Use of the International Classification of Functioning, Disability and Health as a framework for analyzing the Stroke Impact Scale-16 relative to falls

Abstract Objective: To determine if subscores based on grouping Stroke Impact Scale 16 (SIS-16) items according to International Classification of Functioning, Health and Disability (ICF) components are more accurate in identifying individuals with a history of falls than the total SIS-16 score. Design: Case series. Subjects: 43 community-dwelling people with chronic stroke. Methods: Participants were grouped based on six month fall history (no fall versus one or more falls). The SIS-16 items were categorized as belonging to the Body Structure and Function (BSF), Activity (ACT) or Participation (PART) component of the ICF. SIS-16 total score and ICF component subscores were analyzed for their association with falls. Receiver Operating Characteristic Curves were (ROC) analyzed. Results: There were significant differences between groups on SIS-16 total (p = 0.006), BSF (p = 0.041) and ACT (p = 0.003) scores. The BSF and ACT component subscores had the highest specificity (0.91) and sensitivity (0.80), respectively, for categorizing participants according to fall history. The BSF + ACT component subscore demonstrated greater accuracy than the total SIS-16 for identifying people with falls (area under the curve = 0.78). Conclusion: The ICF may be a useful model for analysis of fall screening tools for people with chronic stroke. ICF component subscores are more accurate than the SIS-16 total score for this purpose.