Abstract
Evidence for clinical screening and intervention for depression in cancer and the effect of this intervention on cancer prognosis is suboptimal. This study substantialized a complete model with universal screening and intervention for major depressive disorder (MDD) and explored its effect on survival in patients. This longitudinal study recruited cancer patients routinely screened for MDD with a two-stage model. Data including sex, age, cancer diagnosis, first diagnosis date, date of death, cancer stage, and MDD diagnosis and treatment were collected from medical records and the national registration system for cancer. Kaplan−Meier’s survival analysis and the Cox proportional hazards regression model were applied to analyze the effects of associated factors on survival. Further subgroup analysis for 14 types of cancer primary site was also performed. Overall, the hazard for patients adhering to psychiatric treatment for MDD before cancer diagnosis was not statistically different from that for patients without MDD (hazard ratio (HR) = 1.061, 95% CI: 0.889−1.267, p = 0.512). The hazard for patients adhering to psychiatric treatment after cancer diagnosis was significantly lower than that for patients without MDD (HR = 0.702, 95% CI: 0.607−0.812, p < 0.001). Those who were diagnosed with MDD after cancer diagnosis and adhered poorly to psychiatric treatment had the greatest hazard (HR = 1.829, 95% CI: 1.687−1.984, p < 0.001). The effect of intervention for MDD varied across different primary cancer types.