A cluster analysis approach for the determination of a fall risk level classification

Falls among elderly people have massive social and economic impact. Gait impairments correlated with loss of physical functions are the primary common causes. Today, even if gait deviations between healthy young individuals and elderly ones have been deeply investigated, no standardize fall risk classification have been really established to facilitate fall risk management and prevention. Therefore, the core of this study is to implement a statistic approach to determine a fall risks classification i.e. normal gait, abnormal gait without risk of falling and abnormal gait with risk of falling. In this paper, the method based on a cluster analysis to set up this fall risk level classification is presented. Based on a limited number of easily accessible biomechanics predictors, a fall risk level can be determined and help care providers to earlier and better prevent fall.