Corpora: Summary: Doctor-Patient Consultations

From: Mark G Lee (M.G.Lee@cs.bham.ac.uk)
Date: Fri Oct 26 2001 - 13:28:26 MET DST

  • Next message: German Rigau Claramunt: "Corpora: Mapping WordNet1.6 to WordNet1.7"

    Hello

    Here is a summary of my earlier question regards corpora of
    doctor-patient interactions.

    Thanks to :

    Matthew Purver Kings College London
    Harold Somers UMIST
    Francis Rock University of Roehampton
    Bill Mann SIL International
    David Lee University of Lancaster
    Nick Smith University of Lancaster
    Yorick Wilks University of Sheffield
    Lex Olorenshaw SLT
    Jochen Leidner University of Cambridge
    David Oakey University of Birmingham

    What follows are selected quotes from everybody and a few references:

    Matthew Purver/David Lee/Harold Somers all suggested the British
    National Corpus which contains about 100 examples of short (300900
    words) medical consultations in GP surgeries or hospitals, already
    annotated for POS tags and some other aspects.

    David Lee writes

    >A concordance on 'pain' reveals 68 hits in 38 unique files among these
    >119. This means that you potentially have a corpus of 38 files (with a
    >total of 34,010 words) to work with (though whether the pain is
    >'chronic' or not you'll have to judge for yourself). You can find these
    >doctor-patient files using my BNC Indexer (see
    >http://www.davidlee00.freeserve.co.uk/corpus_resources.htm or the Web
    >Indexer at
    >http://www.comp.lancs.ac.uk/computing/research/ucrel/bncindex/).

    Francis Rock/Nick Smith/Harold Somers/Jochen Leidner all suggest looking
    at Jenny Thomas & Andrew Wilson's work

    Jochen writes:

    > Jenny Thomas <http://www.bangor.ac.uk/ling/staff/els01a/jenny.htm>
    > is probably one of the main author to read if you're interested in
    > doctor-patient interaction. During her time at Lancaster, Jenny did
    > two studies investigating communication with respeact to cancer patients:
    >
    > - Evaluating the Quality of Communication within the Lancaster
    > Cancer-Care System (1989-93) and
    >
    > - A Comparative Study of Patient Satisfaction with Cancer Treatment in
    > Rural and Urban Practices within the Lancaster District Health
    > Authority (1992-94).
    >
    > There are a couple of paper's around that document these efforts,
    > including descriptions how corpus-based methods have been applied.
    >

    Harold also writes:

    >Several other similar corpora have been collected, e.g. by Thomas and
    >Wilson (1996), Wynn (1999). Other researchers have collections of tape-
    >recordings (See for example
    >ww2.mcgill.ca/Psychiatry/transcultural/primary.html) and there are even
    >conferences dedicated to the analysis of doctorpatient discourse (For
    >example, the Conference on Medical Interaction, 18-20 October 2000, at
    >the University of Southern Denmark, Odense. See
    >http://www.conversation-analysis.net/Conferences/Medical/program_doc-pat.htm.

    >Thomas, J. and A. Wilson. 1996. Methodologies for studying a corpus of
    >doctor-patient interaction. In J. Thomas and M. Short (eds) Using
    >Corpora for Language Research, London: Longman, pp. 92109.

    >Wynn, R. 1999. ProviderPatient Interaction: A Corpus-Based Study of
    >DoctorPatient and StudentPatient Interaction.< Kristiansand, Norway:
    >Hyskoleforlaget.

    Bill Mann writes

    >I am just now studying a book on this: The Discourse of Medicine, by
    >Elliot G. Mischler, a 1984 book from Ablex. He discusses the art of
    >studying medical interviews, and in particular he identifies a number
    >of flaws in past work. He studies several actual cases, useing a
    >Conversational Analysis approach.

    >I found it useful at the framework and criticism levels, and there are
    >several dialogues in there that I will probably use to exercise a
    >theory.

    David Oakey writes

    >I saw a paper ("Just Checking: Questions and Social Roles") by Geoff
    >Thompson at the Questions conference in Liverpool a couple of years ago
    >where he looked at evidence from his own (I think) corpus of doctor-patient
    >consultations. He referred to his paper Thompson G. 1999 "Acting the part:
    >lexico-grammatical choices and contextual factors." in M. Ohadessy (ed.) Tex<italic>t
    >and Context in Functional Linguistics.</italic> Amsterdam: Benjamins, 103-126.

    Yorick Wilks pointed me to Bonnie Webbers work on the TraumAID project
    http://www.cis.upenn.edu/~traumaid/

    and also (with Alison Cawsey & Ray Jones)

    A. Cawsey, B. Webber & R. Jones
    Natural Language Generation in Healthcare, Journal of the American Medical Informatics Association 4(6) November - December 1997, pp. 473-482

    Lex Olorenshaw suggests looking at

    http://smi-web.stanford.edu/projects/miigweb/

    (also available at http://www.dai.ed.ac.uk/homes/bonnie/bonnie.html)

    Thanks to everybody for the advice,

    Best regards

    Mark

    Dept of Computer Science | mgl@cs.bham.ac.uk
    University of Birmingham | http://www.cs.bham.ac.uk/~mgl



    This archive was generated by hypermail 2b29 : Fri Oct 26 2001 - 13:51:05 MET DST