P097 Using a smartphone app to better detect and characterise real-time patient-reported flares in rheumatoid arthritis

Abstract

Abstract Background/Aims Flares are an intrinsic part of the rheumatoid arthritis (RA) disease course and may impact clinical and patient outcomes, but the clinical assessment is limited by patient recall. Consumer technologies make it possible to capture and explore patient-reported flares in near real time. We used smartphone app data to explore the frequency of patient-reported flare weeks and their associations with various summary features of daily symptoms reported during the preceding week. Methods We used data from the Remote Monitoring of Rheumatoid Arthritis (REMORA) study. Patients tracked daily symptoms (pain, fatigue, function, sleep, coping, physical and emotional wellbeing) on a 0-10 scale and weekly flares (y/n) on the REMORA app. A flare week was defined as the seven days leading up to the weekly flare-question where the patient answered “yes”. We summarised the number of patient-reported flare weeks. Symptom scores in flare and non-flare weeks were compared using the Wilcoxon rank sum test. For each week prior to a flare question, we calculated three summary features for daily symptoms: 1) mean, 2) variability, and 3) slope. Mixed effects logistic regression models quantified associations between flare weeks and each summary feature. Results Twenty patients tracked symptoms over three months. 60% were female, all but one were white British, and mean age was 56.9±11.1 years. The median number of days in the study was 81 (IQR 80, 82). We included 168 participant weeks. 15/20 participants reported at least one flare week, with 54 patient-reported flare weeks in total. Participants reported a median of two flare weeks (IQR 0.75, 3.5) each. In paired analyses, mean symptom scores were significantly higher (difference on average 0.67 [SD 0.22]) in flare weeks compared to non-flare weeks except for sleep (p < 0.05). Variability was slightly larger in flare weeks, but only significantly different from non-flare weeks for emotional wellbeing. For slope, there was an increase for all symptoms in flare weeks, although only the slopes of pain, physical wellbeing, and coping were statistically significant. Univariate modelling showed that mean scores and slope in the week preceding the flare were important for the likelihood of a flare occurring, but the association with variability was less convincing. Due to high correlation between the daily symptoms, we could not establish which single symptom or summary feature had the strongest association with patient-reported flares. Conclusion In our RA cohort, self-reported flares were frequent. Flare weeks were broadly associated with higher scores (for mean, variability, and slope) across a range of daily symptoms in the preceding week. The correlation between daily symptoms made it impossible to disentangle the contribution of individual symptoms to the flare experience. Future analysis of daily symptoms may allow us to predict imminent flares, opening opportunities for just-in-time interventions. Disclosure J. Gandrup: None. S.N. van der Veer: None. J. Mcbeth: None. W.G. Dixon: Consultancies; WGD has received consultancy fees from Abbvie and Google unrelated to this work.