The Stroke Impact Scale (SIS), developed by Duncan et al. (1999), is a comprehensive measure of health outcomes for stroke populations that incorporates meaningful dimensions of function and health related to quality of life into a single questionnaire that is completed as either a self-report or by proxy (Lai, Studenski, Duncan, & Perera, 2002). Several versions are available. The most recent, SIS 3.0, comprises 59 items that assess the following 8 domains of performance: (1) strength, (2) hand function, (3) mobility, (4) activities of daily living (ADLs), (5) emotion, (6) memory, (7) communication, and (8) social participation, as well as a physical domain that combines the first 4 domains to better examine the person’s physical attributes affected by stroke. A 16-item version is also available that was developed on the basis of the 2.0 version and consists of 7 ADL/instrument ADL (IADL) items, 8 mobility items, and a single hand function item that was designed to quickly address the physical aspects of stroke (Edwards & O’Connell, 2003). Each item of the SIS measures is scored according to 3 possible 5-point Likert scales (1 to 5), where higher scores suggest better performance. The assessment employs a unique scoring algorithm where the maximum raw score (295) is computed through a mathematical equation to give a final score where both can be completed in less than 15 minutes.
A study of 229 stroke patients found good internal consistency of the SIS 3.0 subscales for hand function (α = 0.82), mobility subscale (0.86), ADL (0.79), emotion (0.79), memory (0.87), communication (0.87), and participation (0.75) with the lowest being strength at 0.63 (Nichols-Larsen, Clark, Zeringue, Greenspan, & Blanton, 2005). Instrument reliability of the 2.0 version found that Cronbach coefficients of internal consistency were also good with a range of α= 0.83 to 0.90 and when stroke impact was measured over time internal consistency of all domains ranged from α= 0.70 to 0.92, with the exception of emotion at 0.57, suggesting that the SIS in its present form is a stable instrument (Duncan et al., 1999). The same study showed that each of the 8 domain scales had good criterion validity and that the measures of disability (i.e., mobility and ADL/IADL), in particular, had good correlation coefficients in the range of 0.82 to 0.84 with other established measures such as the Barthel Index (BI) and Functional Independence Measure scale (Duncan et al., 1999). Multiple regression analysis showed that three factors, physical function, emotion, and participation, explained 45% of the variance of the patient’s global assessment of recovery (Duncan et al., 1999). Research by Lai et al (2002) showed that the SIS 3.0 was able to capture persisting difficulties in the performance of stroke patients who otherwise might have been considered functionally independent according to measures such as the BI or the modified Rankin Scale. For example, individuals with BI scores of 95 are usually considered recovered, however, Lai et al. (2002) showed that stroke patients who achieved a BI score of 95 continued to have residual disability and impaired quality of life as noted by SIS scores, highlighting its ability to detect impairment where other global assessments cannot.
The assessment is available in several formats as well as proxy and self-administered versions. There is also a good amount of research in support of its use in clinical practice. No special training or certifications are needed to administer the exam and there is an excellent interactive website maintained by the University of Kansas Medical Center pertaining to the measure where one can download a copy as well as supporting material along with access to exclusive SIS databases of statistical and scoring information.
A study of 377 subjects found that proxy-rated assessors tended to score patients as more severely affected than patients scored themselves on the SIS-16 version and in 7 of 8 domains of the full SIS 3.0. Five of those domain scores were considered statistically significant and although, intra-class correlation coefficients between proxy and patient ranged from 0.50 to 0.83 it was shown that this bias tended to increase as the severity of the patient’s condition increased (Duncan et al., 2002).
During assessment the subject is asked to evaluate how stroke has impacted their health by rating questions relative to three possible scales depending on the question such as (5) a lot of strength to (1) no strength at all; (5) not difficult at all to (1) extremely difficult; or (5) none of the time to (1) all of the time. An analog scale that explores the subjective level of recovery where (0) is no recovery to (100) full recovery is also completed. The SIS scoring algorithm is as follows:
|In the past 2 weeks, how difficult was it to use your hand that was most affected by your stroke to…|
A. Carry heavy objects (i.e., bag of groceries)
B. Turn a doorknob
C. Open a can or jar
D. Tie a shoe lace
E. Pick up a dime