Development of the Japanese version of the stroke stigma scale: a validity and reliability assessment

Abstract

Background: The stigma perceived by many post-stroke persons hinders their social lives. A scale to measure stigma is needed to identify social problems related to stigma, and to evaluate effectiveness of interventions. Objectives: This study aimed to develop a Japanese version of the Stroke Stigma Scale (SSS-J), and confirm its utility by examining reliability and validity. Methods: Eighty community-dwelling post-stroke individuals were enrolled at six sites. After translating the scale into Japanese using back translation methods, psychometric properties of the rating scale, internal scale validity, and reliability were examined to fit the Rasch model. Criterion-related validity, construct validity, and test-retest reliability were examined using total scores transformed to logit. For test-retest reliability, 30 participants completed the SSS-J twice, one week apart. Results: Rasch analysis showed that the SSS-J had the best fit with 15 items on a 3-category rating scale. Item difficulty logits were -2.01 to 2.21, person ability logits were -4.69 to 0.62 (mean, -1.41), person reliability coefficient was 0.71 (separation index, 1.58), and item reliability coefficient was 0.96 (separation index, 5.04). For criterion validity, Spearman’s rank correlation coefficient with the Center for Epidemiologic Studies Depression Scale was 0.51 (p < 0.001). For construct validity, Spearman’s rank correlation coefficients with each subscale of the Stroke Impact Scale ranged from -0.36 to -0.16 (p = 0.002–0.126). For test-retest reliability, the intra-class correlation coefficient was 0.64 (p < 0.001). Conclusions: The SSS-J adapted to the Rasch model was reliable and valid. This scale can be used to quantitatively measure stigma among community-dwelling post-stroke persons in Japan.