Can we predict the risk of falls in elderly patients with instability?

Objective: The aim of this study is to determine whether clinical and instrumental examination of balance can predict the risk of falls in elderly patients with instability. Methods: Study design: It is a case control study, with cases defined by falls in the last year, developed in a third level university hospital. Patients: Seventy patients aged 65 years or more who met at least one of the following inclusion criteria: (a) at least one fall in the last year; (b) spend more than 15 s during the timed up and go test (TUG); (c) a score of less than 68% average balance in the sensory organisation test (SOT) of the computerised dynamic posturography (CDP); or (d) at least one fall in the CDP-SOT. Intervention: TUG test, CDP-SOT, CDP centre of gravity balancing (CG) and limits of stability (LOS), Dizziness Handicap Inventory (DHI) test and short FES-I test. Main outcome measures: Number of steps and time (TUG), average balance and use of sensorial information (CDP-SOT), speed and directional control (CDP-CG and LOS), DHI score and short FES-I score. Results: Comparing subjects without falls (non-fallers) vs subjects with at least one fall (fallers) in the last year, fallers obtain worse scores than non-fallers in condition 2 (p = 0.043) and use of somatosensory information (p = 0.039). Comparing subjects with five falls or less (non-multiple-fallers) vs subjects with more than five falls (multiple-fallers), multiple-fallers obtain worse scores than non-multiple-fallers in overall balance (p = 0.023), condition 6 (p = 0.036), directional control (swaying (p = 0.006) and LOS (p = 0.023)) and short FES-I score (p = 0.007). Conclusion: The three most useful parameters for identifying unstable elderly patients at particularly high risk of repeated falls are mean balance in the CDP SOT, directional control of CDP LOS and short FES-I score. (C) 2014 Elsevier Ireland Ltd. All rights reserved.