Novel Continuous Predictive Wearable Agitation Monitor and Alert Using Actigraphy and AI for Hospitalized Older Adults to Prevent Delirium and Dementia for Clinical Use and Harmonization of Research

Agitation in the hospital setting is a common life-threatening condition for confused patients at risk for vital medical equipment dislodgement and can result in delirium that can worsen or incite Alzheimer’s Disease and Related Dementias (ADRD). Agitation is a predictor of the use of physical and chemical restraint which are independent risk factors for delirium. Restraint and sedation are also utilized as agitation prevention measures. Older adults are especially vulnerable to deliriogenic prophylactic agitation intervention since they are at higher risk of developing associated complications that can worsen dementia, and are often perceived to be most likely to exhibit confused behavior and therefore more likely recipients of these harmful measures. For this reason, it is vital to closely monitor agitation levels to deliver a timely reliable score for implementation of non-harmful preventive and de-escalating strategies before it is necessary to implement prolonged harmful measures, and to limit the use of ongoing deliriogenic interventions at the earliest opportunity. Current standard of care is observational spot-check agitation scoring every 1-12 hours; however, scoring by bedside staff has significant interrater reliability deficiency, and the result is often prophylactic and prolonged restraint and sedation as a protective and convenience measure. With continuous agitation monitoring, the need for prolonged prophylaxis is greatly reduced since there is now a real-time monitor and alert system to report agitation level. The monitor will also allow de-escalation of harmful interventions with speed and the safety that any changes will be detected immediately. To ensure high satisfaction among users, objective monitor output is matched with the current globally-accepted subjective agitation/sedation scoring system. The goal of this Fast Track project is to develop and test a novel wearable continuous agitation monitor with actigraphy and predictive AI for translation of patient movements through a novel Agitation Dictionary including prediction of escalating agitation through a) training novel machine learning models (software), b) creating a device operating system (firmware), c) designing and miniaturizing sensor subsystem (hardware), d) developing an enclosure with user interface (mechanical device), e) developing an external system interface (enterprise connectivity system), f) laying the groundwork for institution connectivity systems (middleware), and g) performing clinical validation studies in ICU older adults, for reporting both over-agitation and over-sedation to standardize scores, speed intervention, prevent harmful prophylactic measures, and mandate accountability, the lack of which is a major gap in complying with guideline targets set for patient agitation, safety, delirium, sedation, and mobility. Successful completion of project aims has a high likelihood of satisfying market pull and interested partners (see LOS section) with 1) a novel production-ready agitation monitor for improved agitation intervention, 2) harmonization of research studies through standardization of data measurement and 3) positioning of HDmedical and AgiMon™ to become the standard of care for agitation monitoring.