I don't know anything about Knowledge Finder. What I do know is that Medline
is terrible about assigning study types to publications. Partly this may due
to the fact that the assignments are based on what is described in the
abstract; however, even just reading the abstracts I find controlled studies
not listed as such, and clinical trials not listed as such. I have even
found non-randomized trials listed as randomized; and likewise, randomized
trials not listed as randomized. Also, many articles listed only as reviews
are actually original trials that include a review of the topic. The people
doing these assignments over the years have done a very poor job, even
within the few broad categories Medline purports to recognize. The system is
broken and totally untrustworthy. Is case-control even a category that
Medline claims to use? Unfortunately, the only solution is to scan the
abstracts yourself and identify the type of study you want. Even then you
will include some studies that are not the type you want and miss some that
are, because the abstracts are incomplete or misleading.
David L. Doggett
-----Original Message-----
From: Simon, Steve, PhD [mailto:[log in to unmask]]
Sent: Thursday, December 28, 2000 11:31 AM
To: [log in to unmask]
Subject: A beginner's question about Medline
I'm trying to conduct a Medline search for case-control studies in the
Journal of Andrology using software called Knowledge Finder that our
hospital provides. When I select case-control from the MeSH thesaurus, it
gives me a bunch of studies that are retrospective, and highlights the word
"retrospective". Some of these studies are clearly not case-control studies.
Under publication types, I can't find an appropriate selection. I could
limit my search to randomized controlled trials or clinical trials, but not
case-control (or for that matter observational or cohort designs).
Why can't I limit my search to just case-control designs? Is this a
limitation of the Knowledge Finder software or of the Medline database
itself? Or am I just missing something obvious? If it is a limitation of
Knowledge Finder, what software should I try to lobby for?
Steve Simon, [log in to unmask], Standard Disclaimer.
STATS: STeve's Attempt to Teach Statistics. http://www.cmh.edu/stats
|