Modifiers in Effects of Combined Pharmacotherapy and Psychotherapy versus Pharmacotherapy Alone for Adult Depression: An Individual Participant Data Meta-Analysis

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Introduction: It is not clear which individual-patient- and study-level variables modify the effects in the comparison of combined treatment versus pharmacotherapy alone for adult depression. Methods: Randomized trials comparing combined treatment with pharmacotherapy alone were identified on September 1, 2024, and individual participant data (IPD) requested from the authors. One-stage IPD meta-analyses were conducted to estimate effects on depressive symptoms at posttreatment and follow-up, response, remission, and reliable improvement and to assess effect modifiers. Results: Of 75 eligible studies, 31 (41.33%) provided their IPD (3,703 participants). Combined treatment was more effective in reducing depressive symptoms at posttreatment (SMD = 0.45, 95% confidence interval [95% CI]: 0.27–0.64), 6- (SMD = 0.55, 95% CI: 0.25–0.85) and 12-month follow-up (SMD = 0.22, 95% CI: 0.02–0.43). The effect size was smaller in studies that provided IPD than in those that did not. No patient-level effect modifiers were identified. No difference between the two arms was found in studies using SSRIs (as opposed to other antidepressants) and recruiting specific target groups (versus unselected adult populations), and the difference between combined treatment and pharmacotherapy alone was larger if pharmacotherapy was not administered adequately, but these findings were not confirmed in sensitivity analyses. Several patient-level characteristics predicted higher depressive symptoms at posttreatment, regardless of treatment condition. Conclusion: Combined treatment outperforms pharmacotherapy alone regardless of individual characteristics. Thus, its delivery does not have to be limited to specific patient subgroups. However, IPD were available for only 41% of eligible studies, thus limiting this sample’s representativeness.

This publication has 102 references indexed in Scilit: