Drug therapy is one of the main treatment approaches for depression, and a consideration of patient preferences can effectively improve patient compliance with medication. Several studies have been conducted to explore the medication choice preferences of patients with depression, but there is a lack of studies to systematically collect and summarize the medication choice preferences of patients with depression.

To systematically review the studies related to the application of discrete choice experiment (DCE) and best-worst scaling (BWS) to the medication choice preferences of patients with depression, in order to provide reference for clinical rational drug use of patients with depression and future studies of preferences.

CNKI, Wanfang Data Knowledge Service Platform, VIP, PubMed, Web of Science and EmBase were systematically searched for the studies related to the application of DCE and BWS to the medication choice preferences of patients with depression from inception to 2021-12-31, using "DCE" "BWS" "depression" "discrete choice experiment" "best-worst scaling" as keywords on January 2022. Data were extracted from the included literature, and the PREFS checklist was used for the quality assessment of all studies. The attributes included in the study were classified into three categories of outcome, process and cost, and their relative importance on the impact of medication preferences of patients was assessed.

A total of 7 papers were included, all of which were studies related to DCE, involving 47 attributes, among which 41 were classified as outcome attributes, 5 were classified as process attributes, and 1 was classified as a cost attribute. A total of 9 most important attribute results were obtained from the 7 studies, the outcome attribute was considered most important 8 times and the cost attribute was considered most important 1 time. Among the outcome attributes, adverse effects were considered most important 5 times and effectiveness was considered most important 3 times. Based on the PREFS checklist, one study received a score of four and six studies received a score of three. Most studies need further improvement in terms of respondents and findings.

Outcome attributes are most important in the medicine taking process of patients with depression, which needs attention of clinicians and policy makers. There is room for improvement in the dimensions of variance reporting and experimental design, it is recommended for further improvement of research design in terms of respondents interpretation of findings and experimental design in future research, in order to provide more high-quality evidence for research of medication preferences of patients with depression.