DEVELOPMENT OF THE DUTCH VALUE SET FOR THE PROMIS PREFERENCE SCORE (PROPR)
Author(s)
Michiel Luijten, PhD1, Ellen Elsman, PhD1, Janel Hanmer, PhD, MD2, G. Ardine de Wit, PhD3, Judith Bosmans, PhD4, Benjamin Schalet1.
1Amsterdam UMC, Amsterdam, Netherlands, 2University of Pittsburgh, Pittsburgh, PA, USA, 3RIVM / Vrije Universiteit Amsterdam, 3720BA Bilthoven, Netherlands, 4Vrije Universiteit Amsterdam, Amsterdam, Netherlands.
1Amsterdam UMC, Amsterdam, Netherlands, 2University of Pittsburgh, Pittsburgh, PA, USA, 3RIVM / Vrije Universiteit Amsterdam, 3720BA Bilthoven, Netherlands, 4Vrije Universiteit Amsterdam, Amsterdam, Netherlands.
OBJECTIVES: Preference-based quality of life instruments are used in cost-effectiveness analysis of healthcare interventions. The PROMIS-Preference measure (PROPr) is a preference-based summary score developed in the US. We aimed to establish a PROPr value set reflecting the health preferences of the Dutch population for 7 core PROMIS domains.
METHODS: We applied standard gamble methodology used in the US and pilot tested for comprehensibility. We replaced the item “I felt unhappy” with “I felt sad” (similar content and item parameters) as the former displayed DIF between the Netherlands and US. Subsequently, we applied standard gamble to elicit health preferences in an adult sample representative of the Dutch general population recruited through a panel agency. We applied isotonic regression with linear interpolation to obtain single-attribute scoring functions for each domain. We calculated mean values of disutility corner states. A multiplicative multi-attribute utility (scoring) function subsequently estimates utilities for PROMIS outcomes on a scale where 0 is the utility of being dead and 1 is the utility of full health.
RESULTS: Based on pilot testing (N=21), we rewrote the instructions to improve readability and refined the visual presentation of the standard gamble tasks. To elicit preferences, we recruited a sample representative of the Dutch general population (N=1139). Sleep Disturbance and Depression had one minor monotonicity violation in the single-attribute domain scoring function. Mean disutility corner state values were: Social Roles (0.403), Sleep Disturbance (0.407), Fatigue (0.415), Physical Function (0.447), Cognitive Function (0.467), Depression (0.504), and Pain (0.608). (Higher values indicate more weight in the final multi-attribute utility function). We are currently estimating the multi-attribute utility function to establish the Dutch PROPr value set; results will be presented at the conference.
CONCLUSIONS: This study results in a societal preference-based scoring system for the Dutch population, using the 7 PROMIS health domains. This Dutch PROPr value set will enhance cost-effectiveness analyses across diverse medical fields using a single PROMIS Profile measure.
METHODS: We applied standard gamble methodology used in the US and pilot tested for comprehensibility. We replaced the item “I felt unhappy” with “I felt sad” (similar content and item parameters) as the former displayed DIF between the Netherlands and US. Subsequently, we applied standard gamble to elicit health preferences in an adult sample representative of the Dutch general population recruited through a panel agency. We applied isotonic regression with linear interpolation to obtain single-attribute scoring functions for each domain. We calculated mean values of disutility corner states. A multiplicative multi-attribute utility (scoring) function subsequently estimates utilities for PROMIS outcomes on a scale where 0 is the utility of being dead and 1 is the utility of full health.
RESULTS: Based on pilot testing (N=21), we rewrote the instructions to improve readability and refined the visual presentation of the standard gamble tasks. To elicit preferences, we recruited a sample representative of the Dutch general population (N=1139). Sleep Disturbance and Depression had one minor monotonicity violation in the single-attribute domain scoring function. Mean disutility corner state values were: Social Roles (0.403), Sleep Disturbance (0.407), Fatigue (0.415), Physical Function (0.447), Cognitive Function (0.467), Depression (0.504), and Pain (0.608). (Higher values indicate more weight in the final multi-attribute utility function). We are currently estimating the multi-attribute utility function to establish the Dutch PROPr value set; results will be presented at the conference.
CONCLUSIONS: This study results in a societal preference-based scoring system for the Dutch population, using the 7 PROMIS health domains. This Dutch PROPr value set will enhance cost-effectiveness analyses across diverse medical fields using a single PROMIS Profile measure.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE763
Topic
Economic Evaluation, Methodological & Statistical Research, Patient-Centered Research
Disease
No Additional Disease & Conditions/Specialized Treatment Areas