I think that this is an excellent idea albeit a mite ambitious. One of the disappointing aspects of ebm at the moment is the paucity of the evidence - I have done a tremendous number of "live sessions" of just the focusing and searching components and would estimate retrieval of relevant evidence (i.e. an exact match of focused question to retrieved item as about 1in 3 or 1 in 4). To illustrate - as replication of the famous Ellis study I have conducted over 300 searches on both Cochrane and Medline for Problem Intervention pairs in A&E (i.e. real presentations) - if your requirement is very forgiving (i.e. right problem right intervention) you may find one or two relevant items but otherwise it is very difficult to match the articles retrieved to the exact question (i.e. it will have different outcomes, different comparisons, answer a broader question or answer a more specific question).
One important confounding factor is what I call "Muir Gray's law of inverse sexiness" (reported in Muir Gray's book Evidence Based Healthcare but based on work by Franckel and West) that there are relatively plenty of articles about less common conditions (articles per 1000 deaths and discharges) but few about the most prevalent (CJD at the top, varicose veins at the bottom)
The usefulness of the literature is increased immeasurably once you have access to the full text rather than just the abstract. However, this usually means that your question is more likely to be covered by the background or discussion rather than the experimental results. (so not truly ebm!)
For real time I would go for a "CAT-challenge" where in a big medical library with 300+ journal titles a group aims to produce a Critically Appraised topic within a half-day.
Sorry if all this sounds pessimistic. Notwithstanding my above comments I do think you should give it a go! If I was in your situation I would stage manage it a little bit more by broadening the base of candidate questions and then selecting the most likely to succeed. With my experience of literature searching I would suggest that the ideal topic would have the following characteristics:
1. It would be a therapy question
2. It would preferably be about a single pharmaceutical intervention (i.e. not adjuvant therapy)
3. There would be a well-established previously gold standard comparison (i.e. so most of the trials were likely to compare the same intervention-comparison pairs) and the intervention of interest must have been in existence for at least three years (excluding very new drugs)
4. We would brainstorm a number of possible outcome measures and be satisfied with retrieval of any or some of these.
5. We would get prior agreement that our ultimate decision would be based on best available evidence not on ideal evidence and that "no decision" is not an option. (i.e. to stress EBM is about pragmatic clinical decision making).
My classic exemplar would be one of the new drugs for schizophrenia, beta-interferon for multiple sclerosis etc.
I appreciate that this is beginning to sound like prestidigitation rather than real time EBM but I think the nature of the event you describe requires some degree of likelihood of success.
Please let the list know how you get on!
Andrew
----------
From: Ati Yates
Sent: 09 December 1998 15:00
To: evidence-based-health
Cc: [log in to unmask]; [log in to unmask]; [log in to unmask]; [log in to unmask]; [log in to unmask]
Subject: Teaching EBM in real time
Hello list members,
We have recently given a symposium on EBM and critical appraisal at a
national meeting of CL psychiatrists here in the US. It generated quite a
bit of interest and we are planning our proposal for a follow up at next
year's meeting. A big part of the fun in EBH work as I see it is its
continued push for further development.
I wonder if anyone has experimented with doing a "live" EBH teaching
symposium in which--(after a short, clear, concise and well structured
"basics of EBM--the 5 step process" introduction with large lettered
handouts showing the essential steps given to participants):
(1) the audience is stimulated to provide a patient scenario and
question (as Scott Richardson has taught us in uproarious and energetic
style);
(2) then we, the panel (in our case 3 for now) with laptop, LCD and
continued audience participation further formulate the question as needed;
...and (3) --deep breath--do a hierarchial search---> Cochrane on CD, Best
Evidence, anything else developing out there? followed by either (a) a Pub
Med (or other) modem connected search by ourselves followed by a librarian
on the panel who does the thing for us AFTER we have exhausted our growing
but still clinician-level capabilities
OR (b) a PubMed connected search via modem by ourselves AND a
librarian at a remote location who is kept privy to the entire process by
real time email who can then help us out by working out a higher level of
the search which we can then put in.
We might well be able to engineer 2 modem connections in a
hotel/conference site, but won't have a roomful of computers at our
disposal. We could have 2 LCDs and screens--one for communication with
librarian and one for searching.....(smile)
One aim of this, of course (other than to entertain and at the same time
avoid losing too much sleep or hair before or after the event) is to
improve the understanding of medical librarian consultation and make sure
the "show can go on" from there.
As for the next steps--the critical appraisal and application to patient,
we could be flexible and go several ways---if an available article is
discovered it could be faxed to us and we could copy for all during coffee
break and we could also be ready with another case and article if this
didn't happen.
I would highly value any ideas or experience you might have and will send
back a compilation/summary to the list. Would also greatly value any list
discussion.
thank you very much,
Ati Yates
______________________
Ati Yates, M.D.
Internal Medicine and Psychiatry
Michigan State University
Mailing address: 6092 Beechwood Drive,
Haslett, MI 48840
Phone: W 517 353 4362
H 517 339 5037
Fax: 517 339 5569
2nd Fax: 517 432 3603
--------------------------------------------e-mail: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|