Outcome measures correlated with falls in nursing home residents-A pilot study

Most fall risk and normal values are collected from the community-dwelling population, which is not representative of nursing home residents. The purpose of this study was twofold: 1) to determine the relationship of clinical impairment and activity limitation variables to the number of falls in nursing home residents; and 2) to determine the amount of variability that can be explained for the number of falls from these predictor variables. Seventeen active nursing home residents (83.7 ± 11.7 years) consented to participate. Mini-Mental Status Examination (MMSE), lower extremity handheld dynamometry, ankle plantar flexion (PF)/dorsiflexion (DF) active range of motion (AROM), hand grip strength, gait speed (GS), Timed Up and Go (TUG), and 5 Times Sit-to-Stand (5TSTS) were recorded in a single visit. Regression analysis was performed to identify the better clinical outcome tool to determine falls. This was followed by a stepwise multiple regression model to predict the criterion variable-number of falls. Of the clinical impairment measures collected, significant correlations with past falls include the following: right DF AROM (-0.436; p = 0.040) and right DF strength (-0.504; p = 0.023). Of the activity limitation measures collected, significant correlations with past falls include the following: 5TSTS (0.585; p = 0.007); TUG time (0.475; p = 0.027); and GS (0.457; p = 0.032). The stepwise multiple regression model explained 59% of the variance using right DF AROM, right DF strength, 5TSTS, and TUG time. These measures are benchmarks for the community dwelling population. The present study indicates that these measures might also be useful in determining fall risk screening for ambulatory nursing home residents.