Rasch Analysis of the Stroke Impact Scale-16.

IMPORTANCE:Because the psychometric evidence supporting the use of the Stroke Impact Scale-16 (SIS-16) is limited, an examination of its psychometric properties is warranted. OBJECTIVE:To evaluate the psychometric properties of the SIS-16 using Rasch analysis. DESIGN:Secondary data analysis of responses to the SIS-16 from the published Field Administration of Stroke Therapy-Magnesium Trial database. SETTING:Emergency medical system agencies and acute care receiving hospitals. PARTICIPANTS:A total of 1,010 people with stroke. OUTCOMES AND MEASURES:We examined the item difficulty hierarchy, item fit, person-item match, separation index, person reliability coefficient, and ceiling and floor effect of the SIS-16. RESULTS:The item "climb a flight of stairs" was the most difficult, and "sit without losing balance" was the easiest. Four items misfit the Rasch partial-credit measurement model. Overall, the average patient ability estimate of 2.1 logits (SD = 2.0) was higher than the average item difficulty estimate of 0.0 logits (SD = 1.1). With a separation index of 2.85, the SIS-16 can differentiate people into 4.1 statistically distinct strata. The person reliability coefficient was .89. Given that 24.2% (n = 244) obtained the maximum score and 0.2% (n = 2) obtained the minimum score, the SIS-16 had a ceiling effect in this patient sample. CONCLUSIONS AND RELEVANCE:Results partially support the validity and clinical use of the SIS-16 in subacute stroke clinical settings. Further research is warranted to examine the psychometric properties of the SIS-16 in patients with chronic stroke. What This Article Adds: Our results partially support the use of the SIS-16 in clinical and research settings.