The administration of hyperbaric oxygen did not improve outcome in mild carbon monoxide poisoning and worsened outcome when given twice to more severely poisoned patients.

Clinical Bottom Line:
1. No evidence for benefit of hyperbaric oxygen in the treatment of carbon monoxide poisoning without loss of consciousness.
2. Worse outcome for severe poisoning with a regimen using two hyperbaric oxygen sessions versus a single session .
Appraised by: Mike Bennett, Dept of Diving and Hyperbaric Medicine, Prince of Wales Hospital
Sydney, 14 December 2012

Clinical Scenario: A patient presented with carbon monoxide poisoning for treatment.
Three-part Question: In patients with carbon monoxide poisoning, does the administration of hyperbaric oxygen, compared to normobaric oxygen (mild to moderate poisoning) or one hyperbaric oxygen session versus two (severe poisoning), result in any reduction in the rate of neurological impairment one month later?
Search Terms: hyperbaric oxygenation, carbon monoxide
First author:

The Study:
Non-blinded concealed randomised controlled trial with intention-to-treat. Two separate randomisation schedules and treatment options. All were adult patients exposed to carbon monoxide (no other toxicities) and presenting for treatment within 12 hours of exposure. Minimum carboxyhaemoglobin levels of 10% in smokers and 5% in non-smokers. Trial A: Only transient impairment or loss of consciousness not witnessed by medical staff (mild to moderate). Trial B: confirmed coma (severe)
Control groups A.(N = 86; 74 analysed): Normobaric 100% oxygen for six hours. B.(N = 101; 80 analysed): Oxygen at 2ATA for two hours then 1ATA for 4 hours.
Experimental groups A.(N = 93; 79 analysed): Oxygen at 2ATA for two hours then 1ATA for 4 hours. B. (N = 105; 90 analysed): Oxygen at 2ATA for two hours, then 1ATA for 4 hours, then 2ATA for a further two hours at 12 hours after the first HBO.

The Evidence:Trial A.
Time to Outcome
Control groups
Hyperbaric groups
Relative risk reduction
Absolute risk reduction
1 month
95% CI:

-30% to 21%
-0.18 to 0.12
NNT = 6 to INF NNH = 8 to INF

Trial B
Time toOutcome
Relaitiverisk reduction
Absoluterisk reduction
1 month
95% CI:

-52% to -9%
-0.30 to -0.06
NNH = 3 to 16

1. Interim analysis (?planned) resulted in stopping recruitment because of these results.
2. Results here are by intention to treat but excluding missing data. Sensitivity analysis for best and worst case scenario for allocation of missing data does not alter the results in trial A, but best case scenario suggests a non-significant difference in trial B.
3. The diagnosis was usually on the basis of history and on COHb level when available.
4. The primary outcome was obtained by self-assessment questionnaire and a physical examination by a physician or a GP blinded to therapy.
5. This trial probably began immediately following that of Raphael (also reported in this section).

Expiry date: April 2020
1. Annane D, Chadda K, Gajdos P, Jars-Guincestre M-C, Chevret S, Raphael J-C. Hyperbaric oxygen therapy for acute domestic carbon monoxide poisoning: two randomised controlled trials. Intensive Care Medicine 2011; 8658:414-419.