A Comprehensive Fall Prevention System for Memory Care: Final Feasibility and Randomized Controlled Study

Abstract In the USAlzheimer s diseaseADis the single most expensive diseasethe only disease in the top six for which the number of deaths is increasingThe greatest cost contributors are frequent hospitalizationswhere falls are the largest culpritand frequent need for assistance with the activities of daily livingFall safety systems show the potential to reduce costs and increase quality of care by reducing the likelihood of emergency eventse gdetecting falls before a fracture occurs and reducing the number of repeat fallsUnfortunatelycurrent safety devices require wearable or sensor technology not suitable for individuals with dementia and incapable of showing caregivers how falls occurOur goal is to extend the impacts of SafelyYou Guardianan online fall detection system with off theshelf wall mounted cameras to automatically detect falls for patients with AD and related dementiasADRDenabled by a human in the loopHILThe HILwho can monitor several facilities at a timeconfirms the fall detection alerts provided by our artificial intelligence algorithmsand places a call to the facilitiesso an intervention can happen within minutes of the fall detectionas opposed to hours afterwhen the next scheduled visit to the room of the patient happensSubsequentlyan Occupational TherapistOTreviews the fall videos to make recommendations on how to re organize the patient spaceinterventionto prevent future fallsWe leverage our HIL paradigmin which our deep learninga subfield of artificial intelligenceapproaches identify and pre filter falls well enough to leave the last check to a humanwho will call the facilities in case of detected safety critical eventsfallsThe present Fast Track NIH SBIR project leverages the already recruitedpatients in our partnermemory care facilitiesrecruited through our previous IRB approved pilotswhich we leveragePilotWe demonstrated the feasibility of the system by collecting proof of concept datacontainingacted falls of healthy subjects and showed accurate fall detectionPilotWe demonstrated acceptance of privacy safety tradeoffs by patientsfamily and staffthrough the collection ofmonths of video data at WindChime of Marina memory care facilityfrom the Integral Senior Living networkin which we identifiedtotal hours of fall dataThis ledto clinical benefits including anfall reduction through the intervention of an OccupationalTherapistOTto re organize patient spacePilotWe demonstrated scalability and acceptance by deploying the system infacilitiestotalingpatients monitored by our systemofflineno HIL interventionPilotongoingWe demonstrated the ability to perform onlinereal timefall detectionwithreal time intervention of the HIL through our partner company Magellan Solutionswhichprovides themonitoring service for the facilitiesand confirmed the decrease of thenumber of fallswith thepatientsIn this projectwe will develop a comprehensive fall detection and prevention system for memory care called SafelyYou Comprehensive CareWe operate in memory care facilitieswhich for the purpose of our work includes skilled nursing and assisted livingIn Phase Iwe seek to establish the technical feasibility and preliminary clinical benefit of sit to stand detectionpassive gait monitoringand realtime fall detection on blurred videoWe hypothesize that sit to stand detection will remove the need for bed alarmspassive gait monitoring will allow identification of high risk individualsallowing interventions to be proactively appliedand applying all techniques on blurred video will increase opt in rate for those with privacy concernstodayopt in leading to existing facility wide outcomesIn Phase IIwe seek to examine clinical outcomes using a waitlist randomized control study design with sample size ofparticipantsWe hypothesize that when compared to standard of carewaitlist control groupSafelyYou Comprehensive Care will reduce fall ratesreduce number of severe falls that require ER care and hospitalizationsand be more cost effective Narrative The goal of this Fast Track SBIR project is to extend the impact of SafelyYou Guardiana realtime fall detection system using off the shelf wall mounted cameras to develop a comprehensive fall detection and prevention system for memory care called SafelyYou Comprehensive CareSafelyYou Guardian is an online fall detection system with off the shelf wall mounted cameras to automatically detect falls for patients with AD and related dementiasADRDenabled by a Human in the LoopHILIn Phase Ithis Fast Track SBIR project will perform a first trial to establish the technical feasibility and preliminary clinical benefit of sit tostand detectionpassive gait monitoringand real time fall detection on blurred videoWe want to demonstrate that sit to stand detection can remove the need for bed alarmsthat passive gait monitoring will enable identification of high risk individualsallowing interventions to be proactively appliedand that applying all techniques on blurred video will increase opt in rate for those with privacy concernstodayopt in leading to existing facility wide outcomesIn Phase IIthrough apatient trialwe will examine clinical outcomes of the system using a waitlist randomized control studyWe want to demonstrate that when compared to standardof carewaitlist control groupSafelyYou Comprehensive Care will reduce fall ratesreduce number of severe falls that require ER care and hospitalizationsand be more cost effectiveThe project leverages apatient population in ourpartner facilitiesassisted living and skill nursingrecruited through our previous IRB approved pilotsPast results demonstrated andecrease in the fall rate in a single facility over the course of a three months period