POS0803 A LONGITUDINAL ANALYSIS OF OUTCOMES OF SYSTEMIC LUPUS ERYTHEMATOSUS PATIENTS COMMENCING ANTIMALARIAL MEDICATION USING A MULTISTATE MODEL APPROACH

Background: Antimalarial medications (AM), such as hydroxychloroquine and chloroquine, are essential in treating systemic lupus erythematosus (SLE), effectively preventing disease flares, reducing mortality, and minimizing organ damage with minimal toxicity [1, 2]. Despite these benefits, AM non-adherence remains a significant issue, with rates between 25% and 57% over five years [3]. Prior studies on medication adherence in SLE have predominantly used prevalent cohort, traditional cohort designs, assessing adherence within fixed time windows or modelling time to first non-adherence. These approaches may be subject to survivorship bias and fail to capture the dynamic and recurrent transitions between adherence, non-adherence, and death states, potentially leading to biased inferences [4]. Therefore, there is a pressing need for incident population-based cohorts and longitudinal studies to comprehensively model the complex dynamics between AM adherence, non-adherence, and death in SLE. Objectives: To identify factors influencing transitions between AM adherence, non-adherence, and death in a population-based incident cohort of SLE patients [5]. Methods: Using British Columbia health administrative data, we conducted a retrospective cohort study of all incident SLE patients who took AM between 1997 and 2023. Eligible SLE patients in the cohort were age ≥ 18, with two principal SLE diagnoses from any care visits at least two months apart within two years, without any SLE diagnoses seven years before the first diagnosed ICD code. During the study period, three states were observed: AM adherence, non-adherence, and death (Figure 1). We used reversible multistate models to characterize transitions between states of adherence, non-adherence, and death. The "AM-adherence" state starts from the prescription date and lasts until the dosage runs out. The "non-adherence" state is defined as ≥ 3 consecutive months without AM use after the supply depletion. Transitions occurred bilaterally between adherence and non-adherence states and unidirectionally from both adherence and non-adherence to death. Baseline covariates inlude demographics, socioeconomic status, comorbidities, and medication use. Results: The cohort included 7,399 incident SLE patients, with a median age of 43, 86.0% female, and 11.0% residing in rural areas. Medication use included NSAIDs (25.4%), glucocorticoids (23.6%), and cardiovascular drugs (16.6%), with 21.5% reporting prior hospitalized infections. Among 20,559 adherence transitions, 99.2% transitioned to non-adherence, 0.75% to death, and 0.01% had no event and were censored due to the end of the study. From 20,402 non-adherence transitions, 64.5% returned to adherence, 4.07% transitioned to death, and 31.4% had no event and were censored. Regarding transitions from adherence to non-adherence, rural residence was associated with a decreased risk (Hazard Ratio [HR] = 0.938; 95% Confidence Interval [CI]: 0.885–0.994), while alcohol use disorder increased the risk (HR = 1.465; 95% CI: 1.109–1.935). Higher-income also decreased the risk (HR = 0.929; 95% CI: 0.877–0.984). For transitions from adherence to death, both older age and rural residence were associated with increased risks (HR = 1.059; 95% CI: 1.046–1.073, and HR=1.664; 95% CI: 1.114–2.485, respectively), with glucocorticoid use showing a reduced although not statistically significant mortality risk (HR = 0.681; 95% CI: 0.450–1.031). In transitions from non-adherence to adherence, NSAID use increased the likelihood of returning to adherence (HR = 1.080; 95% CI: 1.031–1.131), while older age (HR = 0.995; 95% CI: 0.994–0.997), depression (HR = 0.922; 95% CI: 0.861–0.988), and a higher Charlson Comorbidity Index (HR = 0.972; 95% CI: 0.945–0.999)) were also associated with lower likelihoods. Male gender (HR = 1.544; 95% CI: 1.282–1.858), older age (HR = 1.090; 95% CI: 1.083–1.096), lower income (HR = 0.633; 95% CI: 0.506–0.792), and frequent outpatient visits (HR = 1.014; 95% CI: 1.008–1.020) were associated with an increased risk of transitioning from non-adherence to death. Conclusion: Multistate models identified key risk factors influencing transitions among adherence, non-adherence, and death. Socioeconomic status, comorbidities, and medication use significantly affected adherence and mortality outcomes. Targeted interventions for low-income patients, addressing alcoholism, managing depression, and optimizing comorbid condition treatments may improve adherence and reduce mortality. Additionally, promoting the appropriate NSAIDs use can facilitate transitions from non-adherence back to adherence. Clinicians should closely monitor antimalarial medication adherence to improve patient outcomes. REFERENCES: [1] Lee SJ, Silverman E, Bargman JM. The role of antimalarial agents in the treatment of SLE and lupus nephritis. Nat Rev Nephrol. 2011 Dec;7(12):718–29. [2] Ruiz-Irastorza G, Ramos-Casals M, Brito-Zeron P, Khamashta MA. Clinical efficacy and side effects of antimalarials in systemic lupus erythematosus: a systematic review. Annals of the Rheumatic Diseases. 2010 Jan 1;69(01):20–8. [3] Mehat P, Atiquzzaman M, Esdaile JM, AviÑa-Zubieta A, De Vera MA. Medication Nonadherence in Systemic Lupus Erythematosus: A Systematic Review. Arthritis Care Res (Hoboken). 2017 Nov;69(11):1706–13. [4] Survivorship Bias in Performance Studies | The Review of Financial Studies | Oxford Academic [Internet]. [cited 2024 Nov 18]. Available from: https://academic.oup.com/rfs/article/5/4/553/1590264?login=true. [5] Breheny P. Multistate models and recurrent event models. Figure 1Multistate model for observed transitions between AM adherence, non-adherence, and death in SLE patients. Acknowledgements: NIL . Disclosure of Interests: None declared . © The Authors 2025. This abstract is an open access article published in Annals of Rheumatic Diseases under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Neither EULAR nor the publisher make any representation as to the accuracy of the content. The authors are solely responsible for the content in their abstract including accuracy of the facts, statements, results, conclusion, citing resources etc.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.