Malarkey, David DNP Project

David Malarkey, DNP, RN, PHN, CEN

 DNP: California State University, Fresno 

Project Title: Preventing High-Risk Falls in the Emergency Department: Implementation of the MEDFRAT Fall Risk Assessment Tool

View Presentation

Background: Emergency department falls remain a significant patient safety concern in acute care settings. Fall risk assessment tools have been developed for inpatient populations; however, ED-specific instruments that account for transient risk factors are limited. Traditional fall risk assessment tools, such as the Schmid tool, were designed for inpatient hospital stays and may not adequately capture the dynamic, short-term risk states common in emergency department patients, such as intoxication, altered mental status, and post-seizure states.

Purpose: This quality improvement project evaluated the implementation of the Memorial Emergency Department Fall Risk Assessment Tool (MEDFRAT) and compared risk identification accuracy with the current standard Schmid tool using paired chart-audit data in an emergency department setting.

Methods: This study is a quality improvement project using paired comparison of retrospective chart audit data. Implementation was guided by Lippitt's Phases of Change. The baseline assessment period was January 1 through June 30, 2025. Staff training was conducted in July and early August 2025. The post-implementation evaluation window was August 15 through December 15, 2025. Chart audits were extracted (n = 1,285) and de-duplicated to 1,203 unique encounters; 1,189 had complete paired tool documentation for the primary comparison. Both tools were scored for each encounter using existing documentation. McNemar's chi-square test was used to compare discordant pairs between the two tools.

Results: With a high-risk threshold score of ≥3, MEDFRAT classified 662 of 1,189 encounters (55.7%) as at-risk compared with 339 of 1,189 (28.5%) using Schmid, a difference of 27.2 percentage points. Discordant classification favored MEDFRAT, with 348 encounters identified as at-risk by MEDFRAT only, compared with 25 by Schmid only. McNemar's chi-square test revealed a statistically significant difference between the tools: χ²(1) = 277.97, p < .001; matched-pairs odds ratio = 13.92, 95% confidence interval: 9.28 to 20.89. Fall rates fluctuated throughout the study period and were classified as rare events; identification of at-risk encounters was the primary outcome.

Conclusions: Findings demonstrate that ED-specific fall risk assessment tools significantly outperform inpatient-derived tools in identifying patients with transient risk states. MEDFRAT identified substantially more at-risk encounters, supporting the implementation of ED-specific screening as a primary prevention signal. These findings suggest that replacing inpatient-derived tools with ED-specific instruments creates critical opportunities for targeted fall prevention and improved patient safety in emergency department settings. A future study spanning at least one year is recommended to determine the sustained impact on fall rates and to validate these findings across multiple ED settings.