P063 Beyond the clinical trials: real-world retention rate and use of digital interventions in patients with inflammatory arthritis

Abstract

Abstract Background/Aims The use of digital tools to deliver healthcare interventions has increased rapidly. Inadequate engagement and retention are two major barriers to widespread and long-term adoption of electronic patient-reported outcomes (ePROs) for monitoring in clinical practice. Whilst retention rates are high in research studies, these are often short-term (<12months) and not representative of routine clinical practice. The reasons for variable engagement remain unclear. We present an analysis of engagement and retention of digital intervention in a routine rheumatology setting. The research aims to explore determinants of long-term sustainability of app-based intervention to collect ePROs in patients with inflammatory arthritis (IA). Methods This is a secondary analysis of real-world data obtained in clinical settings from IA patients who attended routine appointments in the rheumatology clinic at North Bristol NHS Trust between 2018 and 2022. Patients were invited to use a smartphone-based symptom diary (Living With) between appointments to track disease course by reporting RAPID3, HAQ-DI, or self-assessing joint pain/swelling. Statistical tests were performed using Stata16.1. Maximum elapsed days of the app use, number, and time between reports were explored as explanatory variables. Associations between demographic and disease parameters in different user subpopulations were explored by Poisson and tobit regression models. Results were stratified post-hoc by gender, age, diagnosis, and baseline disease activity. The research has ethical approval. Results 673 patients provided ≥1 report of any ePROs. 68% were female, mean age 53.7±13.9 years. 60% had rheumatoid arthritis; 24% - psoriatic arthritis and 17% - other IA. 613(91%) patients provided a RAPID3 measure of disease activity. Among them, 82(12%) participants reported only baseline RAPID3 and were classified as “lost to follow-up”. 531(79%) were identified as “follow-up” users. No difference between these subpopulations considering gender, age, IA type, and baseline arthritis activity was identified in the adjusted Poisson regression model. Retention rates reduced most in the first six months from a baseline of 91% to 38% at month 6 and subsequently to 24% at month 12, whereafter plateaued, highlighting a group of long-term users. Older patients (aged 60-69) were more likely to be long-term users compared to younger age (≤50) participants, demonstrating longer app engagement 379 [169; 645] vs 256 [90; 515] days, p < 0.006. Patients aged 60-69 and ≥70 provided more reports (18 [7; 46] and 25 [8; 62]) compared to younger users (8 [4;18], p < 0.0001). There was no association between the length or frequency of digital tool use and gender, baseline arthritis activity, and disability. Conclusion This analysis offers insights into what might be expected regarding engagement and the long-term sustainability of app-based interventions in routine clinical settings. Understanding the real-world performance and factors that could potentially improve user retention or dropout rates is critical to realizing the promise of digital interventions. Disclosure I. Biliavska: None. E. Lenguerrand: None. J. Tobias: None. P. Hamann: Consultancies; provided consultancy for Living With Ltd Software Company. Royalties; has limited royalty agreement with Living With Ltd Software Company for the development of the smartphone app LivingWith. Honoraria; have received honoraria from Gilead and AbbVie Pharmaceuticals for the production of training materials on remote monitoring for patients with arthritis..