(PS5-31) Rubber Meets the Roadblock: A Mixed Methods Review of a Quality Improvement Project Using Wearable Devices for Inpatient Measurement

Abstract Text: This quality improvement (QI) project explored the use of wearable technology to track inpatient sleep and activity, aiming to enhance psychiatric care. While wearable technologies offer new opportunities to support clinical decision-making, real-world implementation remains limited. Using a wrist-worn device, we collected accelerometer-derived metrics and integrated them with electronic health record (EHR) data to produce automated reports of inpatient sleep and activity for clinicians on an inpatient psychiatric unit. Between August 2023 and August 2024, we produced 135 reports, conducted 13 interviews with clinical staff, and collected data on patient participation and report generation times. During the first phase of the project, 155 patients were admitted, 88 (56.8%) were offered a wearable, and 68 (77.3%) agreed to wear one. Of the 20 patients who declined, the most common reason cited was concerns about tracking and data privacy (4/11, 36.4%). Reports were considered useful, yet in need of further refinement. Interview data shows the QI project struggled to achieve buy-in on the unit among key clinical staff, and the pace of clinical care demands quicker, more reliable delivery of patient data than was initially possible with this prototype. Technical improvements during the project allowed reports to be returned in 3–5 hours. The primary use case for our report was in discrepancies between patient and nursing subjective reports on sleep. All interviewed psychiatrists wanted the reports to appear in the EHR. Our findings suggest that implementation of wearable technologies on an inpatient psychiatric unit can add value, though need be further tailored with regards to staff buy-in; the speed, clarity, and reliability of the product; and integration with existing clinical workflows. Ideally, reports would be simple to understand and offer daily readouts for sleep duration, sleep quality, and time in bed vs. time asleep, with options for more granular data as needed. They would also populate in the EHR, be available immediately after patient wake with sleep values from the night before, and require only a single click to access. Beyond this, we would note that while significant technical improvements were made to the system for report generation during our project, we encountered our greatest difficulties in attempts to affect attitudinal and behavioral change among unit staff. We believe this had to do both with general resistance to change as well as a sense among unit staff that our reports were not urgently requested by clinicians, and thereby the work they were asked to do in offering wearables to and collecting wearables from patients was not worth prioritizing among their many responsibilities. While psychiatrists were greatly consulted in developing this product, future efforts are likely to benefit from engaging all staff involved in implementation to create a system valued across different roles and improve adoption in clinical practice.