Modeling Keyword Search Strategy: Analysis of Pharmacovigilance Specialists' Search of MedDRA Terms

Autor: Romaric, Marcilly, Laura, Douze, Cedric, Bousquet, Sylvia, Pelayo
Přispěvatelé: CHU Lille, Evaluation des technologies de santé et des pratiques médicales - ULR 2694 (METRICS), Université de Lille-Centre Hospitalier Régional Universitaire [Lille] (CHRU Lille), Centre d'Investigation Clinique - Innovation Technologique de Lille - CIC 1403 - CIC 9301 (CIC Lille), Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lille-Centre Hospitalier Régional Universitaire [Lille] (CHRU Lille), Laboratoire d'Informatique Médicale et Ingénierie des Connaissances en e-Santé (LIMICS), Université Paris 13 (UP13)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Sorbonne Université (SU)
Jazyk: angličtina
Rok vydání: 2019
Předmět:
Zdroj: Studies in Health Technology and Informatics
Studies in Health Technology and Informatics, 2019, 257, pp.298-302. ⟨10.3233/978-1-61499-951-5-298⟩
ISSN: 0926-9630
1879-8365
DOI: 10.3233/978-1-61499-951-5-298⟩
Popis: International audience; In the information retrieval task, searching and choosing keywords to form the query is crucial. The present study analyzes and describes the keywords' search strategy into a thesaurus in the field of pharmacovigilance. Two ergonomics experts shadowed 22 pharmacovigilance specialists during their daily work. They focus on the strategies for searching and choosing MedDRA terms to build pharmacovigilance queries. Interviews of four pharmacovigilance specialists completed the observations. Results highlight that, for unusual or complex searches, pharmacovigilance specialists proceed iteratively in three main phases: (i) preparation of a list of terms and of evaluation criteria, (ii) exploration of the MedDRA hierarchy and choice of a term, and (iii) evaluation of the results against the criteria. Overall, the search and the choice of keywords within a thesaurus shares similarity with the information retrieval task and is closely interwoven with the query building process. Based on the results, the paper proposes design specifications for new interfaces supporting the identification of MedDRA terms so that pharmacovigilance reports searches achieve a good level of expressiveness.
Databáze: OpenAIRE