RT Journal Article
SR Electronic
T1 Establishing digital biomarkers for clinical trials. (P6.181)
JF Neurology
JO Neurology
FD Lippincott Williams & Wilkins
SP P6.181
VO 90
IS 15 Supplement
A1 Neil Thomas
A1 Nora Mattek
A1 Thomas Riley
A1 Phelps Witter
A1 Christina Reynolds
A1 Johanna Austin
A1 Nicole Sharma
A1 Jennifer Marcoe
A1 Jeffrey Kaye
YR 2018
UL http://n.neurology.org/content/90/15_Supplement/P6.181.abstract
AB Objective: This study uses a home-based pervasive computing system to identify changes in meaningful outcomes in typical patients with mild cognitive impairment (MCI) or early Alzheimer’s disease (AD) and compares outcome measures of the automated system to conventional clinical outcome measures while patients transition on or off typical AD-related treatments.Background: Conventional trials methodology for testing dementia treatments is limited due to an assessment paradigm that acquires data in an episodic, infrequent manner and relies on subjective self- and proxy-based reports. Home-based pervasive computing and sensing systems provide a means to acquire objective, high-frequency, longitudinal data generating ecologically valid outcome measures (also known as digital biomarkers) directly related to in-the-moment cognitive and functional status.Design/Methods: This natural history, pragmatic proof of concept clinical trial enrolls MCI or early-stage AD patients living with a care partner. A pervasive sensing and computing system is deployed in each couple’s home continuously providing data regarding key outcomes: physical activity and mobility, sleep, computer use, driving and medication adherence. Each participant undergoes a conventional neurological and cognitive battery at baseline, and weekly on-line reports on changes to their health and medications. Conventional tests are compared to the continuous data.Results: Eight homes are currently enrolled: 3 participants with MCI and 5 with AD. Overall 98.6% of weekly surveys were completed by participants with cognitive impairment and 3 of 8 participants required assistance from their care partner. Daily activity (step count), medication adherence, total time in bed and number of nighttime awakenings are reliably derived from algorithms using data from the sensing and computing system.Conclusions: Outcome metrics comprising multiple functional and health related domains impacted by cognitive impairment can be successfully collected and analyzed from multi-person homes. The approach may ultimately reduce trial durations, sample size needs and reliance on episodic clinic-based assessment.Disclosure: Dr. Thomas has nothing to disclose. Dr. Mattek has nothing to disclose. Dr. Riley has nothing to disclose. Dr. Witter has nothing to disclose. Dr. Reynolds has nothing to disclose. Dr. Austin has nothing to disclose. Dr. Sharma has nothing to disclose. Dr. Marcoe has nothing to disclose. Dr. Kaye has nothing to disclose.