Smoke in the cabin: carbon monoxide and cyanide

A regional aircraft is parked on the apron when an electrical fire starts behind a panel. Dark smoke fills the cabin for a few minutes before evacuation. The pilot and several passengers arrive at the airport emergency department. The pilot has a severe headache, nausea and is confused about dates. A 60-year-old female passenger is drowsy, with low blood pressure and soot in her nose. Pulse oximetry shows 98 in both. The passenger's blood gas shows lactate of 11 mmol/L. The scene is fictional and was written for teaching.

This text uses the scenario as a clinical reasoning setting. The trap lies in the most comfortable number in the room: a saturation near 100 seems to rule out lack of oxygen, and the team may discharge the pilot and treat the passenger only as shock. In fire smoke, carbon monoxide and cyanide act together and hide tissue hypoxia. The text is not a summary of aviation regulations. If you play the case of the day on NEXO, you will recognize the challenge: the clue that reassures does not always measure what you think.

Key points

• Fire smoke in an enclosed space exposes to carbon monoxide and cyanide at the same time.
• A common pulse oximeter does not distinguish carboxyhemoglobin from oxyhemoglobin.
• Carboxyhemoglobin measurement requires co-oximetry and, if done late, underestimates exposure.
• Lactate above 8 mmol/L in a fire patient suggests cyanide poisoning.
• 100% oxygen by non-rebreather mask is the first treatment for everyone.
• Hydroxocobalamin is the antidote used when cyanide is suspected.

Why this is a reasoning scenario

The case brings together several victims, an enclosed environment and an instrument that seems objective. A review on inhalation injury describes that, in addition to heat, smoke carries toxins that act throughout the body, and that the management of carbon monoxide and cyanide remains controversial (Dries and Endorf, 2013). There is no single sign that confirms the diagnosis.

In an aircraft cabin, smoke accumulates in a small space, and lining and wiring materials burn together. The post on emergency in commercial flight covers on-board care. Here the question is different: what the ground team needs to think about when receiving the victims.

What the scene teaches

The pilot and the passenger were in the same environment but show different pictures. He has headache and confusion; she has drowsiness, hypotension and very high lactate. The difference suggests that there is not a single poisoning with different intensities, but two toxins with different weights in each person. The post on signs, symptoms and history shows how comparing patients exposed to the same event helps separate mechanisms.

Two toxins, two mechanisms

Carbon monoxide and cyanide arrive together but attack different points of oxygen transport and use. A review of inhalation injury summarizes the difference: carbon monoxide displaces oxygen from hemoglobin and also binds to cytochrome oxidase in the mitochondria, while cyanide binds only to cytochrome oxidase (Walker et al., 2015).

Carbon monoxide: the blood that does not deliver

Carbon monoxide's affinity for hemoglobin is about 200 times greater than that of oxygen. The binding shifts the oxyhemoglobin dissociation curve to the left and changes its shape, and oxygen delivery to tissues falls (Dries and Endorf, 2013). A didactic review on the curve recalls that the left shift means hemoglobin that holds oxygen more tightly and releases less in the tissues (Collins et al., 2015).

The result is blood that appears full of oxygen and delivers little. The brain and heart, which consume a lot of oxygen, feel it first. The post on from the basic cycle to the clinic shows how the dissociation curve studied in the early years explains a confused patient with normal saturation.

Cyanide: the cell that does not use

Cyanide does not affect hemoglobin. It blocks the respiratory chain inside the mitochondrion, and the cell comes to depend on metabolism without oxygen. That is why lactate rises. The review by Lawson-Smith and colleagues describes that, in mild cases, the symptoms are headache, nausea, dizziness, anxiety, altered mental status and tachypnea; in severe cases, dyspnea, bradycardia, hypotension and arrhythmia (Lawson-Smith, Jansen and Hyldegaard, 2011).

The passenger in the case has drowsiness, hypotension and soot in her nose. This set weighs in favor of cyanide, together with the carbon monoxide she certainly also inhaled.

How the condition presents

Carbon monoxide poisoning has a broad spectrum. A clinical review describes a presentation that ranges from headache and dizziness to coma and death (Rose and colleagues, 2017). Headache, nausea and confusion, as in the pilot, are common and nonspecific symptoms. On an ordinary day, they would suggest migraine, dehydration or stress.

The context changes everything. After a fire, headache and confusion are poisoning until proven otherwise. Cyanide has mild symptoms that overlap with those of carbon monoxide, and so the clinical picture alone does not separate the two (Lawson-Smith, Jansen and Hyldegaard, 2011).

Signs of inhalation injury

Soot in the nose and mouth indicates that the person breathed in a large amount of smoke. Walker and colleagues describe that the diagnosis of inhalation injury long relied on nonspecific examination findings and on bronchoscopy, and that the severity seen on bronchoscopy relates to airway inflammation (Walker and colleagues, 2015). The passenger needs airway assessment, in addition to treatment of the toxins.

The heart suffers too

Carbon monoxide damages the myocardium. In a Korean registry of 850 poisoned adults, about 12 in every 100 had cardiomyopathy on echocardiography. An initial troponin I above 0.01 ng/mL had a sensitivity of 95 in every 100 and a negative predictive value of 97 in every 100 for this injury, and myocardial dysfunction was associated with a worse course (Kim and colleagues, 2020). Electrocardiogram and troponin are part of the assessment of victims with symptoms.

The pulse oximeter trap

The number 98 on the oximeter is the most dangerous clue in the case. A review on pulse oximetry explains that conventional devices distinguish only two substances, reduced hemoglobin and oxyhemoglobin, and assume that abnormal hemoglobins, such as carboxyhemoglobin, are absent (Jubran, 2015). The same review lists carboxyhemoglobin among the limitations of the method.

In practice, the device reads hemoglobin bound to carbon monoxide as if it were bound to oxygen. Saturation appears normal while part of the hemoglobin carries no oxygen at all. In cyanide poisoning, the situation is even more misleading: the blood may in fact be well oxygenated, because the problem lies in the cell that does not use the oxygen.

A bias with a name

Trusting the number that reassures is a way of closing the case too early. The post on examples of heuristics in clinical reasoning shows how a useful shortcut becomes an error when the context changes. The post on hypobaric hypoxia in flight shows the opposite case, in which the pulse oximeter works well because the hemoglobin is normal.

Tests: what to order and how to read them

The first decision is to order the right tests, and not to trust the oximeter. The table summarizes the role of each one in the case.

Test	What it shows	Care when reading	Source
Conventional pulse oximetry	Apparent saturation	Reads carboxyhemoglobin as if it were oxyhemoglobin	Jubran, 2015
Co-oximetry	Measured carboxyhemoglobin	Done late, it underestimates the real exposure	Dries and Endorf, 2013
Lactate on blood gas	Metabolism without oxygen	Above 8 mmol/L in a fire patient suggests cyanide	Lawson-Smith, Jansen and Hyldegaard, 2011
Blood cyanide	Confirmation of poisoning	The result takes hours and should not delay the antidote	Lawson-Smith, Jansen and Hyldegaard, 2011
Troponin and electrocardiogram	Myocardial injury	A normal initial troponin makes cardiomyopathy unlikely	Kim and colleagues, 2020
Bronchoscopy	Airway injury	Severity related to inflammation	Walker et al., 2015

Carboxyhemoglobin: time matters

Carboxyhemoglobin falls quickly when the person leaves the smoke and receives oxygen. Dries and Endorf describe a half-life of about 250 minutes in ambient air, which falls to 40 to 60 minutes with 100% oxygen. Because of the interval between exposure and collection, levels measured on arrival at the department do not reflect the real poisoning (Dries and Endorf, 2013).

This means that a low level, drawn after an hour of oxygen mask in the ambulance, does not rule out significant poisoning. The level must be read together with the exposure time, the oxygen time and the symptoms.

Lactate: the cyanide clue

Blood cyanide measurement does not help at the time of the decision. Lawson-Smith et al. explain that the result takes hours to be ready and that cyanide has a half-life of about 1 hour in the blood (Lawson-Smith, Jansen and Hyldegaard, 2011). The same review proposes suspecting cyanide in a fire victim with neurological signs, soot in the mouth or sputum, or metabolic acidosis with lactate above 8 mmol/L. A lactate of 10 mmol/L is described as a sensitive and specific indicator.

The passenger has a lactate of 11 mmol/L, drowsiness and soot. She meets these criteria. The post on pretest probability shows why the same lactate weighs more in a fire victim than in a patient with sepsis.

Differential diagnosis

The differential in smoke victims includes four mechanisms that can coexist. Simple hypoxia, from lack of oxygen in the air during the fire, improves quickly outside the smoke. Carbon monoxide poisoning reduces oxygen delivery. Cyanide poisoning prevents the use of oxygen. Inhalation injury from heat and irritants compromises the airway. Walker et al. recall that combustion products such as carbon dioxide act as simple asphyxiants and that the fire environment is already poor in oxygen (Walker et al., 2015).

Outside the smoke

Causes that do not come from the fire must also be remembered. A hypotensive and drowsy passenger may have trauma from the evacuation, extensive burns with volume loss or a prior disease decompensated by stress. The confused pilot may have hit his head on the way out. None of these hypotheses excludes the poisoning, but all of them change management. The post on differential diagnosis shows how to keep the list open without losing priority.

What separates the pictures in this case

In the pilot, headache, nausea and confusion with little change in lactate suggest a predominance of carbon monoxide. In the passenger, the very high lactate and hypotension suggest added cyanide. In both, co-oximetry helps, but does not decide on its own.

Management: oxygen for all, antidote for some

The first step is the same for all symptomatic victims: 100% oxygen by non-rebreather mask with reservoir, or by tube if the airway is compromised. Lawson-Smith et al. place 100% oxygen and life support as the basis of cyanide treatment (Lawson-Smith, Jansen and Hyldegaard, 2011). For carbon monoxide, oxygen shortens the half-life of carboxyhemoglobin from about 250 to 40 to 60 minutes (Dries and Endorf, 2013).

Hydroxocobalamin

When cyanide is suspected, the antidote does not wait for confirmation. Walker et al. describe hydroxocobalamin as the most available antidote: it binds to cyanide and forms cyanocobalamin, which is not toxic and is excreted in the urine. The standard dose is 5 g intravenously over 15 minutes, and a second 5 g dose may be given in severe toxicity or poor response (Walker et al., 2015).

The same review recalls that hydroxocobalamin turns the skin and urine red and may interfere with colorimetric tests. It is worth drawing blood for tests before the infusion, when this does not delay treatment.

Hyperbaric oxygen

The role of the hyperbaric chamber is the most discussed point. A randomized trial compared three chamber sessions within 24 hours, with hyperbaric oxygen or with normobaric oxygen and room air. Cognitive sequelae at 6 weeks were less frequent in the hyperbaric group, 19 of 76 versus 35 of 76 (Weaver et al., 2002). The review by Dries and Endorf, on the other hand, concludes that the data do not support the routine use of hyperbaric oxygen in victims of inhalation injury outside clinical trials (Dries and Endorf, 2013).

Lawson-Smith et al. cite the recommendation of the American society of hyperbaric medicine to treat carbon monoxide poisoning complicated by cyanide with hyperbaric oxygen (Lawson-Smith, Jansen and Hyldegaard, 2011). The data on the chamber in cyanide poisoning are scarce: in a Danish study with 37 patients poisoned by carbon monoxide, the blood cyanide concentration did not change after hyperbaric oxygen (Lawson-Smith et al., 2010). In practice, the decision depends on severity, on access to a chamber and on the patient's stability for transport. The hypotensive passenger needs to be stable before any transfer.

Sequelae and follow-up

Improvement in the emergency department does not close the case. Rose et al. describe that a significant share of carbon monoxide poisoning survivors are left with long-term neurological and affective sequelae, and that these deficits do not necessarily track the blood carbon monoxide level (Rose et al., 2017). The authors attribute the damage to the gas's effects on mitochondrial respiration, on the cell's energy use, on inflammation and on free radical formation.

This dissociation between the blood level and the damage in the brain has a practical consequence. A pilot with a modest carboxyhemoglobin on arrival may still have memory, attention or mood problems weeks later. In the trial by Weaver et al., cognitive sequelae were measured by neuropsychological tests at 6 weeks and at 12 months, and differences between the groups appeared at both time points (Weaver et al., 2002). A review on neurocognitive sequelae describes that they may persist from the poisoning or appear after an apparent recovery, with onset between 2 and 40 days, even in those who received hyperbaric oxygen (Ning et al., 2020).

What to agree on at discharge

At discharge, it is worth explaining to the victims that forgetfulness, difficulty concentrating, mood changes or a returning headache are reasons for a new assessment. For the pilot, the return to work goes through the medical examiner, who may request a cognitive assessment before clearing the flight. Those who had myocardial injury need cardiology follow-up, because myocardial dysfunction was associated with a worse course (Kim et al., 2020).

The post on prognostic reasoning addresses this question that comes after the diagnosis: what to expect in the coming months and how to follow up.

Step-by-step reasoning in the case

For the pilot, the sequence goes like this. He was in the smoke and has headache, nausea and confusion. A normal oximeter rules nothing out. He receives 100 percent oxygen by mask, and co-oximetry, blood gas with lactate, electrocardiogram and troponin are collected. Carboxyhemoglobin is read together with the time since exposure and the time on oxygen. He does not fly again the same day, and the assessment of sequelae is left for follow-up.

For the passenger, the sequence is more urgent. She has drowsiness, hypotension, soot and a lactate of 11 mmol/L. She receives 100 percent oxygen, immediate airway assessment, venous access and hydroxocobalamin without waiting for the cyanide measurement. Then come co-oximetry, cardiac tests and the discussion about the hyperbaric chamber according to stability.

Deciding under uncertainty

For neither of the two victims is the team certain at the moment of the decision. The antidote is given based on probability, not on confirmation. The post on diagnostic uncertainty discusses how to act when the confirmatory test arrives too late to change management. The post on clinical decision-making shows how to weigh the risk of treating unnecessarily against the risk of not treating.

A pause before clearing

The most dangerous moment is when the pilot is discharged. He improves with oxygen and wants to leave. A diagnostic pause before discharge forces you to ask whether the heart was assessed, whether carboxyhemoglobin was read at the right time and whether someone explained the signs of delayed sequelae.

Common mistakes

The first mistake is trusting the oximeter. It does not measure carboxyhemoglobin and, in cyanide poisoning, it can show a true and useless saturation (Jubran, 2015).

The second mistake is reading carboxyhemoglobin without the clock. A level drawn after an hour of oxygen underestimates the exposure (Dries and Endorf, 2013).

The third mistake is waiting for the cyanide level to give the antidote. The result takes hours, and the lactate together with the clinical picture is already enough for the suspicion (Lawson-Smith, Jansen and Hyldegaard, 2011).

The fourth mistake is treating the passenger only as shock. Fluids and vasopressors do not fix a cell that cannot use oxygen.

The fifth mistake is forgetting the heart. Myocardial dysfunction is common and is associated with a worse outcome (Kim and colleagues, 2020).

The sixth mistake is discharging without talking about sequelae. Neurological and affective deficits can appear in survivors and do not necessarily follow the blood carbon monoxide level (Rose and colleagues, 2017). A pilot who feels fine the next day still needs clear guidance on what to watch for in the following weeks and a return to flying arranged with the medical examiner.

How NEXO helps train this reasoning

To use the case of the day as training, try three habits. Before the first guess, ask what each measuring instrument can and cannot see. With each clue, check whether it changes management or only confirms what you already thought. After the result, note which number reassured you too early.

Three exercises with cases

Ask what the test measures. For each test in the case, write in one sentence what it actually detects.

Link mechanism and management. Write the mechanism of the main hypothesis and the treatment that follows from it.

Compare the victims. When several people go through the same event, compare their pictures and ask what explains the difference.

The post about case-based learning in clinical practice shows how cases help cement this habit.

The limits on using the game are described on the NEXO economy page. If you research the teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information under curation and under creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Can a pulse oximeter show normal saturation in carbon monoxide poisoning?

Yes. A conventional oximeter distinguishes only oxyhemoglobin and reduced hemoglobin and assumes that carboxyhemoglobin is absent (Jubran, 2015).

Can a low carboxyhemoglobin level on arrival hide a large exposure?

Yes. Because of the interval between exposure and collection, the levels measured on arrival do not reflect the real poisoning (Dries and Endorf, 2013).

Does lactate help raise suspicion of cyanide?

Yes. In a fire victim, lactate above 8 mmol/L suggests cyanide, and 10 mmol/L is described as a sensitive and specific indicator (Lawson-Smith, Jansen and Hyldegaard, 2011).

Can hydroxocobalamin be given before confirmation?

Yes. The cyanide level takes hours, and the review by Lawson-Smith and colleagues proposes deciding by the clinical signs and by lactate (Lawson-Smith, Jansen and Hyldegaard, 2011).

Can carbon monoxide damage the heart?

Yes. In a registry of 850 poisoned adults, about 12 in every 100 had cardiomyopathy on echocardiography (Kim and colleagues, 2020).

Final checklist

• I considered carbon monoxide and cyanide in every victim of smoke in an enclosed space.
• I was suspicious of the reading from a standard pulse oximeter.
• I ordered co-oximetry and noted the time since exposure and the time on oxygen.
• I ordered a blood gas panel with lactate.
• I gave 100 percent oxygen by non-rebreather mask to all symptomatic victims.
• I assessed the airway when there was soot or a burn.
• I gave hydroxocobalamin based on suspicion, without waiting for the cyanide level.
• I drew labs before the infusion when that did not delay the antidote.
• I ordered an electrocardiogram and troponin.
• I discussed the hyperbaric chamber according to severity and stability.

Want to practice your reasoning with progressive clues? Play the case of the day.

References

1. Collins JA, Rudenski A, Gibson J, et al. Relating oxygen partial pressure, saturation and content: the haemoglobin-oxygen dissociation curve. Breathe (Sheff). 2015. PMID 26632351.
2. Dries DJ, Endorf FW. Inhalation injury: epidemiology, pathology, treatment strategies. Scand J Trauma Resusc Emerg Med. 2013. PMID 23597126.
3. Jubran A. Pulse oximetry. Crit Care. 2015. PMID 26179876.
4. Kim JS, Ko BS, Sohn CH, et al. High-Sensitivity Troponin I and Creatinine Kinase-Myocardial Band in Screening for Myocardial Injury in Patients with Carbon Monoxide Poisoning. Diagnostics (Basel). 2020. PMID 32326299.
5. Lawson-Smith P, Jansen EC, Hilsted L, et al. Effect of Hyperbaric Oxygen Therapy on whole blood cyanide concentrations in carbon monoxide intoxicated patients from fire accidents. Scand J Trauma Resusc Emerg Med. 2010. PMID 20550698.
6. Lawson-Smith P, Jansen EC, Hyldegaard O. Cyanide intoxication as part of smoke inhalation--a review on diagnosis and treatment from the emergency perspective. Scand J Trauma Resusc Emerg Med. 2011. PMID 21371322.
7. Ning K, Zhou YY, Zhang N, et al. Neurocognitive sequelae after carbon monoxide poisoning and hyperbaric oxygen therapy. Med Gas Res. 2020. PMID 32189667.
8. 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. PMID 27753502.
9. Walker PF, Buehner MF, Wood LA, et al. Diagnosis and management of inhalation injury: an updated review. Crit Care. 2015. PMID 26507130.
10. Weaver LK, Hopkins RO, Chan KJ, et al. Hyperbaric oxygen for acute carbon monoxide poisoning. N Engl J Med. 2002. PMID 12362006.