Presentation (CTI)

OBJECTIVES: Patient preference studies help to elucidate what is important to patients, and/or their willingness to trade-off on specific benefits and risks. However, heterogeneity in preferences is common, and quantitative approaches to investigating heterogeneity provide specific insights into the “who” and “what” but no information on “why.” Understanding the rationale for heterogeneity is increasingly valuable as treatment options and associated benefits and risks expand, while the use of PP information for decision-making concurrently increases. We present a novel method helpful in elucidating individuals’ rationale for stated preferences and deriving insights into the drivers of the heterogeneity that is often observed in stated preference studies.
METHODS: Following completion of a DCE survey (total n=300 severe asthma patients; n=247 clinicians) qualitative interviews were conducted with a subset of 25 patients and 25 clinicians. Individual-specific relative attribute importance (RAI) estimates were generated for each participant; interviews targeted participants whose preferences were either most divergent from, or most similar to, the average sample preferences. During the interviews, participants were presented with individual-level RAI figures and probed on the reasons for their preferences.
RESULTS: Presenting individual-specific RAI figures effectively facilitated conversations around individuals’ preferences and the associated rationale. The greatest insights were achieved when asking participants about attributes they considered to be comparatively more important than the average. This approach also provided insights into the range of rationales for preferences, the contributions of personal experiences to preferences, attribute interpretation, and discrepancies between participants’ views of the DCE choice context and their broader real-world decision-making.
CONCLUSIONS: Post-DCE qualitative interviews can provide valuable insights into participants’ rationale behind preferences that have been estimated using quantitative approaches. These qualitative data may be particularly valuable when preferences are highly heterogenous and/or the rationale for specific preferences is poorly understood.