POPULATION PREFERENCES FOR MODE OF ADMINISTRATION IN EARLY ALZHEIMER'S DISEASE THERAPY: UTILITY DECREMENTS ESTIMATED BY A DISCRETE CHOICE EXPERIMENT IN JAPAN
Author(s)
Ataru Igarashi, PhD1, Mie Kasai Azuma, MSc, RPh, PhD2, Mayaka Tani, MPH2, Fujio Matsuyama, MS3, Naoki Tashiro, BS3, Tatsunori Murata, MS, PhD3, Masahiko Mori, BSc2, Kiyoyuki Tomita, MMA2.
1Graduate School of Business Administration, Keio University, Yokohama, Japan, 2Eisai Co., Ltd., Tokyo, Japan, 3CRECON Medical Assessment Inc., Tokyo, Japan.
1Graduate School of Business Administration, Keio University, Yokohama, Japan, 2Eisai Co., Ltd., Tokyo, Japan, 3CRECON Medical Assessment Inc., Tokyo, Japan.
OBJECTIVES: To quantify treatment preferences of the Japanese general population for early Alzheimer's disease (AD) as utility decrements (disutilities), covering both clinical attributes and treatment-related burden, with particular focus on different modes of administration, namely subcutaneous injection (SC) and intravenous infusion (IV), using a discrete choice experiment (DCE).
METHODS: A web-based DCE was administered to 2,194 Japanese adults (aged 40-79 years), yielding 65,820 observations. Health states were defined by five attributes: survival years, memory, mood, ability to perform housework, and treatment-related burden. The treatment-related burden attribute captured route of administration, visit frequency, administration time, and setting (self-administration versus administration at a healthcare facility). A conditional logit model was estimated with survival years as a continuous variable and other attributes as interactions with survival years. Each level's disutility was obtained by dividing its coefficient by the survival-year coefficient. Subgroup (age, sex, caregiving experience, contact with patients) and sensitivity analyses by response quality were performed.
RESULTS: The relative importance of attributes was mood (29.9%), memory (27.2%), treatment-related burden (25.5%), and housework ability (17.5%). Both modes of administration produced disutilities relative to a hypothetical reference of monthly outpatient visits without injection: weekly SC -0.339 and biweekly IV -0.388, indicating a lower burden for SC than for IV (disutility difference 0.049). The disutility associated with treatment-related burden remained directionally stable across subgroups and sensitivity analyses.
CONCLUSIONS: Among the Japanese general population, treatment burden for early AD differed by mode of administration, with SC showing a lower burden than IV. As these utility estimates reflect general-population preferences, they can inform value assessment of early AD therapies that differ in administration burden.
METHODS: A web-based DCE was administered to 2,194 Japanese adults (aged 40-79 years), yielding 65,820 observations. Health states were defined by five attributes: survival years, memory, mood, ability to perform housework, and treatment-related burden. The treatment-related burden attribute captured route of administration, visit frequency, administration time, and setting (self-administration versus administration at a healthcare facility). A conditional logit model was estimated with survival years as a continuous variable and other attributes as interactions with survival years. Each level's disutility was obtained by dividing its coefficient by the survival-year coefficient. Subgroup (age, sex, caregiving experience, contact with patients) and sensitivity analyses by response quality were performed.
RESULTS: The relative importance of attributes was mood (29.9%), memory (27.2%), treatment-related burden (25.5%), and housework ability (17.5%). Both modes of administration produced disutilities relative to a hypothetical reference of monthly outpatient visits without injection: weekly SC -0.339 and biweekly IV -0.388, indicating a lower burden for SC than for IV (disutility difference 0.049). The disutility associated with treatment-related burden remained directionally stable across subgroups and sensitivity analyses.
CONCLUSIONS: Among the Japanese general population, treatment burden for early AD differed by mode of administration, with SC showing a lower burden than IV. As these utility estimates reflect general-population preferences, they can inform value assessment of early AD therapies that differ in administration burden.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
P4
Topic
Patient-Centered Research, Study Approaches
Topic Subcategory
Health State Utilities
Disease
Geriatrics, Neurological Disorders, No Additional Disease & Conditions/Specialized Treatment Areas