Abstract Background/Aims To explore the relationship of clinical and psychological factors with adherence based on a patient self-reported questionnaire from a prospective longitudinal study of patients receiving TNFi biologic treatment for rheumatoid arthritis, recruited since 2008. Methods Analysis of data collected from a prospective longitudinal UK study (Biologics in Rheumatoid Arthritis Genetics and Genomics Study Syndicate [BRAGGSS]) of patients with moderate to severe RA commencing TNFi biologic treatment. Clinical data including Disease Activity Score-28 (DAS28), serological status (Rheumatoid Factor positivity) and body mass index (BMI) were collected. Patients also completed self-reported questionnaires, the health assessment questionnaire (HAQ), hospital anxiety and depression scale (HAD), beliefs about medicine questionnaire (BMQ), brief illness perception score (B-IPQ) and a self-reported adherence questionnaire. Adherence was defined as any deviation from prescribed medication regime such as altering the dose, forgetting a dose, stopping treatment, missing a dose, and taking less than instructed. Associations between demographic and psychological variables and adherence were evaluated using regression models and a multivariable logistic regression model was used to further evaluate associations, controlling for BMI. Results 1,513 patients (75.7% female, age 58.1 [IQR 16.0], disease duration 9.6 ± 9.6 years and pre-treatment DAS28 scores 5.7 ± 0.8) were included. 70% of patients self-reported adherence to their medication. When comparing the different TNFi treatments, etanercept had the lowest levels of adherence at the 3 and 6 month time-points. Adherence was higher in those with older age (β = 0.03, p< 0.01), higher baseline CRP levels (β = 0.01, p= 0.03) and higher necessity beliefs (β = 0.04, p= 0.03). Adherence was lower in those with higher tender joint counts at baseline (β=-0.03, p= 0.04) and 3 months (β = 0.03, p= 0.02) (Table I). These associations persisted after controlling for the effects of BMI using a multivariable logistic model. Conclusion Self-reported adherence was 70% in this BRAGGS population. Factors associated with non-adherence included higher tender joint counts and lower necessity beliefs pre-treatment, which may allow targeting of interventions to optimise adherence in these groups. Disclosure T.K.Y. Chong: None. N. Nair: None. A.W. Morgan: None. J.D. Isaacs: None. A.G. Wilson: None. K.L. Hyrich: Honoraria; Abbvie. Grants/research support; K.H has grants from BMS and Pfizer. D. Plant: None. A. Barton: Grants/research support; A.B has received funding/speaker fees from BMS, Pfizer and Galapagos.
Read full abstract