2 Nachteile von PubMed

["Oliver Obst" <obsto@uni-muenster.de>]

Zur Info, O.Obst
------- Forwarded Message Follows -------
Date:    Thu, 29 May 1997 16:15:21 -0400
From:    "MIGNON ADAMS, PHILADELPHIA COLLEGE OF PHARMACY AND SCIENCE"
         <adams@SHRSYS.HSLC.ORG>
Subject: Why anyone would choose to pay for Medline data...

At 03:19 PM 5/26/97 -0500, you wrote:
>A quick update.  The PubMed URL is now:
>
>http://www.ncbi.nlm.nih.gov/PubMed/
>
>This should remain stable.  The numbered www4 and www3 sites were still
>testing pre-launch.  Why anyone should want to access the Medline data from
>a subscription and/or fee-based service beats me.


----------------------------------------------------------------------------
---------------

A recent poster inquires why anyone would pay for Medline access. The reply
to this is that you pay for the search engine. The web-based database above
is a perfect example of you-get-what-you-pay-for.

The PubMed database, available  free "at least till the end of the year,"
does not map to MESH subject headings. If you use the MESH term field
(supposing that you know offhand the correct MESH heading), the term will
be automatically exploded. If you search  "all fields," the term is NOT
exploded. These two crucial missing features mean that searches done in
PubMed will  probably miss many, many pertinent articles. (Subheadings,
says the documentation, are automatically included; if you want to use
them, you must enter them as part of your search strategy).

To illustrate: I searched for tricyclic antidepressants in both PubMed and
in Ovid's Medline (generally considered to be one of the best Medline
search engines).

Here are my results:

                     PUBMED

searching under:
tricyclics:      453 citations


tricyclic
antidepressants  2765

antidepressive
agents,
tricyclic, (MESH
term field)              4283

antidepressive
agents,
tricyclic (all fields)  5073


OVID

Searching under either "tricyclics" or "tricyclic antidepressants"
automatically mapped me to the correct MESH term, "antidepressant agents,
tricyclic." I exploded the term, as both a minor and major MESH term, chose
all subheadings, and got:

 *********19,811 citations***************

The results really should not have been this dramatic (I suspect some
problems with PubMed's programming, or maybe they don't really have
MEDLINE back to 1966 like they say they do), but medical librarians
testing out the "free" Medline search engines on the Internet have
consistently found them to be inferior.

If you only want a few citations on a topic, and you have a simple search,
then sure, go ahead and search the free Medlines. If you want a
comprehensive search, or a complex one with a number of terms, then you
ought not to depend on inferior search engines.  Remember, you can have
cheap, fast, or good, but only 2 of these at a time.



*******************************************************************************
  Mignon Adams
  Director of Library Services and Information Technology
  Philadelphia College of Pharmacy and Science              Voice: 215-596-8791
  600 S. 43rd St.                                             Fax: 215-596-8760
  Philadelphia, PA 19104                        Internet: adams@shrsys.hslc.org

*******************************************************************************