On several occasions I have called errors to the attention of the
NLM indexers.  I remember at least 2 occasions that the subject headings did
not reflect something in the title even though there was an almost exact
match between the two (e.g. PAH in the titile but no MeSH heading for
Polycyclic Hydrocarbons, Aromatic).  They did make those corrections--I went
back and checked at a later date.  They are usually quite responsive when I
have pointed out obvious errors (like leaving off the first author) and have
sent me a nice response like the one below:

Thank you for reporting a possible error in the National Library of
Medicine's (NLM) MEDLINE database. We appreciate your concern and time taken
to help us maintain the quality of the database....

        I think they are willing to take any requests/suggestions into
consideration since it affects the quality of their database.  They may not
always agree, but as you point out, an indexer on one day may see things
differently than the original indexer.

        I try to send emails (lots easier than writing a letter like in the
past) any time I run across something the I feel needs to be looked at more
closely.  If we want a quality database to search, shouldn't we help
maintain it to ensure that quality?

Auburn Steward, MLIS, AHIP
Center for Toxicology & Environmental Health
4301 W. Markham--Slot 767
Hendrix Hall--Suite 100A
Little Rock, AR 72205
501-614-2835 (fax)
[log in to unmask]

> -----Original Message-----
> From: Webster Library [SMTP:[log in to unmask]]
> Sent: Thursday, March 25, 1999 1:40 PM
> To:   [log in to unmask]
> Subject:      Re: CHAT: Data Cited In Journal Abstracts Often Do Not Match
> Dat a In Text (fwd)
> In respnse to T. Scott Plutchak'd interesting comment on how he was
> instructed to index at NLM:
> ............I have often wondered why a term or
> key word will appear in the title or the abstract, but not be indexed for
> the corresponding Mesh (and thus missed unless a textword search in done).
> It seems to me, however, that when the author is an MD writing for other
> MDs, he might have a good reason to call attention to a word or a term.
> Most person who ask for searches love to specify "key words" and want to
> see those in the title or the abstract.  Over the years I had to explain
> over and over again about controlled vocabulary, and the (very possible)
> subjectivity of the indexing experts at NLM, etc.  Has anyone ever "test"
> the indexers by presenting the same article at different times under
> different work load/conditions, to see how consistent the results? Or
> pitting an idexer against the author?
> Dalia Kleinmuntz                                          847/570-2664
> Webster Library                                      FAX: 847/570-2926
> Evanston Hospital
> 2650 Ridge Ave
> Evanston IL 60201                                 [log in to unmask]
> ______________________________________________________________________
> "..the secret of the care of the patient is in caring for the patient"
>                                                 - Francis W. Peabody
>                                                      (1881-1927)