| Abstrak/Abstract |
Medication non-adherence in patient with diabetes has been associated with poor blood glucose control. A self-reported questionnaire consisted of 4 items Morisky Green Levine Adherence Scale (MGLS) has been developed and validated as well as widely used because of its convenience for patients. The objectives to the study were to examine the psychometric properties including internal consistency, test-retest reliability, convergent validity, and
known-groups validity of the MGLS among patients with diabetes in Indonesia. A cross sectional study was conducted to collect the data, using a convenience sampling. A total of 250 patients with diabetes at four primary health centers in Yogyakarta Province were recruited and interviewed, during September to December 2017. Four items MGLS, sociodemographic data, and medical records were reviewed for data analysis. Internal consistency reliability was moderate (Cronbach’s ? =0.651), as well as the test-retest reliability of the MGLS was moderate (Spearmans’ rank correlation = 0.425; p<0.001). Regarding convergent validity, the MGLS had a moderate correlation with the 5 items Medication Adherence Report Scale (r=0.58, p<0.01). MGLS had excellent score in known group validity, showed a significant association between MGLS and fasting blood glucose level (p<0.001). The sensitivity, specificity, positive predictive value, and negative predictive value of the MGLS were 65.5%, 46.7%, 46% and 63.5% respectively. The four items MGLS is satisfactory tool in measuring medication adherence among patients with diabetes in Indonesia, but future research could target the reasons of non-adherence and investigating its use in other chronic disease populations.
|
