Key points are not available for this paper at this time.
Background: Rheumatoid arthritis (RA) considerably impacts patients' lives. Mobile health (mHealth), the use of smartphone apps in healthcare, is one approach to explore this impact and to support patients in their ability to manage it, even between clinic visits. However, mHealth interventions involve specific challenges, including variable user engagement. Objectives: We aimed to describe engagement with a smartphone app for patient education and remote monitoring in RA, and to explore if apps can provide insight into the dynamics of patient-experienced disease impact. Methods: App-based Education and Goal-setting in RA (AEGORA) is an ongoing multicenter, open-label, randomized controlled trial studying an mHealth self-management intervention to improve self-efficacy and remotely monitor disease impact among patients with RA 1. The 16-week intervention is accessible via a study version of the commercially available app Sidekick Health and consists of patient education, goal-setting, lifestyle support and remote reporting of the Rheumatoid Arthritis Impact of Disease (RAID) instrument. Participants were randomized 2:1:1 to usual care or intervention group A/B (access to the study app with weekly/monthly prompts to complete the RAID, respectively). We described engagement as the proportion of participants with ongoing activity in the study app beyond each time (retention), the number of days participants used the app, and the proportion of RAID prompts that were completed. Finally, we described the temporal evolution of RAID scores reported in the app. Results: In total, 122 patients were recruited, 60 of whom were randomized to the study app (n = 30 in both Groups A and B). The mean age was 58 (SD 12) years, the mean disease duration 11 (SD 9) years, disease activity was well-controlled DAS28-CRP mean 2.3 (SD 0.9), RAID mean 3.6 (SD 2.4), and the majority were women (63%) and seropositive for RF/ACPA (72%). Overall, 33/60 (55%) patients had attained at least a bachelor's level tertiary education, but only 18/60 (30%) had any prior experience with mHealth. Overall, retention steadily declined to 43% by week 16, trending towards higher retention in Group A (Figure 1). Participants used the app for a median (IQR) of 19 (65) days, with 19/60 (32%) using the app on ≥ 56/112 days and 13/60 (22%) on ≥ 84/112 days. In total, 292/600 (49%) of prompted RAIDs were completed, with no clear difference between groups A and B (47% and 53%, respectively, p = 0.26). Reported RAID scores remained stable over time at the group level, but with important inter- and intra-individual variation (Figure 2). In 21/205 (10%) pairs of consecutively completed RAID scores, intra-individual variations exceeded both the minimal clinically important worsening of ≥3/10 or ≥50% and the threshold value of 2/10 corresponding with an acceptable symptom state. Among patients who completed consecutive RAIDs at least once, 14/40 (35%) experienced ≥1 such episode over the 16-week intervention. When a subsequent score was available, RAID scores eventually returned to their initial value in 4/11 (36%) of these patients. Conclusion: Remote monitoring of the RAID via a smartphone app appears to be particularly useful at the individual level, with one-in-three patients reporting ≥1 episode of clinically important worsening over 16 weeks. Although engagement declined over time, 43% of participants were still using the app at the end of the 16-week intervention. REFERENCES: 1 Doumen M, et al. Effectiveness and feasibility of a mobile health self-management intervention in rheumatoid arthritis: study protocol for a pragmatic multicentre randomised controlled trial (AEGORA). Trials 2023; 24:697. Colored lines show RAID scores (no imputation) from individual participants. The black line shows their overall group-level evolution. Group A received weekly prompts to complete the RAID in the study app, Group B received these monthly (weeks 1, 5, 9 and 13). Acknowledgements: We thank Gerda Vandevoorde for her help with data entry. Disclosure of Interests: None declared.
Doumen et al. (Sat,) studied this question.