Evaluating Emergency Department Transfers for Falls in Skilled Nursing Facilities: A Single-Site Observational Study of Witnessed and Unwitnessed Falls.

OBJECTIVES

Falls are a leading cause of morbidity and mortality among older adults, with residents of skilled nursing facilities (SNFs) at a particularly high risk. Although emergency department (ED) transfers after falls are common, they carry risks of delirium, infection, and care fragmentation. The objective of this study is to compare ED transfer rates between witnessed and unwitnessed falls in SNFs.

DESIGN

Single-site, observational study.

SETTING AND

PARTICIPANTS

A 120-bed SNF in the greater Roanoke area, serving both long-term and short-term rehabilitation residents. Data were collected prospectively on 212 fall events between November 2023 and April 2025.

METHODS

Falls were reported through a structured workflow integrating QR (Quick Response) code-based logging and REDCap documentation. Reports included witnessed status, fall location, injury documentation, and whether emergency medical services (EMS) were activated. Incidence rates were calculated as the number of falls per bed-days. Descriptive statistics and Fisher exact tests were used to compare fall characteristics (witnessed status, injury presence, and fall location). The primary outcome was transfer to the ED.

RESULTS

Of 212 fall events (4.1 falls per 1000 bed-days), 81.1% were unwitnessed and 83.0% occurred in resident rooms. Witnessed falls were more common in common areas (50.0%) compared with in-room falls (12.5%) (OR 6.91; P < .001). EMS was activated in 9.9% of all cases. Witnessed falls were more than twice as likely to prompt EMS compared with unwitnessed falls (17.5% vs 8.1%; OR 2.21; P = .145). Injuries were reported in 17.0% of falls, and EMS was called in 53.8% of these. Head trauma was the strongest driver of transfer, present in all transported injuries (OR 490.5; P < .001).

CONCLUSION AND IMPLICATIONS: Unwitnessed falls represented the majority of events and were associated with lower rates of ED transfer. This highlights the need for more objective data to reduce uncertainty about fall mechanisms and potential injuries, particularly when head strike is not directly observed. Head injuries were the strongest predictor of transport. These findings underscore the diagnostic challenges surrounding unwitnessed falls and emphasize the importance of standardized postfall protocols, decision-support tools, and passive fall detection technologies to guide ED transfer decisions in SNFs.