Hyperbaric oxygen for carbon monoxide poisoning

Abstract

Background: Poisoning with carbon monoxide (CO) remains an important cause of accidental and intentional injury worldwide. Several unblinded nonrandomized trials have suggested that the use of hyperbaric oxygen (HBO) prevents the development of neurological sequelae. This has led to the widespread use of HBO in the management of patients with carbon monoxide poisoning.

Objectives: To assess the effectiveness of hyperbaric oxygen (HBO) compared to normobaric oxygen (NBO) for the prevention of neurologic symptoms in patients with acute carbon monoxide poisoning.

Search strategy: We searched MEDLINE (1966-present), EMBASE (1980-present), and the Controlled Trials Register of the Cochrane Collaboration, supplemented by a manual review of bibliographies of identified articles and discussion with recognized content experts.

Selection criteria: All randomized controlled trials involving non pregnant adults acutely poisoned with carbon monoxide, regardless of severity, with adequate or unclear allocation concealment were examined. Trials with a score of 3 out of 5 or higher on the validity instrument of Jadad were included in the primary analysis.

Data collection and analysis: Two reviewers independently extracted from each trial information on the number of randomized patients, types of participants, the dose and duration of the intervention, and the prevalence of neurologic symptoms at follow-up. A pooled odds ratio (OR) for the presence of neurologic symptoms at one month follow-up was calculated using a random effects model.

Main results: Six randomized controlled trials were identified. The trials were of varying quality. Three trials employing different doses of NBO and HBO were included in the primary analysis. The severity of CO poisoning was inconsistent between trials. At one month follow-up after treatment, symptoms possibly related to carbon monoxide poisoning were present in 81 of 237 patients (34.2%) treated with HBO, compared with 81 of 218 patients (37.2%) treated with NBO (O.R. for benefit with HBO 0.82; 95% CI 0.41-1.66).

Reviewer's conclusions: There is no evidence that unselected use of HBO in the treatment of acute CO poisoning reduces the frequency of neurological symptoms at one month. However, evidence from the available randomized controlled trials is insufficient to provide clear guidelines for practice. Further research is needed to better define the role of HBO, if any, in the treatment of carbon monoxide poisoning. This research question is ideally suited to a multicentre, randomized, double-blind controlled trial.