Comparative Responsiveness of Preference-Based Health-Related Quality of Life, Social Care and Wellbeing Measures in the Context of Multiple Sclerosis.

Abstract

Objectives: To provide evidence on the responsiveness of social care and wellbeing preference-based measures (PBMs) compared to health-related quality of life PBMs in the context of multiple sclerosis (MS).

Methods: The ICEpop CAPability measure for Adults (ICECAP-A) and Adult Social Care Outcomes Toolkit (ASCOT) were completed online in September 2019, March 2020, September 2020, via the UK MS Register. Responses were linked to EQ-5D-3L and MS Impact Scale-Eight Dimensions (MSIS-8D) values, and to MS Walking Scale-12 (MSWS-12), Hospital Anxiety and Depression Scale (HADS) and Fatigue Severity Scale (FSS) scores. Responsiveness was assessed in relation to minimal important differences on MSWS-12, HADS and FSS between timepoints, using mean change scores, t-tests, standardised effect sizes, standardised response means and multivariable regression analyses.

Results: Data from 1,742 people with MS were available for analysis. When using standardised values, MSIS-8D showed the greatest responsiveness and EQ-5D-3L the least. In contrast, when absolute utility values were used, EQ-5D-3L performed similarly to MSIS-8D and better than ICECAP-A and ASCOT. Standardised regression analyses indicated the MSIS-8Ds to be the most responsive, followed by the ASCOT, ICECAP-A, and EQ-5D-3L.

Conclusions: The ICECAP-A, ASCOT and MSIS-8D were more responsive than the EQ-5D-3L in the context of MS when compared using standardised scores. The increased responsiveness of EQ-5D-3L when absolute values were used seems an artefact of the wide-ranging scale of this measure. This illustrates how the maximum potential range of values for a given PBM tariff could influence whether an intervention is found to be cost-effective.