OA41 Exploring patterns of patient engagement with electronic patient reported outcome measures in inflammatory arthritis using latent class growth modelling

Abstract

Abstract Background/Aims Routine collection of electronic patient reported outcome measures (ePROMs) can facilitate dynamic disease activity monitoring in patients with inflammatory arthritis (IA). In a real-world cohort of patients with IA, we describe patient characteristics associated with ever engaging with ePROMs, evaluate whether there are distinct groups of patients with similar engagement trajectories over time and identify patient characteristics associated with these groups. Methods We conducted a prospective observational study based on routinely collected data from a remote monitoring platform (RMP) at Guy’s and St Thomas’ NHS Foundation Trust in London, UK, between 18 December 2018 and 15 August 2022. Patients with rheumatoid arthritis (RA), psoriatic arthritis (PsA) and axial spondyloarthritis (AxSpA) were sent text message requests to complete 4-weekly disease specific PROMs. A patient was defined as an engager if they completed at least one PROM episode during the study period of 68 weeks. Logistic regression (LR) was used to identify associations between baseline characteristics and any PROM engagement. Latent class growth modelling (LCGM) was used to determine PROM engagement trajectories, with multinomial LR exploring baseline characteristic associations with latent trajectory class membership. Results Of 1,203 patients on RMP in the study period, 1,129 met the inclusion criteria, of which 91.5% (95% CI 89.7% - 93.0%) completed at least one PROM. Mean age was 49 years and 64% were female. 60% had RA, 24.4% had PsA and 15.7% had AxSpA. Overall engagement was relatively high throughout, with continued engagement >60% over the remaining 68 weeks. Older patients (adj OR 0.98, 95% CI 0.96 - 0.99), those from more deprived areas (adj OR 0.80, 95% CI 0.68 - 0.96) and those who attended <80% of their outpatient visits in the preceding 3 years (adj OR 0.48, 95% CI 0.26 - 0.89) were less likely to complete a PROM. Four latent trajectory classes were identified: consistent engagers (38.1%), high-variable engagers (24.3%), low-variable engagers (19.5%) and never & dis-engagers (18.1%), with a mean (SD) overall response rate per class of 96.5% (7.1), 74.1% (12.1), 35.6% (13.8) and 7.5% (12.0) respectively. Compared to the consistent engager class, the two variable engager classes were significantly more likely to be younger at baseline, whilst the never & dis-engager class was more likely to be from a deprived area. Having an IA type of PsA and previous clinic non-attendance was associated with membership of the three lower engaging classes. Conclusion We have described key patient characteristics that influence engagement with ePROMs in a real-world cohort of patients with IA, and patterns of engagement over time. Overall engagement was very high with >90% of patients enrolled engaging with PROM completion. There is still scepticism among clinicians about remote monitoring’s ability to work, and our findings should be reassuring: patients do engage with ePROMs over time. Disclosure N. Arumalla: None. Z. Yang: None. S. Norton: None. M. Crooks: None. A. Khan: None. R. Fitzgerald: None. R. Gilligan: None. E. Lindberg: None. T. Tabibi: None. Y.L. Man: None. S. Subesinghe: None. J.B. Galloway: Consultancies; Abbvie, Biovitrum, BMS, Celgene, Chugai, Galapagos, Gilead, Janssen, Lilly, Pfizer, Novartis, Roche, Sanofi, Sobi, UCB. Honoraria; Abbvie, Biovitrum, BMS, Celgene, Chugai, Galapagos, Gilead, Janssen, Lilly, Pfizer, Novartis, Roche, Sanofi, Sobi, UCB. T. Garrood: Shareholder/stock ownership; Serac healthcare. Honoraria; Abbvie, UCB. Grants/research support; Versus Arthritis, Pfizer, Gilead, Guy's and St Thomas' Charity, NHSX. Other; Fresnius-Kabi.