Background/Purpose: Adherence to disease modifying anti-rheumatic drug (DMARD) therapy is suboptimal in patients with rheumatoid arthritis (RA). Efficient, low-cost measures are required for optimal monitoring of medication adherence in the rheumatology clinic. Self-report tools are the most efficient and cost-effective measures available. Recently, a 5-item version of the Compliance Questionnaire Rheumatology (CQR5) was developed from the original 19-item version and was validated in a group of RA patients¹. We also recently demonstrated that the CQR5 exhibits excellent test-retest reliability and significant correlation with medication beliefs in people with RA in the short-term², however long-term stability of the questionnaire remains unknown. Therefore, in the same group of patients, we investigated 1) test-retest reliability of the CQR5 and 2) correlation of the CQR5 with medication beliefs in the long term.

Methods: RA patients (disease duration ≥ 1 yr) taking at least one DMARD prescription were randomly selected from a rheumatology outpatient clinic database. Patients were assessed at baseline and three months. At each visit, medication adherence was assessed with the CQR5. Each item on the CQR5 was scored on a four point Likert scale (1 = strongly disagree, 4 = strongly agree). Scores for each item were then summed into a total score which varied between 0 and 20. Higher scores indicated greater adherence. Medication Beliefs were evaluated using the Beliefs about Medicines Questionnaire (BMQ)³, which examines beliefs around concern, necessity, harm and overuse of taking medications (high scores reflect strong beliefs). Intraclass correlation coefficients (ICC) were used to evaluate test-retest reliability, and bivariate correlations (Pearson’s r) were performed to determine relationships between medication adherence measured by CQR5 and medication beliefs measured with the BMQ.

Results:   100 RA patients, [age, mean (SD) = 60.75(12.67) yrs], were recruited and 10 dropped out of the study by three months. In this sample (n=90), the CQR5 demonstrated good test-retest reliability (ICC=0.73, 95% confidence interval = 0.61-0.81; Cronbach’s α = 0.84). At 3 months, adherence measured by CQR5 also had a significant positive correlation with BMQ necessity scores (r = 0.67, p<0.01) and significant negative correlations with BMQ concerns (r = -0.24, p<0.05), harm (r = -0.39, p<0.01), and overuse (r = -0.44, p<0.01) scores.

Conclusion: The CQR5 demonstrates long-term stability by maintaining its reliability and ability to predict medication beliefs over time. A more in-depth investigation of the predictive validity of the CQR5 using multiple chain imputation and multivariate analysis is currently underway.

References:

1) BMC Musculoskeletal Disorders. 2013; 14:286 2) Arthritis & Rheumatology. 2015; 67(suppl 10): Abstract # 2649 3) Psychology & Health. 1999; 14: 1

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/long-term-stability-of-the-5-item-compliance-questionnaire-rheumatology-as-a-measure-of-adherence-in-patients-with-rheumatoid-arthritis/