Pediatric Carbon Monoxide Toxicity

Updated: Dec 28, 2015
  • Author: Samara Soghoian, MD, MA; Chief Editor: Timothy E Corden, MD  more...
  • Print


Carbon monoxide (CO) is a colorless, odorless, tasteless, and highly poisonous gas produced from the incomplete combustion of organic matter, including fossil fuels. It is the most frequent agent of toxic exposure in North America. [1]

See Clues on the Skin: Acute Poisonings, a Critical Images slideshow, to help diagnose patients based on their dermatologic presentations.

According to the Centers for Disease Control and Prevention (CDC), approximately 15,000 visits to emergency departments (EDs) and around 500 deaths are caused by unintentional, non-fire-related carbon monoxide exposures alone each year. [2] Between 2004 and 2006, the highest estimated rate of ED visits for unintentional, non-fire-related carbon monoxide exposure in any age group was for children younger than 5 years (11.6 cases per 100,000). [3]

In a review using data from the National Poison Data System (NPDS), the CDC found that between 2000 and 2009, an average of 23.2 CO exposures per 1 million population (range, 19.7-25.3) were reported annually. [4] CO exposures were slightly more common in females than in males (23.0 versus 20.9 per million). The most commonly exposed age groups were those younger than 17 years (25.7 per million) and those between the ages of 18 and 44 years (19.4 per million). In 2014, the American Association of Poison Control Centers reported 12,478 single exposures, of which 12,068 are listed as unintentional. [5]

Intentional CO poisoning is far more often fatal than unintentional exposure is. Overall, between 1999 and 2012, CO poisoning was the cause of 6,136 unintentional, nonfire-related deaths. CO poisoning was the second highest cause of nonmedical poisoning deaths.{ref5) [2]

Toxic exposures to CO are most frequently the result of house fires or the use of fuel-burning heating appliances or poorly maintained generators. From 2004-2006, the most common cause of unintentional, non-fire-related CO exposures in the United States was home heating systems (16.4%), followed by motor vehicles (8.1%), and the highest percentage of exposure occurred during the winter months from December to February. [3]

CO intoxication may also result from inhaling methylene chloride, a volatile liquid found in degreasers, solvents, and paint removers. Most of the adsorbed vapor is exhaled unchanged, but up to one third is metabolized in the liver to CO.

Because methylene chloride can be stored in tissues, it is released and metabolized gradually; thus, methylene chloride inhalation elevates CO concentrations in blood and tissues for more than twice as long as direct CO inhalation does. Prolonged exposure to methylene chloride (up to 8 hour) can produce CO concentrations in blood that exceed 8%.

People who smoke cigarettes may have baseline carboxyhemoglobin (COHb, or HbCO) concentrations as high as 10%, and their susceptibility to toxic effects from inadvertent exposure to other sources of CO may be heightened.

Acute CO toxicity may cause asphyxia, myocardial dysfunction, and a full spectrum of peripheral and central nervous system (CNS) effects. Symptoms are generally nonspecific and protean. Therefore, if a history of exposure is not given or suspected, this disease is extremely difficult to diagnose.

In fact, CO toxicity is frequently misdiagnosed as a simple headache or viral syndrome. Accordingly, a high index of suspicion must be maintained, particularly during the winter months, when faulty heating systems and enclosed spaces make CO poisoning more common than it is at other times.



Carbon monoxide (CO) exerts its toxic effects through a combination of tissue asphyxia and inflammatory activity. Hypoxia occurs from 3 primary mechanisms: CO diminishes the oxygen carrying capability of hemoglobin, decreases the uptake of bound oxygen into tissues, and impairs the mechanisms of cellular respiration.

CO readily crosses capillary membranes in the lungs and binds the heme moiety on the erythrocyte hemoglobin complex with an affinity 200-300 times greater than that of oxygen. This binding drastically reduces the number of binding spots available for oxygen transport. The amount of oxygen that is able to bind hemoglobin in the setting of CO exposure is proportional to the partial pressure of oxygen (PO2) in respired air and can be increased by giving supplemental oxygen.

CO also shifts the oxyhemoglobin dissociation curve to the left, inhibiting the release of bound oxygen to tissues. In addition, approximately 10-15% of absorbed CO binds to extravascular proteins directly, with a variety of results. For example, CO dissolved in plasma is known to cross capillary membranes and bind myoglobin, reduced cytochromes, guanylate cyclase, and nitric oxide (NO) synthase. This process also decreases the number of binding sites available for oxygen in select tissues, further contributing to hypoxia.

The interplay of all these effects causes tissues in the person exposed to CO to have an oxygen tension lower than that due to simple hypoxia alone. Because of the high affinity of CO for hemoglobin, even low ambient levels of CO can lead to clinically significant toxicity over long exposures.

CO directly impairs aerobic metabolism in tissues by poisoning the mitochondrial electron-transport chain. It does so by binding mitochondrial cytochromes, preventing the binding and subsequent reduction of oxygen at the end of the cycle. The process of oxidative phosphorylation cannot be completed, and the mitochondria, instead of making water and adenosine triphosphate (ATP), make destructive oxygen free radicals.

Tissue hypoxia and oxidative stress account for most of the pernicious effects of CO in the body. Hypoxic stress in patients with CO poisoning is increased because of CO’s effects on mitochondrial electron transport and cellular respiration. Although the acutely toxic effects of CO are primarily due to hypoxia, activation of inflammatory processes plays a major role in CO poisoning, particularly in the development of neurologic damage.

Inflammatory and immune-mediated mechanisms contribute to the development of the systemic inflammatory response syndrome (SIRS) and delayed neurologic sequelae (DNS). Animal models demonstrated that CO causes perivascular changes in the central nervous system (CNS) that cause neutrophil sequestration and activation in the brain.

Reactive oxygen species released by these cells then cause brain lipid peroxidation. Byproducts of peroxidation alter myelin basic protein (MBP) in the presence of CO, affecting immunologic recognition of MBP and starting a cascade of autoimmune activity against cerebral proteins.



Carbon monoxide (CO) is produced by the incomplete combustion of organic matter and fuels (eg, gas, oil, wood, and charcoal). Therefore, fires are the major sources of exposure and toxicity. In 2004, the Consumer Product Safety Commission reported that about 64% of deaths from unintentional CO poisoning were due to exposures in the home. [6] The most common cause of unintentional, non-fire-related CO exposure is malfunctioning household heating appliances used in poorly ventilated rooms.

The incidence of CO poisoning increases after environmental disasters in which heating and electrical systems are destroyed. For example, after hurricanes Katrina and Rita in 2005, 78 cases of nonfatal CO poisoning and 10 deaths were reported in affected counties in Alabama and Texas. Nearly all cases were due to gasoline-powered back-up generators being run outside but near the home’s air conditioner, through which CO was drawn into the home. During hurricane Sandy in the northeastern United States in 2012, federal agencies recorded an increase in CO poisoning, in large part because power outages resulted in improper use of heating equipment, such as generators and grills. [7]

Workers with a high risk of exposure to CO include forklift operators, attendants of underground parking garages, and mechanics. Open-air exposure leading to CO toxicity is not uncommon among motor boat enthusiasts, and it has been reported in children riding in the back of pickup trucks.

Fatal carbon monoxide poisonings due to cooking fumes are reported among climbers and polar explorers. An experimental study in Norway showed that a kerosene camping stove used inside a closed tent for 2 hours raised ambient CO levels enough to cause a mean carboxyhemoglobin (COHb) level of 21.5% and clinically significant hypoxia in healthy volunteers. [8]



Estimates of the frequency of carbon monoxide (CO) exposure vary widely. The scope of the problem is difficult to assess, because patients with mild CO exposure may not seek medical attention and because CO poisoning is frequently misdiagnosed. The Centers for Disease Control and Prevention (CDC) estimated that in 2001-2002, about 15,200 people were treated for unintentional non-fire-related CO exposure in US emergency departments (EDs) annually. [2] Most exposures occur during the winter months.

Age-related demographics

The very young and very old are the groups most susceptible both to CO exposure and to the pernicious effects of exposure.

The CDC reported that in the United States in 2001-2003, children younger than 4 years had the highest incidence of unintentional CO exposure but the lowest death rates from carbon monoxide poisoning. [2] The risk of death from CO poisoning increased with age. The case-fatality rate was 0.6% for children younger than 4 years, which increased to 5.5% in adults aged 55-64 years. In 2001-2003, 23% of all deaths from CO poisoning occurred in adults older than 65 years. [2]

Human and animal data suggest that CO is both teratogenic and strongly associated with fetal death. Neonates and in utero fetuses are most vulnerable to CO poisoning, for a number of reasons. Fetal hemoglobin binds CO with greater affinity than adult hemoglobin does, resulting in increased tissue hypoxia at similar carboxyhemoglobin (COHb) levels. The fetus also has a low baseline oxygen tension (PO2), and fetal COHb levels at equilibration are 10-15% higher than maternal levels.

Sex-related demographics

CDC analysis of data from the National Vital Statistics Program and National Electronic Injury Surveillance System All Injury Statistics for 2001-2003 showed that males and females were equally likely to seek care in the ED for non-fire-related carbon monoxide exposure and poisoning. [2]

However, males were 2.7 times more likely than females to die of CO toxicity, with a case-fatality rate increased by 2.3 times (where the rate was defined as the number of CO-related deaths divided by the sum of CO exposures times 100). [2] Reasons for this discrepancy are not clear.

Race-related demographics

All races are equally susceptible to the physiologic effects of CO poisoning, but non-English-speaking immigrants and minority racial groups are at increased risk of exposure. For example, the unintentional death rate related to CO is 20% higher among African Americans than among whites.

The difference is primarily due to economic disparities among racial and ethnic groups in the United States. Members of economically disadvantaged social groups are most likely to be involved in house fires, to have malfunctioning or no CO detectors in the home, to have faulty indoor heating systems, and to work in high-risk settings.


Clinical Presentation

Patient history

Carbon monoxide (CO) has been called the great imitator because of the protean symptoms it produces. As a result, CO poisoning is frequently misdiagnosed.

Patients may complain of any number of vague symptoms. In one series, headache was the most frequent complaint (37.5%), followed by dizziness (18%), nausea (17.3%), loss of consciousness (7.7%), shortness of breath (6.7%), and loss of muscle control (3.5%).

Symptoms may be attributed to a viral syndrome, migraine or tension headache, anxiety attack, hyperventilation syndrome, or a nonspecific illness. Contemporaneous development of symptoms among several persons from the same location should alert the clinician to the possibility of CO exposure.

Unless the patient is brought in from the scene of a fire, a high index of suspicion must be maintained to make the diagnosis. An important clue is the finding of similar complaints among people who work or live together, particularly during the winter months, when heaters are on and when windows tend to be closed.

Anyone working with combustion engines or combustible gasses indoors should be considered to be at high risk. CO poisoning should also be considered in patients presenting with vague somatic complaints after a natural disaster, when generator use is common.

If CO poisoning is suspected or diagnosed, attempt to determine the source, the duration of exposure, the amount of time elapsed since the patient was withdrawn from the source, and the occurrence of any neurologic symptoms (eg, syncope, seizure, altered mental status, vertigo, or focal neurologic deficits).

Physical examination

Although burns, singed facial hair, and oropharyngeal soot clearly suggest CO exposure, there are few physical findings specific to CO poisoning.

The classic cherry-red skin and retinal discoloration is seen only with the most severe cases of CO poisoning and is generally a postmortem finding. The skin is most likely to be pale, cyanotic, or mottled because of hypoxia and cardiac depression. Bright red retinal veins are a relatively sensitive and early but often overlooked finding in moderate-to-severe cases of CO poisoning. Ophthalmic examination may reveal flame-shaped retinal hemorrhages or papilledema.

CO causes myocardial depression and dysrhythmias. Animal models suggest that shock, if present, is more likely due to vasodilation.

Vital signs may reflect tachypnea, hypoxia, tachycardia, hypotension or hypertension, and mild hyperthermia. Pulse oximetry may remain in the normal range despite cyanosis and tissue hypoxia because the wavelengths produced by carboxyhemoglobin (COHb) and oxyhemoglobin are read similarly by these machines. CO poisoning typically produces a pulse oximetry gap (ie, the difference between the pulse oximeter reading and the spectrophotometrically measured oxyhemoglobin saturation).

Rales may be a sign of noncardiogenic pulmonary edema.

Neurologic and neuropsychologic symptoms frequently occur in the setting of acute CO toxicity and are the most frequent long-term consequence of poisoning. Severe cases of CO poisoning are often characterized by serious neurologic abnormalities, including low Glasgow Coma Scale (GCS) scores and seizures.

Overall, memory disturbances, including both anterograde and retrograde amnesia, are the most common neurologic abnormalities. Other signs include lethargy, stupor, coma, gait disturbance, movement disorders, apraxia, agnosia, tics, vestibular dysfunction, hearing and visual loss, rigidity, brisk reflexes, emotional lability, frank psychosis, and impaired judgment and cognitive function.


Differential Diagnosis

The differential diagnosis of pediatric carbon monoxide (CO) toxicity includes the following:

  • Acute abdomen
  • Acute chest syndrome
  • Alcohol intoxication
  • Altered mental status
  • Anxiety
  • Cardiac arrhythmia
  • Cyanide toxicity
  • Drug abuse
  • Drug overdose
  • Hyperventilation syndrome
  • Meningitis
  • Migraine
  • Psychiatric conditions (eg, delirium, acute psychosis)
  • Pulmonary embolism
  • Shock lung
  • Vascular headache
  • Viral syndrome

Laboratory Studies

A number of laboratory studies may be helpful in working up patients with carbon monoxide (CO) poisoning.

Arterial or venous blood gas evaluation is indicated. The cornerstone of diagnosis is measurement of the blood carboxyhemoglobin (COHb) concentration by means of CO-oximetry. However, levels measured in the emergency department (ED) are not well correlated with extent of exposure, symptoms, or morbidity and mortality. COHb concentrations in the normal or undetectable range rule out exposure or poisoning.

Most measurements done in the hospital setting are based on direct spectrophotometric measurement of COHb concentrations using specific blood gas analyzers. Venous samples are adequate because venous COHb concentrations accurately reflect arterial concentrations.

Handheld, noninvasive CO-oximetry monitors have been tested and appear to be generally accurate. [9] As they become more widely available in both prehospital and ED settings, the use of specific COHb values for the diagnosis and risk stratification of patients with CO poisoning may change.

Infants with persistent fetal hemoglobin may have falsely elevated COHb measurements. Fetal hemoglobin may remain as high as 30% of total hemoglobin at age 3 months.

The partial pressure of oxygen (PO2) should remain normal after CO exposure. However, oxygen saturation may be falsely elevated if it is calculated from the PO2, as is common with many blood gas analyzers, rather than directly measured.

Arterial blood gas values are also used to assess the patient’s acid-base status and help guide resuscitation efforts. Lactate and base deficit may be correlated with duration of exposure and resultant cellular hypoxia. However, the authors know of no studies conducted to examine their prognostic values of these variables.

Unexplained metabolic lactic acidosis suggests cyanide exposure. Measurement of methemoglobin levels is also indicated in the setting of cyanosis with a low oxygen saturation but a normal PO2.

A complete blood count (CBC) should be obtained to evaluate the hemoglobin concentration. Anemia further reduces total arterial oxygen content.

A complete metabolic panel allows the clinician to calculate the anion gap in patients with acidosis and to assess renal function in moderate-to-severe cases that may be complicated by rhabdomyolysis.

Urinalysis and a serum creatine kinase (CK) determination should be ordered to assess the extent of muscular damage and to rule out rhabdomyolysis.

In addition to basic laboratory tests, cardiac enzyme measurements should be performed when patients have chest pain, risk factors for myocardial infarction, or notable CO exposures. The incidence of ischemic cardiac insult after CO poisoning is high, even in young, healthy patients.

Baseline coagulation parameters should be evaluated in severely poisoned patients at risk for systemic inflammatory response syndrome (SIRS) with multiple organ dysfunction syndrome (MODS) and disseminated intravascular coagulation (DIC).

Measurement of the ethanol level and urine toxicologic screening may be useful in ruling out other causes of altered mental status.


Imaging Studies

Patients with evidence of hypoxia or any respiratory embarrassment should undergo chest radiography to evaluate for other causes of respiratory impairment. Changes such as a ground-glass appearance, perihilar haze, peribronchial cuffing, and intra-alveolar edema imply a worsened prognosis.

Computed tomography (CT) of the head may reveal hypoattenuation of the globus pallidus and white matter within hours of carbon monoxide (CO) poisoning. Positive CT scan findings are generally predictive of neurologic complications.

Magnetic resonance imaging (MRI) is more sensitive than CT scanning but is difficult to perform on an emergency basis. Neither CT nor MRI yields findings specific for CO poisoning.

Positron emission tomography (PET) and single-photon emission CT (SPECT) are the most sensitive tests for ischemic brain injury, but the findings are nonspecific, and the studies are even more difficult to perform than MRI is.


Other Tests

Electrocardiography (ECG) should be performed in all patients with notable carbon monoxide (CO) exposure or with risk factors for acute myocardial infarction. Sinus tachycardia is the most common abnormality. Arrhythmias may occur secondary to hypoxia, ischemia, or infarction. Myocardial injury may exist in children with CO poisoning even in the absence of abnormal ECG findings. [10]

Acute myocardial infarction may occur even with low levels of CO exposure in patients with cardiovascular disease. Myocardial infarction is common among patients with moderate-to-severe CO poisoning.


Treatment & Management

Failure to diagnose carbon monoxide (CO) poisoning, thereby potentially allowing the patient to return to a source of toxic exposure, is the primary pitfall of care.

Failure to explain the risk of delayed neurologic problems, even in cases of apparently mild or asymptomatic exposure, may be an issue.

Failure to transfer a patient with moderate-to-severe CO poisoning or a history of neurologic impairment after exposure may be considered failure to meet the standard of care, even if the use of hyperbaric oxygen (HBO) therapy is controversial.

Prehospital care

Patients should immediately be removed from the source of CO exposure and given supplemental high-flow oxygen by means of a nonrebreather face mask. They should be kept calm and still to avoid exertion; increased oxygen demand exacerbates symptoms. Comatose patients and patients with severely altered mental status should be intubated for airway protection.

Cardiac monitoring should be started as soon as possible because of the high incidence of dysrhythmias and cardiac arrest. If possible, emergency medical system (EMS) personnel should try to estimate the total time of exposure and the time elapsed since the patient was removed from the source.

Emergency department care

Attention to the ABCDs of resuscitation is the mainstay of emergency care for the patient with CO intoxication.

Obtunded, comatose, or severely hypoxic patients should be intubated for airway protection. All patients with suspected or confirmed CO exposure should be given 100% oxygen until they are asymptomatic and the carboxyhemoglobin (COHb) concentration is below 10%.

Cardiac monitoring should be started immediately, and 12-lead electrocardiography (ECG) should be performed as soon as possible.

Pulse oximetry readings may be falsely elevated in the setting of COHb because light absorption is nearly the same for COHb as for oxyhemoglobin. Arterial or venous blood gas analysis with CO-oximetry should be done to measure the COHb concentration directly, to determine the degree of hypoxia, and to monitor the patient’s acid-base status.

The half-life of COHb is about 320 minutes (5.3 hours) while the person is breathing room air. This decreases to 30-90 minutes with 100% oxygen and falls to 15-23 minutes at 2.5-2.8 atm with 100% oxygen. These numbers can be used to estimate the duration of treatment for particular patients.

If mild symptoms do not resolve or if severe symptoms are present, HBO therapy should be strongly considered. Specific indications for HBO therapy include a history of seizure or syncope, coma, altered mental status or confusion, an abnormal neurologic examination (particularly if any cerebellar signs are present), a COHb level higher than 25%, or fetal distress in pregnancy.

However, a Cochrane Database of Systematic Reviews study that examined randomized trials comparing HBO therapy with normobaric oxygen treatment concluded that the existing data were not sufficient to determine whether administration of HBO reduces the incidence of adverse neurologic outcomes. [11] Additional studies are needed to better define the use of HBO therapy in the treatment of adult, nonpregnant patients with acute CO poisoning.

Caution should be exercised in treating acidosis because low pH shifts the oxyhemoglobin dissociation curve to the right, increasing oxygen uploading to tissues. Acidosis should improve with oxygenation. Cyanide poisoning should be suspected in cases of severe or recalcitrant acidosis. A serum lactate concentration higher than 8 mmol/L should immediately raise the suspicion of cyanide toxicity. If concomitant cyanide and CO toxicity is suspected, treat the patient with sodium thiosulfate alone.

The methemoglobinemia produced by amyl nitrite also shifts the oxyhemoglobin curve to the left, worsening hypoxia at the tissue level.

Special concerns in pregnancy

Particular caution must be exercised when one treats a pregnant patient with potential CO exposure. Although the mother may appear well, the developing fetus is at risk for hypoxia, even with nontoxic maternal COHb levels.

CO shifts the oxygen-hemoglobin dissociation curve to the left. In fetuses, this effect is even more pronounced than in adults. In addition, fetal hemoglobin binds CO with more avidity that adult hemoglobin does, and the normal partial pressure of oxygen (PO2) is lower in the fetal circulation than in adult circulation. These factors all make the fetus more vulnerable to hypoxia than children and adults. HBO therapy should be strongly considered for pregnant patients.

In the pregnant patient, the lag time for uptake and elimination of CO between the mother and the fetus is considerable. Fetal COHb levels change little during the first hour of maternal intoxication, then increase slowly over the first 24 hours. Fetal COHb levels may peak after maternal levels decline.

The half-life of fetal COHb is 7-9 hours during washout with room air. Maternal supplementation with 100% normobaric oxygen reduces the half-life to 3-4 hours. The half-life of fetal COHb during HBO treatment is not known.


Patients with moderate-to-severe cases of CO poisoning should be admitted to a medical intensive care unit (ICU). A cardiologist should be consulted when patients have evidence of cardiac compromise. A neurologist should be consulted, at least for patient follow-up; delayed neurologic symptoms are relatively common.

Consultation for HBO therapy may be warranted; there is evidence suggesting (though not proving [11] ) that this approach does improve long-term neurologic outcome. [12, 13] If the patient has any mental status changes or a history of neurologic impairment, an immediate consultation for HBO therapy should be sought. This may require transfer to another center after the patient’s condition is stabilized.

Although the use of HBO therapy in preventing mortality from CO poisoning is still debated, it is now the standard of care for moderate-to-severe CO poisoning in patients with neurologic impairment, acidosis, severe hypoxia, myocardial dysfunction, or systemic inflammatory response syndrome (SIRS); in pregnant patients with symptomatic poisoning; and in pregnant patients with asymptomatic poisoning whose COHb levels are higher than 15%.

The nearest HBO therapy center can be located by calling the Divers Alert Network (DAN) at 1-800-446-2671 or 1-919-684-2948 (Monday-Friday, 9 AM to 5 PM [Eastern time]). For nonemergency medical questions, call 1-919-684-2948; for emergencies, call 1-919-684-8111.

Further inpatient care

Patients with carbon monoxide (CO) poisoning should be admitted to the hospital if they have persistent mild symptoms, if they have any history of neurologic impairment (syncope, seizure, amnesia, unresponsiveness) after exposure, if they have risk factors for or evidence of acute coronary syndrome (ACS), or if admission is needed for other reasons.

Most patients requiring admission should be monitored in a telemetry or intensive care setting. Asymptomatic pregnant patients may not require admission, but they should be observed for a period of fetal monitoring.

Admitted patients should be watched for development of the following syndromes:

  • Disseminated intravascular coagulation (DIC)
  • SIRS or multiple organ dysfunction syndrome (MODS)
  • Acute tubular necrosis (ATN)
  • Interval CO syndrome, including leukoencephalopathy of the subcortical white matter with ischemic damage to basal ganglia and hippocampus

Although severe CO poisoning may result in any of the postinjury syndromes listed above, patients with concomitant trauma, burns, intoxication, inhalation injury, or serious premorbid illness are at increased risk.

Further outpatient care

Asymptomatic patients with COHb concentrations lower than 10% may be discharged home after observation. Patients with only mild symptoms may be safely discharged home after 4 hours of treatment with 100% oxygen if their symptoms completely resolve in that time. A physician should reevaluate all discharged patients within 24-48 hours because symptoms may recur or be delayed.

Follow-up must be ensured because delayed sequelae are relatively common. All discharged patients must be warned that some patients with no symptoms and low CO levels may still have interval CO injury. Delayed neurologic symptoms typically occur within 1-2 weeks of the initial exposure but are possible as long as 1 month afterward. Improvement of neurologic function with HBO therapy given as long as 20 days after injury has been reported.


Rates of both intentional and unintentional CO poisoning have declined precipitously in the United States since the enactment of improved vehicle-emissions policies in the 1970s. Unintentional CO exposure in the home is by far the most commonly reported cause of poisoning.

Important precautions include the following:

  • CO detectors with audible alarms should be installed in all homes, and the batteries should be changed regularly
  • Before buying a CO alarm, check to ensure it is listed with Underwriter’s Laboratories (UL) standard 2034 for Single and Multiple Station Carbon Monoxide Detectors, or check the package or owner’s manual to ensure that the detector and alarm meets the requirements of the International Approval Services (IAS) 6-96 standard
  • Carbon monoxide detectors and alarms are also available for boats and recreational vehicles; the Recreation Vehicle Industry Association requires CO detectors and alarms to be installed in motor homes and in recreational vehicles that have or that are outfitted for a generator
  • All fuel-burning appliances should be properly installed, maintained, and operated; a qualified technician should inspect furnaces, water heaters, and gas pipes each year, and fireplace chimneys and flues should also be cleaned and checked yearly
  • Unvented fuel-burning space heaters should be used only if someone is awake to monitor them and only if windows or doors are slightly opened to allow ventilation of the space
  • Automobile exhaust systems should be inspected regularly for defects, and tailpipes should be examined for blockages (which are especially common in the winter, when snow may accumulate and become impacted in them)
  • Vehicles or fuel-burning appliances should never be left running in an enclosed space or outside an open window where exhaust can be drawn into an enclosed space
  • A charcoal grill, hibachi, lantern, or camp stove should never be used inside a tent or camper


Variations in clinical severity, laboratory values, and outcomes all limit prognostic accuracy.

Cardiac arrest, coma, metabolic acidosis, and extremely high carboxyhemoglobin (COHb) levels are associated with poor outcomes. Abnormalities on computed tomography (CT) are associated with persistent neurologic impairment. Approximately 480 unintentional deaths and 2000 suicides are due to carbon monoxide (CO) poisoning each year in the United States.

Certain groups are more susceptible than others to the toxic effects of CO. Morbidity and mortality risks are increased in fetuses; infants; young children; people older than 65 years; people who smoke; and patients with heart disease, pulmonary disease, or anemia.

In one study, 37% of patients treated for moderate-to-severe CO poisoning with hyperbaric oxygen (HBO) therapy nonetheless had acute myocardial injury. [14] Patients were relatively young (mean age, 47.2 years), and rates of previous cardiovascular disease were low (6.5% for previous myocardial infarction, 2.6% for a previous revascularization procedure, and 3% for a history of congestive heart failure). About 38% of patients with acute myocardial infarction after CO poisoning died (24% overall) at a median follow-up of 7.6 years.

More than 50% of patients with severe CO poisoning develop encephalopathy within 1 month of injury.


Patient Education

Patients who smoke should be counseled to avoid smoking for 1-2 weeks after exposure because additional carbon monoxide (CO) from tobacco smoke may increase the risk of toxicity and sequelae.

Although exercise increases hypoxic stress in patients with acute CO poisoning, exposure to CO alone is not an indication for bed rest.

For additional information on CO toxicity, see the Web sites of the National Capital Poison Center and the American Association of Poison Control Centers.