Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Hyperbaric oxygen for acute carbon monoxide poisoning.

Weaver LK, Hopkins RO, Chan KJ et al.

N Engl J Med 2002; 347:1057-67. [abstract]

Reviewed by Mike Fiore, MD, Children's Hospital of Michigan, Detroit MI

Review posted June 21, 2003

---

I. What is being studied?:

The study objective:

To compare the rate of cognitive sequelae in patients with carbon monoxide poisoning treated with hyperbaric oxygen with the rate in those treated with normobaric oxygen.

The study design:

A prospective, double-blind, randomized clinical trial of patients referred to a single academic center from multiple emergency departments in three states from November 1992 through February 1999

The patients included:

All patients with documented exposure to carbon monoxide (elevation of carboxyhemoglobin level or the ambient CO concentration) or obvious exposure to CO, and clinical signs or symptoms (loss of consciousness, confusion, headache, malaise, fatigue, forgetfulness, dizziness, visual changes, nausea, vomiting, cardiac ischemia, or metabolic acidosis [calculated base excess -2mmol/L or lower or lactate > 2.5 mmol/L]). If CO level < 10 percent, patient eligible if CO poisoning was the only plausible diagnosis.

The patients excluded:

> 24 hours since CO exposure had ended, < 16 years age, if they were moribund, informed consent unattainable, pregnancy, previous enrollment in a RCT on CO poisoning, referring physician insisted upon hyperbaric treatment.

The interventions compared:

All patients received 3 sessions in a monoplace chamber at intervals of 6 to 12 hours. The first session was started within 24 hours of CO exposure. Patients in the hyperbaric group were exposed to 100% oxygen at 3 atm pressure (for 75 minutes, then 2 atm for an additional 75 minutes) during the first chamber session and then 100% oxygen at 2 atm pressure during the second (150 minutes) and third chamber session (120 minutes). Patients in the normobaric group were exposed to 1 atm pressure for all three chamber sessions. They received 100% oxygen at 1 atm during the first session and room air at 1 atm during the second and third chamber sessions.

The outcomes evaluated:

A battery of neuropsychological tests including tests of general orientation, digit span, Trail-Making (Parts A and B), digit-symbol, block design, and story recall was administered immediately following the first and third chamber sessions and at 2 weeks, 6 weeks, 6 months and 12 months.

Primary outcome was the incidence of cognitive sequelae six weeks after randomization. Cognitive sequelae were considered to be present if, at six weeks 1) any T score for a neuropsychological subtest was more than 2 SD below the mean of demographically corrected standardized T scores, or 2) two or more T scores for subtests were more than 1 SD below the mean, or 3) patient reported difficulties with memory, attention, or concentration and the T score on any neuropsychological subtest was more than 1 SD below the mean.

Secondary outcomes were neuropsychological test scores obtained after third chamber session (including those conducted at 6 weeks after CO poisoning), self-reports of symptoms of CO poisoning at six weeks, scores on the Geriatric Depression Scale, the Katz index of daily living activities, the SF-36 at two and six weeks, and the results on the neurologic examination after the third chamber session.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes. This was a double blinded, randomized trial. Randomization was stratified according to loss of consciousness, the interval between end of CO exposure and entrance into the chamber (< 6 hours or ≥ 6 hours), and age (< 40 years or ≥ 40 years). Treatment group assignments were computer generated in blocks of 6.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

460 patients were screened for study participation. 332 of those met eligibility requirements (128 excluded). Of the 332 patients, 152 were enrolled in the study (180 patients declined). 76 patients were assigned to the hyperbaric O2 group, 76 to the normobaric group. 3 patients in the hyperbaric group had incomplete follow-up (no 2 week data in 2, no 6 week data in 1), and 5 in the normobaric group had incomplete follow-up (no 2 week data in 1, no 6 week data in 2, and no 2 week or 6 week data in 2.)

Were patients analyzed in the groups to which they were randomized?

Yes. For the primary outcome variable (the incidence of cognitive sequelae 6 weeks after randomization) the results were analyzed according to the intention to treat principle. Failure to complete the chamber sessions was more in the hyperbaric group compared to the normobaric group (18.4% vs. 3.9%). These patients were still analyzed in the groups to which they were randomized, regardless of the duration of actual treatment.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

Yes. Patients and treating physicians were blinded to the treatment arms. Only the respiratory therapists, who were not involved in the analysis of data, were aware of the treatment group assignments. They operated the chamber controls and monitored the pressure gauges visible only to them. The therapists maintained separate confidential records of the chamber sessions. The investigators performing the neuropsychological tests as well as those that analyzed the data were unaware of the randomization.

4. Were the groups similar at the start of the trial?

The patients were not clinically nor statistically different in age, gender, language, educational level, suicide attempts, symptoms, intubation, need for hospital admission, source of carbon monoxide, interval between exposure and initiation of treatment, or CO levels initially and prior to starting treatment. However, cerebellar dysfunction before treatment was more common in the normobaric oxygen group than in the group assigned to hyperbaric therapy (15 % vs 4%, p = 0.03). The presence of cerebellar dysfunction before treatment (regardless of treatment group) was associated with a higher incidence of cognitive sequelae.

5. Aside from the experimental intervention, were the groups treated equally?

Yes. Details of treatment other than the chamber sessions and neuropsychological tests are not provided but the blinded nature of the study suggests that groups were treated similarly.

III. What were the results?

1. How large was the treatment effect?

The trial was stopped prematurely after the third (of 4 scheduled) interim analyses because the hyperbaric oxygen was judged to be efficacious. 152 patients had been enrolled, 76 patients randomized to each group. Cognitive sequelae at six weeks were less frequent in the hyperbaric-oxygen group (25%) than in the normobaric-oxygen group (45%), p=0.007.

Primary outcome variable: Cognitive sequelae at 6 weeks post exposure

		95% Confidence Intervals
Risk without therapy (baseline risk):X	35/76 = 0.46	0.34 - 0.58
Risk with therapy: Y	19/76 = 0.25	0.14 - 0.35
Absolute Risk reduction: X - Y	0.21	0.06 - 0.36
Relative Risk: Y/X	0.25/0.46 = 0.54	0.341 - 0.855
Relative Risk Reduction: [(X-Y)/X] x 100	0.46	0.141 - 0.657
Number needed to treat	5	3 - 16

Secondary outcomes: The T scores for neuropsychological tests did not differ significantly between treatment groups. (p=0.31). The T scores of both groups showed improvement from immediately after the third chamber session to six weeks (p<0.001), with similar rates of improvement in both groups (p=0.62). Patients treated with hyperbaric oxygen reported fewer difficulties with memory (p=0.004). No significant differences were found between groups for depression, Katz index of daily living, or subscales of the SF-36. Nystagmus after the third chamber session was more frequent in the patients treated with hyperbaric oxygen than normobaric oxygen (12% vs 2.7%, p=0.05). Cognitive sequelae at 6 (p=0.02) and 12 months (p=0.04) were less frequent in the hyperbaric-oxygen group.

2. How precise was the estimate of the treatment effect?

See confidence intervals on above table.

The 95% CI for the ARR ranges from 0.06 to 0.36. If the true effect were only a 6% ARR, then the NNT is 16. This may mean treating a lot of patients with a potentially risky therapy - for an intubated patient, requiring transport and the potential temporary loss of access to the patient while they are in the chamber - for a "soft" endpoint (soft compared to mortality or severe neurologic disability).

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

The incidence of delayed neurologic sequelae may be less in children, although this is difficult to assess because different types of neuropsychological tests would have to be used for children of different ages (1,2). Although the data presented in the study is convincing, it may or may not apply to children. Therefore the findings cannot be applied directly to my patients' care.

2. Were all clinically important outcomes considered?

Yes. The study evaluated a large battery of neuropsychological tests. However, cognitive dysfunction was considered to be present in some patients if two of the neuropsychiatric tests had scores < 1 SD from the mean, or if there was a subjective complaint of memory loss with 1 test < 1 SD below the mean. Scores < 1 SD below the mean may still be considered to be in the normal range.

3. Are the likely treatment benefits worth the potential harms and costs?

Risks associated with hyperbaric oxygen therapy include seizures, barotrauma, anxiety, and oxidative stress (3). Seizures are probably the most severe complication with a reported incidence of 1 in 10,000 patients. None occurred in this study. Seizures are reportedly easy to treat with reduction of oxygen tension and benzodiazepines with no neurologic sequelae. Other risks include those inherent in the transport of patients to an institution where hyperbaric therapy is available. Most cases of carbon monoxide poisoning in children occur during house fires. These children may have associated burn injuries requiring resuscitation and burn wound care. Whether these children would ultimately benefit from hyperbaric therapy is unknown.

Other risks not commented on in this study were the risks of transporting a critically ill patient to a hyperbaric facility, difficulties monitoring patients during the hyperbaric sessions, and the lack of access to the patients while in the chamber.

In this study, failure to complete the chamber sessions was more common in the hyperbaric-oxygen group than in the normobaric-oxygen group (18.4% vs. 3.9%, p=0.005). The primary complications were anxiety and aural barotrauma. The cost in the US for an average 90 minute hyperbaric-oxygen treatment is $300-400. The number needed to treat to prevent one patient from developing cognitive sequelae is 5.

References

1. Choi IS: Delayed neurologic sequelae in carbon monoxide poisoning. Arch Neurol. 1983; 40:433 [abstract]
2. Kim JK, Coe CJ: Clinical study on carbon monoxide intoxication in children. Yonsei Med J. 1987;28:266 [citation]
3. Tibbles PM, Edelsberg, JS: Hyperbaric-Oxygen Therapy. N Engl J Med 1996; 334:1642-48. [citation]

Comments

Back to the EB Journal Club Index