Decision tree for ward admissions of older patients at the emergency department after a fall

Falls are a prevalent issue for the older population, and for the healthcare system in terms of emergency department (ED) access and hospitalizations. There is still a lack of knowledge and guidelines, however, regarding the need to hospitalize older patients accessing the ED after a fall. In the present study, we aimed to analyze the factors and the decisional process that led to older patients accessing the ED after a fall being admitted to hospital or discharged. The study sample included 2144 older people who accessed the ED after a fall. For each patient, we obtained information on the nature of the fall and the related injuries, previous falls, dementia and ongoing medical therapies. As the outcome variable, we considered the indication for ward admission after the ED visit. Of the 2144 individuals who accessed the ED after a fall, 38% had at least one fracture, and 40.1% were admitted to a ward. The decision tree obtained using the chi-squared automatic interaction detection algorithm showed that the indication for ward admission could be accurately predicted (risk estimate 0.205) by just five factors, namely: presence and severity of fall-related injuries, reportedly suspicious fall dynamics, use of anticoagulants, polypharmacy, and dementia. The need for ward admission in older patients who access the ED after a fall seems to be determined not only by the severity of fall-related injuries, but also by the fall dynamics and the individual's clinical complexity.