Prediction model for medication adherence using a medication event monitoring system in recurrent major depressive disorder

To investigate the risk factors associated with nonadherence to antidepressive drugs in patients with recurrent major depressive disorder (MDD). A total of 847 patients undergoing maintenance treatment for recurrent MDD were prospectively enrolled. One year after discharge, patients' adherence to the prescribed antidepressants was tracked over a 30-day period using the medication event monitoring system. Low adherence was identified in 30.7% of cases. Patients with more than three exacerbations had a 2.040-fold higher risk of low adherence ( P < 0.025). Those with drug concentrations below or above the recommended therapeutic range had a 2.096-fold ( P < 0.025) and 2.361-fold ( P < 0.05) increased risk of low adherence. Patients rating their depression severity from mild-to-severe showed a trend toward increased risk of low adherence, with odds ratios (ORs) of 2.020 (NS), 4.644 ( P < 0.025), and 5.347 ( P < 0.025). Patients reporting mild to severe side effects exhibited higher risks of low adherence, with ORs of 2.212 (NS), 3.993 ( P < 0.05), and 10.965 ( P < 0.001), respectively. Conversely, older age and Drug Attitude Inventory-10 scores greater than 0 were positive predictors of adherence. A prognostic index greater than or equal to 0.800 indicated a high risk of developing low adherence. A predictive model was established to assess adherence after 1 year of maintenance treatment for recurrent MDD. Patients at high risk of low adherence could be promptly identified and closely monitored, enabling physicians to develop targeted strategies to improve adherence.