Using a Medical Intranet of Things System to Prevent Bed Falls in an Acute Care Hospital: A Pilot Study

Background: Hospitalized patients in the United States experience falls at a rate of 2.6 to 17.1 per 1000 patient-days, with the majority occurring when a patient is moving to, from, and around the bed. Each fall with injury costs an average of US $14,000.

Objective: The aim was to conduct a technology evaluation, including feasibility, usability, and user experience, of a medical sensor-based Intranet of things (IoT) system in facilitating nursing response to bed exits in an acute care hospital.

Methods: Patients 18 years and older with a Morse fall score of 45 or greater were recruited from a 35-bed medical-surgical ward in a 317-bed Massachusetts teaching hospital. Eligible patients were recruited between August 4, 2015 and July 31, 2016. Participants received a sensor pad placed between the top of their mattress and bed sheet. The sensor pad was positioned to monitor movement from patients' shoulders to their thighs. The SensableCare System was evaluated for monitoring patient movement and delivering timely alerts to nursing staff via mobile devices when there appeared to be a bed-exit attempt. Sensor pad data were collected automatically from the system. The primary outcomes included number of falls, time to turn off bed-exit alerts, and the number of attempted bed-exit events. Data on patient falls were collected by clinical research assistants and confirmed with the unit nurse manager. Explanatory variables included room locations (zones 1-3), day of the week, nursing shift, and Morse Fall Scale (ie, positive fall history, positive secondary diagnosis, positive ambulatory aid, weak impaired gait/transfer, positive IV/saline lock, mentally forgets limitations). We also assessed user experience via nurse focus groups. Qualitative data regarding staff interactions with the system were collected during two focus groups with 25 total nurses, each lasting approximately 1.5 hours.

Results: A total of 91 patients used the system for 234.0 patient-days and experienced no bed falls during the study period. On average, patients were assisted/returned to bed 46 seconds after the alert system was triggered. Response times were longer during the overnight nursing shift versus day shift (P=.005), but were independent of the patient's location on the unit. Focus groups revealed that nurses found the system integrated well into the clinical nursing workflow and the alerts were helpful in patient monitoring.

Conclusions: A medical IoT system can be integrated into the existing nursing workflow and may reduce patient bed fall risk in acute care hospitals, a high priority but an elusive patient safety challenge. By using an alerting system that sends notifications directly to nurses' mobile devices, nurses can equally respond to unassisted bed-exit attempts wherever patients are located on the ward. Further study, including a fully powered randomized controlled trial, is needed to assess effectiveness across hospital settings.