Dear Mary (and the EBH List)
The Cochrane Library includes a function that allows you to reanalyse the
presented data using odds ratios, risk ratios or risk differences, and to
switch between fixed and random effects analyses. There are two routes by
which you can access this:
With a Cochrane Review selected on the screen
(a) either use the structure on the LHS of the screen of page down until
you reach the "Analyses" and display the forest plot for the comparison and
outcome you are interested in, then click on the highlighted text in the
figure legend "Show Statistical Analyses"
(b) select "Figures (fullsize)" from the structure on the LHS of the screen
and display the forest plot for the comparison and outcome you are
interested in, then click on the highlighted text in the figure legend
"Show Statistical Analyses"
Once on the statistical analyses screen there are drop downs to allow you
to select an alternative statistic to that chosen by the reviewer for analysis.
However, there may be good reasons why reviewers have chosen a particular
statistic - for example reviews rarely choose the risk difference statistic
as this is least likely to be consistent across the studies.
Any comments on whether this feature of the Cochrane Library is valuable,
and how it could be made more accessible would be gratefully received
BW
Jon Deeks
Centre for Statistics in Medicine
Oxford
At 13:34 20/02/2006 +0000, Olive Goddard wrote:
>Dear Colleagues,
>
>Would you be kind enough to respond to Mary directly if you can help
>her.
>
>All good wishes,
>
>Olive
>
>Olive Goddard
>Centre and Editorial Manager
>Centre for Evidence-Based Medicine
>Department of Primary Health Care
>Old Road Campus, Headington
>Oxford, OX3 7LF
>.....................................................................
>Tel: +44 (0)1865 226991 email: [log in to unmask]
>Fax: +44 (0)1865 226845 web: www.cebm.net
>Mobile: 0773 484 2403 web: www.cebmh.com
>
> >>> "Mary Publicover" <[log in to unmask]> 02/20/06 12:01 pm
> >>>
>Dear Olive
>
>I wonder if you could refer this question to someone who could help
>me?
>
>
>
>I'd be grateful for some advice about using CATmaker to appraise
>Cochrane systematic reviews.
>
>
>
>The CATmaker software requires an odds ratio, Cochrane reviews seem to
>(mostly?) use relative risk. It's easy enough to calculate the odds
>ratio from the event rates but:
>
>1. I don't know how to get the confidence intervals, and if it's a
>complex formula it may not be practical for everyday use.
>2. Is it appropriate to use the pooled figures. Do these take
>account adequately of the different weightings of the different
>studies
>ie high variance for a small study.
>
>
>
>Are there any work arounds you would suggest eg using a single study?
>We
>tend to use RR and ORs interchangeably for smaller RRs (ie <3) Is this
>justified?
>
>
>
>Sorry if these are stupid questions- and complicated. I'd like to get
>CATmaker in more general use in our Trust, just to record and give
>prominence to appraisal work going on. I want to make it as easy as
>possible for people (and easier for me of course...)
>
>
>
>Thanks
>
>
>
>Mary
>
>
>
>Mary Publicover
>
>Trust Clinical Librarian
>
>0121 627 5846
>
>x8746
>
>
>
>
>DISCLAIMER:
>This e-mail and any attachments hereto contains proprietary
>information, some or all of which may be confidential or legally
>privileged. It is for the exclusive use of the intended recipient(s)
>only. If an addressing or transmission error has misdirected this e-mail
>and you are not the intended recipient(s), please notify the author by
>replying to this e-mail. If you are not the intended recipient you must
>not use, disclose, distribute, copy, print, or rely on this e-mail or
>any attachments, as this may be unlawful.
________________________________________________________________
[log in to unmask]
Jon Deeks
Senior Medical Statistician
Centre for Statistics in Medicine
Wolfson College Annexe
Linton Road, Oxford OX2 6UD
UK
Phone +44-(0)-1865-284403
Fax +44-(0)-1865-284424
________________________________________________________________
|