Background Adherence to urate lowering therapy (ULT) is associated with better gout outcomes. Medication Possession Ratio (MPR)1 is commonly used in administrative database studies to assess adherence. The Morisky medication adherence scale (MMAS-8)2, an 8-item patient questionnaire (0-8), has not been studied in gout.
Objectives To evaluate the validity of the MMAS-8 in determining adherence to ULT in a multi-ethnic Asian gout cohort.
Methods Adults with gout fulfilling the ACR criteria on ≥9 months of ULT were recruited from rheumatology clinics in a tertiary hospital in Singapore. Baseline and 6-monthly clinical data, MMAS-8, Gout Assessment Questionnaire v2.0, and serum urate levels were collected. Variable MPR over 9-15 months preceding the MMAS-8 survey was calculated from dispensing records. This reflected the proportion of time a patient had access to medications over the analyzed period1. Construct validity (Spearman's correlation), reliability (intra-class correlation coefficient, ICC) and internal consistency (Cronbach's alpha) of the MMAS-8 was assessed. Validity of MMAS-8 as a measure of adherence was evaluated against MPR. MMAS-8 scores of 8, 6-7.9 and <6 were used as cutoffs for high, medium and low adherence, respectively. MMAS-8 score was analyzed both as an overall score and as 2 factors derived from exploratory factor analysis3.
Results Of the 91 patients included, 92.3% were male, 72.5% Chinese and 26.4% Malay. The mean age at baseline was 52.8 (SD 17) years. The mean BMI was 29.7 (SD 7.7), 73.7% had at least completed secondary school, and 45% had ≥3 comorbidities. The mean duration of ULT at baseline was 2.5 (SD 2.6) years and mean number of medications other than ULT was 3.3 (SD 3.4). Mean MPR and MMAS-8 were 96.3% (SD 19.0) and 6.17 (SD 1.8). High and medium adherence was seen in 24.4% and 37.7%, respectively. Internal consistency of MMAS-8 was good (α=0.725) and test-retest reliability (n=18, 20% sample) was satisfactory (ICC 0.702 (95%CI 0.362-0.877)). There was no relationship between MMAS-8 and MPR (r=0.069, p=0.521). In addition, MMAS-8 scores did not correlate with baseline urate levels (r=-0.023, p=0.83) or the absolute change in urate levels between baseline and final visits (r=-0.042, p=0.69). MMAS-8 scores were not influenced by education level, number of comorbidities, degree of polypharmacy or frequency of gout flares. However, increasing age was associated with higher MMAS-8 scores (β=0.256, p=0.015). Patients who reported “Fear of medication side effects” had lower MMAS-8 scores (p<0.005). Findings were similar when analysing MMAS-8 as 2 factor scores instead of an overall score.
Conclusions MMAS-8 does not appear to be a valid measure of adherence to ULT in our cohort, despite satisfactory internal consistency and reproducibility. Other feasible methods of reporting ULT adherence should be evaluated in patients with gout.
Kozma CM, et al. Patient Prefer Adherence (2013), 7:509–516.
Morisky DE, et al. Journal of Clin Hypertens (2008), 10:348–54.
Wang J, et al. Journal of Clinical Gerontology & Geriatrics (2013), 4: 119-122.
Disclosure of Interest None declared