Quantitative and qualitative assessment to predict fall events for older inpatients at a tertiary referral hospital

Abstract

Abstract Purpose Falls among older patients are one of the most common adverse events in hospitals. Many hospitals have established and adopted their own multifaceted fall assessment tool. There has not been much research on fall assessment tools’ validation by incorporating various factors as covariates. We aimed to validate a fall risk assessment tool for older inpatients who are admitted to a tertiary referral hospital. Methods This study used electronic medical record (EMR) of inpatients aged ≥60 years who were admitted to Nagoya University Hospital from January 1st, 2020 to December 31st, 2020 were included in the study. On admission, the patients’ background information and fall risk assessment were extracted retrospectively from the EMR. Fall events within 30 days of hospitalization were the outcome, which were prospectively recorded by the nurses. We examined the association between fall risk assessment scores and the outcome as well as the C-statistics to examine if adding the score improved the models. Results A total of 10,353 patients were included in the study. The number of fall event was 224. The hazard ratio (95% confidence interval) for the fall risk assessment score was 1.30 (1.18–1.45) in multivariable Cox models. By adding the risk assessment scores, C-statistics of the multivariable models improved by 0.031 (0.007–0.054). Conclusions Even after adjusting for numerous quantitative factors, we validated the usefulness of the risk assessment tool in predicting fall events, demonstrating the utility of a qualitative tool assessed by healthcare providers in predicting fall events.