Eliciting Health State Preferences With Analytical Hierarchy Process: An Illustration With EQ-5D-5L
Author(s)
Clement Cheuk Wai Ng, BSc, MPH, PhD, Annie Wai-ling Cheung, MPhil, Amy Yuen-Kwan Wong, MSc, Eliza Lai-Yi Wong, PhD.
JC School of Public Health and Primary Care, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, China.
JC School of Public Health and Primary Care, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, China.
OBJECTIVES: In the current EQ-VT protocol, discrete choice experiment (DCE) heavily depends on participants imagining hypothetical health states with five health-related quality of life (HRQoL) dimensions combined. Alternatively, the analytical hierarchy process (AHP) allows participants to compare dimensions in a pairwise manner, and may potentially impose lower cognitive burden on the respondents. The study aims to explore the applicability of AHP in EQ-5D-5L health state preference elicitation.
METHODS: A cross-sectional survey (n=504) was conducted in 2022-2023 based on the gender and age group composition of the Hong Kong SAR, China community. The respondents compared the EQ-5D five dimensions in pairs based on a scale of 1 (equal importance) to 9 (extreme importance), followed by five absolute scoring tasks to assign a 0 (worst HRQoL) to 100 (perfect HRQoL) value to each level of the five dimensions. The AHP attribute matrix was derived with Saaty’s principal right eigenvector method, and the attribute weightings were aggregated with arithmetic means. AHP health state values were computed by combining the pairwise comparisons and absolute scoring tasks. Predictive ability of the AHP was assessed by evaluating the preference agreement between the AHP health state values against health state values derived from the EQ-5D-5L(Hong Kong) value sets among the EQ-VT 196 health state pairs.
RESULTS: In terms of the EQ-VT health state pairs, AHP showcased high predictive ability in matching the preferred health states exhibited by the EQ-5D-5L-HK value set (92.9%). Similar findings were also revealed in the sub-group comparisons between gender (female: 91.84%; male: 94.9%), age group (<50: 94.4%; 50: 90.8%), and education attained (tertiary education: 92.9%; secondary education or below: 91.8%).
CONCLUSIONS: Findings uncovered may hint AHP as an useful alternative in health state preference elicitation. Further research may explore its full potential in settings with limited resources and biased samples.
METHODS: A cross-sectional survey (n=504) was conducted in 2022-2023 based on the gender and age group composition of the Hong Kong SAR, China community. The respondents compared the EQ-5D five dimensions in pairs based on a scale of 1 (equal importance) to 9 (extreme importance), followed by five absolute scoring tasks to assign a 0 (worst HRQoL) to 100 (perfect HRQoL) value to each level of the five dimensions. The AHP attribute matrix was derived with Saaty’s principal right eigenvector method, and the attribute weightings were aggregated with arithmetic means. AHP health state values were computed by combining the pairwise comparisons and absolute scoring tasks. Predictive ability of the AHP was assessed by evaluating the preference agreement between the AHP health state values against health state values derived from the EQ-5D-5L(Hong Kong) value sets among the EQ-VT 196 health state pairs.
RESULTS: In terms of the EQ-VT health state pairs, AHP showcased high predictive ability in matching the preferred health states exhibited by the EQ-5D-5L-HK value set (92.9%). Similar findings were also revealed in the sub-group comparisons between gender (female: 91.84%; male: 94.9%), age group (<50: 94.4%; 50: 90.8%), and education attained (tertiary education: 92.9%; secondary education or below: 91.8%).
CONCLUSIONS: Findings uncovered may hint AHP as an useful alternative in health state preference elicitation. Further research may explore its full potential in settings with limited resources and biased samples.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
MSR228
Topic
Economic Evaluation, Methodological & Statistical Research, Patient-Centered Research
Topic Subcategory
PRO & Related Methods
Disease
No Additional Disease & Conditions/Specialized Treatment Areas