A LONGITUDINAL COMPARISON OF THE 2014 AND 2018 CHINESE EQ-5D-3L VALUE SETS IN PATIENTS WITH ACUTE ISCHEMIC STROKE: FINDINGS FROM THE CASTOR STUDY
Author(s)
Yiru Ma, MSc,1, Li Pingping, PhD1, Hongchao Li, MSc, PhD2.
1School of International Pharmaceutical Business, China Pharmaceutical University, Nanjing, China, 2School of International Pharmaceutical Business, China Pharmaceutical University；Center for Pharmacoeconomics and Outcomes Research, China Pharmaceutical University, Nanjing, China.
1School of International Pharmaceutical Business, China Pharmaceutical University, Nanjing, China, 2School of International Pharmaceutical Business, China Pharmaceutical University；Center for Pharmacoeconomics and Outcomes Research, China Pharmaceutical University, Nanjing, China.
OBJECTIVES: Two EQ‑5D‑3L value sets (3L2014 and 3L2018) are available for health utility estimation in China. However, their longitudinal agreement to clinical measures in acute ischemic stroke (AIS) patients remains unexplored. This study aimed to assess the absolute and change-score agreement between the two value sets, evaluate temporal trends in systematic bias, and compare their associations with the modified Rankin Scale (mRS) and the National Institutes of Health Stroke Scale (NIHSS).
METHODS: 9,978 AIS patients from the Chinese Acute Ischemic Stroke Treatment Outcome Registry (CASTOR) were evaluated at admission (V1), discharge (V2), 90-day (V3), and 360-day since admission (V4). Agreement was examined via intraclass correlation coefficient (ICC), Bland-Altman plots, and effect sizes (ES), with subgroup analyses by baseline mRS, NIHSS, and TOAST. Change‑score agreement was examined using ICC on health transitions. Linear mixed models (LMM) and censored Tobit regressions were used to explored bias trajectories and clinical‑scale associations.
RESULTS: The overall mean utility was significantly higher for 3L2018 than for 3L2014 (0.831 vs. 0.705). Although the ICC indicated good relative agreement (0.843), the width of the 95% limits of agreement (0.360) exceeded the established minimal important difference (MID) of 0.074, and the ES across visits ranged from 0.402 to 0.552. Subgroup analyses indicated that the agreement decreased with greater stroke severity. The mean utility difference narrowed from V2 to V4 (0.177 vs. 0.090). The two sets showed strong agreement on health changes (ICC ≥0.885). LMM showed that higher baseline NIHSS widened the utility gap between two value sets, an effect attenuated over time. In Tobit models, mRS exhibited a stronger association with 3L2018, whereas NIHSS showed a stronger association with 3L2014.
CONCLUSIONS: The two Chinese value sets are not interchangeable in AIS. Choice of value set may influence cost-utility ratios and should be a key sensitivity analysis source in future stroke-related economic models.
METHODS: 9,978 AIS patients from the Chinese Acute Ischemic Stroke Treatment Outcome Registry (CASTOR) were evaluated at admission (V1), discharge (V2), 90-day (V3), and 360-day since admission (V4). Agreement was examined via intraclass correlation coefficient (ICC), Bland-Altman plots, and effect sizes (ES), with subgroup analyses by baseline mRS, NIHSS, and TOAST. Change‑score agreement was examined using ICC on health transitions. Linear mixed models (LMM) and censored Tobit regressions were used to explored bias trajectories and clinical‑scale associations.
RESULTS: The overall mean utility was significantly higher for 3L2018 than for 3L2014 (0.831 vs. 0.705). Although the ICC indicated good relative agreement (0.843), the width of the 95% limits of agreement (0.360) exceeded the established minimal important difference (MID) of 0.074, and the ES across visits ranged from 0.402 to 0.552. Subgroup analyses indicated that the agreement decreased with greater stroke severity. The mean utility difference narrowed from V2 to V4 (0.177 vs. 0.090). The two sets showed strong agreement on health changes (ICC ≥0.885). LMM showed that higher baseline NIHSS widened the utility gap between two value sets, an effect attenuated over time. In Tobit models, mRS exhibited a stronger association with 3L2018, whereas NIHSS showed a stronger association with 3L2014.
CONCLUSIONS: The two Chinese value sets are not interchangeable in AIS. Choice of value set may influence cost-utility ratios and should be a key sensitivity analysis source in future stroke-related economic models.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
PCR160
Topic
Clinical Outcomes, Patient-Centered Research, Real World Data & Information Systems
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
Cardiovascular Disorders (including MI, Stroke, Circulatory)