# Comparing the EQ-5D-5L and stroke impact scale 2.0 in stroke patients: an analysis of measurement properties

## TL;DR

This study compares two tools for measuring quality of life in stroke patients and finds that one is more suitable for capturing their health outcomes.

## Contribution

The study provides a direct comparison of the EQ-5D-5L and Stroke Impact Scale 2.0 in stroke patients, focusing on their measurement properties.

## Key findings

- The EQ-5D-5L has better feasibility and less ceiling effects compared to the Stroke Impact Scale 2.0.
- Both measures show known-groups validity but struggle with emotional domains.
- Neither tool is highly suitable for transient ischemic attack (TIA) patients.

## Abstract

Stroke has evolved to become a chronic disease and a major public health challenge. To adequately capture the full disease burden of stroke patients, the assessment of health-related quality of life (HRQoL) and thus the performance of respective measures is increasingly relevant. The aim of this analysis was to compare the measurement properties of two self-report instruments, the EQ-5D-5L and the Stroke Impact Scale 2.0.

The data used for the analysis was derived from a quasi-experimental case management study for mildly to moderately affected incident stroke and transient ischemic attack (TIA) patients aged ≥ 18 in Germany. Data was collected patient-individually at 3, 6 and 12 months after initial stroke. The EQ-5D-5L and SIS 2.0 were compared in terms of feasibility, ceiling and floor effects, responsiveness and known-groups validity (Kruskal-Wallis H and Wilcoxon rank-sum test).

A response for all three follow-ups is available for n = 855 patients. The feasibility of the EQ-5D-5L is determined as good (completion rate: 96.4–96.6%, ≥ one item missing: 3.2 − 3.3%), whereas the SIS 2.0 is moderately feasible (overall completion rate: 44.9–46.1%, ≥ one item missing in domains: 4.7 − 28.7%). The SIS 2.0 shows substantial ceiling effects in comparable domains (physical function: 10.4 − 13%, others: 3.5–31.3%) which are mainly larger than ceiling effects in the EQ-5D-5L index (17.1–21.5%). In terms of responsiveness, the EQ-5D-5L shows small to moderate change while the SIS 2.0 presents with moderate to large responsiveness. The EQ-5D-5L index, mobility, usual activities and Visual Analogue Scale show known-groups validity (p < 0.05). Content-related domains of the SIS 2.0 show known-groups validity as well (p < 0.05). However, it is compromised in the emotion domain in both measures (p > 0.05).

The EQ-5D-5L seems to be slightly more suitable for this cohort. Nonetheless, the results of both measures indicate limited suitability for TIA patients. Large-scale studies concerning responsiveness and known-groups validity are encouraged.

The study was registered in the German Clinical Trials Register, retrospective registration on 21.09.2022. Registration ID: DRKS00030297.

The online version contains supplementary material available at 10.1186/s12955-024-02252-z.