For those who have seen patients who have suffered a stroke, it is very clear that stroke is a life-changing disabling disease. Its high prevalence makes stroke the leading cause of disability in adults from the industrialized world. Measuring the degree of disability after a stroke is, however, less clear. In this paper, researcher from England tested the American-born, 8-dimension Stroke Impact Scale (SIS) and applied to the British population. The goals of the investigators were to evaluate the SIS applicability in the UK and to propose a shortcut of the SIS that might increase the response rate without. Participants were recruited from several mainly urban general practices and health questionnaires were sent (which included the SIS and the generic health questionnaire EuroQoL EQ-5D).
Only a third of all mailed surveys were retuned (N=151) to the investigators. The final sample consisted of predominantly men (58 %), British (81 %), 75 years or older (39 %) fully retired from works (56 %) that have had their stroke on average 7.3 prior to the survey. The mead disability score for the physical domain was 73.5/100, with internal consistency reliability > 85, but only half of the patients surveyed responded to all 8 dimensions items. The investigators added on all eight dimensions and created an index reflective of all items. Also, they chose individual components of each domain based on the highest correlation with the total score for that domain to create a short form SIS index. They found a high correlation among the original and shorten version of the indexes. Furthermore, the total disability scores were highly correlated.
Intuitively, one might think that this shorten version of the SIS might improve response rates without sacrificing the quality of what it portends to measure. Would it be the case? Is response rate directly related to the complexity and/or length of the questionnaire? It might be, but further validation might be required.