Framing Matters! Implications of Loss Aversion for Communicating Benefits to Patients in Preference Studies (and Beyond)
Author(s)
Hui Lu, PhD1, Joshua Coulter, MA2, Travis Gould, PhD3, Lisa Wilcox, PhD4, Carmine Colavecchia, PhD5, Stephanie Christopher, PhD6, Álvaro Gutiérrez, PhD7, Brett Hauber, PhD8, Jennifer Whitty, PhD9.
1Senior Research Associate, Evidera, Cambridge, United Kingdom, 2Pfizer, Inc, Grand Rapids, MI, USA, 3Pfizer Inc, New York, NY, USA, 4Pfizer Inc, Kirkland, QC, Canada, 5Pfizer, Inc., New York, NY, USA, 6Pfizer, Inc., Remote, NY, USA, 7Evidera, Remote, United Kingdom, 8Pfizer, New York, NY, USA, 9Evidera, London, United Kingdom.
1Senior Research Associate, Evidera, Cambridge, United Kingdom, 2Pfizer, Inc, Grand Rapids, MI, USA, 3Pfizer Inc, New York, NY, USA, 4Pfizer Inc, Kirkland, QC, Canada, 5Pfizer, Inc., New York, NY, USA, 6Pfizer, Inc., Remote, NY, USA, 7Evidera, Remote, United Kingdom, 8Pfizer, New York, NY, USA, 9Evidera, London, United Kingdom.
OBJECTIVES: Discrete choice experiments (DCEs) are widely used in health economics to quantify patient and caregiver preferences, informing healthcare decision making. DCEs assume rational, well-formed preferences. Prospect Theory suggests that choices are reference-dependent and subject to loss aversion, meaning individuals place greater weight on losses than on equivalent gains. We investigate whether framing a benefit as a loss or gain impacts preference estimates in a DCE and discuss implications for interpreting preference data and communicating benefits to patients.
METHODS: A web-based DCE was conducted among adults with hemophilia (n = 194) and caregivers of children with hemophilia (n = 169) in the United States and United Kingdom. The primary efficacy attribute—the change in annual bleeds—was anchored to a reference point representing the average change in bleed rate of a patient on treatment from baseline. Attribute levels included both perceived gains (up to three fewer bleeds) and losses (up to two more bleeds) relative to this reference. A mixed logit model was used for primary analysis.
RESULTS: The analysis revealed a statistically significant asymmetry between the desire to reduce annual bleeds and the desire to avoid additional annual bleeds. Both adults and caregivers demonstrated statistically significant loss aversion. Among adults, the disutility of one additional bleed was 1.64 times greater than the utility of one fewer bleed. Among caregivers, the loss aversion ratio was even higher at 2.90.
CONCLUSIONS: These findings are consistent with Prospect Theory and provide evidence of reference-dependent preferences and loss aversion. Researchers should carefully consider framing when designing a DCE. Accounting for behavioral patterns may enhance the interpretation of preference data and identify scenarios where benefits may be overestimated or underestimated, depending on individuals’ baseline. Emphasizing loss avoidance may be more effective in communicating the benefits of treatment and could impact treatment adherence.
METHODS: A web-based DCE was conducted among adults with hemophilia (n = 194) and caregivers of children with hemophilia (n = 169) in the United States and United Kingdom. The primary efficacy attribute—the change in annual bleeds—was anchored to a reference point representing the average change in bleed rate of a patient on treatment from baseline. Attribute levels included both perceived gains (up to three fewer bleeds) and losses (up to two more bleeds) relative to this reference. A mixed logit model was used for primary analysis.
RESULTS: The analysis revealed a statistically significant asymmetry between the desire to reduce annual bleeds and the desire to avoid additional annual bleeds. Both adults and caregivers demonstrated statistically significant loss aversion. Among adults, the disutility of one additional bleed was 1.64 times greater than the utility of one fewer bleed. Among caregivers, the loss aversion ratio was even higher at 2.90.
CONCLUSIONS: These findings are consistent with Prospect Theory and provide evidence of reference-dependent preferences and loss aversion. Researchers should carefully consider framing when designing a DCE. Accounting for behavioral patterns may enhance the interpretation of preference data and identify scenarios where benefits may be overestimated or underestimated, depending on individuals’ baseline. Emphasizing loss avoidance may be more effective in communicating the benefits of treatment and could impact treatment adherence.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
PCR99
Topic
Economic Evaluation, Methodological & Statistical Research, Patient-Centered Research
Topic Subcategory
Patient Behavior and Incentives
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, Rare & Orphan Diseases