Activated charcoal

Why this drug is interesting

For twenty years the teaching was a single sentence from the 2005 AACT/EAPCCT position paper: single-dose activated charcoal should not be given routinely, and there is no evidence it improves clinical outcome; if given at all, give it within one hour.² It was one of the few genuinely memorable rules in toxicology, and it hardened in practice into something the position paper never said — that charcoal after an hour is pointless.

In 2026 the Clinical Toxicology Recommendations Collaborative — a joint project of three international clinical toxicology societies — published a 57-page systematic review and modified-Delphi guideline covering 43 poisons or poison categories.¹ It does not soften the one-hour rule so much as replace the framework around it — though its own conclusion is carefully bounded: charcoal beyond one hour "in selected poisons", not in everyone.

This page reports the 2026 recommendations, checked against the full text. It is a summary of a 57-page consensus document, not a substitute for it — and in the UK, not a substitute for TOXBASE or a call to the National Poisons Information Service.

Pharmacology

Mechanism

Activated charcoal is carbon processed to an enormous internal surface area — published figures for pharmaceutical grades span roughly 950–2,000 m² per gram. Poison molecules adsorb onto that surface by van der Waals forces and remain bound within the gut lumen, so they are excreted rather than absorbed. It is not absorbed itself, has no systemic pharmacology, and no dose–response relationship in the usual sense: what matters is the charcoal-to-poison ratio in the gut at the time it is given.

That ratio explains most of the clinical rules. Charcoal fails for metals and small polar molecules because they adsorb poorly. It fails in massive ingestion because the ratio collapses. And it works late where the poison is still in the stomach — a pharmacobezoar, a modified-release preparation, or a drug burden exceeding its own solubility.

Enhanced elimination

Multiple-dose activated charcoal does something different from decontamination. By maintaining a concentration gradient across the gut wall, and by interrupting enterohepatic and enteroenteric recirculation, it can increase clearance of drug that has already been absorbed — sometimes described as gastrointestinal dialysis. This only works for drugs with the right properties: low volume of distribution, low protein binding, long half-life, and continued adsorption in the gut.

Dose, and the intervals

These are good practice statements in the 2026 guideline, so for the first time the dose is in the guideline rather than inherited from custom:¹

• "The dose of AC for single-dose AC and additional-dose AC is 50 g in adults or 1 g/kg (up to the adult dose) in children."
• "The dose of AC for additional-dose AC can be administered to complete decontamination at any point in time following single-dose AC when there is evidence or suspicion of ongoing absorption."
• "The multiple-dose AC regimen begins with the same AC dose as for single-dose AC and repeats with that dose every 4 h, or a half-dose every 2 h."

Special populations

• Children: the adult indications and contraindications are "generally applicable" (2D).¹
• Pregnancy and the elderly: the evidence was sparse and of very low quality; the workgroup reached consensus statements rather than finding data, and named this an area for future research.¹
• Morbid obesity and post-bariatric surgery: individualised risk assessment; explicitly flagged as needing research.¹
• Patients on maintenance medication that charcoal would adsorb and that cannot be given by another route: individualised risk assessment.¹

The 2026 recommendations — where charcoal has no role

The workgroup concluded there is no role for activated charcoal in poisoning from:¹

• Metals and metalloids — grouped in Table 4 as lead, arsenic, copper, caesium and mercury, plus iron as its own row. All strong recommendations against
• Alcohols — ethanol, and the toxic alcohols methanol and ethylene glycol. All strong recommendations against
• Lithium — strong recommendation against
• Metformin — but see below: this one is a weak suggestion against, not a strong one

The 2026 recommendations — where charcoal is appropriate

Single-dose (decontamination)

Activated charcoal is appropriate after ingestion of:¹

• Cardiovascular — antidysrhythmics (types I and III, not discussed individually), beta-adrenergic antagonists, calcium channel blockers, cardiac glycosides, disopyramide, quinidine
• Neuropsychiatric — bupropion, carbamazepine, lamotrigine, moclobemide, phenobarbital, phenytoin, selective serotonin reuptake inhibitors, tricyclic antidepressants, valproic acid, venlafaxine
• Analgesics and antipyretics — ibuprofen, paracetamol, salicylates, opioids, colchicine
• Anti-infectives and antimalarials — chloroquine, dapsone, isoniazid, quinine
• Endocrine and metabolic — sulfonylureas, theophylline
• Anticoagulants — factor Xa inhibitors, warfarin
• Other — cocaine, cyanide, diphenhydramine, methotrexate, organophosphorus insecticides, paraquat, thallium

The additional dose

The guideline introduces a new concept: an additional dose to complete gastrointestinal decontamination, distinct from a multiple-dose regimen for enhanced elimination. It is appropriate after ingestion of carbamazepine, paracetamol, paraquat, phenobarbital, salicylates, thallium, theophylline, valproic acid and verapamil — poisons that may remain in the gastrointestinal tract for prolonged periods.¹

Multiple-dose (enhanced elimination)

Multiple-dose activated charcoal for enhanced elimination is appropriate in poisoning with carbamazepine, cardiac glycosides, colchicine, dapsone, phenobarbital, phenytoin, thallium and theophylline.¹

Dose thresholds — the part that decides most cases

The single most useful thing in the guideline is not the timing at all: it is that Table 4 gives an ingested-dose threshold for each poison, below which charcoal is not recommended. The lists of "appropriate" poisons above are conditional on crossing these. Doses shown are the lowest intervention dose; TD = maximal therapeutic doses, IR/MR = immediate/modified release.¹

Timing

The guideline's timing statement is the heart of the change:¹

The maximum time post-ingestion for which activated charcoal administration is recommended differs for each poison and different formulations. According to an individualized risk assessment, activated charcoal is appropriate up to 6 h post-ingestion for many poisons.

Clinical Toxicology Recommendations Collaborative, 2026¹

And beyond six hours: "If ongoing absorption is suspected, which may occur, for example, with pharmacobezoar formation, certain modified-release preparations, or when drug burden exceeds the limits of solubility, then activated charcoal can be administered beyond 6 h post-ingestion for gastrointestinal decontamination."¹

Note also what has not changed: there is still no randomised evidence that charcoal improves clinical outcome. The 2005 paper's absorption data — a mean reduction of about 47% at 30 minutes, 40% at 60 minutes, 16.5% at 120 minutes and 21% at 180 minutes with at least 50 g — remain the underlying pharmacological reality.² Earlier is still better; later is no longer nothing.

The six-step algorithm

Figure 1 of the guideline sets out a general approach. In sequence, each "no" ending the process:¹

1. Is the ingested dose expected to cause significant toxicity? If no → no charcoal.
2. Is at least one poison adsorbable to charcoal? If no → no charcoal. The figure's non-adsorbable list: alcohols, ions (e.g. potassium), most metals and metalloids, corrosives, petroleum distillates.
3. Is charcoal recommended or suggested at the current time post-ingestion? If no → no charcoal.
4. Are there no contraindications present or expected? If yes → give a single dose.
5. Is an additional dose recommended for at least one poison, with no contraindications? If yes → give an additional dose. Triggers: dose clinically significant for severe toxicity, or evidence of delayed absorption (modified-release formulation, rising poison concentrations).
6. Is multiple-dose charcoal recommended, with no contraindications? If yes → give multiple doses. Only where there is evidence of enterohepatic and/or enteroenteric circulation.

Airway, aspiration and intubation

The 2026 guideline devotes a section to the question that generates most ED conflict — whether to intubate a drowsy overdose patient in order to give charcoal. It adopts four explicit good-practice statements, and they are conditional rather than a blanket answer either way.¹

The reasoning is quantitative. Across three studies of over 2,200 poisoned patients, intubation carried rates of hypotension 1.5–11.8%, desaturation 3.4–7.1% and cardiac arrest 0.4%, with adverse events apparently commoner in children. The risk of aspiration after charcoal given via an already-secured airway is, by contrast, 1–4%.¹

The corollary matters as much as the prohibition. Where intubation is already clinically indicated — a compromised airway, the need for haemodialysis or extracorporeal support, or transfer to another institution — the risk–benefit shifts in favour of giving charcoal, because the procedural risk has already been accepted for another reason.¹

The 2005 position remains true and is worth stating alongside: unless the patient has an intact or protected airway, charcoal is contraindicated.² What 2026 adds is that the fix for an unprotected airway is not automatically a tube.

Contraindications — graded, not conventional

An earlier version of this page listed these as conventional practice with no source, because the 2005 paper gave only the airway rule and there is no UK label. The 2026 guideline grades them explicitly. Note the evidence level behind every one is D — very low; the strength reflects consensus, not data.¹

Practical use in the ED

Dose and administration

Adult

50 g — the dose used in the absorption studies underpinning the recommendations²

Child

1 g/kg, commonly capped at 50 g

Route

Orally if the patient will drink it; nasogastric tube if not and the airway is secure

Palatability

Serve cold, in an opaque cup with a lid and a straw. Most refusals are about the appearance, not the taste

Vomiting

Common. Consider an antiemetic before, not after

Contraindications and cautions

• Unprotected airway in a patient who is not going to be intubated for another reason.¹² Note the guideline's own nuance below — this is a reason not to give charcoal, not automatically a reason to intubate.
• Risk of gastrointestinal perforation (strong against, 1D); absent peristalsis (weak against, 2D). See the graded table above.
• Likely to need upper gastrointestinal endoscopy (weak against, 2D) — which is the properly framed version of the old corrosives rule: charcoal blackens the mucosa and obstructs the view.
• Hydrocarbon ingestion, where the aspiration risk dominates.
• Consider whether an antidote given orally will itself be adsorbed — the classic example being oral acetylcysteine regimens, not used in the UK where the intravenous route is standard.

UK supply

Critical appraisal

Charcoal has always occupied an awkward place: strong pharmacological plausibility, good volunteer data on absorption reduction, and no randomised evidence of benefit on any patient-centred outcome. The 2026 guideline does not resolve that. It is a Delphi consensus built on a systematic review, and its recommendations are expert opinion about appropriateness, not demonstrations of efficacy. Nothing in it shows that a patient given charcoal at five hours does better than one who is not.

What it does do is repair a distortion. The one-hour rule was a pragmatic simplification that became a prohibition, and it led to charcoal being withheld in exactly the situations where it retained most promise — massive ingestions, modified-release preparations, and drugs with slow gut transit. Replacing a single number with 43 poison-specific appraisals is less memorable and more honest.

One conflict-of-interest note worth reading in the original: the author list includes an employee of SERB Pharmaceuticals (Medical Affairs) — a supplier in the antidote market and the marketing authorisation holder for Cyanokit. The paper's disclosure statement reads: "No potential conflict of interest was reported by the authors." Both facts are in the paper, a page apart, and readers can weigh them. It does not invalidate a 22-author, three-society consensus, and the funding statement reports none.

References

1. 1
   Hoegberg LCG, Gosselin S, Buckley NA, Wood DM, Shepherd G, Hanley J, Bates N, St-Onge M, Caravati EM, Smith SW, Shadnia S, Gudjonsdottir G, Jiranantakan T, Johnson J, Olson KR, Bédry R, Eyer F, Tse ML, Chan WL, Stolbach A, Lang E, Hoffman RS. Recommendations from the Clinical Toxicology Recommendations Collaborative on the administration of activated charcoal in acute oral overdose. Clinical Toxicology (Philadelphia) 2026;64(6):419–475. PubMed 41906697 · doi:10.1080/15563650.2025.2609807. Systematic review and two-round modified Delphi across 43 poisons, 22 authors, three sponsoring societies. Full text read 4 September 2026 — source of the dose and interval good practice statements, the graded contraindications, the Table 4 dose thresholds and GRADEs, the Figure 1 algorithm, the four intubation statements, the metformin adsorption-capacity reasoning and the limitations. Per-poison time bands are colour-coded in Table 4 and are not reproduced on this page.
2. 2
   Chyka PA, Seger D, Krenzelok EP, Vale JA. Position paper: single-dose activated charcoal. American Academy of Clinical Toxicology and European Association of Poisons Centres and Clinical Toxicologists. Clinical Toxicology 2005;43:61–87. EAPCCT PDF. The superseded framework. Source of the time–absorption figures and the unprotected-airway contraindication.
3. 3
   National Poisons Information Service. TOXBASE. The UK operational source for poison-specific decontamination advice and charcoal timing. NPIS telephone advice: 0344 892 0111.