Preventing Elderly Falling Through Machine Learning

P. Hardigan
Nova Southeastern University,
United States

Keywords: elderly falling medications


In the United States (U.S.), falling is a leading cause of injury and trauma-related hospital admissions among adults, aged 65 years and older. Falling is also a prominent reason associated with injury-related mortality in older adults. The direct medical costs tied to fall-related injuries is about $2 billion for fatal falls, with non-fatal falls costing approximately $19 billion, as these patients are twice as likely to be hospitalized as those admitted for other medical conditions.Demographic research estimates that by 2030, the population of individuals who are 65 years of age or older will double and by 2050 the population of individuals who are 85 years of age or older will quadruple.With the aging population growing and the high incidence and costs tied to falling, fall prevention has become a national public health priority. While falling has been on the rise among older adults, falls are a preventable geriatric syndrome, with screening being the first step in reducing fall risks. Studies have demonstrated that falling can be reduced up to 40 percent by performing fall assessments and addressing a patient’s risk factors with proper management, lifestyle adjustments and/or pharmaceutical or medical interventions. Our recent research efforts have demonstrated the ability to prevent injurious falls, using a novel machine-learning algorithm that we created, implemented, and patented. This allows clinicians to screen for individual fall risk, with suggestions for appropriate interventions. A significant part of our fall-prevention program incorporates a critical risk factor—specific drugs and drug dosages. This algorithm not only was able to accurately predict a patient’s risk of falling but demonstrated a cost benefit by lower healthcare utilization, reducing hospitalization and providing overall economic savings. Specifically, we reduced injurious falls and saved health care providers $2.40 for every $1.00 invested. Our long-term goal builds on the foundation we developed with the creation of this machine-learning algorithm by expanding this approach into a computerized clinical decision support system (CCDSS), which clinicians can use to offer tailored, specific suggestions for fall prevention to their patients. CCDSS is a technology that uses patient-specific data to provide relevant medical knowledge at the point of care. It is considered to be an important quality improvement intervention, and the implementation of CCDSS is growing substantially.