Hyperbaric oxygen for carbon monoxide poisoning: do we know the answer

January 27, 2011, 10:55 pm

★★★½☆

Hyperbaric oxygen therapy for acute domestic carbon monoxide poisoning: two randomized controlled trials. Annane D et al. Intensive Care Med 2010 Dec 2. [Epub ahead of print]

Abstract

Despite several large studies investigating the role of hyperbaric oxygen (HBO) therapy in treating carbon monoxide (CO) poisoning, we still don’t know definitively in what situations it is effective in reducing the incidence of long-term neurological dysfunction following exposure.  Many toxicologists would recommend HBO in these patients if they experienced loss of consciousness (LOC) at any time — with the important caveat that transporting a seriously ill patient (for example, one with active cardiac ischemia) to another facility for HBO is probably ill-advised.

This large study (385 patients) actually encompassed two trials.  In the first, patients with transient loss of consciousness (defined as normal consciousness at time of rescue but syncope or “he/she could not recall what happened”) were randomized to 6 hours normobaric oxygen (NBO) or 4 hours of NBO and 1 HBO session.  Patients with initial coma (Glasgow coma score <8) were randomized to 4 hours NBO plus one HBO session (2 ATA) or 4 hours of NBO plus two HBO sessions.  Follow-up extended to 1 month after treatment, with a self-assessment questionnaire and a “thorough” physical examination by a blinded “intensive care physician qualified in neurology” or the patient’s general practitioner. No neuropsychiatric testing was performed.

The authors found that in CO-exposed patients who met their definition of “transient loss of consciousness”, there was no advantage of HBO over NBO. In patients with coma, 2 HBO sessions were associated with worse outcomes than one HBO session.

When reading this study, it is important to pay close attention to details and definitions, and to realize that it doesn’t really answer (or address) the key question: does treating a CO-exposed patient who at any time lost consciousness with HBO improve long-term neurological outcome. In the first group (“transient loss of consciousness”), the composite entry criteria (syncope or amnesia for the event) does not isolate those with documented LOC.  This subset was not analyzed separately.

In the second trial (comatose patients), there was no study arm that omitted HBO – the comparison was between one HBO session and two.  So all patients who were comatose at time of rescue received at least one treatment with HBO.

Some other points:

Since the study was conducted at Raymond Poincare Teaching Hospital and Hyperbaric Center, it is apparent that the difficult issue of transporting potentially unstable patients to another facility was not a factor.  There is some evidence that episodes of hypotension will worsen neurological outcome in patients with severe CP toxicity.  It is not unreasonable to assume that the sick CO patient would be better off carefully monitored in an ICU than bouncing around in the back of an ambulance en route to treatment modality whose effectiveness has not yet been demonstrated. (Don’t even get me started on the misguided idea of sending a patient by helicopter for HBO.)

Patients who received HBO were given 10 mg IM diazepam before each session.  It is not clear whether this had any effect on outcomes.

The follow-up criteria seem rather ill-defined.  An unspecified “thorough physical examination” by some random general practitioner is, shall we say, somewhat less than rigorous.

Comments are closed.