Mean pooled differences in total effects and costs at 6 months follow-up and results of cost-effectiveness and cost-utility analyses after 6 months follow-up for the main analysis and for the sensitivity analyses.

*Models with costs as dependent variable adjusted for gender, costs in the previous three months and baseline severity of depression. Models with adherence as dependent variable adjusted for gender and age. Models with reduction of symptoms (50% or over reduction in PHQ-9) as dependent variable adjusted for gender, comorbidities and presence of major depression. Models with quality adjusted life years (QALY) as dependent variable adjusted for gender, age and baseline quality of life.

* 以医疗成本为因变量的校正模型，校正了性别、前三个月医疗成本及抑郁症基线严重程度。以治疗依从性为因变量的校正模型，校正了性别与年龄。以症状缓解（PHQ-9抑郁症状量表评分降低50%及以上，Patient Health Questionnaire-9）为因变量的校正模型，校正了性别、合并症及重性抑郁症患病情况。以质量调整生命年（Quality Adjusted Life Years，QALY）为因变量的校正模型，校正了性别、年龄及基线生活质量。