Psychometric comparisons of four disease-specific health-related quality of life measures for stroke survivors

Objective: To examine psychometric properties of four stroke-specific health-related quality of life (HRQoL) measures, including original Stroke-Specific Quality of Life Scale (12-domain SSQoL), modified 8-domain SSQoL, Stroke Impact Scale (SIS 3.0), and modified SIS-16 focused on physical domains. Design and Setting: Prospective repeated measures study conducted in rehabilitation and wards in hospitals. Subjects: Study cohort was recruited with 263 patients in the first administration and 121 in the second administration, an average of two weeks later. To investigate discriminant validity, the same number of patients (i.e., 52) was grouped for each of 3 levels of stroke severity. Main measures: Outcome measures, including National Institutes of Health Stroke Scale, Mini-Mental State Examination, and Barthel Index. Patients completed HRQoL self-reports. Results: Domains of four measures showed (1) good reliability, except 12-domainSSQoL family roles (Cronbach's alpha = 0.68) and personality domains (Cronbach's alpha = 0.65) and SIS 3.0 social participation (ICC=0.67) domain; (2) acceptable precision, except 12-domain SSQoL family role domain and SIS 3.0 social participation domain; (3) good convergent validity, except 12-domain SSQoL/8-domain SSQoL vision domain (r = 0.19), (4) good discriminant validity, except 12-domain SSQoL and 8-domain SSQoL thinking domains (P = 0.365); and (5) acceptable floor effects and strong ceiling effects. The 12-domain SSQoL and 8-domain SSQoL met scaling assumptions better than SIS 3.0 and SIS-16. Conclusions: Four measures showed acceptable psychometric properties with some domains slightly less satisfactory. Overall, use of 8-domain SSQoL and SIS 3.0 are feasible for clinical practice to monitor HRQoL of stroke survivors.