Should HBO Be the Standard of Care for CO Poisoning?


You are working in a rural ER in a mountain resort town. A family of four is brought in because they have headache, nausea and malaise without syncope or chest pain. Following a thorough history and physical examination you quickly note the cause of the illness is carbon monoxide (CO) poisoning from a faulty heat generator. The closest HBO chamber is 300 miles away.

DISCLAIMER: I am certified in HBOT and Medical Toxicology. My bias before covering this controversial topic is that I want it to work. Unfortunately, the evidence does not support its benefit in this setting and I cannot ethically recommend it.

You are working in a rural ER in a mountain resort town. A family of four is brought in because they have headache, nausea and malaise without syncope or chest pain. Following a thorough history and physical examination you quickly note the cause of the illness is carbon monoxide (CO) poisoning from a faulty heat generator. The closest HBO chamber is 300 miles away.

The father is a retired family practitioner, a non-smoker, and has a CO level of 28%. All patients are breathing 100% oxygen via non-rebreather mask. All symptoms have resolved and neurological examinations and ECGs are normal. The remainder of the family CO levels are all less than 10%.

Is HBO therapy beneficial in this patient? The physician requests an evidence-based review of the pros and cons of HBO therapy (HBOT).

Treatment of the CO-poisoned patient begins with supplemental oxygen and aggressive supportive care. When occult CO poisoning is discovered, other potential patients should be cautioned and evacuated until the source is identified and the environment is non-hazardous.

High-flow oxygen therapy should be administered immediately to treat hypoxia resulting from CO poisoning and to accelerate eradication of CO from the body. Whether this oxygen should be given under increased pressure with HBO, or under ambient pressures (normobaric oxygen) is a subject of debate.

Multiple sources recommend hyperbaric oxygen therapy in the setting of carbon monoxide poisoning: (1)  Neurologic findings (coma, seizure, focal findings), (2) Pregnancy with CO-Hgb levels > 15–20% and (3)  History of loss of consciousness.

Nine prospective, randomized, controlled trials compared HBOT with NBO for CO poisoning. Five of these studies showed a benefit of HBOT, and four of the studies did not. The data and conclusions drawn from these studies highlight the controversy surrounding the utility of HBOT. A Cochrane Review in 2009 by Juurlink DN, et al, stated “Existing randomized trials do not establish whether the administration of HBO to patients with carbon monoxide poisoning reduces the incidence of adverse neurologic outcomes.” In the Cochrane Review the pooled analysis did not suggest a benefit from HBOT (OR for neurological deficits 0.78, 95%CI 0.54 to 1.12, p=0.18), significant methodological and statistical heterogeneity was apparent among the trials, and this results should be interpreted cautiously.

Advocates of HBOT reference the study by Weaver et al., considered by many to be the most methodologically rigorous. I agree that it is excellent and well designed study but significant questions exist. Weaver et al., combined patients enrolled in a previously reported randomized controlled trial (RCT) with patients who did not participate in that RCT. The 91 patients in the latter group had not been enrolled in the RCT for various reasons, including refusal to participate and presenting too late after CO exposure. Obvious questions include: Did these 91 patients represent every single patient excluded from initial randomization? Were provisions for repeated neuropsychiatric testing and data collection despite exclusion from randomization approved in the original protocol by the institutional review board? Was an incentive offered for participation, and was this incentive the same as that for subjects enrolled in the original RCT?

Weaver et al., advised hyperbaric oxygen (HBO) therapy for symptomatic CO-poisoned patients aged 36 years and older. However, combining the two patient groups represents post hoc analyses (acknowledged by the authors) of a nonrandomized trial, or quasi-experimental design. Hence, results cannot represent evidence any stronger than those from an observational study with inherent biases and both known and unknown confounders. Was the cutoff of 36 years chosen a priori or was it chosen post hoc during statistical analyses?

Weaver et al., wrote that “If the patient complained of memory, attention, or concentration difficulties, the required neuropsychological test decrement was decreased to greater than one standard deviation below the mean of demographically corrected standardized T scores on any one subtest.” Using such a liberal cutoff increases the sensitivity at the expense of specificity. As the authors recognized, subjective complaints such as poor memory and concentration can result from causes other than CO poisoning. We noticed no mention of alcoholism, alcohol intoxication, or secondary gain issues, such as litigation. Were these considered or adjusted for?

The most recent study is very convincingly negative. Annane D et al.,  performed two prospective randomized trials on two parallel groups involving 385 patients with acute domestic CO poisoning. Patients with transient loss of consciousness (trial A, n = 179) were randomized to either 6 h of normobaric oxygen therapy (NBO; arm A0, n = 86) or 4 h of NBO plus one HBO session (arm A1, n = 93). Patients with initial coma (trial B, n = 206) were randomized to either 4 h of NBO plus one HBO session (arm B1, n = 101) or 4 h of NBO plus two 2 HBO sessions (arm B2, n = 105). The primary endpoint was proportion of patients with complete recovery at 1 month. In trial A, there was no evidence for a difference in 1-month complete recovery rates with and without HBO [58% compared to 61%; unadjusted odds ratio, 0.90 (95% CI, 0.47-1.71)]. In trial B, complete recovery rates were significantly lower with two than with one HBO session [47% compared to 68%; unadjusted odds ratio, 0.42 (CI, 0.23-0.79)]. In patients with transient loss of consciousness, there was no evidence of superiority of HBO over NBO. In comatose patients, two HBO sessions were associated with worse outcomes than one HBO session.

It should be noted that HBOT is not entirely risk-free. Patients commonly complain of painful barotrauma affecting the ears and sinuses, and patients with claustrophobia are often unable to tolerate the close confines of a monoplace hyperbaric chamber. Less common risks include oxygen toxicity, seizures, pulmonary edema and hemorrhage, decompression sickness including pneumothorax and nitrogen emboli, and fire hazard. The only absolute contraindication to HBOT is an untreated pneumothorax. Relative contraindications include claustrophobia, otosclerosis or other scarring of middle ear, bowel obstruction, significant chronic obstructive pulmonary disease particularly with bullae formation, and requirement of care beyond what can be provided in a monoplace chamber. In addition, if the patient requires emergency intervention (ie, defibrillation) while undergoing HBOT, ~ 5 minutes are required to decompress the patient safely before interventions can proceed. A risk-benefit analysis should be considered for each individual patient, depending on other concomitant medical needs, and discussed with the patient or family.

Currently the Accreditation Council of the Undersea & Hyperbaric Medical Society lists less than 100 accredited HBOT centers in the USA with the vast majority in urban and coastal centers. The transfer of this patient to a center with accreditation seems prudent if indicated but I would not support the transfer in this case.

In summary, the conclusion that HBO should be recommended to patients aged 36 years or older is hypothesis-generating at best. I do not think that EPs should consider HBO therapy the standard of care in the setting o
f CO poisoning. Physicians should adhere to the American College of Emergency Physicians guideline that was based on a review of the RCT by Weaver and colleagues as well as the other evidence. The guideline states that HBO is a therapeutic option for CO-poisoned patients, but its use cannot be mandated, and that no clinical variables identify a subgroup of CO-poisoned patients for whom HBO is most likely to provide benefit or cause harm.


  • Weaver LK, Valentine KJ, Hopkins RO. Carbon monoxide poisoning: risk factors for cognitive sequelae and the role of hyperbaric oxygen. Am J Respir Crit Care Med 2007;176:491–497.
  • Weaver LK, Hopkins RO, Chan KJ, Churchill S, Elliott CG, Clemmer TP, Orme JF Jr, Thomas FO, Morris AH. Hyperbaric oxygen for acute carbon monoxide poisoning. N Engl J Med 2002;347:1057–1067.
  • Juurlink DN, Buckley N, Stanbrook MB, Isbister G, Bennett MH, McGuigan M. Hyperbaric oxygen for carbon monoxide poisoning. Cochrane Database of Systematic Reviews 2005, Issue 1. Art. No.: CD002041. DOI: 10.1002/14651858.CD002041.pub2
  • Annane D, Chadda K, Gajdos P, Jars-Guincestre MC, Chevret S, Raphael JC. Intensive Care Med. 2011 Mar;37(3):486-92. Epub 2010 Dec 2. Hyperbaric oxygen therapy for acute domestic carbon monoxide poisoning: two randomized controlled trials.




  1. You article has a clear bias against HBO – blatant in your tone. The exaggerations of HBOT side effects are ludicrous. You mentioned DCS as a risk – that is laughable. The well established, occasional, rarely NOT minor, almost always preventable (and very rarely irreversible) side effects of HBO are also not relevant in the context of arguing the medical effectiveness of the treatment Your tone and bias raise the concern of interpretation and selective reference in your review of the studies.

    Personally, I am biased towards HBO as an effective modality for certain indications so I won’t offer opinions however, I will ask this question:

    Why do you treat an acute case of CO poisoning with 100% 02? Why not 50% or just stick the PT on a nasal cannula at 2 lpm.. or any supplementary O2 at all?

    Your answer may, in fact, support HBO as an immediate treatment for CO poisoning. Some would argue that it would be negligence to intentionally not provide HBO, if available, to an acute PT.

  2. I am board certified in Emergency Medicine and EMS. I also practice HBO and have additional training in diving medicine. The case presented here, of a man with a CO level of 28%, with a normal EKG (presumably normal troponin) who is asymptomatic and 300 miles from an HBO chamber could reasonably be managed with normobaric oxygen via NRB until he clears his CO. There are risks and costs of driving/flying him 300 miles. My inclination is that absolute levels alone are not that helpful, particularly because a prolonged exposure at moderate levels may be worse than a brief exposure at high levels. The only asymptomatic patient that I believe mandates a stronger consideration for CO level based treatment is the pregnant female and pediatrics, as they (fetus/ped) are not reliable historians and it is difficult to get a reliable neuropsych exam.

Leave A Reply