Characteristics of patients hospitalized for falls: electronic health records analysis

Abstract Aim Falls are common adverse events, often leading to hospitalization with adverse effects on the physiological and psychological health of patients. Although risk factors for falls are well examined, it is difficult to identify fallers through their electronical health records (EHR). This study investigates whether EHR contain characteristics that indicate a fall before hospitalization. Subject and methods We evaluated 500 EHR of patients hospitalized through the Department of Emergency Medicine of Jena University Hospital. We considered electronically available characteristics including free texts and ICD-10-GM codes. Inpatients with and without a fall event previous to their hospitalization were compared for significant differences in EHR characteristics. Therefore, we applied the Mann–Whitney U test and the exact Fisher test, respectively, with a two-sided significance level of 0.05 and without correction for multiple testing. Results It was possible to identify significant differences between fallers and non-fallers by gender (66.3% women), age (median: 84 years), living situation prior to hospital admission (20.5% from nursing home, 41.1% living alone) and type of discharge. Besides other risk factors, ICD-10-GM codes included injuries, especially fractures and superficial injuries, but also cognitive deficits, incontinence, and physical limitations were more prevalent with fallers than non-fallers. Conclusion This study identified characteristics that can support identification of fall patients through EHR. Such information should be tested in other hospitals and is expected to enhance patient safety and utilization of healthcare resources.