Rationale:

• HBOT, 100% oxygen delivered at 2 to 4 times atmospheric pressure in a hyperbaric chamber, markedly increases the dissociation and elimination of CO from the body, reducing intracellular CO exposure.

Evidence:

• Several case series (21; 31) suggested improved outcomes in CO poisoning with HBOT, but four subsequent randomized trials described have yielded conflicting results.
• In a randomized trial of 152 patients with severe, acute CO poisoning, patients received either HBOT for one session of 100% oxygen at 3 atmospheres for 150 minutes followed by two sessions of 100% oxygen at 2 atmospheres for 120 minutes, or NBOT of 100% oxygen at 1 atmosphere. Patients treated with HBOT had less cognitive sequelae at 6 weeks (25% vs. 46%, P=0.007) and 12 months (18% vs. 33%, P=0.04). The improvements in cognitive sequelae were related to high-level functioning such as memory, attention, and concentration skills. Simple neurologic functions such as the ability to communicate and perform activities of daily living were not different between the two groups. The initial level of HbCO or duration of CO exposure were not correlated with delayed neurologic symptoms (44).
• In a randomized trial of HBOT for 120 minutes (2.8 atmospheres for 30 minutes followed by 2 atmospheres for 90 minutes) compared with NBOT in 65 patients who presented with CO toxicity within 6 hours of exposure and no history of loss of consciousness or cardiac instability, all symptoms resolved after oxygen therapy. Immediately after oxygen therapy there was no difference in neurologic testing results or symptoms between the NBOT and HBOT groups. Delayed neurologic symptoms or new abnormalities on neurologic testing were detected in 23% of the NBOT patients, with day of onset from 3 to 10 days after exposure. These symptoms or abnormalities resolved from 25 to 77 days after exposure. HBOT-treated patients did not experience delayed neurologic symptoms or new abnormalities on neurologic testing, yielding a significant 95% CI of 8.2% to 38.4% compared with NBOT. The initial level of HbCO or duration of CO exposure were not correlated with delayed neurologic symptoms (45).
• One study of 191 patients with acute CO poisoning were randomly assigned to receive 100% oxygen at 2.8 atmospheres or 1.0 atmospheres for 60 minutes each day for 3 days. Patients who did not return to normal after 3 days were treated for 3 additional days. Overall mortality (3%), neurologic symptoms at discharge (71%), and neurologic symptoms 1 month after exposure (62%) were no different between the two treatment groups. Delayed neurologic sequelae developed in 5% of the HBOT patients but none of the NBOT patients (46).
• Another randomized trial of 629 CO-poisoned adults who had been poisoned at home within 12 hours of therapy found no difference in acute or delayed neurologic outcome between HBOT of oxygen at 2 atmospheres followed by 4 hours of NBOT, compared with 6 hours of NBOT. Potential criticisms of this study are the long time from exposure to oxygen therapy and the use of only a single session of HBOT (47).

Comments:

• The use of HBOT remains controversial and the recommendations are based largely on consensus reading of conflicting data. Patients who are comatose or pregnant and children are always treated; to randomly assign such patients to placebo in an RCT would be considered unethical. There are no data or consensus about treating those with neurologic symptoms that resolve quickly.
• Side effects of HBOT are not common and include claustrophobia, seizures (2% to 3%), patient isolation in the chamber, and trauma to the tympanic membrane (48).
• HBOT should only be performed by experienced personnel at hyperbaric therapy centers.
• Use of HBOT involves additional risks associated with transportation to a hyperbaric center, often complicated by limited equipment and personnel. Patients who are critically ill also run the risk of isolation and suboptimal monitoring while isolated in the hyperbaric chamber.