Carboxyhaemoglobin (COHb) Interpreter
Carboxyhaemoglobin (COHb) Interpreter
Band a co-oximetry carboxyhaemoglobin percentage from a non-smoker's baseline to a life-threatening level, and see why the number tracks the clinical picture so poorly.
Carboxyhaemoglobin level
1 input → % COHbCarboxyhaemoglobin 18% by co-oximetry in a non-smoker given high-flow oxygen on arrival
How the level is read
Half-life ≈ 4–5 hours on room air, ≈ 1 hour on high-flow oxygen
Hyperbaric oxygen generally considered from a level of at least 25%
- co-oximetry
- the only routine way to measure it — standard pulse oximetry cannot distinguish carboxyhaemoglobin from oxyhaemoglobin and reports a falsely reassuring saturation
- half-life
- roughly 4–5 hours breathing room air and about 1 hour on high-flow oxygen, which is why a level drawn after a long transport understates the exposure
- 25%
- the level from which hyperbaric oxygen is generally considered, alongside clinical criteria such as loss of consciousness, neurological signs, cardiac ischaemia and pregnancy
Worked example
Carboxyhaemoglobin 18% by co-oximetry in a non-smoker given high-flow oxygen on arrival
18.0% lies between 10 and 20 → mild poisoning
Well above the 2–3% non-smoker baseline and above the 8–10% seen in smokers
On high-flow oxygen the half-life is about 1 hour, so the level measured on arrival is lower than the level at the scene
The banding describes the number, not the patient: symptoms, syncope, neurological signs and ECG changes drive management
Carboxyhaemoglobin percentage and typical features
| COHb | Interpretation | Typical features |
|---|---|---|
| < 2–3% | Normal for a non-smoker | Endogenous haem breakdown only |
| Up to 8–10% | Consistent with smoking | Usually asymptomatic |
| 10 – 20% | Mild poisoning | Headache, dizziness, nausea, poor concentration |
| 20 – 30% | Moderate poisoning | Vomiting, breathlessness, impaired judgement |
| > 30% | Severe poisoning | Neurological impairment, syncope, arrhythmias |
| 50 – 60% | Frequently life-threatening | Coma, seizures, cardiovascular collapse |
Why the measured level can mislead
| Situation | Effect on the number |
|---|---|
| Pulse oximetry used instead of co-oximetry | Carboxyhaemoglobin is read as oxyhaemoglobin, so the saturation looks normal |
| High-flow oxygen started before the sample | Half-life falls to about 1 hour, so the level drops quickly |
| Long transport or delayed presentation | A low level does not exclude significant poisoning |
| Smoker | Baseline of up to 8–10% before any exposure |
| Pregnancy | Foetal haemoglobin binds carbon monoxide more avidly, so the foetus is at risk at maternal levels that look modest |
Why a co-oximetry number needs its context
Carbon monoxide binds haemoglobin with an affinity some two hundred times that of oxygen, and the resulting carboxyhaemoglobin both removes oxygen-carrying capacity and shifts the dissociation curve so that the remaining oxygen is released less readily. It must be measured by co-oximetry on a blood gas sample. Standard pulse oximetry cannot distinguish carboxyhaemoglobin from oxyhaemoglobin and reports a falsely reassuring saturation in a patient who is significantly poisoned, which is the single most important pitfall in the assessment.
A non-smoker sits below 2 to 3% from endogenous haem breakdown alone, and smokers commonly run up to 8 to 10%. Between 10 and 20% patients report headache, dizziness and nausea; above 30% expect neurological impairment, syncope and arrhythmias; and levels of 50 to 60% are frequently life-threatening. Those bands describe the number rather than the patient, and the correlation between concentration and both immediate symptoms and delayed neurological sequelae is poor.
The level also moves fast. The half-life of carboxyhaemoglobin is roughly 4 to 5 hours breathing room air and about 1 hour on high-flow oxygen, so a sample taken after a long transport with a reservoir mask in place can be a fraction of the level at the scene. A low result therefore does not exclude significant poisoning, and the time of the sample and whether oxygen had been started must be recorded alongside it. The decision to treat is clinical.
Hyperbaric oxygen is generally considered from a level of at least 25%, with loss of consciousness, neurological signs, cardiac ischaemia, metabolic acidosis and pregnancy weighing alongside the number. Pregnancy is a special case: foetal haemoglobin binds carbon monoxide more avidly than adult haemoglobin, so the foetus is at risk at maternal levels that look modest. Poisoning management is time-critical and directed by a poisons centre or clinical toxicology service, and this interpreter supports that advice rather than replacing it.
Frequently asked questions
What is a normal carboxyhaemoglobin level?
Below 2 to 3% in a non-smoker, from endogenous haem breakdown. Smokers commonly run up to 8 to 10%, so a smoker begins any exposure from a higher baseline.
Why can I not use pulse oximetry to detect carbon monoxide poisoning?
Standard pulse oximetry cannot distinguish carboxyhaemoglobin from oxyhaemoglobin, so it reports a normal-looking saturation in a poisoned patient. Carboxyhaemoglobin has to be measured by co-oximetry on a blood sample.
Does a low carboxyhaemoglobin level exclude poisoning?
No. The half-life is roughly 4 to 5 hours on room air but only about 1 hour on high-flow oxygen, so a level drawn after treatment and transport can be far below the level at the scene. The diagnosis is clinical.
When is hyperbaric oxygen considered?
Generally from a level of at least 25%, and also for loss of consciousness, neurological signs, cardiac ischaemia, severe acidosis or pregnancy. The level alone does not decide it.
Why is pregnancy treated differently, and who decides?
Foetal haemoglobin binds carbon monoxide more avidly than adult haemoglobin, so the foetus is at risk at maternal levels that look modest. Poisoning management is time-critical and directed by a poisons centre or clinical toxicology service, which this tool supports rather than replaces.
Related calculators
References
- Rose JJ, Wang L, Xu Q, et al. Carbon monoxide poisoning: pathogenesis, management, and future directions of therapy. Am J Respir Crit Care Med. 2017;195(5):596–606.
- Weaver LK. Carbon monoxide poisoning. N Engl J Med. 2009;360(12):1217–1225.
- Carbon monoxide poisoning. In: Merck Manual Professional Edition. Merck & Co.