And probably the reason for this is that quite a few folks in the EBH community (including me, for what it's worth) believe that one should avoid assessing therapy using observational studies whenever possible. Since most physicians make treatment decisions purposively, patients who receive one therapeutic alternative in practice are likely to be very different from those who receive another. Statistical adjustment for such differences is often difficult, especially because it is hard to think of, much less control for every dimension across which patients may differ. Thus, the first question in the original Users Guide for therapy was whether the study was an RCT. Hence one should only be searching for observational studies of a particular treatment for a particular problem if there are no passably methodologically sound RCTs of that treatment for that problem available.
Roy M. Poses MD
Center for Primary Care and Prevention
111 Brewster St.
Pawtucket
RI 02860
401 729-3400
fax 401 729-2494
email: [log in to unmask]
________________________________
From: Evidence based health (EBH) on behalf of Mick Arber
Sent: Thu 10/28/2004 1:25 PM
To: [log in to unmask]
Subject: Re: Study design filters for therapy observational studies
Hi Paul,
Thanks very much for your reply.
I have taken a look at these before, but I'm not sure how well they would work to retrieve therapy papers, as oppose to etiology or prognosis. What do you think?
There is a therapy filter on PubMed, but it doesn't seem designed to retrieve observational studies (or all cohort / case-control / case-series).
Thanks for your help,
best wishes,
Mick
Mick Arber, Information Specialist
BMJ Knowledge
BMJ Publishing Group, BMA House
Tavistock Square, London. WC1H 9JR
http://www.clinicalevidence.com
http://www.besttreatments.co.uk
Tel: 020 7383 6385 Fax: 020 7383 6242
|