Corpus of Doctor-Patient Consultations from Ahus
The Corpus of Doctor-Patient Consultations from Ahus is a unique corpus of transcribed dialogue in Norwegian between doctors and patients in different types of consultations at Akershus University Hospital (Ahus). The audio files are not available in the corpus due to their sensitive nature.
The transcriptions in the corpus are based on video recordings of consultations between doctors and patients/relatives at Akershus University Hospital (Ahus) in 2007 and 2008. The material was collected in connection with a study on the effect of a course in communication for doctors. From the hospital's 249 doctors, 103 were selected at random, of whom 71 agreed to participate.
The doctors represent all non-psychiatric clinical disciplines (internal medicine, surgery, orthopedics, gynecology, pediatrics, neurology, ear-nose-throat, anaesthesiology) at the hospital. Each doctor participated in up to eight recordings, each with a different patient, totalling at 497 recordings. The material was collected in outpatient clinics, emergency rooms and wards. Patients were asked to participate while waiting for the doctor, with 94% giving their consent. Informed consent was obtained from both doctors and patients.For 380 of the recordings, additional consent was obtained to store video recordings until 2020, for further studies. The remaining videos were deleted in 2011. The Regional Ethics Committee for Medical Research has since approved lasting storage of transcriptions for research.
The study was lead by Professor Pål Gulbrandsen, who collaborated with Professor Arnstein Finset on its planning. Both are based at the Faculty of Medicine, University of Oslo. Almost 90% of the material was collected by the then fellow, now PhD Bård Fossli Jensen, the rest was collected by Gulbrandsen.
The transcription work was organized by the Text Laboratory and started with the 113 consultations that were to be deleted in 2011. All of these transcriptions are now part of version 2 of the corpus. Version 2 (June 2015) contains 220 consultations and over 950,000 words.
The consultations are orthographically transcribed with Norwegian Bokmål using a transcription guide written in collaboration with Professor Jan Svennevig, Department of Linguistics and Scandinavian Studies, UiO.
Read the Transcription Guide in Norwegian here.
The transcriptions are tagged morphologically with a Treetagger trained on proofread output from the Oslo-Bergen tagger. The tagger was first developed for the NoTa Corpus. and has achieved 96.9 % (10-fold cross validation performance), see
Nøklestad, Anders and Åshild Søfteland (2007). Tagging a Norwegian Speech Corpus. NODALIDA 2007 Conference Proceedings.
Refer to the corpus as follows:
Gulbrandsen, Pål, Arnstein Finset, og Bård Fossli Jensen. 2013. Lege-pasient-korpus fra Ahus. Tekstlaboratoriet, ILN, UiO.
NB! Gulbrandsen, Finset og Fossli Jensen want the users of the corpus to send a message to them that the corpus is used, and in what kind of research. Where it is natural to draw medical or psychological expertise into the research, Gulbrandsen, Finset and Fossli Jensen should be asked whether they wish to participate, before eventually seeking expertise elsewhere. Contact: pal.gulbrandsen by medisin.uio.no
The recordings and the first study were funded by Helse Øst (later Helse Sør-øst) and Ahus. The Norwegian Medical Association and The Norwegian Directorate of Health financed a pilot study that made the project possible. The transcriptions are funded by the University of Oslo.
These people have so far been involved in building the corpus::
Professor Pål Gulbrandsen, professor Arnstein Finset and dr. med. Bård Fossli Jensen (Faculty of Medicine, University of Oslo), professor Jan Svennevig (Department of Linguistics and Scandinavian Studies, UiO) and professor Janne Bondi Johannessen, language engineers Kristin Hagen and Michal Kosek, transcribers Benedikte Homme Holstad, Christine A. Nikolaisen, Helene Aarem, Ida Stang, Kjersti Wictorsen Kola, Kristine M. Stensland, Sine Halkjelsvik Bjordal, all from the Text Laboratory, Department of Linguistics and Scandinavian Studies, UiO.
1.Gulbrandsen P, Fossli Jensen B, Finset A. Endring i mestringstillit hos sykehusleger etter kurs i klinisk kommunikasjon. Tidsskr Nor Legeforen 2009; 129: 2343-6.
2. Gulbrandsen P, Fossli Jensen B. Post recruitment confirmation of informed consent by SMS. J Med Ethics 2010; 36: 126-8.
3. Fossli Jensen B, Gulbrandsen P, Dahl FA, Krupat E, Frankel RM, Finset A. Effectiveness of a short course in clinical communication skills for hospital physicians: results of a crossover randomized controlled trial (ISRCTN22153332). Patient Educ Couns 2011; 84; 163-9.
pal.gulbrandsen at medisin.uio.no
tekstlab-post at iln.uio.no