Winter is coming… Think CO toxicity!

Much has been written on the topic of Carbon Monoxide Toxicity. I will not reiterate that CO is a colorless, tasteless, odorless gas, nor will I restate that it is among the top killers of toxins worldwide. I may tell you, however, that whenever a carbon based material undergoes incomplete combustion, CO is generated. Another tidbit commonly discussed in this context is that dichloromethane (CH2Cl2) ingestion will lead to high serum CO concentrations from metabolic conversion in the liver. There have been countless laboratory, animal and human studies which elucidate aspects of the multifactorial pathophysiology of CO toxicity which show effects from displacement of oxygen, causing hypoxic injury, binding to cytochrome, myoglobin and other proteins causing disruption in cellular respiration or function, lipid peroxidation and inflammatory effects with oxygen free radical generation, etc. most commonly manifesting in neurological, neuropsychiatric and cardiac dysfunction1. I am going to cut to the chase, as this is a blog and not a stuffy scientific manuscript, and discuss the presentation, diagnosis, treatment and outcome of patients with CO exposure and toxicity. Be warned, there is much controversy, a lot of data- some of which is contradictory- and even more anecdotal case studies. There are also medical-legal overtones and regional differences in preferred treatment modalities.

CO 1
Figure 1: Electron Transport Chain and the Effect of CO.

Case: 54 yo woman presenting with headache, nausea and blurred vision for 7-10 days. She stated that her contractor hired to re-work a botched job on her boiler advised her to go the emergency department because the CO level in her house was “sky-high”. After seeing another physician and hours after leaving her house, she presented to the emergency department. The boiler was faulty for about 2 weeks. Her examination was normal. Her labs and ECG were normal. Her COHgb concentration was 4.

Presentation: Predicting which patient has CO toxicity purely based on presenting signs and symptoms is not possible. The most common presenting complaints of nausea, headache, dizziness, and blurred vision are non-specific to CO toxicity and often vague. The diagnosis is easily suspected in the patient arriving to the emergency department from a house fire, in a coma, with no clear injuries (although one must also suspect other toxins related to combustion of synthetic plastics in this scenario, mainly cyanide), the more common presentation is that of someone presenting with symptoms of nausea and headache with a lower level of exposure from a furnace with faulty exhaust system. Since this usually occurs in autumn and winter months and coincides with the increase in flu-like illness, CO toxicity from hidden sources can be overlooked and missed. With this premise, investigators at Brown used a non-invasive device to measure carboxyhemoglobin concentrations using pulse CO-oximetry (SpCO), in all patients presenting to the emergency department at Anderson Emergency Center and showed that, in this cohort of undifferentiated patients, 1 in 10,000 had occult exposure to CO2. Among these patients, some had clear sources of CO exposure upon further history (charcoal grill in the kitchen, etc.) and some had hidden sources (e.g. one patient’s apartment building had a faulty water heater exhaust system with environmental CO levels of 350 ppm, measured by the fire department upon notification from the emergency department, endangering all of the occupants)3. Therefore, I advocate universal screening for CO exposure in all patients (another vital sign). Since the device that measures SpCO coexists with pulse-oximetry and takes an extra 30 seconds to display SpCO, this can easily be incorporated into triage vital signs or even performed in the pre-hospital setting3. We have identified false positive SpCO measurements, therefore, elevated levels should be confirmed with blood COHgb concentration measurements. Since, my academic focus is in disaster medicine, I would be remiss if I did not state that rates of CO toxicity are markedly increased after natural disasters such as ice storms, hurricanes and floods (mainly because of disruption of standard power sources and heating systems. Survivors resort to burning fossil fuels or running generators indoors without proper exhaust systems. I personally witnessed this, when I was deployed with the RI-1 Disaster Medical Assistance Team in the aftermath a severe ice storm in northern New York in 1998).

Diagnosis: Any evidence of exposure to high levels of CO (greater than 80 ppm for 8 hours) coupled with signs and symptoms of toxicity confirms the diagnosis of CO toxicity. Often elevated SpCO or lab confirmed COHgb concentrations are present but are not necessary. Measurement of environmental CO may also be used in determining exposure. Since CO is eliminated from the blood even while breathing ambient air, after a period time, measures of SpCO and COHgb may be misleadingly low. The total body dose of CO is also thought to be important. One may be exposed to low levels of CO for a prolonged period of time, saturating all body compartments (there is no laboratory measurement for tissue carbon monoxide, however there is investigation into markers of CO toxicity), with mildly elevated blood COHgb concentrations, while others may have a short period of exposure to a high concentration of CO with markedly elevated blood COHgb concentrations. The role of [CO concentration exposed*exposure time] area under curve remains understudied (difficult to study since these variables are often elusive in patients presenting with toxicity) and may have important treatment implications. When evaluating carboxyhemoglobin concentration results, there are several factors to consider. One: how long has the patient been breathing fresh air or oxygen (ETT, NRB, nasal cannula) and two: is the patient a smoker? CO is eliminated from the blood with zero or first order pharmacokinetic depending on the oxygen partial pressure5. The “normal” COHgb concentration for smokers differs from non-smokers. It is not clear what role time from last cigarette, length of smoking or amount of smoking is related to the “normal” COHgb. This is an ongoing topic of research in our department. Some textbooks and articles and laboratory norms use 10% as the upper level of normal in this population but this is arbitrarily assigned.

CO 2
Figure 2: CO O2-HgB Dissociation Curve.

Treatment: Once a diagnosis is made, or yet, even while the diagnosis is entertained, treatment with oxygen should be initiated without delay. Typically, initially, this is done with a tight fitting non-rebreather mask with high flow oxygen. Then a decision has to be made regarding the duration and modality of oxygen therapy. There are three options to choose from: 1-Hyperbaric oxygen (HBO) (I will not go into the mechanisms of action of HBO therapy) 2-Inpatient prolonged normobaric oxygen (NBO) and finally 3- ED short-term NBO. Presenting signs and symptoms, patient age, pregnancy and blood COHgb concentrations will guide the decision making process. The controversy surrounding HBO vs. NBO treatment for CO toxicity is long-standing and fierce. While there is data in this domain, including several blinded and placebo controlled trials (which have shown split and contradictory results), there has not been a consensus on treatment options. The following two manuscripts exemplify the controversy and I would advise enthusiasts to read each individual controlled trial to reach their own conclusion6,7. My take on the issue after scouring the literature and taking care of patients with CO toxicity both with HBO and NBO is this: 1- If a patient presents with LOC, confusion, alteration in mental status or any focal neurological findings (including gait or documented vision problems); or if the patient has chest pain, shortness of breath, any ischemic ECG changes or conduction abnormalities or elevated troponin levels; or if the patient is <16 years of age; or pregnant I would suggest HBO regardless of carboxyhemoglobin concentration with documented exposure to CO. 2- If there are other symptoms consistent with CO toxicity (nausea, headache, vague weakness, etc.) and if the history suggests prolonged exposure over days to weeks to CO or if there is significant elevation in COHgb) I suggest admission to the hospital with NBO. NBO in these patients should be administered for for 6-8 hours after all symptoms have resolved. 3- In patients with not so elevated COHgb levels, mild or resolving symptoms with a history that suggests only transient exposure to CO, I suggest ED NBO for 4 hours after symptom resolution (admit to the hospital if symptoms do not resolve). Every patient should have extensive neuropsychiatric evaluation as soon as possible and serially for 6 months after exposure. Now, of course, concomitant trauma, co-ingestion or co-toxicity with other agents such as cyanide should be treated (suicide attempts should prompt further toxicological evaluation as con-ingestion is common). Traumatic injuries, burn or smoke inhalation or other illness may interfere with HBO treatment and in these circumstances NBO treatment should be given and treatment of other life threatening injury or illness should take precedence over HBO. Despite the controversy and conflicting literature, there have been lawsuits brought against Emergency Physicians for failing to treat CO toxicity with HBO. (

Outcome: Patients with CO toxicity are at risk for delayed neurological, psychiatric, neuropsychiatric, and cardiac sequelae days to weeks after exposure. In a cohort of patients treated for CO toxicity with HBO, there was an increased incidence of cardiac morbidity and mortality years after exposure8. The goal of treatment either with HBO or NBO is to prevent these sequelae. Most controversy regarding treatment modalities centers around if HBO is better than NBO in preventing delayed sequelae. All patients with CO toxicity should be referred for close follow-up and regular long-term monitoring by their primary care physician.

Case Outcome:

The patient received NBO treatment with a tight fitting NRB mask. She was admitted to the hospital for prolonged NBO treatment. She received a total 30 hours of NBO and was discharged the following day after symptoms resolved.


  1. Suner S and Jay G. Carbon monoxide has direct toxicity on the myocardium distinct from effects of hypoxia in an ex-vivo rat heart model. (2008). Academic Emergency Medicine 2008; 15(1):59-65.
  2. Suner S, Partridge R, Sucov A, et al. Noninvasive Pulse CO-Oximetry Screening in the Emergency Department Identifies Occult Carbon Monoxide Toxicity.  Emerg Med.2008; 34(4): 441-450.
  3. Chee K.J., Nilson D., Partridge R., Hughes A., Suner S., Sucov A., Jay G. Finding Needles in a Haystack: A Case Series of Carbon Monoxide Poisoning Detected Using New Technology in the Emergency Department Clin Toxicol. 2008 Jun; 46(5):461-9.
  4. Nilson D., Partridge R., Suner S., Jay G. Non-invasive Carboxyhemoglobin Monitoring: Screening Emergency Medical Services Patients for Carbon Monoxide Exposure. Prehosp Disaster Med. 2010;25(3):253–256
  5. Jay GDMcKindley DS Alterations in pharmacokinetics of carboxyhemoglobin produced by oxygen under pressure. Undersea & Hyperbaric Medicine : Journal of the Undersea and Hyperbaric Medical Society, Inc. 1997, 24(3):165-173.
  6. Hampson NB, Piantadosi CA, Thom SR, Weaver LK. Practice Recommendations in the Diagnosis, Management, and Prevention of Carbon Monoxide Poisoning. Am J Respir Crit Care Med 2012; 186(11):1095–1101.
  7. Scheinkestel CD, Jones K, Myles PS, Cooper JD, Millar IL, Tuxen DV. Where to now with carbon monoxide poisoning? Emergency Medicine Australasia (2004) 16, 151–154.
  8. Henry CR, Satran D, Lindgren B, Adkinson C, Nicholson CI, Henry TD. Myocardial Injury and Long-term Mortality Following Moderate to Severe Carbon Monoxide Poisoning. 2006;295(4):398-402.

Graphics and formatting: Dr. Jordan Wolfe

2 thoughts on “Winter is coming… Think CO toxicity!

  1. Nice piece, Selim. CO can be such a deadly, sneaky toxin. One of my high school friends had CO poisoning and long term brain damage from sitting in a car with a tail pipe blocked by snow. Years ago I worked in a sleepy community hospital in upstate NY in the winter, and was awoken during my night shift by a family who were all coming in with “the flu”. I almost missed it, but inquired about their heating, and got COHb levels that were still elevated. More recently, my 21 yo son cleaned ashes out of our woodstove and left the metal bucket in the closed family room. There was still some combustion in the pail, and CO was released. Luckily, the CO alarm in the room blurted out “DANGER CARBON MONOXIDE, LEVEL 75%” My son and his friend were feeling headache and nausea. It cleared rapidly with opening the door. So, beware of the sneaky gas assasin. -BZ

    • Thanks for the personal anecdotes. I did not get into it but obviously all must have CO detectors on every floor in the home. Doesn’t have to be a fancy one. They save lives. One area emergency physicians can make an impact is to effect changes in policy and legislation to mandate these regardless of when the home was built.

Leave a Reply