Affiliations: Department of Physical Therapy, School of Health and Rehabilitation Sciences, MGH Institute of Health Professions, CNY, Boston, MA, USA | Baylor University Medical Center, Dallas, TX, USA
Note:  Corresponding author: Marianne Beninato, DPT, PhD, Department of Physical Therapy, School of Health and Rehabilitation Sciences, MGH Institute of Health Professions, 36 First Ave, CNY, Boston, MA 02129, USA. Tel.: +1 617 726 3128; E-mail: firstname.lastname@example.org
Abstract: Background and purpose: The Stroke Impact Scale 16 (SIS-16) total score (TotSIS-16) as a self report of physical function has previously been associated with falls, but the relationship of individual SIS-16 items to fall history has not been examined. The purpose of this study was to examine the accuracy of individual SIS-16 items compared with TotSIS-16 in identifying falls history in people with chronic stroke. Methods: Forty-three people were categorized according to prior 6 month fall history as having either No falls (NF; n = 23) versus any fall (AF; one or more fall; n = 20). Group comparisons (AF versus NF) were made for individual SIS-16 item scores and the TotSIS-16 to identify items significantly related to fall history. Significant items were further analyzed for sensitivity (Sn) and specificity (Sp) of specific cutoff scores for identifying individuals with fall history. Results: TotSIS-16 and items related to balance, mobility and urinary continence had high Sp (0.70–0.96) useful for ruling in a history of falls while only the item “Balance while Walking” had Sn (0.89) useful for ruling out falls. Discussion and conclusions: Individual SIS-16 items may be useful for ruling falls in (high Sp) or out (high Sn), depending on the clinical need. One or two items used together may be more accurate than the TotSIS-16 score. Study of a larger sample is needed to confirm these results.