ASSESSING RECURRENT FALL RISK OF COMMUNITY-DWELLING, FRAIL OLDER VETERANS USING SPECIFIC TESTS OF MOBILITY AND THE PHYSICAL PERFORMANCE-TEST OF FUNCTION

Background. The purpose of this prospective cohort study was to determ ine if order individuals at risk for recurrent falls are best identifi ed by mobility or functional assessments. Methods. Eighty-four communi ty-dwelling, frail male veterans, mean age of 75.5 years (SD = 7.33), participated. The history of recurrent fails was determined by self or proxy report in a clinical interview. Mobility assessments included t he Modified Gait Abnormality Rating Scale (GARS-M), stride length, and walking velocity; functional performance was determined using the Phy sical Performance Test (PPT). The clinical usefulness of the measures was described by determining the sensitivity and specificity of each m easure using the history of recurrent falls as a standard. Results. St epwise logistic regression analysis of the data indicated that the GAR S-M (p < .01) and the PPT (p < .01) were the most important predictors of recurrent fall risk. The sensitivity and specificity of the measur es used were: GARS-M, 62.3% and 87.1%; PPT, 79.3% and 71.0%; walking s peed, 71.7% and 74.2%; and stride length, 63.2% and 77.4%. Together th e GARS-M and PPT demonstrated the highest sensitivity of 90.6% and the highest specificity of 87.1% based on a subject testing positive on a i least one test. Conclusion. Used independently and in combination, t he GARS-M and the PPT were clinically useful measures in screening for older individuals at risk for recurrent falls.