Causes of drug discontinuation in patients with major depressive disorder in China

Background: This research was designed to investigate patient-reported and doctor-reported reasons for the discontinuation of pharmacological treatment in Chinese patients with major depressive disorder (MDD), which was part of the National Survey on Symptomatology of Depression (NSSD) from 2014 to 2015. Methods: This cross-sectional study included 649 patients who had discontinued antidepressant medications and 711 patients who had remained on them, selected from a group of 3516 candidates who have had at least one depressive episode. Differences in the two groups' sociodemographic factors, clinical characteristics, medication use, and self-reported reasons for drug discontinuation were compared via Student's t-test or chi-square test. Logistic regression analysis was then used to determine the association of all non-subjective dichotomous and ordinal categorical variables, including the additional 63 items of our physician-evaluated symptomatic assessment, with drug compliance. Results: Compared to the spontaneous drug discontinuation (SDD) group, the drug adherence (DA) group had significantly lower rates of the following: family history of mental disease (9.0% vs 13.6%), highest level of education achieved being post-graduate or above (1.6% vs 4.7%), smoking (5.8% vs 9.7%), and other health problems (33.9% vs 42.4%) (p's<0.05). On the other hand, first-episode depression (48.5% vs 21.9%) and taking of mood stabilizer(s) (8.3% vs 5.6%) were higher in the former group than in the latter (p's< 0.05). Logistic Regression Analysis showed that five symptoms, such as depressed mood, were correlated positively with SDD, while another six symptoms, such as psychomotor retardation, were correlated negatively with it. The receiver operating characteristic (ROC) curve of this model yielded an area under the curve (AUC) of 0.701 (95% CI, 0.673-0.729). Notably, there were three main reasons given by patients in the DA group as to why they discontinued their medication(s): (1) concern about long-term side effects (36.1%), (2) no perceived need for taking said medication(s) long-term (34.2%), and (3) believing oneself to have been cured completely (30.0%). Conclusions: The aforementioned factors may affect patient compliance and elicit maladaptive thinking even from patients with good educational backgrounds, increasing the risk of drug discontinuation. Compliance of pharmacological treatment might be improved by increasing clarification and elucidation of different symptom clusters to the patient and combating the main reasons for drug discontinuation.