Host: Centre for Primary Care, Institute of Population Health
Speaker: Caroline Vass (Centre for Health Economics, University of Manchester)
About the event:
Abstract:
Objective: This study aimed to identify all published health studies that have reported the use of qualitative methods to inform the design and/or the interpretation of discrete choice experiments (DCEs).
Methods:
DCEs were identified from a published systematic review and an updated systematic search of the Medline database. Studies were classified depending on the quantity of qualitative research reported, and those with ‘extensive’ qualitative aspects were critically appraised using a newly developed structured tool.
Results:
A total of 277 articles (254 studies) were identified and included in the review. Of the 254 studies: 111 (44%) did not report the use of any qualitative methods; 114 (45%) reported minimal data on the use of qualitative methods; 29 (11%) reported, or cited, the extensive use of qualitative methods. A variety of applications of qualitative research were identified, including the selection of attributes and/or levels (n=95, 66%), piloting the DCE (n=26, 18%) and understanding respondents’ decision making processes (n=4, 3%). Popular qualitative research methods included focus groups (n=64, 45%) and interviews (n=108, 76%), with semi-structured interviews (n=25, 17%), structured interviews (n=5, 3%), cognitive debriefing techniques (n=12, 8%) being the most common approaches.
Conclusions:
Guidelines for the conduct of DCEs, such as Lancsar and Louviere (2008), recommend the use of qualitative research methods to inform attribute and level selection and assessment of respondents’ comprehension of the choice tasks. This study indicates the implementation of these recommendations is not consistently reported by authors. A key limitation of this systematic review is that it cannot ascertain whether the absent or minimal reporting of qualitative research reflects absent or minimal qualitative research in practice, or rather that the reporting of this work is poorly captured within the published papers.
All welcome. No booking necessary
