Ascertaining Medication Use and Patient-Reported Outcomes via an App and Exploring Gamification in Patients With Multiple Sclerosis Treated With Interferon β-1b: Observational Study (Preprint)

BACKGROUND The BETACONNECT autoinjector and myBETAapp app were designed to support patients with multiple sclerosis receiving interferon β-1b and are an ideal platform for digital observational studies. A recent pilot study in Germany demonstrated the feasibility of using the app to recruit patients, obtain informed consent, and evaluate medication-taking behavior over 6 months. OBJECTIVE This study aims to describe medication-taking behavior for 1 year in patients with multiple sclerosis receiving interferon β-1b based on data collected from the app and to provide information on patient-reported outcomes (PROs). The optional use of the cognitive training tool PEAK (Peak, formerly Brainbow Ltd) is included to test the feasibility of gamification in this setting. METHODS A prospective and retrospective, exploratory, digital, observational cohort study was conducted among users of the app in Germany. Invitations to participate were sent to patients’ apps between February and May 2019. Participants provided electronic informed consent. Injection-related data from consenting patients’ devices were collected prospectively for 1 year following the consent date and retrospectively for ≤1 year from the first day of use (if historical data were available). Participants also completed three electronic PRO instruments every 3 months: the EuroQol 5-Dimension, 5-Level questionnaire (EQ-5D-5L); the Treatment Satisfaction Questionnaire for Medication (TSQM; version II); and a questionnaire on satisfaction with treatment support (on a server accessed via an emailed hyperlink). All patients were offered optional access to the professional version of PEAK. RESULTS Of 1778 registered app accounts (May 2019), 79 patients (4.44%) provided informed consent; 62 (3.49%) were eligible for inclusion in the prospective analysis, of whom, 60 (97%) also had retrospective data. The mean age of the 62 participants was 43.2 (SD 11.5) years and 41 (66%) were women. Compliance over the 1-year prospective observational period (primary end point) was high (median 98.9%, IQR 94.3%-100%) and similar among men and women. Persistence and adherence (coprimary end points) decreased from 85% (53/62) and 74% (46/62), respectively, at 6 months to 76% (47/62) and 65% (40/62), respectively, at 12 months; both were higher in men than in women. A retrospective analysis showed similar patterns. The PRO questionnaires were answered by 79% (49/62) of the participants at baseline and 50% (31/62) of them at month 12. Women had more severe problems in some EQ-5D-5L dimensions (mobility, usual activities, and pain/discomfort) and lower median convenience scores on the TSQM (version II) than men. At month 12, 84% (26/31) of the patients were satisfied or very satisfied with the app. PEAK was used by 67% (14/21) of men and 49% (20/41) of women. CONCLUSIONS This study showed high compliance and decreasing persistence and adherence over 1 year and demonstrated the feasibility of including remotely completed electronic PRO instruments in digital observational studies.