PMS74 – Validation of the French Translated Version of the Osteoporosis Specific Morisky Medication Adherence Scale (OS-MMAS) in France

The OS-MMAS is a disease-specific 8 item self-reported measure of adherence for osteoporosis patients. This study evaluated the measurement properties of its French translated version. A cohort of women aged 55 or older, with post-menopausal osteoporosis (PMO) residing in France and treated with daily or weekly oral bisphosphonates (OBPs) were selected based on their historical data from the French Longitudinal Patient Database; following their visit to one of the participating practices. Eligible patients were given an OS-MMAS questionnaire for completion at home. Internal consistency was evaluated using the Cronbach’s coefficient and construct validity using confirmatory one factor analysis (CFA). Convergence validity was assessed using agreement between patients’ medication procession ratio (MPR) and their total score from OS-MMAS. To assess reproducibility, a subset of respondents received a second questionnaire approximately 3 months later. Reproducibility was tested using the intra-class correlation (ICC). From a network of 1200 French general practices, 117 participated; 687 eligible women visited these practices during the period 20-Jul-2012 to 31-Dec-2012. Of these women, 218 (mean 73.2 years; SD= 8.1) completed ≥1 OS-MMAS. The 6 and 12 month MPR was less than 80% for 25.7% patients and 59.1% patients, respectively. Cronbach’s alpha coefficient was 0.76 and varied between 0.71 and 0.79 with the deletion of one item. The Chi-Square test of association between MPR (<80%, >80%) and OS-MMAS scores (<6; 6-8; 8+) was statistically significant at 6 months (p= 0.025), but not at 12 months (p-value= 0.059). CFA reported a standardized root mean square residual of 0.06. The ICC on 70 patients was 0.24 with 95% confidence interval [0.01-0.45]. This study demonstrated acceptable agreement between classification of patient’s adherence to OBPs assessed using French translated OS-MMAS and that using historical 6 months MPR; but also indicated that the construct of the OS-MMAS could be improved.