Key points are not available for this paper at this time.
Fall is a devastating critical event with high incidence and mortality rate among older adults in the United States. Due to the stochastic nature of falls, accurate modeling of fall occurrences is important to reduce and prevent falls. Existing approaches mainly focused on population-level modeling and evaluating fall risk based on static and aggregate measures, such as average fall frequency during a certain time period. To account for individual heterogeneity of fall occurrences and provide instantaneous assessment of fall risk over time, this paper proposes a personalized fall risk assessment model. Bayesian estimation is performed to achieve simultaneous fall risk assessment of multiple individuals. Different performance indices are further presented to comprehensively evaluate fall characteristics of individuals over time. A real case study based on fall records collected from Florida assisted living facilities is provided to illustrate the proposed work and demonstrate its effectiveness. The proposed work will facilitate personalized and proactive interventions and managements for long-term care services improvement.
Chen et al. (Sun,) studied this question.