Carbon monoxide (CO) poisoning is one of the main causes of unintentional poisoning death in the United States. Hyperbaric oxygen (HBO₂) has been advocated as a therapy for CO poisoning, but recommendations for its use vary. Appropriate indications cited include loss of consciousness, altered mental status, pregnancy, and high carboxyhemoglobin levels.

A subcommittee of the American College of Emergency Physicians (ACEP) reviewed the literature to determine whether adults presenting to the emergency department with acute symptomatic CO poisoning should be treated with HBO₂ therapy, and whether clinical or laboratory criteria can identify patients with CO poisoning who are likely to benefit from this therapy.

One randomized controlled trial (RCT) comparing HBO₂ therapy with normobaric oxygen in 152 patients found that HBO₂ therapy was associated with a 21.1 percent reduction in the rate of neurologic sequelae after six weeks, and a 14 percent reduction after 12 months. A randomized study involving 65 patients found that 0 percent of those treated with HBO₂ had neurologic sequelae compared with 23 percent of those treated with ambient pressure oxygen. However, an RCT with 191 participants found no difference in adverse neurologic outcomes between those who received HBO₂ therapy and those who did not (74 and 68 percent, respectively), although most patients had severe poisoning. Finally, an unblinded study with 343 participants found that 32.1 percent of patients treated with HBO₂ reported neurologic symptoms one month after treatment compared with 33.8 percent of those who received mask oxygen.

None of the trials prospectively designated subgroups of patients for separate analysis. No study reported separate outcome data about loss of consciousness or abnormal mental status, and all trials excluded pregnant women. Three of the studies found no difference in outcomes in patients with high and with low carboxyhemoglobin levels, regardless of treatment used.

The committee determined that HBO₂ is a therapeutic option for patients with CO poisoning, but that its use cannot be mandated. It found that no clinical variables, including carboxyhemoglobin levels, identify patients for whom HBO₂ therapy is most likely to provide benefit or cause harm.