DEVELOPING AN INSTRUMENT TO ELICIT TREATMENT PREFERENCES FOR IDIOPATHIC PULMONARY FIBROSIS: ENGAGING PATIENTS THROUGH COGNITIVE INTERVIEWS
Candidate for the Lee B. Lusted Student Prize Competition
Method: We conducted a series of cognitive interviews among patients with IPF, a rare and universally fatal condition for which there is currently no approved therapy. Interviews involved patients completing a draft survey instrument. Additionally, prospective and retrospective probes were used to engage patients. The study was guided by applying a checklist spanning ten items, including confirming that respondents are: i. understanding the attributes; ii. understanding attribute levels; iii. understanding the study purpose; iv. accepting the elicitation format; v. finding supporting information relevant; vi. making trade-offs; vii. not adversely influenced by framing effects; viii. not making assumptions; ix. not overburdened; and x. accepting of the overall survey. Data was drawn from the completed surveys, the interviewers’ field notes and digital recordings of the interviews. Concordant and discordant themes were identified.
Result: In all, 15 interviews were completed, each lasting approximately 30 minutes. Respondents tended to be male, aged between 65-85 years. Participants differed in both time since diagnosis and severity of disease. During the interviews, data were collected to support all items on the cognitive issue checklist. Participants were engaged and appreciative of their role as a partner in this research. Applying the checklist, we found that patients accepted the attributes, levels, and overall survey. Both best-worst and discrete-choice formats were presented and were accepted by most participants. Best-worst was reportedly easier to answer as there is less information to evaluate. Reducing the number of attributes was suggested as a way to ease respondent burden. Participants were willing to make tradeoffs among the attributes. No framing effects or inappropriate assumptions were observed, although clarification of the supporting information and instructions, particularly with respect to perspective, was needed.
Conclusion: Good research practices for the application of cognitive interviews in the development of patient-reported outcomes, but it is unclear how they should be used in the development of stated-preference instruments. Through the development and application of a novel checklist, we have demonstrated the role of cognitive interviews in engaging patients in the process of developing a stated-preference instrument.