Use of smartphone-based remote assessments of multiple sclerosis in Floodlight Open, a global, prospective, open-access study

Abstract

Floodlight Open was a global, open-access, digital-only study designed to understand the drivers and barriers in deployment and use of a smartphone app in a naturalistic setting and broad study population of people with and without multiple sclerosis (MS). The study utilised the Floodlight Open app: a 'bring-your-own-device' solution that remotely measures a user's mood, cognition, hand motor function, and gait and postural stability via smartphone sensor-based tests requiring active user input ('active tests'). Levels of mobility of study participants ('life-space measurement') were passively measured. Study data from these tests were made available via an open-access platform. Data from 1350 participants with self-declared MS and 1133 participants with self-declared non-MS from 17 countries across four continents were included in this report. Overall, MS participants provided active test data for a mean duration of 5.6 weeks or a mean duration of 19 non-consecutive days. This duration increased among MS participants who persisted beyond the first week to a mean of 10.3 weeks or 36.5 non-consecutive days. Passively collected life-space measurement data were generated by MS participants for a mean duration of 9.8 weeks or 50.6 non-consecutive days. This duration increased to 16.3 weeks/85.1 non-consecutive days among MS participants who persisted beyond the first week. Older age, self-declared MS disease status, and clinical supervision as part of concomitant clinical research were all significantly associated with higher persistence of the use of the Floodlight Open app. MS participants performed significantly worse than non-MS participants on four out of seven active tests. The findings from this multinational study inform future research to improve the dynamics of persistence of use of digital monitoring tools and further highlight challenges and opportunities in applying them to support MS clinical care.

Figure 1

Overview of the Floodlight Open study workflow. Consent could be withdrawn at any time during the study, resulting in the termination of further data collection from the withdrawn participant. All data provided up to the withdrawal of consent were retained and used for study purposes. The Floodlight Open app was developed for iOS- and Android-based smartphones and made available for download in 17 countries (see Supplementary Table S1 for release dates in each country). ID identification, iOS iPhone operating system.

Figure 2

Funnel plot depicting the proportion of self-declared MS participants that progress from one active test (or PRO) to another active test, or drop out, during the fixed test sequence. Participants who never completed a full fixed test sequence were included. DaS Draw a Shape, DMQ Daily Mood Questionnaire, IPS Information Processing Speed, MS multiple sclerosis, PRO patient-reported outcome, SBT Static Balance Test, UTT U-Turn Test.

Figure 3

Persistence among participants. Persistence was defined as one of the following: at least any one test of the fixed sequence performed per week, at least one 2MWT performed per week, at least one full fixed test sequence performed per week, or having provided life-space measurement data in a given week. Participants are ordered along the x-axis starting with those with the highest persistence (self-declared MS, n = 1350; self-declared non-MS, n = 1133). The green shading indicates the exact number of days that the participant was adherent in the specific week. 2MWT Two-Minute Walk Test, MS multiple sclerosis.

Figure 4

Weekly persistence among self-declared MS and non-MS participants by country in the first 20 weeks. Persistence of self-declared (a) MS and (b) non-MS participants per week (at least one test of the fixed sequence performed per week) across countries with at least 20 participants; numbers of participants shown are those who performed at least one test of the fixed sequence. MS multiple sclerosis.

Figure 5

Distribution of test performance data for self-declared MS and non-MS participants. The distribution of performance data for each active test and life-space measurement is depicted with violin plots with low-degree smoothing applied. Note that for the DaS Test, the maximum average Hausdorff distance recorded for the non-MS population is 0.68. 2MWT Two-Minute Walk Test, DaS Draw a Shape, DMQ Daily Mood Questionnaire, IPS Information Processing Speed, MS multiple sclerosis, s seconds, SBT Static Balance Test, UTT U-Turn Test.