Margaux Poleur1, Barbara Willekens2, Bertrand Degos3, Damien Ricard4, Vincent Van Pesch5, Alexis Tricot6, Laurie Médard1, Hui Li6, Emilie Lommers7, Mona Michaud4, Paul Strijbos8, James Overell8, Céline Cluzeau6, Damien Eggenspieler6, Laurent Servais9
1University department of neurology, Citadelle Hospital of Liège, 2Department of Neurology, Antwerp University Hospital, 3Neurology Department, Avicenne Hospital, APHP, Hôpitaux Universitaires de Paris-Seine Saint Denis (HUPSSD), 4Service de Neurologie, Service de Santé des Armées, Hôpital d’Instruction des Armées de Percy, 5Cliniques Universitaires Saint-Luc, UCLouvain, 6Sysnav, 7Department of neurology, Centre Hospitalier Universitaire de Liège, 8Hoffman-La Roche, 9MDUK Oxford Neuromuscular Centre, John Radcliffe Hospital
Objective:
To validate the feasibility and reliability of digital outcomes derived from wearable Digital Health Technology (wDHT) that measure real-life ambulation in people with Multiple Sclerosis (pwMS). We investigated several ambulation digital outcomes, and their correlation with MS gold standard outcome measures including Expanded Disability Status Scale (EDSS) and timed 25-foot walk (T25FW).
Background:
The wDHT is a magneto-inertial device specifically designed for continuous assessment of patients with neurological diseases (ActiMyo®). We had previously demonstrated accuracy of the measurements (with centimetric precision) in MS. One of the measures studied, 95th centile of stride velocity (SV95C), is the first digital clinical outcome measure qualified as a primary endpoint by the European Medicines Agency (for Duchenne Muscular Dystrophy).
Design/Methods:
Participants were included in a longitudinal multisite study and were asked to wear the wDHT device for an initial period of three months then after 1 year. EDSS, 9-hole peg test (9HPT), T25FW, 6-minute-walking-test (6MWT) and Berg-Balance Scale (BBS) were performed at both timepoints. Reliability was assessed by comparing first and second half the recording period using inter-class correlation (ICC). Correlations were estimated using Spearman's correlation coefficient. We studied the SV95C and the maximal distance covered by patients (WD90C).
Results:
We included pwMS aged from 22 to 65 years (median 48 years), with mild to moderate impairment (EDSS range 0-5.5, median 3). 74/78 patients (95%) provided at least 50 hours of recording during the initial period. ICC for SV95C and SD90C was 0.99 and 0.83, respectively. Correlations of SV95C and WD90C with EDSS (-0.66 -and -0.53
respectively) and T25FW (-0.66 -and -0.57
respectively) at baseline were moderate. Other correlations and the 1-year evolution will be presented.
Conclusions:
The results demonstrate the feasibility and reliability of recording using wDHT and the internal and external consistency of SV95C, and WD90C as potential digital outcome measures in MS.