Skip to content
hyperbaric.app
emergencyevidencepoisoning

Hyperbaric Treatment for Carbon Monoxide Poisoning: Who Needs It?

Carbon monoxide poisoning is treated with 100% oxygen first. When hospitals add hyperbaric oxygen, what the trials show and how referral criteria differ.

September 14, 2026
10 min read
by Hyperbaric.app

This content is for informational purposes only and does not constitute medical advice. Always consult a healthcare professional. Read full disclaimer

Carbon monoxide (CO) poisoning is treated first with high-flow 100% oxygen. Hospitals add hyperbaric oxygen therapy (HBOT), breathing pure oxygen in a pressurised medical chamber, for selected patients: typically those who lost consciousness, have neurological or heart symptoms, or are pregnant. The aim is to reduce the risk of lasting memory and brain problems, but the trials disagree, so guidelines differ on exactly who should be treated.

If you think someone has carbon monoxide poisoning

  1. Get everyone out into fresh air now. Stop using the appliance you suspect only if you can do it without delay, and do not go back in.
  2. Call emergency services (for example 911, 112 or 999) if anyone has trouble breathing, chest pain, confusion, weakness or has passed out. Tell them you suspect carbon monoxide.
  3. Get checked even if you feel better outside. The NHS advises that anyone with carbon monoxide poisoning needs a hospital check. In the US, Poison Control (1-800-222-1222) can also advise.

Hyperbaric treatment is arranged by the hospital. Do not try to find a chamber yourself.

How Does Carbon Monoxide Poison You?

Carbon monoxide is a colourless, odourless gas produced when fuel burns incompletely. It binds to haemoglobin, the oxygen carrier in red blood cells, about 250 times more strongly than oxygen does, according to a 2017 review in the American Journal of Respiratory and Critical Care Medicine. The result, carboxyhaemoglobin (COHb), leaves less room for oxygen, so tissues are starved of it.

That is not the whole story. The 2025 clinical policy from the American College of Emergency Physicians (ACEP) describes a cascade of inflammatory and immune damage at the cellular level, with the heart and brain most affected. The Undersea and Hyperbaric Medical Society (UHMS) notes that the COHb level measured in hospital does not correlate with who goes on to develop neurological or cognitive problems.

Common symptoms listed by the CDC are headache, dizziness, weakness, upset stomach, vomiting, chest pain and confusion. Because these can look like flu, CO poisoning is easy to miss, especially in colder weather. Infants, older people and anyone with heart disease, anaemia or breathing problems are more likely to become ill.

How Fast Does Oxygen Clear Carbon Monoxide?

Oxygen pushes CO off haemoglobin, and more oxygen pushes it off faster. The figures below are the approximate half-life of COHb, the time for the level to fall by half. Published estimates vary.

What the patient breathesApproximate COHb half-life
Room airAbout 5 hours (320 minutes)
100% oxygen by mask at normal pressure74 minutes on average, ranging from 26 to 148 minutes, in a study of 93 poisoned patients
Hyperbaric oxygenAbout 20 minutes, though up to 42 minutes has been reported in clinical practice

Room air and hyperbaric figures come from the ACEP policy and the 2017 review; the normal-pressure oxygen figure is from a 2000 study in Chest.

Faster CO removal is not the main argument for HBOT, because everyone already receives 100% oxygen straight away. The debate is about brain injury. UHMS points to animal studies in which hyperbaric oxygen, but not oxygen at normal pressure, improved mitochondrial function, limited lipid peroxidation and reduced white blood cell adhesion to damaged blood vessels. Those are animal findings, which is why human trials matter.

What Are Delayed Neurological Sequelae?

Some people seem to recover, then develop new problems 2 to 40 days after the exposure, according to ACEP. These delayed neurological sequelae range from poor concentration and memory to seizures and a Parkinson's-like syndrome, and can include depression or psychosis. The 2017 review estimates that long-term neurocognitive deficits affect 15% to 40% of patients.

A 2007 analysis of 163 CO-poisoned patients who did not receive HBOT found cognitive problems at six weeks in 68 (42%). Risk was higher in people aged 36 or older and with exposures lasting 24 hours or more, and trial subgroups also pointed to loss of consciousness and COHb levels of 25% or higher.

If memory, concentration, mood or movement problems appear in the weeks after a CO exposure, see a doctor and mention the poisoning.

What Do the Trials Show?

This is one of the most argued-over questions in hyperbaric medicine. Randomised trials have pointed in different directions.

The main positive trial. In a 2002 double-blind trial in the New England Journal of Medicine, 152 patients with symptomatic CO poisoning received either three hyperbaric sessions within 24 hours or one normobaric oxygen session plus two sham chamber sessions. Six weeks later, cognitive problems affected 25.0% of the hyperbaric group and 46.1% of the normobaric group. The difference was still present at 12 months. The trial was stopped early at an interim analysis.

Trials that found no benefit. A 1999 Australian double-blind trial compared daily treatments at 2.8 ATA with the same sessions at normal pressure, with both groups on high-flow oxygen in between. Hyperbaric patients did worse on some neuropsychological tests, and the authors concluded they could not recommend HBOT. In a 2011 French study of 385 patients with domestic CO poisoning, people with brief loss of consciousness had similar full-recovery rates at one month with or without one hyperbaric session (58% vs 61%). Among patients who had been in a coma, two hyperbaric sessions did worse than one (47% vs 68% full recovery).

Systematic reviews. A Cochrane review of six trials with 1,361 participants found two positive and four negative. The pooled result did not show a significant benefit, but the authors stressed that every trial had design or analysis flaws: the positive trials were stopped early, which can inflate effects, and the negative ones had low power or poor follow-up. Their conclusion was that the existing trials "do not establish" whether HBOT reduces adverse neurological outcomes.

The latest US emergency medicine view. ACEP's 2025 policy reviewed studies published since its 2017 policy and found only lower-grade evidence. It concluded that HBOT "may provide a modest benefit, especially in memory impairment," and issued a Level C recommendation (its weakest grade): in symptomatic CO poisoning, "selected patients may benefit from HBO2 treatment based on severity of symptoms and availability (distance and time)." It also noted that trials varied widely in timing, pressure, number of sessions, blinding and whether comatose patients were included, which helps explain the conflicting results.

Who Is Referred for Hyperbaric Oxygen?

There is no single international rule. The main published positions overlap but differ in detail:

SourceWho should be considered for HBOT
European Committee for Hyperbaric Medicine, 2016 consensusAnyone with altered consciousness or neurological, cardiac, respiratory or psychological signs, whatever the COHb level. All pregnant women, whatever their symptoms. Not recommended for people without symptoms seen more than 24 hours after exposure ended.
US expert review, 2017Serious poisoning: loss of consciousness, ischaemic heart changes, neurological deficits, significant metabolic acidosis, or COHb above 25%.
Polish Medical Society toxicology position, 2025Neurological or cardiac symptoms or metabolic acidosis that persist despite normobaric oxygen, regardless of COHb. Always in pregnancy. Ideally within 6 hours.
ACEP clinical policy, 2025Selected symptomatic adults, based on severity and how far away a chamber is. The policy does not cover children, pregnancy or presentations more than 24 hours after exposure.

The common threads are loss of consciousness, neurological signs, heart involvement and pregnancy. Some guidance uses a COHb cut-off such as 25%, while the European consensus bases its advice on symptoms, whatever the COHb level. Timing also matters: UHMS states that the greatest benefit occurs in patients treated with the least delay.

HBOT carries risks too. ACEP lists middle-ear barotrauma, oxygen toxicity seizures, claustrophobia and the dangers of transferring an unwell patient, sometimes more than 50 miles, to reach a chamber. For a mildly poisoned patient, the European consensus considers either 12 hours of normobaric oxygen or HBOT reasonable.

Where Is Emergency Hyperbaric Treatment Given?

Emergency HBOT is given in chambers equipped to look after acutely ill patients, and US public health guidance advises emergency teams to decide on it in consultation with a toxicologist, a hyperbaric facility or a poison centre. Few centres can take these patients. A 2016 survey of US hyperbaric programmes found that only 43 of 361 (11.9%) had the equipment, such as ventilators and intravenous infusion pumps, and staff to treat high-acuity patients. The European consensus lists CO poisoning among its accepted indications with its strongest (Type 1) recommendation; see our overview of hyperbaric oxygen therapy indications.

The trials above used medical chambers at around 2 to 3 ATA. Inflatable "mild" wellness chambers run at much lower pressure (a typical FDA 510(k) summary lists 1.3 ATA, using air, for altitude sickness) and are not a substitute. Our guide to hard-shell vs soft-shell chambers explains the difference.

Our directory can help you see which hyperbaric facilities are near you, which is useful for follow-up questions or for knowing what exists in your region. In an emergency, though, the decision and the transfer belong to the hospital team.

Sources

This article is for information only and is not medical advice. Suspected carbon monoxide poisoning is an emergency: get to fresh air and contact emergency services.

Back to all articles