Abstract
Introduction
Carbon monoxide (CO) poisoning is a significant public health concern in the UK, primarily resulting from the unintentional exposure to CO, a colourless, odourless and highly toxic gas produced by incomplete combustion of carbon-containing fuels. The majority of cases in the UK occur in domestic settings, with faulty boilers and inadequately ventilated gas appliances being the most frequently implicated sources (Gentile et al., 2022). Seasonal trends are notable, with higher incidence in autumn and winter due to increased use of these heating systems (Ghosh et al., 2015).
Although the annual number of fatal unintentional non-fire-related CO poisonings in England and Wales has declined over recent decades, deaths and hospital admissions persist, particularly among older adults and males, and often in private dwellings, vehicles or outbuildings (Ghosh et al., 2015).
Clinical presentation is variable, ranging from mild symptoms such as headache and nausea to severe neurological and cardiac complications, and diagnosis is frequently delayed due to non-specific symptoms (Chenoweth, 2016) and clinicians often can confute the diagnosis with a viral infection given the similarities and high prevalence in autumn/winter.
Who gets carbon monoxide poisoning?
Males are disproportionately affected, accounting for approximately 64–82% of fatal and non-fatal cases, with a particular predominance in deaths occurring outside the home, such as in vehicles, garages or outbuildings (Close et al., 2022). Older adults, especially those aged over 80 years, have the highest rates of hospital admission and fatality, with risk increasing steadily with age (Ghosh et al., 2015).
Socioeconomic deprivation and rurality are significant risk factors: individuals living in highly deprived or rural areas experience higher rates of hospital admission for unintentional non-fire-related CO poisoning (Roca-Barceló et al., 2020). Ethnic minority populations, particularly those residing in areas with higher proportions of Asian or Black residents, also show increased standardised risk of hospitalisation (Roca-Barceló et al., 2020) although the underlying reason for this is not clear from the literature.
Most exposures occur in domestic settings, with faulty boilers and inadequate ventilation being common sources, and incidents are more frequent in autumn and winter due to increased use of heating appliances (Close et al., 2022)
What symptoms does carbon monoxide poisoning cause?
The clinical spectrum of CO poisoning is broad and can involve multiple organ systems. Early symptoms often include headache, dizziness, weakness and nausea, but, as exposure increases, patients may develop confusion, disorientation, visual disturbances and presyncope/syncope. Cardiovascular manifestations such as tachycardia, tachypnea, angina, arrhythmias and pulmonary oedema are common, particularly in those with underlying cardiac disease, and these may progress to hypotension and cardiovascular collapse in severe cases (Ernst and Zibrak, 1998).
Neurological involvement is prominent with seizures, coma and collapse occurring in severe poisoning. Notably, delayed neuropsychiatric sequelae including cognitive impairment, memory loss, gait disturbance, urinary incontinence and movement disorders such as parkinsonism can develop days to weeks after the initial exposure especially in older adults (Rose et al., 2017). Chronic or subacute exposures may result in subtle neurobehavioural deficits, such as reduced attention, impaired manual dexterity and decreased visual perception, even at lower carboxyhaemoglobin levels (Raub and Benignus, 2002)
Other systemic effects include muscle necrosis, acute renal failure, skin lesions, and changes in auditory and visual perception. The classic finding of cherry-red lips is rare and not a reliable diagnostic sign (Ernst and Zibrak, 1998). In pregnant patients, there is increased risk for both maternal complications and foetal injury, including foetal death and developmental disorders (Raub et al., 2000)
The variability and non-specificity of symptoms, especially in the context of seasonal viral illnesses, underscore the importance of maintaining a high index of suspicion for CO poisoning in at-risk populations and environments.
What investigations for carbon monoxide poisoning?
The diagnosis is established by a combination of clinical suspicion, exposure history and measurement of carboxyhaemoglobin levels. The primary investigation is the measurement of carboxyhaemoglobin in blood, which can be performed on either venous or arterial samples; venous sampling is generally sufficient for diagnosis, while arterial sampling may be used if assessment of acid–base status is also required (Ernst and Zibrak, 1998).
Non-invasive pulse CO-oximetry devices are increasingly used in emergency settings for rapid screening, but confirmatory diagnosis should rely on laboratory co-oximetry, especially in cases with high clinical suspicion or when non-invasive results are equivocal (Koyuncu, 2020). Expired air CO measurement and ambient air sampling at the scene can support the diagnosis, particularly in prehospital or mass exposure scenarios, but are not definitive (Moss et al., 2025).
Additional investigations include electrocardiography and cardiac biomarkers to assess for myocardial injury because CO poisoning can precipitate cardiac ischaemia or arrhythmias. In cases of altered mental status or focal neurological deficits, neuropsychological testing may be performed to document cognitive impairment, but neuroimaging is reserved for ruling out alternative diagnoses rather than confirming CO poisoning
What is the treatment for carbon monoxide poisoning?
The recommended treatment for CO poisoning is immediate administration of 100% oxygen, preferably via a non-rebreather reservoir face mask or (if necessary) through an artificial airway. This intervention accelerates the dissociation of CO from haemoglobin, thereby reducing tissue hypoxia and the half-life of carboxyhaemoglobin (Ernst and Zibrak, 1998). Oxygen therapy should be continued until the carboxyhaemoglobin level falls below 5% and the patient is asymptomatic (Weaver, 2009). High flow oxygen must be administered irrespective of normal SpO2 readings because pulse oximetry cannot distinguish carboxyhaemoglobin from oxyhaemoglobin, giving falsely reassuring results. CO binding to haemoglobin both reduces oxygen carrying capacity and impairs oxygen release to tissues by shifting the oxyhaemoglobin dissociation curve to the left. Administering 100% oxygen competitively displaces CO from haemoglobin, dramatically reducing its half-life, and must be continued until COHb levels on venous blood gas/arterial blood gas (VBG/ABG) have normalised.
Hyperbaric oxygen therapy (HBOT) is considered for patients with severe poisoning, such as those with loss of consciousness, persistent neurologic deficits, cardiovascular dysfunction, severe metabolic acidosis or markedly elevated carboxyhaemoglobin levels (commonly higher than 25%), as well as pregnant patients due to foetal risk, but the evidence for long-term benefit remains mixed (Weaver, 2009). HBOT should ideally be initiated within 6 hours of exposure in eligible patients however this therapy is not possible for patients who are being mechanically ventilated due to practicalities.
Supportive care includes monitoring for cardiac ischaemia with ECG and cardiac biomarkers because CO poisoning can precipitate myocardial injury. Airway protection and ventilatory support are indicated for patients with altered mental status or respiratory compromise. There are no approved antidotes or pharmacologic therapies beyond oxygen and supportive care.
The source of exposure must be identified and eliminated to prevent recurrent poisoning. All exposed individuals should be evaluated for symptoms and treated as indicated (Weaver, 2009).
What should front-line clinicians know?
In summary, front-line clinicians should recognise that CO poisoning is a common, potentially fatal and often underdiagnosed condition that typically arises from domestic exposures, especially faulty boilers and poorly ventilated appliances, with a marked increase in cases during colder months.
The clinical presentation is highly variable and non-specific, ranging from headache, dizziness and nausea to confusion, chest pain, syncope and coma; symptoms may mimic viral illnesses, which can delay diagnosis particularly in the winter months. Delayed neurological sequelae, including cognitive impairment and movement disorders, can develop days to weeks after exposure. This can often be confounded and mistaken for persistent viral illness.
Diagnosis requires a high index of suspicion and confirmation by blood carboxyhaemoglobin measurement using laboratory co-oximetry because standard pulse oximetry is unreliable. Carboxyhaemoglobin levels do not reliably correlate with symptom severity or prognosis, and so clinical assessment and exposure history remain paramount (Gentile et al., 2022)
Immediate treatment is 100% oxygen via non-rebreather mask or artificial airway, which accelerates CO elimination and should be continued until carboxyhaemoglobin is lower than 5% and the patient is asymptomatic. Cardiac monitoring and investigation for myocardial injury are essential because CO poisoning can precipitate ischaemia and arrhythmias.
Clinicians should ensure the source of exposure is identified and eliminated, and all potentially exposed individuals are assessed. Public health measures, including patient education on CO alarms and appliance maintenance, are critical for prevention (Weaver, 2009).
