Purpose Medication non-adherence occurs frequently in individuals with systemic lupus erythematosus (SLE); proportions as high as 43–75% have been documented, depending on the methods of assessment.¹ The aim of this study was to determine factors that impact adherence to medications in a Swedish SLE population.

Methods A cross-sectional study was conducted using survey data collected from individuals diagnosed with SLE from two Swedish tertiary referral centres (Karolinska and Örebro University Hospitals). The survey comprised validated questionnaires, including the 19-item Compliance Questionnaire of Rheumatology (CQR-19) that assesses rheumatic disease-specific medication adherence and the generic Medication Adherence Self-Report Inventory (MASRI), with which we assessed adherence to glucocorticoids (GCs) and antimalarial agents (AMAs) separately. The study participants were asked to report their beliefs in medications using the Beliefs about Medicines Questionnaire (BMQ), which contains one part dealing with the patient’s specific medications and another part for medications in general. Health-related quality of life (HRQoL) was assessed with the SLE-specific LupusQoL and the generic EQ-5D-5L. Disease activity and organ damage were assessed with the Systemic Lupus Activity Questionnaire (SLAQ) and Self-Administered Brief Index of Lupus Damage (SA-BILD), respectively. Non-adherence was defined as <80% adherence according to CQR or MASRI. The CQR and the MASRI were investigated for linear relationship with Spearman’s rank correlation test. Odds ratios and 95% confidence intervals were estimated using logistic regression, with non-adherence as the outcome variable. Predictors that were investigated included country of birth, disease duration, living situation, employment status, education level, BMI, smoking status, HRQoL, and beliefs in medications. The odds ratios were adjusted for age, sex, damage, and disease activity.

Results Two hundred and five patients participated in our survey study; 45.9% (N=94) were on five medications or more i.e., polypharmacy. Most patients (66.8%) were non-adherent to their medications when assessed with CQR. However, only 6.6% and 6.3% were non-adherent to AMA or GCs respectively according to the MASRI. Adherence levels assessed by CQR showed a moderate linear relationship with those assessed using MASRI for both AMA (ρ=0.47; P<0.001) and GCs (ρ=0.34; P<0.001). The average age was 52, 86% were female and most of the patients were on AMAs (Table 1). Belief in the necessity of a specific medication and positive beliefs in medications in general were associated with adherence to GCs (Table 2). Concerns regarding specific medications and believing that medications are generally overused and harmful had a negative association with overall medication adherence. The other predictors were not associated with medication adherence.

Conclusions Our findings show that patients’ beliefs in medications may overall impact medication non-adherence. Adherence levels based on MASRI and CQR were not strongly correlated, suggesting that these two instruments capture different aspects of adherence.

Reference

1. Mehat, et al. 2017.