Attribute Development in Health-Related Discrete Choice Experiments: A Systematic Review of Qualitative Methods and Techniques to Inform Quantitative Instruments.

ABSTRACT

OBJECTIVES: This review sought to identify the qualitative methods and techniques that researchers have used in the past decade to develop attributes and inform health-related discrete choice experiments (DCEs) surveys from a patient perspective. METHODS: The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for reporting systematic reviews. An adapted appraisal tool following guidelines for reporting qualitative research for quantitative instruments and criteria for attribute development in DCEs was applied for quality assessment and data extraction. A narrative approach was used to synthesize data. This examination included consideration of issues pertaining to sampling, data collection, data analysis, attribute list reduction, wording, methodological adaptations to capture patient preferences, and testing the pre-experimental design decisions of the DCE survey. RESULTS: Of 8505 articles identified for abstract screening, 680 were included for full-text screening, 36 of which met the inclusion criteria. Practices to improve methodological robustness included pre-data collection materials to inform instruments, data collection methods specific for decision-making scenarios, purposeful selection of data analysis methods to address the research question, and participants' involvement in reducing the list of attributes. Examples of methodological adaptations for patients were noted. CONCLUSIONS: DCEs have the potential to become a mixed-method approach in which the qualitative phase informs a reduced list of attributes for a survey, serves the predesign decisions of the experiment by testing trade-offs, overlapping, understandability, face, and content validity and provides explanations of the quantitative results. Establishing guidelines for using qualitative methods for DCE attribute development may help to broadly enhance the methodological robustness of DCEs.