Clinical studies

Since HBO was first used as a treatment in CO poisoning in 196061, treatment guidelines have been developed on the basis of clinical experience. Numerous uncontrolled studies reported lower mortality and morbidity in patients treated by HBO15-18. In the late 1980s, HBO treatment came under criticism due to the lack of prospective controlled studies supporting its use in CO poisoning.

This last 15 years, 6 prospective randomised trials have been reported comparing HBO and NBO for CO poisoning19-23,61. Of the 6 trials, 4 demonstrated better clinical outcome among patients receiving HBO while 2 showed no effect.

The first randomized study by Raphael et al19 included 343 patients without loss of consciousness, treated either by HBO (2 ata for 60 min) or NBO (1 ata for 6 hours). There were no significant differences in PNM at one month after poisoning. A negative conclusion was drawn from 286 patients with loss of consciousness treated with one or two HBO sessions. This study was criticized for using overly broad inclusion criteria; an inadequate HBO regimen; inappropriate time of assessment; and unsuitable outcome measures.

A second prospective trial was performed by Ducasse et al20 and included 26 non-comatose patients treated by HBO or NBO. Evaluation was done by clinical examination, electroencephalogram and cerebral blood flow response to acetazolamide. A significant benefit at 3 weeks was found in the HBO treated group. Limitations of this study included its small sample size and the use of surrogate outcome measures.

A third study by Thom et al21 included 60 patients with mild CO poisoning, specifically excluding those with a history of unconsciousness or cardiac disturbances. After randomization, patients were treated by HBO (2.8 ata for 30 minutes, followed by 2 ata for 90 minutes) or NBO (until symptom relief). Patients were followed with neurological testing. Persistent neurological manifestations were found in 7 of 30 (23 %) patients treated by NBO and in none of the patients treated by HBO (p<0.05). Neurological manifestations persisted for an average of 6 weeks and often interfered with normal daily activities. The trial was stopped early due to the obvious treatment effect but has consequent limitations of having a small sample size. In addition it was not double-blinded and a relatively large number of patients were lost to follow-up.

A fourth randomized trial performed by Scheinkestel et al22 enrolled 191 CO poisoned patients of different severity. Patients were treated either by

HBO (1 session a day, 3 ata, 60 minutes with intervening NBO for 3 to 6 days) or NBO for 3 to 6 days. Outcome measures were clinical evaluation and neuropsychological testing after treatment and one month after. No beneficial effect was found. There were numerous flaws: firstly, most of the CO poisonings were suicide attempts with associated ingestion of alcohol and other drugs that may have interfered with the results of neuropsychological testing. Secondly, neither the HBO nor NBO protocols were consistent with current practice. The HBO group received only 7 % more oxygen than the NBO group. The small difference in oxygen dosing would be expected to disguise an HBO effect in favour of an unattainable long normobaric regimen requiring admission that would have been more costly than repetitive HBO therapy over a shorter period of time. Finally, less than half of the patients completed the one-month follow-up.

These conflicting results have fed the controversy. Recently, however, a randomized, prospective, double-blinded study by Weaver et al61 has reported unequivocal beneficial results of HBO treated patients. One hundred and fifty two patients were included and randomized to receive either HBO (3 sessions in 24 hours) or NBO. An extensive neuropsychological test battery was performed at 6 weeks and one year. Cognitive sequelae were significantly lower at 6 weeks in the HBO treated patients than in NBO treated patients (25 % versus 46 %, p<0.07). The beneficial effect persisted at one year (4 vs. 15 %, p<0.04).

A post hoc analysis of the results of this study26 showed the beneficial effect of HBO was present in the following patients: age over 50 years; an episode of loss of consciousness during exposure; COHb levels over 25 %; and metabolic acidosis (BE lower than -2mEq/l). HBO had no advantage to NBO in patients without these criteria.

Another prospective randomized trial has been published but only as an abstract23. It provided an interim analysis of a prospective multicenter study. Five hundred and seventy five non-comatose patients were randomized to receive either HBO (1 session, 2.5 ata, for 90 minutes) or NBO (12 hours). Follow-up was done at 1, 3, 6, 9 and 12 months. A significant difference in favour of HBO existed at 3 months (8.7 % vs 15.2 %, p<0.016). The difference lessened at 6 months and disappeared at 1 year. However the reduction in morbidity within the first six months is important to consider and would have great economic impact on the ability.

In summary, following the recent Cochrane's review62 we conclude that HBO is probably not always required for CO poisoning but rather should be guided by severity: Certain sub-groups of patients seem to benefit more from HBO than NBO - particularly cases with neurological involvement (coma, loss of consciousness or objective neurological manifestations). Psychometric testing might be useful in identifying those with minor neurological involvement who are nevertheless at risk for PNM, but may be overlooked in the absence of specific screening.

0 0

Post a comment