Reason for withdrawal from publication
This review was from The Cochrane Library, Issue 10, 2010 as the review authors were not able to update the review due to time constraints.
The editorial group responsible for this previously published document have withdrawn it from publication.
Feedback
Haemophilus influenzae oral vaccination
Summary
I was concerned about the reporting, analysis and interpretation of the results of the included studies. Specifically the problems I noted were:
1. The numerical values in the abstract don't seem to tally with those in the Forest plot. I'm not sure that giving the WMD to 3 decimal places is justified. 2. What exactly do the numerical values of the WMD difference represent. Is it really infectious episodes? Values in the 20's and 30's in the 3 month follow‐up don't seem compatible with this. 3. Is the meta‐analysis correct? The chi‐square value is astronomical ‐ the analysis needs to be checked. 4. The conclusions need to reflect the variation in results with some studies being favourable and some unfavourable. It is not clear to me that the vaccine is effective from the data presented.
Hope these brief observations help improve the review.
I certify that I have no affiliations with or involvement in any organisation or entity with a direct financial interest in the subject matter of my criticisms.
Reply
Response from authors. The comments made relate to the first issue of this review published in Issue 1, 2001. Since the original publication all the statistics have been verified and clarified. The most instructive response is in the translation of the results into clinical practice, particularly in relation to the frequency of exacerbations and the ability to reduce exacerbations for three months following the taking of the oral vaccine.
As noted in the review, the estimated rate ratio is 0.666 (95% CI 0.500 to 0.887) for zero to three months after vaccination. Assuming a background rate of 0.6 episodes per person, this translates to a reduction of about one episode for every five persons vaccinated (95% CI 3.3 to 14.7). Over zero to six months the estimated rate ratio is less extreme and is not statistically significant at 0.83 (95% CI .67 to 1.03). The background rate is higher, however, at about one episode per person in the first six months, and the number needed to treat is about six persons vaccinated to prevent one episode (95% CI 3 to infinity). The implication for zero to three months vaccination has now been included in the results section with a comment also inserted in the 'Implication for practice'.
Ruth Foxwell Allan Cripps Keith Dear
Contributors
Chris Hyde (21/09/2000)
What's new
| Date | Event | Description |
|---|---|---|
| 14 April 2008 | Amended | Converted to new review format. |
History
Protocol first published: Issue 3, 1998 Review first published: Issue 1, 2000
| Date | Event | Description |
|---|---|---|
| 2 February 2006 | New search has been performed | Searched conducted. |
| 14 March 2003 | New search has been performed | Searched conducted. Review reinstated issue 3, 2003 |
| 6 November 2000 | Amended | This review was withdrawn after comments were received pointing out inconsistencies between the data and the text. The reviewers are currently working on revisions. |
| 27 October 1998 | New search has been performed | Review first published Issue 3, 1998 |
Withdrawn from publication for reasons stated in the review