin (biguanide)

Metformin — mechanism, indication, MALA, and the UKPDS-34 evidence

Metformin is a biguanide and first-line oral hypoglycaemic for type 2 diabetes. Its molecular target is AMP-activated protein kinase (AMPK) — activation of this cellular 'energy sensor' suppresses hepatic gluconeogenesis (the dominant effect), increases peripheral (muscle) glucose uptake, and reduces intestinal glucose absorption and insulin resistance. Crucially, metformin does NOT stimulate insulin secretion — so it carries no risk of hypoglycaemia when used alone. Its place in therapy was cemented by UKPDS-34 (1998), which showed that in overweight patients with newly-diagnosed T2DM, metformin reduced myocardial infarction, diabetes-related death and all-cause mortality more than sulfonylurea or insulin — the only monotherapy in that trial to show a clear macrovascular benefit. The feared complication is metformin-associated lactic acidosis (MALA) — rare in healthy users but occurring in renal failure, hypoxia, sepsis, dehydration or acute kidney injury when metformin accumulates and inhibits mitochondrial Complex I, blocking gluconeogenesis and trapping lactate. Presentation: profound high-anion-gap metabolic acidosis (pH often <7.1, lactate often >10-15) with hypotension and AKI. The Cochrane review (Salpeter 2010) estimated MALA incidence at ~3-6 per 100,000 patient-years — far lower than historically feared — but mortality remains high (~30-50%) when it occurs. Treatment: stop metformin; aggressive supportive care; haemodialysis (removes both metformin and lactate, and corrects the acidosis). Metformin is withheld perioperatively, around iodinated contrast, and whenever eGFR <30 mL/min.[3][11][12]

Metformin pharmacokinetics and the lactic-acidosis risk profile

PropertyValue
MechanismAMPK activation → ↓hepatic gluconeogenesis (dominant), ↑muscle glucose uptake
HypoglycaemiaNone (insulin-independent)
Weight effectWeight-neutral / modest weight loss
EliminationRenal (100%, unchanged) — accumulates in renal failure
Half-life~6 h
Standard dose500 mg BD up to 1 g TDS (max 2-3 g/day)
ContraindicationeGFR <30; hold perioperatively / around contrast / acute illness; avoid in severe hepatic failure, hypoxia
MALA incidence~3-6 / 100,000 patient-years (Cochrane)[11]
MALA treatmentStop metformin; supportive; haemodialysis (clears metformin + lactate, corrects acidosis)[12]

Sulfonylureas

Sulfonylureas — the K-ATP channel, hypoglycaemia, and the octreotide antidote

Sulfonylureas (gliclazide, glibenclamide/glyburide, glipizide, glimepiride) lower glucose by directly closing the ATP-sensitive K+ channel (K-ATP) on the pancreatic β-cell — specifically binding the SUR1 subunit. Channel closure depolarises the cell, opens voltage-gated Ca2+ channels, and triggers insulin granule exocytosis — independently of glucose. The therapeutic downside follows directly from this: because insulin release is uncoupled from the blood glucose, hypoglycaemia is the principal adverse effect, especially with glibenclamide (long half-life, active metabolites, renally excreted), in the elderly, in renal/hepatic impairment, and after alcohol or missed meals. Sulfonylurea-induced hypoglycaemia in the elderly is a major cause of hospital admission and is associated with increased mortality. In overdose, sulfonylurea hypoglycaemia is prolonged and refractory (24-48 h) because the drug continues to stimulate insulin release — giving IV dextrose alone worsens the cycle (glucose → more insulin → deeper hypoglycaemia). The antidotal treatment is octreotide (a somatostatin analogue, 50-100 mcg SC every 6-8 h), which directly inhibits pancreatic insulin secretion, breaking the cycle; diazoxide is an alternative. Patients need 24 h observation minimum (glibenclamide has caused delayed hypoglycaemia beyond 48 h).[13]

Sulfonylurea agents — risk profile comparison

AgentHalf-lifeEliminationHypoglycaemia riskNotes
Gliclazide~10-12 hHepatic (inactive metabolites)Lowest of classPreferred in elderly/renal; modified-release formulation
Glipizide~2-4 h (effect 12-24 h)HepaticModerateShorter half-life favours safety
Glimepiride~5-9 hHepatic + renalModerateOnce daily; some insulin-sensitising effect
Glibenclamide (glyburide)~10 h (active metabolites)Renal (40%) + hepaticHIGHEST — prolonged; avoid in elderly/renalClassically implicated in severe overdose
[1]

DPP-4 inhibitors (gliptins)

DPP-4 inhibitor pharmacology — the incretin axis

  1. PHYSIOLOGY: after oral glucose, gut L- and K-cells release the incretin hormones GLP-1 and GIP, which amplify glucose-stimulated insulin secretion (the 'incretin effect' — oral glucose evokes more insulin than IV glucose). Native GLP-1/GIP are rapidly degraded (1-2 min) by the enzyme dipeptidyl peptidase-4 (DPP-4). [1]

  2. MECHANISM of gliptins: sitagliptin, linagliptin, vildagliptin, saxagliptin competitively inhibit DPP-4, prolonging the half-life of endogenous incretins → enhanced glucose-dependent insulin secretion. Because the effect is glucose-dependent (only amplifies insulin release when glucose is high), hypoglycaemia is uncommon and the drugs are weight-neutral. [1]

  3. SAFETY: the TECOS trial (2015) established that sitagliptin is cardiovascularly neutral in T2DM with established CV disease — no increase in CV events, heart failure or pancreatitis. (Saxagliptin and alogliptin showed a signal for heart-failure hospitalisation in their CVOTs, so are used with caution in HF.) [1]

  4. ADVERSE EFFECTS: pancreatitis (rare but reported — stop and avoid if history), arthralgia, bullous pemphigoid, upper-respiratory symptoms. Dose-reduce in renal impairment (sitagliptin, saxagliptin); linagliptin is hepatically cleared and needs no renal dose adjustment. [1]

  5. ICU CONTEXT: generally held in acute critical illness; low glycaemic volatility makes them attractive when restarting post-ICU.[9]

SGLT2 inhibitors (gliflozins)

SGLT2 inhibitors — mechanism, CV/renal benefit (EMPA-REG, CANVAS, CREDENCE), and euglycaemic DKA

SGLT2 inhibitors (dapagliflozin, empagliflozin, canagliflozin, ertugliflozin) block the sodium-glucose cotransporter-2 (SGLT2) in the early proximal tubule, the transporter responsible for reabsorbing ~90% of filtered glucose. Inhibition causes glycosuria (~60-80 g glucose excreted per day), a modest diuresis/natriuresis, and a fall in blood glucose that is independent of insulin — hence low hypoglycaemia risk. Beyond glucose-lowering, the class has transformed cardiorenal medicine: EMPA-REG OUTCOME (empagliflozin, 2015) showed a striking 38% reduction in cardiovascular mortality and reduced heart-failure hospitalisation; CANVAS (canagliflozin, 2017) confirmed CV benefit and renal protection (at the cost of a small ↑amputation/fracture signal); CREDENCE (canagliflozin, 2019) was the first trial to show reduced kidney-failure progression in diabetic nephropathy (30% ↓composite renal endpoint). The mechanism of CV/renal benefit is multifactorial: osmotic diuresis (↓preload), ↓blood pressure, ↓glomerular hyperfiltration (restoring tubuloglomerular feedback by increasing distal Na delivery), improved cardiac energetics (ketone utilisation), and weight loss. The signature danger is euglycaemic diabetic ketoacidosis — because the drugs lower glucose independent of insulin, ketogenesis can proceed while the glucose stays deceptively normal (often <14 mmol/L), precipitated by surgery, fasting, acute illness, alcohol, or insulin dose reduction. Presentation is atypical; check ketones and a venous blood gas in any unwell SGLT2-treated patient. Stop SGLT2 inhibitors 3-4 days before elective surgery and during acute illness.[4][5][6][10]

Cardiovascular/renal outcome trials of SGLT2 inhibitors

TrialDrugPopulationKey resultCitation
EMPA-REG OUTCOME (2015)EmpagliflozinT2DM + established CVD38% ↓ CV death; ↓HF hospitalisation; ↓all-cause mortalityZinman[4]
CANVAS Program (2017)CanagliflozinT2DM + high CV risk↓composite CV events; ↓HF hospitalisation; renal protection; ↑amputation signalNeal[5]
CREDENCE (2019)CanagliflozinT2DM + diabetic nephropathy30% ↓ kidney-failure composite (ESKD, doubling creatinine, renal/CV death) — stopped early for benefitPerkovic[6]
DAPA-HF / EMPERORDapagliflozin / empagliflozinHFrEF (with or without diabetes)↓HF hospitalisation + mortality — class now an HFrEF pillar—

GLP-1 receptor agonists

GLP-1 receptor agonists — incretin mimetics, weight loss, and CV benefit (LEADER, SUSTAIN-6)

GLP-1 receptor agonists (exenatide, liraglutide, semaglutide, dulaglutide, tirzepatide — a dual GIP/GLP-1 agonist) are incretin mimetics: resistant to DPP-4 degradation, they activate the GLP-1 receptor to produce glucose-dependent insulin secretion, suppress glucagon, slow gastric emptying, and increase satiety — producing clinically meaningful weight loss (the reason semaglutide/tirzepatide are now approved for obesity). They are administered subcutaneously (semaglutide also available orally). The CV outcome trials LEADER (liraglutide, 2016) and SUSTAIN-6 (semaglutide, 2016) showed reduction in major cardiovascular events and cardiovascular mortality in T2DM with established/at-high-risk CVD, and renal benefit (reduced albuminuria progression). Adverse effects are predominantly gastrointestinal (nausea, vomiting, diarrhoea — titrate slowly up), with rare pancreatitis and (for semaglutide) early worsening of diabetic retinopathy attributed to rapid glucose lowering. In the ICU they are generally held (delayed gastric emptying complicates enteral feeding; ketoacidosis risk with starvation).[7][8]

Hypoglycaemia management

ICU glycaemic control and hypoglycaemia management: target 6 to 10 mmol per L, variable-rate insulin infusion, stop insulin for hypo, IV dextrose, sulfonylurea protocol
FigureTarget moderate control (about 6–10 mmol/L). Treat hypoglycaemia immediately with IV glucose; for sulfonylureas plan prolonged observation and consider octreotide.
[1]

The one-paragraph answer on hypoglycaemia management — rule of 15, glucagon, 50% dextrose, octreotide

Hypoglycaemia (glucose <4.0 mmol/L; severe = any glucose requiring third-party assistance) is treated by severity. In the conscious patient able to swallow, use the 'rule of 15': give 15-20 g of fast-acting carbohydrate (3-4 glucose tablets, 150-200 mL fruit juice, glucose gel), re-check capillary glucose at 15 minutes, and repeat if <4 mmol/L; once above 4, give a long-acting carbohydrate snack. In the unconscious/NBM/uncooperative patient, give IV dextrose — 50% dextrose 25-50 mL via a large peripheral or central vein (50% is irritant — extravasation causes tissue necrosis; many ICUs prefer 10% glucose 100-250 mL via central line for safety), or glucagon 1 mg IM/SC when IV access is delayed (glucagon stimulates hepatic glycogenolysis — works within 10-15 min but is ineffective when glycogen is depleted — alcohol misuse, starvation, chronic liver disease). For sulfonylurea-induced hypoglycaemia give octreotide 50-100 mcg SC every 6-8 h (a somatostatin analogue that directly suppresses pancreatic insulin release, breaking the glucose→insulin→hypoglycaemia cycle that IV dextrose perpetuates) and observe for 24-48 h; diazoxide is an alternative. Always identify and treat the cause and, after recovery, reduce the precipitating insulin/sulfonylurea to prevent recurrence.[1][13]

Hypoglycaemia treatments — agent, route, onset, caveat

TreatmentDose / routeOnsetWhen to useKey caveat
Oral fast carbohydrate (rule of 15)15-20 g PO (glucose tablets, juice, gel)10-15 minConscious, can swallowRecheck at 15 min; follow with long-acting carb
IV dextrose 10%100-250 mL IV5-10 minICU first-line — central line, saferLarge volume; preferred over 50% by many units
IV dextrose 50%25-50 mL IV5 minRapid correction, large vein/centralExtravasation → tissue necrosis; rebound in sulfonylurea
Glucagon1 mg IM/SC10-15 minNo IV access, out-of-hospitalIneffective if glycogen-depleted (alcohol, starvation); causes vomiting (aspiration risk if unconscious)
Octreotide50-100 mcg SC q6-8h30-60 minSulfonylurea-induced refractory hypoglycaemiaDoes not work instantly — bridge with dextrose; observe 24-48 h
Hydrocortisone / glucagon infusionAs indicatedVariableAdrenal insufficiency; refractoryNot first-line
[1]

Stepwise management of severe hypoglycaemia in the ICU

  1. RECOGNISE — glucose <4.0 mmol/L (severe <2.8 or any level with impaired consciousness). Symptoms: autonomic (sweating, tremor, palpitations, hunger) and neuroglycopaenic (confusion, seizures, coma). In ICU, neuroglycopaenic signs may be the only clue in sedated patients. [1]

  2. STOP any insulin infusion / oral hypoglycaemic and send a laboratory glucose to confirm. [1]

  3. GIVE IV DEXTROSE — 50% 25-50 mL via large vein, or 10% 100-250 mL via central line (safer). Recheck glucose in 10-15 min and repeat until >6 mmol/L. [1]

  4. IF NO IV ACCESS — glucagon 1 mg IM/SC (note: ineffective in alcohol or starved states). [1]

  5. IDENTIFY AND REMOVE THE CAUSE — review all insulin/sulfonylurea; check renal function; consider sepsis/adrenal insufficiency/hepatic failure. If sulfonylurea is the cause → octreotide 50-100 mcg SC q6-8h + continuous dextrose infusion + 24-48 h observation. [1]

  6. PREVENT RECURRENCE — once glucose stable >6, start a background glucose substrate infusion if on IV insulin; reduce the insulin dose; reassess the glycaemic-control target (was it too tight?). [1]

  7. DOCUMENT and review — every severe hypoglycaemic event is a critical incident; tight control (NICE-SUGAR lesson) is dangerous.

[1]

SAQ — ICU insulin infusion protocol in a septic post-operative diabetic patient

10 minutes · 10 marks

A 68-year-old man with type 2 diabetes (HbA1c 78 mmol/mol on metformin and gliclazide) is admitted to ICU post-emergency laparotomy for perforated diverticulitis. He is in septic shock on noradrenaline 0.3 mcg/kg/min. His capillary glucose is 18 mmol/L. Outline your insulin infusion protocol, your glycaemic target, and how you will safely transition him to subcutaneous insulin for ward transfer.

[1]

SAQ — Sulfonylurea and metformin overdose: pharmacology and ICU management

10 minutes · 10 marks

A 24-year-old woman (60 kg) is brought to ED 4 hours after ingesting glibenclamide 70 mg (seventy 10-mg tablets) and an unknown quantity of metformin, with alcohol. She is drowsy (GCS 13), glucose 1.4 mmol/L, lactate 7.2 mmol/L, pH 7.18, and transiently responded to 25 g IV dextrose before recurring hypoglycaemia. The registrar asks for your ICU approach.

[1]

Clinical pearls

Clinical pearl

  1. Insulin types are classified by ONSET and DURATION, driven by the rate of dissociation of the subcutaneous hexamer. Native insulin self-associates into hexamers (the storage form); only monomers are absorbed. Rapid analogues (lispro/aspart/glulisine) are engineered to stay monomeric (onset ~15 min); regular/Actrapid must dissociate (onset 30-60 min); NPH/glargine are slowed by protamine or precipitation. Only regular soluble insulin (Actrapid) can be given IV — analogues offer no advantage IV because there is no absorption step.[1]

  2. The ONLY insulin given IV is soluble (regular/Actrapid). It has no absorption step, so onset is minutes and duration ~30-60 min — perfect for infusion. All other types are optimised for subcutaneous pharmacokinetics and have no role IV. [1]

  3. NICE-SUGAR is the pivotal trial of ICU glucose control — it killed 'tight glycaemic control'. The Leuven trial (2001) suggested tight control (4.4-6.1) reduced mortality; the much larger NICE-SUGAR (2009, n=6104) showed the tight arm had higher mortality (27.5% vs 24.9%) driven by severe hypoglycaemia. ICU target is 6-10 mmol/L (moderate). Never aim for a glucose of 4-6 in an ICU patient.[1][2]

  4. In DKA, insulin's job is to switch off KETOGENESIS, not just to lower glucose. Hence the fixed-rate 0.1 U/kg/h (weight-based, not glucose-titrated), continued until ketones clear (<0.6). When glucose falls below ~14, start a glucose substrate rather than reducing the insulin rate — the insulin rate stays up until ketones are suppressed. If you titrate insulin to glucose you will stop it too early and the ketosis will recur.[15]

  5. In HHS, give FLUIDS first and use HALF the DKA insulin dose (0.05 U/kg/h). HHS patients are profoundly dehydrated (deficit ~9-10 L) and have residual insulin (so little ketosis); glucose often falls dramatically with rehydration alone. Aggressive insulin risks osmotic shifts and thromboembolism — the leading cause of HHS mortality.[14]

  6. Potassium falls predictably during DKA treatment — replace it proactively. Acidosis masks total-body K+ depletion (K+ shifts out of cells); insulin drives it back in, and rehydration improves perfusion. Once K+ <5.5, add K+ to the substrate; if baseline K+ <3.5, DEFER insulin and replete first (risk of arrhythmia). [1]

  7. Metformin is the ONLY oral hypoglycaemic that does NOT cause hypoglycaemia (it does not stimulate insulin) and the only one (in monotherapy) to show a mortality benefit (UKPDS-34). Its danger is MALA — profound high-anion-gap acidosis in renal failure/hypoxia/sepsis — treated with haemodialysis (clears metformin AND lactate).[3][12]

  8. Sulfonylurea overdose causes PROLONGED refractory hypoglycaemia — the antidote is octreotide, not just dextrose. Giving dextrose alone stimulates further insulin release and deepens the cycle. Octreotide 50-100 mcg SC q6-8h (a somatostatin analogue) suppresses endogenous insulin and breaks the cycle; observe 24-48 h (glibenclamide especially can cause delayed rebound).[13]

  9. SGLT2 inhibitors cause EUGLYCAEMIC DKA — the glucose can be normal. The drug lowers glucose independent of insulin, so ketogenesis proceeds while glucose looks fine (often <14 mmol/L). Precipitants: surgery, fasting, acute illness, alcohol. In any unwell SGLT2 patient, check ketones and a venous gas, not just glucose. Stop 3-4 days before elective surgery.[10]

  10. The molecular target of sulfonylureas is the SUR1 subunit of the β-cell K-ATP channel — the same channel glucose acts on. Glucose → ↑ATP → closes K-ATP → depolarisation → insulin. Sulfonylureas close K-ATP directly (glucose-independent) — which is why they cause hypoglycaemia. Diazoxide OPENS K-ATP (used for insulinoma and adjunctively in sulfonylurea overdose); octreotide works downstream (inhibits exocytosis). [1]

  11. Glucagon is INEFFECTIVE in alcohol-induced and starved hypoglycaemia. It works by hepatic glycogenolysis — if glycogen is depleted (chronic alcohol misuse, starvation, severe liver disease) there is no glycogen to mobilise. Use IV dextrose instead. Glucagon also causes vomiting — never give to an unconscious/unprotected patient without airway control. [1]

  12. 50% dextrose is VESICANT — extravasation causes tissue necrosis. Use a large-bore, well-placed cannula or a central line, and aspirate to confirm patency. Many ICUs now default to 10% glucose via central line (100-250 mL) for safety, reserving 50% for clearly peripheral emergencies. [1]

  13. Insulin shifts K+ into cells within 15 min — the basis of insulin-dextrose for hyperkalaemia. But the effect is REDISTRIBUTION, not removal — K+ leaks back out within 4-6 h. Always pair with a definitive removal strategy (potassium binder or dialysis). Watch for late hypoglycaemia (1-6 h) — monitor glucose for 4-6 h, especially in renal failure; smaller insulin doses (5 U) are reasonable in high-risk patients. [1]

  14. Every insulin infusion needs a SUBSTRATE and an OVERLAP when stopping. Running insulin without glucose risks hypoglycaemia and ketogenesis; stopping an insulin infusion without background long-acting insulin on board causes rebound hyperglycaemia and ketogenesis within 1-2 h. Give the long-acting insulin and overlap the VRII for 2-4 h before stopping. [1]

  15. High-dose insulin euglycaemia therapy (HIET/GIK) is first-line antidotal therapy for severe calcium-channel-blocker and beta-blocker toxicity — insulin 1 U/kg bolus then 0.5-1 (up to 10) U/kg/h, with concurrent dextrose and K+. It works by a mechanism independent of the blocked channel (enhanced carbohydrate utilisation, intracellular Ca2+ handling) when conventional inotropes fail.[1]

  16. GLP-1 agonists and DPP-4 inhibitors exploit the 'incretin effect' — oral glucose evokes more insulin than IV glucose. The gut releases GLP-1/GIP, which amplify insulin secretion in a glucose-dependent way — which is why both classes have low hypoglycaemia risk. GLP-1 agonists (LEADER, SUSTAIN-6) reduce CV events and cause weight loss; DPP-4 inhibitors (TECOS) are CV-neutral and weight-neutral.[7][8][9]

  17. C-peptide distinguishes endogenous from exogenous insulin. Co-secreted 1:1 with insulin; high in insulinoma (and sulfonylurea), suppressed in exogenous insulin overdose. A key discriminator in the workup of spontaneous hypoglycaemia. [1]

  18. Insulin is an ANABOLIC, storage-promoting hormone — its deficiency causes catabolism (the DKA phenotype). ↑glycogenolysis/gluconeogenesis (hyperglycaemia), ↑lipolysis → ↑free fatty acids → ketogenesis (ketoacidosis), ↑proteolysis (muscle wasting). Insulin reverses all three — which is why it is the cornerstone of DKA treatment, not just a glucose-lowerer.

[1]

Red flags

TIGHT glucose control (4.4-6.1) INCREASES mortality — target 6-10 mmol/L (NICE-SUGAR)

Aggressive 'tight glycaemic control' (target 4.4-6.1 mmol/L), promoted after the Leuven trial, was shown by the multinational NICE-SUGAR trial (n=6104) to increase mortality (27.5% vs 24.9% with moderate control) because of severe hypoglycaemia (<2.2 mmol/L) — which independently predicted death. ICU glucose target is 6-10 mmol/L. Never chase normoglycaemia in the critically ill.[1]

Sulfonylurea overdose = prolonged refractory hypoglycaemia — use OCTREOTIDE, observe 24-48 h

Sulfonylureas (especially glibenclamide) cause continuous endogenous insulin release for 24-48 h. Giving IV dextrose alone perpetuates the cycle (glucose → more insulin). The definitive antidote is octreotide 50-100 mcg SC every 6-8 h (somatostatin analogue — suppresses insulin release). Admit/observe for at least 24 h (longer for glibenclamide) because of delayed rebound hypoglycaemia.[13]

SGLT2 inhibitors cause EUGLYCAEMIC DKA — check KETONES, not just glucose

In any unwell patient on an SGLT2 inhibitor (dapagliflozin, empagliflozin, canagliflozin), suspect euglycaemic DKA — glucose often <14 mmol/L while ketones are high and pH low. Precipitants: surgery, fasting, sepsis, alcohol, insulin dose reduction. Check blood ketones and a venous gas. Treat as standard DKA (fluids, insulin, K) but the glucose may not fall (the SGLT2 keeps excreting it) — start a glucose substrate early. Stop SGLT2 inhibitors 3-4 days before elective surgery.[10]

Metformin-associated lactic acidosis (MALA) — high mortality, treat with haemodialysis

Metformin accumulates in renal failure, hypoxia, sepsis, dehydration and inhibits mitochondrial Complex I → profound high-anion-gap lactic acidosis (pH often <7.1, lactate often >10-15). Presentation: hypotension, AKI, Kussmaul breathing. Treatment: stop metformin; aggressive supportive care; haemodialysis (clears both metformin and lactate and corrects acidosis). Mortality 30-50%. Withhold metformin perioperatively, around iodinated contrast, in acute illness, and whenever eGFR <30 mL/min.[11][12]

DKA fixed-rate insulin infusion drops potassium fast — replace proactively or risk arrhythmia

Insulin drives K+ intracellularly; acidosis correction unmasks total-body depletion. Serum K+ can fall 0.5-1 mmol/L per hour during DKA treatment. Start K+ replacement once serum K+ <5.5 mmol/L (typically 40 mmol/L in the substrate bag); if baseline K+ <3.5, DEFER insulin and replete first (arrhythmia risk). Check K+ every 1-2 h during the FRII.[15]

Glucagon is INEFFECTIVE in alcohol-induced and starved hypoglycaemia

Glucagon works by hepatic glycogenolysis. In alcohol misuse, starvation and severe liver disease glycogen is depleted, so glucagon will fail and may precipitate vomiting (aspiration risk if airway unprotected). Use IV dextrose instead for these patients.[1]

50% dextrose is VESICANT — extravasation causes tissue necrosis

Concentrated dextrose (50%) is highly irritant; extravasation causes severe soft-tissue injury and necrosis. Use a large-bore, well-sited cannula (aspirate to confirm) or a central line, and infuse slowly. Many ICUs now prefer 10% glucose 100-250 mL via central line for routine ICU hypoglycaemia treatment.[1]

Stopping an insulin infusion without background cover causes REBOUND ketogenesis

Insulin has a short half-life (~5 min IV). Stopping a VRII/FRII abruptly leaves the patient with no circulating insulin → rapid rebound hyperglycaemia and ketogenesis within 1-2 h. Always give the long-acting background insulin and overlap the infusion for 2-4 h before stopping. A 'gap' after an insulin infusion is a common cause of recurrent DKA.[15]

Key trials and evidence

NICE-SUGAR — Normoglycaemia in Intensive Care Evaluation and Survival Using Glucose Algorithm Regulation (PMID 19318384)

Study design

Multicentre (42 ICUs, international), randomised, controlled — 6104 mechanically-ventilated adults expected to stay ≥3 days

Population

Critically ill adults (mixed medical/surgical)

Intervention

Intensive glucose control (target 4.5-6.0 mmol/L; 81-108 mg/dL) vs conventional (target ≤10 mmol/L; ≤180 mg/dL)

Primary outcome

Death by 90 days

Key finding

INTENSIVE-control arm had SIGNIFICANTLY HIGHER 90-day mortality (27.5% vs 24.9%; OR 1.14) and dramatically more SEVERE HYPOGLYCAEMIA (6.8% vs 0.5%). Hypoglycaemia independently predicted death.

Clinical bottom line

OVERTURNED the Leuven 'tight control' era. ICU glucose target is MODERATE (6-10 mmol/L) — tight control is dangerous. The single most influential trial in ICU glucose management.

[1]

Leuven (van den Berghe) — Intensive Insulin Therapy in the Critically Ill (PMID 11794168)

Study design

Single-centre (Leuven surgical ICU), randomised, prospective — 1548 patients

Population

Surgical ICU patients (mostly post-cardiac surgery)

Intervention

Intensive insulin (target blood glucose 4.4-6.1 mmol/L, 80-110 mg/dL) vs conventional (target 10-11.1 mmol/L)

Primary outcome

ICU mortality

Key finding

Intensive insulin reduced ICU mortality (4.6% vs 8.0%) — greatest benefit in long-stay (>5 days) patients; also reduced bacteraemia, AKI needing dialysis, polyneuropathy.

Clinical bottom line

Pioneered 'tight glycaemic control' — but single-centre, protocol-driven, enteral-fed patients; could NOT be replicated elsewhere (NICE-SUGAR refuted it). Historical context only.

[1]

EMPA-REG OUTCOME — Empagliflozin, CV Outcomes and Mortality (PMID 26378978)

Study design

Randomised, double-blind, placebo-controlled — 7020 patients

Population

Type 2 diabetes + established cardiovascular disease

Intervention

Empagliflozin (SGLT2 inhibitor) 10 or 25 mg vs placebo, added to standard care

Primary outcome

Composite of major adverse CV events (CV death, non-fatal MI, non-fatal stroke)

Key finding

14% reduction in primary composite; striking **38% reduction in cardiovascular death** and 32% reduction in all-cause mortality; 35% reduction in HF hospitalisation.

Clinical bottom line

First SGLT2 CVOT to show CV/MORTALITY benefit — transformed SGLT2 inhibitors from glucose-lowering drugs into cardioprotective agents.

[1]

CREDENCE — Canagliflozin and Renal Outcomes in Type 2 Diabetes and Nephropathy (PMID 30990260)

Study design

Randomised, double-blind, placebo-controlled — 4401 patients; STOPPED EARLY for efficacy

Population

Type 2 diabetes + chronic kidney disease (eGFR 30-90) + albuminuria, on maximum-tolerated ACE/ARB

Intervention

Canagliflozin 100 mg vs placebo

Primary outcome

Composite of ESKD, doubling of serum creatinine, renal or CV death

Key finding

30% REDUCTION in the primary composite (HR 0.70); significant reduction in ESKD and CV outcomes — first dedicated trial of an SGLT2 inhibitor in diabetic kidney disease.

Clinical bottom line

Established SGLT2 inhibitors as renoprotective — now a foundation of diabetic kidney disease management regardless of glucose.

[1]

LEADER — Liraglutide and Cardiovascular Outcomes (PMID 27295427)

Study design

Randomised, double-blind, placebo-controlled — 9340 patients

Population

Type 2 diabetes + high cardiovascular risk

Intervention

Liraglutide (GLP-1 agonist) 1.8 mg SC daily vs placebo

Primary outcome

Composite of CV death, non-fatal MI, non-fatal stroke

Key finding

13% reduction in primary composite (HR 0.87); significant reduction in CV death (22%) and all-cause mortality (15%); reduced nephropathy progression.

Clinical bottom line

Established GLP-1 agonists as cardioprotective — liraglutide preferred in T2DM with established/at-risk CVD where weight loss and CV benefit are desired.

[1]

TECOS — Sitagliptin and Cardiovascular Outcomes (PMID 26052984)

Study design

Randomised, double-blind, placebo-controlled — 14,671 patients

Population

Type 2 diabetes + established cardiovascular disease

Intervention

Sitagliptin (DPP-4 inhibitor) vs placebo, added to existing therapy

Primary outcome

Composite of CV death, non-fatal MI, non-fatal stroke, unstable angina hospitalisation

Key finding

CV-NEUTRAL — no difference in primary composite, no increase in HF hospitalisation, no increase in pancreatitis. Sitagliptin safe in established CVD.

Clinical bottom line

Reassuring CV safety for DPP-4 inhibitors (sitagliptin specifically); unlike saxagliptin/alogliptin, no HF signal — sitagliptin preferred when a DPP-4 inhibitor is needed in CVD/HF.

[1]

Dosing quick-reference (ICU)

ICU insulin and hypoglycaemic dosing — practical reference

Drug / scenarioTypical doseRouteOnset / durationNotes
Insulin — DKA (FRII)0.1 U/kg/h (titrate to 0.15 if glucose fall slow)IV infusion (Actrapid 50 U/50 mL NS)Onset min / effect while runningContinue until ketones <0.6; start glucose substrate when glucose <14[15]
Insulin — HHS0.05 U/kg/h (after fluids)IV infusionMin / while runningHalf the DKA dose; fluids first[14]
Insulin — VRII / sliding scalePer algorithm (0.5-6 U/h typical)IV infusionMin / while runningTarget 6-10 mmol/L; substrate + K+ alongside
Insulin — hyperkalaemia10 U Actrapid + 25-50 g dextroseIV over 15-30 min15 min / 4-6 hMonitor glucose 4-6 h (late hypoglycaemia); consider 5 U in renal failure
Insulin — CCB/BB toxicity (HIET)1 U/kg bolus → 0.5-1 (up to 10) U/kg/hIV + dextrose ± K+Min / while runningFirst-line antidotal therapy; titrate to haemodynamics
Glucagon — hypoglycaemia1 mgIM/SC10-15 min / 1-2 hIneffective if glycogen-depleted; causes vomiting
IV dextrose — severe hypoglycaemia50% 25-50 mL OR 10% 100-250 mLIV (large vein / central)5-10 min50% is VESICANT; 10% via central line safer
Octreotide — sulfonylurea hypoglycaemia50-100 mcg SC q6-8hSC30-60 min / 6-8 hInhibits insulin release; observe 24-48 h[13]
Metformin (oral)500 mg BD → 1 g TDSPODaysHold periop / eGFR <30 / acute illness
Gliclazide (oral)40-80 mg OD → up to 320 mgPODaysPreferred sulfonylurea (lowest hypoglycaemia)
Sitagliptin (oral)100 mg OD (50 mg if eGFR 30-45)PODaysCV-safe (TECOS); no weight gain
Empagliflozin / Dapagliflozin (oral)10 mg OD (± up-titrate)PODaysStop pre-op / acute illness (euDKA risk)
Liraglutide / Semaglutide (SC)Liraglutide 0.6→1.8 mg OD; semaglutide 0.25→1 mg weeklySCDaysWeight loss + CV benefit; titrate to limit GI effects

References

  1. [1]NICE-SUGAR Study Investigators; Finfer S, Chittock DR, Su SY, et al. Intensive versus conventional glucose control in critically ill patients N Engl J Med, 2009.PMID 19318384
  2. [2]van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients N Engl J Med, 2001.PMID 11794168
  3. [3]UK Prospective Diabetes Study (UKPDS) Group Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK Prospective Diabetes Study (UKPDS) Group Lancet, 1998.PMID 9742977
  4. [4]Zinman B, Wanner C, Lachin JM, et al. Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes N Engl J Med, 2015.PMID 26378978
  5. [5]Neal B, Perkovic V, Mahaffey KW, et al. Canagliflozin and Cardiovascular and Renal Events in Type 2 Diabetes N Engl J Med, 2017.PMID 28605608
  6. [6]Perkovic V, Jardine MJ, Neal B, et al. Canagliflozin and Renal Outcomes in Type 2 Diabetes and Nephropathy N Engl J Med, 2019.PMID 30990260
  7. [7]Marso SP, Daniels GH, Brown-Frandsen K, et al. Liraglutide and Cardiovascular Outcomes in Type 2 Diabetes N Engl J Med, 2016.PMID 27295427
  8. [8]Marso SP, Bain SC, Consoli A, et al. Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes N Engl J Med, 2016.PMID 27633186
  9. [9]Green JB, Bethel MA, Armstrong PW, et al. Effect of Sitagliptin on Cardiovascular Outcomes in Type 2 Diabetes N Engl J Med, 2015.PMID 26052984
  10. [10]Bonora BM, Avogaro A, Fadini GP Euglycemic Ketoacidosis Curr Diab Rep, 2020.PMID 32424730
  11. [11]Salpeter SR, Greyber E, Pasternak GA, Salpeter EE Risk of fatal and nonfatal lactic acidosis with metformin use in type 2 diabetes mellitus Cochrane Database Syst Rev, 2010.PMID 20393934
  12. [12]Roberts DM, Ghannoum M Expanding the evidence for managing metformin poisoning to support decision-making Clin Toxicol (Phila), 2023.PMID 37129225
  13. [13]Glatstein M, Scolnik D, Bentur Y Octreotide for the treatment of sulfonylurea poisoning Clin Toxicol (Phila), 2012.PMID 23046209
  14. [14]Mustafa OG, Haq M, Dashora U, Dhatariya K, et al. Management of Hyperosmolar Hyperglycaemic State (HHS) in Adults: An updated guideline from the Joint British Diabetes Societies (JBDS) for Inpatient Care Group Diabet Med, 2023.PMID 36370077
  15. [15]Sharma A, Rengarajan L, Narendran P, Dhatariya K, et al. Guidelines for the management of diabetes-related ketoacidosis (DKA) have been poorly adopted and implemented, resulting in a lack of improvement in outcomes Diabet Med, 2025.PMID 39928758
  16. [16]Nyenwe EA, Kitabchi AE Evidence-based management of hyperglycemic emergencies in diabetes mellitus Diabetes Res Clin Pract, 2011.PMID 21978840