Patient presentation:
A family of 4 (ages 46, 5, 15 and 17) presented to the Emergency Department for an “unknown gas leak” at their apartment complex. EMS reported 2 elderly patients who were reported dead in their apartment. The patients presented with flu-like symptoms as well as a headache. In the ED, 2 of the 4 were febrile and 3 of the 4 patients were tachycardic but all had normal oxygen saturations on room air without evidence of respiratory distress.
Pathophysiology of Carbon Monoxide Poisoning:
Normal carbon monoxide (CO) levels in healthy nonsmokers are around 1%. The affinity of hemoglobin for CO is over 200 times higher than that of oxygen, thus CO binds with greater affinity than oxygen to hemoglobin to form carboxyhemoglobin (COHb) which reduces the carrying capacity of oxygen in the blood. This causes tissue hypoxia and increases inflammatory mediators that can injure the heart and CNS.
Presenting Symptoms:
Most pulse oximeters cannot differentiate the wavelengths of hemoglobin bound by oxygen and carbon monoxide, thus a normal pulse oximeter saturation should not rule out the diagnosis. There are however pulse - carbon monoxide oximeters that are only useful as a screening test.
C0Hb levels have been classically reported with their corresponding symptoms as seen in the table below. However, many reviews have shown that carboxyhemoglobin levels do not have precise correlation with symptoms and level of poisoning.
Treatment:
Treatment is based on symptoms and high risk versus low risk exposures:
In patients with CO poisoning, only mild symptoms, and no high risk features patient’s should be put on 100% oxygen until symptoms have resolved, COHb levels are <3%, and they should be monitored for at least 6 hours. Patients with no high risk features who have resolution of symptoms with 100% oxygen treatment can be discharged as long as the CO poisoning was deemed not intentional.
Hyperbaric oxygen should be considered in patients with high risk features which include loss of consciousness, seizure, coma, dysrhythmia, severe metabolic acidosis, focal neurological deficits, COHb level >25%, and pregnancy COHb levels >15%.
Prognosis:
There is a feared complication that delayed neuropsychiatric syndrome sequelae (DNS) can occur in cases of delayed treatment for carbon monoxide poisoning. Although the mechanism for this is not understood, it has been proposed that lipid peroxidation may play a role that ultimately functions similarly to ischemia-reperfusion injury.
DNS may present on average at 20 days post-exposure from carbon monoxide poisoning, however can present anywhere from 3 to 240 days after apparent recovery from patients with significant carbon monoxide exposure. The typical presenting symptoms of DNS can range from personality changes, movement disorders, cognitive effects or focal neurologic findings. Of note, levels do not precisely correlate nor do they predict DNS.
Patient Outcome:
In this specific case, the carboxyhemoglobin levels in our patients aged 46, 17, 15 and 5 respectively were 9.4, 3.8,5 and 2.1. All 4 patients were put on 100% 02 via non rebreather with improvements in their carboxyhemoglobin levels within 3 hours (all three children with levels < 1 and the mother with a level of 1.6). They all had normalization of vital signs, and resolution of symptoms making the symptoms most likely due to carbon monoxide poisoning rather than a viral syndrome, however given the fevers in the children and lack of viral swabbing we cannot say for sure if there was not also a viral syndrome at play. Patients were discharged and went to stay at a family member’s home.
Take Home Points
Symptoms of carbon monoxide poisoning are often vague and similar to the flu, in the winter season when heaters are being used and flu is common make sure to ask your patients who present for “flu like symptoms” about home heating systems and other sources of heat or cooking (kerosene heaters, charcoal grills, camping stoves) poorly ventilated garages, use of power washers, and gasoline-powered electrical generators.
It’s important to remember that pulse oximetry cannot be used to screen for carbon monoxide poisoning because it does not differentiate carboxyhemoglobin from oxyhemoglobin. You need to obtain serum carboxyhemoglobin levels.
Other labs that can help determine CO poisoning, or should make you think of CO poisoning in the right clinical situations, are an unexplained anion gap metabolic acidosis, elevated lactate, and elevated troponin without chest pain symptoms. Consider concomitant cyanide poisoning given the situation the patient was in (i.e. residential or industrial fires).
Carboxyhemoglobin levels are useful for determining exposure, but they correlate imprecisely with the degree of poisoning and symptoms. Treatment should be based on high risk features and symptoms.
Considering CO poisoning, making the diagnosis, inquiring about the use of home CO detectors, educating patients about the source and severity of poisoning, and consulting your regional poison center for advice in management are critical actions for the Emergency Physician.
Written by:
Dr. Taylor Wahrenbrock, MD
Emergency Medicine, PGY-1
Cook County Health
Dr. Jennifer Lee MD
Emergency Medicine, PGY-4
Cook County Health
Reviewed by:
Dr. Sean Bryant, MD
Toxikon Consortium
Cook County Health
References:
Dolan, Michael C., et al. "Carboxyhemoglobin levels in patients with flu-like symptoms." Annals of emergency medicine 16.7 (1987): 782-786.
Hampson, Neil B., and Susan L. Dunn CHRN. "Symptoms of carbon monoxide poisoning do not correlate with the initial carboxyhemoglobin level." Undersea & Hyperbaric Medicine 39.2 (2012): 657
Maloney GE. Carbon Monoxide. In: Tintinalli JE, Ma O, Yealy DM, Meckler GD, Stapczynski J, Cline DM, Thomas SH. eds. Tintinalli's Emergency Medicine: A Comprehensive Study Guide, 9e. McGraw Hill; 2020.
Nickson, C. (2020, November 3). Carbon monoxide poisoning. Life in the Fast Lane • LITFL. Retrieved December 14, 2021, from https://litfl.com/carbon-monoxide-poisoning/
Sethuraman, Kinjal N. MD, MPH∗; Douglas, Taylor M. BS†; Bostick, Barbara B. MD‡,§; Comer, Angela C. MPH∥; Myers, Bennett MD§; Rosenthal, Robert E. MD∗ Clinical Characteristics of Pediatric Patients With Carbon Monoxide Poisoning, Pediatric Emergency Care: April 2020 - Volume 36 - Issue 4 - p 178-181 doi: 10.1097/PEC.0000000000001378