Fall Patterns Are Independent Risk Factors for Mortality After Hip Fracture in Older Adults

Abstract

Abstract Hip fracture is becoming a major health problem with high mortality and morbidity in older adults. However, whether specific fall patterns could act as independent risk factors for predicting mortality after hip fracture remains unknown. We aimed to investigate whether fall patterns can serve as an independent risk factor for mortality after hip fracture. Electronic medical records (EMR) of individuals who visited emergency room or admitted to the Severance hospital, Seoul, Korea, between January 2005 to December 2019 were reviewed to categorize fall patterns. Fall patterns were categorized upon review of explanatory description in EMR, using modified classification based on motion analysis of video-captured falls in a prior study. Among 1,991 study subjects (mean age 77 years, 71% women), 211 patients died (10.6%; median survival 296 days). Fall location was divided into home (67.4%) and outdoor (32.6%) with mortality rate of 11.9% and 8.0% (p=0.009), respectively. Fall patterns were specified by “cause of fall” (6 categories; slip [29.6%], trip or stumble [17.5%], etc.) and by “activity at time of fall” (6 categories; walking [54.8%], getting up or rising [14.1%], etc). Among the combinations of both causes and activities, individuals who sustained hip fracture during “incorrect weight shift while sitting down or lowering”(hazard ratio [HR] 3.35, p=0.003), “collapsed during unclassified activity”(HR 2.37, p=0.006), “incorrect weight shift while getting up or rising”(HR 2.13, p=0.003), and “slipped while walking”(HR 1.83, p=0.004) had increased mortality after hip fracture compared to those with outdoor falls, after adjustment for age, sex, and Charlson comorbidity index. Specific fall patterns in individuals who sustained hip fracture predicted excess mortality in older adults, independent of age, sex, and comorbidities. Acknowledgement: We thank Doori Cho of the the SENTINEL (Severance ENdocrinology daTa scIeNcE pLatform) team (4-2018-1215) for the data acquisition process. Conflict of Interest: SB, NH, and YR have nothing to declare.