COMPARATIVE PERFORMANCE OF THE EQ-5D-5L AND PROMIS-29 IN ADULTS WITH INFLAMMATORY BOWEL DISEASE AND IRRITABLE BOWEL SYNDROME
Author(s)
Fatima Al Sayah, PhD1, Shakib Rahman, PhD2, Nicolle Hua, PhD2, Hilary Short, MSc3, Haley Gross, BS2, Jennifer Bennett, BS2, Michelle Buresi, MD, PhD4, Yasmin Nasser, MD, PhD2, Deborah A. Marshall, PhD2.
1Research Associate, University of Alberta, Edmonton, AB, Canada, 2University of Calgary, Calgary, AB, Canada, 3University of Alberta, Edmonton, AB, Canada, 4Alberta Health Services, Calgary, AB, Canada.
1Research Associate, University of Alberta, Edmonton, AB, Canada, 2University of Calgary, Calgary, AB, Canada, 3University of Alberta, Edmonton, AB, Canada, 4Alberta Health Services, Calgary, AB, Canada.
OBJECTIVES: To evaluate and compare the construct validity and responsiveness of the EQ-5D-5L and PROMIS-29 measures in adults with inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS).
METHODS: Using the IMAGINE Chronic Disease Network MAGIC cohort, a prospective observational multicentre study of patients with IBD, IBS, and healthy controls, we analyzed adults in Alberta completing English-language surveys. Cross-sectional analyses included participants with concurrent EQ-5D-5L and PROMIS-29 assessments, while longitudinal analyses used 1-year follow-up data. EQ-5D-5L utilities were derived using Canadian and US value sets, and PROMIS-29 responses were converted to PROPr utilities. Disease-specific measures included the IBS Symptom Severity Scale (IBS-SSS) and IBD Symptom Inventory-Short Form (IBDSI-SF). Additional mental health specific measures were included. Analyses assessed distributional properties, agreement, convergent and divergent validity, known-groups validity, and responsiveness.
RESULTS: A total of 993 participants had complete EQ-5D-5L and PROMIS-29 data (IBD=756; IBS=237). Participants with IBS reported poorer health-related quality of life than those with IBD, with lower mean EQ-5D-5L utility scores (0.813 vs. 0.866) and PROPr scores (0.384 vs. 0.478). A ceiling effect was observed for EQ-5D-5L in IBD (30.4%) compared with IBS (10.0%). EQ-5D-5L and PROPr utilities demonstrated strong correlations (r=0.67-0.73) and good agreement (ICC=0.71-0.78). Strong correlations between conceptually similar domains (r=0.59-0.83) and weak correlations between unrelated constructs supported convergent and divergent validity. Both measures effectively discriminated between groups defined by disease severity, depression, anxiety, stress, and resilience, with PROPr generally demonstrating larger effect sizes. Longitudinal analyses showed that both measures were responsive to changes in health status; however, PROPr was more sensitive to mental health and stress-related changes.
CONCLUSIONS: Both EQ-5D-5L and PROMIS-29 demonstrated good construct validity and responsiveness in adults with IBD and IBS. Despite comparable performance, PROPr was more sensitive to psychosocial health impacts and may be more relevant in populations with significant mental health burden.
METHODS: Using the IMAGINE Chronic Disease Network MAGIC cohort, a prospective observational multicentre study of patients with IBD, IBS, and healthy controls, we analyzed adults in Alberta completing English-language surveys. Cross-sectional analyses included participants with concurrent EQ-5D-5L and PROMIS-29 assessments, while longitudinal analyses used 1-year follow-up data. EQ-5D-5L utilities were derived using Canadian and US value sets, and PROMIS-29 responses were converted to PROPr utilities. Disease-specific measures included the IBS Symptom Severity Scale (IBS-SSS) and IBD Symptom Inventory-Short Form (IBDSI-SF). Additional mental health specific measures were included. Analyses assessed distributional properties, agreement, convergent and divergent validity, known-groups validity, and responsiveness.
RESULTS: A total of 993 participants had complete EQ-5D-5L and PROMIS-29 data (IBD=756; IBS=237). Participants with IBS reported poorer health-related quality of life than those with IBD, with lower mean EQ-5D-5L utility scores (0.813 vs. 0.866) and PROPr scores (0.384 vs. 0.478). A ceiling effect was observed for EQ-5D-5L in IBD (30.4%) compared with IBS (10.0%). EQ-5D-5L and PROPr utilities demonstrated strong correlations (r=0.67-0.73) and good agreement (ICC=0.71-0.78). Strong correlations between conceptually similar domains (r=0.59-0.83) and weak correlations between unrelated constructs supported convergent and divergent validity. Both measures effectively discriminated between groups defined by disease severity, depression, anxiety, stress, and resilience, with PROPr generally demonstrating larger effect sizes. Longitudinal analyses showed that both measures were responsive to changes in health status; however, PROPr was more sensitive to mental health and stress-related changes.
CONCLUSIONS: Both EQ-5D-5L and PROMIS-29 demonstrated good construct validity and responsiveness in adults with IBD and IBS. Despite comparable performance, PROPr was more sensitive to psychosocial health impacts and may be more relevant in populations with significant mental health burden.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MSR7
Topic
Methodological & Statistical Research, Patient-Centered Research
Disease
Gastrointestinal Disorders