Executive Summary

Patient falls are the most common adverse event in U.S. hospitals, are associated with increased hospital length of stay, and cost U.S. hospitals $7 billion annually. A great deal of research and improvement efforts have been dedicated to addressing hospital falls, but they remain a major challenge for healthcare organizations who dedicate significant resources to preventive interventions. The challenge of fall prevention begins with accurately identifying patients at risk for falling so appropriate interventions can be put in place. Traditional tools have been in use for decades, require manual nursing assessment, and, while shown to have good inter-rater reliability, validations of accuracy have produced variable results. More recent options for fall risk assessment include clinical decision support tools that automate the risk assessment. To establish the practice implications of different fall risk assessment tool options, we compare the performance and practice implications of three fall risk assessment tools: the Johns Hopkins Fall Risk Assessment Tool (JHFRAT), Epic’s Risk of Falling (ROF) cognitive computing model, and Kinometrix’s Fall Risk Assessment Solution (K-FRAS). Both the Epic ROF model and the K-FRAS were separately compared to the JHFRAT to evaluate their accuracy. The ROF model ultimately performed only as well as the JHFRAT, while the K-FRAS showed improved sensitivity without an increase in false negatives. This is of great importance when trying to balance a patient safety program with the resources required to provide the needed interventions. The K-FRAS not only reduces documentation, but also improves clinical processes by more accurately identifying fall risk patients, allowing hospitals to direct resource-intensive interventions where they are needed most.