Response to Letter by Helms et al
We thank Dr Helms et al for their interest in our summary.1,2 Readers should note that Graeme Hankey is the editor of the section of the journal in which the summary appears, rather than the principal author as Helms et al suggest. He should not be held responsible for any errors we as authors may have inadvertently committed.
Before addressing the specific comments in the letter, we make the point that the full conduct and interpretation of our review is difficult to convey in the limited space available for this summary. We apologize for any inadvertent misunderstanding that may have arisen in our attempts to summarize the review and refer readers to that document for a better understanding of our conclusions.3 It is important to appreciate that Cochrane reviews are systematic reviews with meta-analysis of the randomized evidence. They are not guidelines for practice where insufficient evidence exists in the randomized literature.
We are pleased the correspondents agree with the statement that “there is no level 1 data to support the use of HBO in the treatment of acute ischemic stroke”.1 More specifically, our conclusions were that “there is insufficient evidence to suggest that HBOT significantly affects outcome after acute ischemic stroke” and that “we cannot be certain a benefit from HBOT has been excluded”. We go on to suggest further studies will need to be carefully planned and raise the issue of the timing of therapy and the appropriate oxygen dose.2 We are somewhat confused by the correspondents’ assertion that the 3 included randomized clinical trials were not adequate to support these conclusions.
Helms et al state, “At this stage of the evidence we feel it is too early to make any conclusions regarding HBO …… and thus no statement should be made about this as-yet untested modality.” It is possible there is cultural misunderstanding here, but we are at a loss to see the purpose in taking such a position, and it seems inconsistent with their own statement referred to in paragraph 3 above. We take the absence of evidence of benefit to indicate only that—not to imply that we have evidence of an absence of effect. We were careful to point out that a possibility of benefit remains and could be explored by future researchers. This may be exactly the position of Helms et al, and if so we can only agree.
Helms et al also criticize our definition of HBOT and specifically the inclusion of pressures <2.0 ATA. We agree that most treatment centers use either 2.0 or 2.4 ATA, but many practitioners who advocate the use of HBOT for neurological problems including stroke, strongly advocate lower pressures.4,5,6 Because there is no sound evidence for a minimum effective pressure for any indication and the exclusion of studies using <2.0 ATA would leave us open to criticism from proponents of “low pressure” HBOT, we felt it was prudent to allow such studies to be included in the review. Our conclusions would not be substantially altered if those studies were excluded—there is no evidence of benefit from randomized clinical trials in this area.
We agree there are significant shortcomings in all 3 published randomized clinical trials. Nevertheless, they do represent the most methodologically trustworthy attempts to assess any therapeutic benefit of HBOT in this area, and we have evaluated them on an a priori set of inclusion criteria. We applaud the authors of those 3 trials for making the effort to illuminate a difficult treatment area and hope in the future that others will perform even better clinical investigations.
Helms et al also point to an error in the text of the full review. The details of treatment pressures for Nighoghossian et al7 are recorded correctly in our table of the characteristics of included studies but were transcribed incorrectly into the text. We are grateful for the correction and will amend the text in the next iteration of this review (due in 2007). One of the strengths of the Cochrane approach is that any such errors can be corrected in the contemporary version of a review, and it is expected that not all readers will take the view that this “calls the entire report into question”. The details for Anderson et al are correctly represented in our review (sham air at 1.5 ATA versus 100% oxygen at 1.5 ATA).8
We look forward to future well-conducted clinical research in this area with careful recruitment of patients experiencing truly acute ischemic stroke, randomization to a range of treatment pressures and the measurement of appropriate clinical outcomes. Until then, we must interpret what is presented to us.
Helms AK, Whelan HT, Torbey MT. Hyperbaric oxygen therapy of acute ischemic stroke. Stroke. 2007; 38: 1137.
Bennett MH, Wasiak J, French C, Kranke P, Schnabel A. Hyperbaric oxygen therapy for acute ischemic stroke. Stroke. 2006; 37: 1953–1954.
Bennett MH, Wasiak J, Schnabel A, Kranke P, French C. Hyperbaric oxygen therapy for acute ischaemic stroke. The Cochrane Database of Systematic Reviews. 2005, Issue 3. Art. No.: CD004954. DOI: 10.1002/14651858.CD004954.pub2.
Neubauer RA, James PB. Cerebral oxygenation and the recoverable brain. Neurol Res. 1998; 20(Suppl 1): 533–536.
Jain KK. Hyperbaric oxygen in acute ischemic stroke. Stroke. 2003 ;34: e153.
Nighoghossian N, Trouillas P, Adeleine P, Salford F. Hyperbaric oxygen in the treatment of acute ischemic stroke: a double-blind pilot study. Stroke. 1995; 26: 1369–1372.
Anderson DC, Bottini AG, Jagiella WM, Westphal B, Ford S, Rockswold GL, Loewenson RB. A pilot study of hyperbaric oxygen in the treatment of human stroke. Stroke. 1991; 22: 1137–1142.