The high global prevalence of depression, together with the recent acceleration of remote care owing to the COVID-19 pandemic, has prompted increased interest in the efficacy of digital interventions for the treatment of depression. We provide a summary of the latest evidence base for digital interventions in the treatment of depression based on the largest study sample to date. A systematic literature search identified 83 studies (N = 15,530) that randomly allocated participants to a digital intervention for depression versus an active or inactive control condition. Overall heterogeneity was very high (I2 = 84%). Using a random-effects multilevel metaregression model, we found a significant medium overall effect size of digital interventions compared with all control conditions (g = .52). Subgroup analyses revealed significant differences between interventions and different control conditions (WLC: g = .70; attention: g = .36; TAU: g = .31), significantly higher effect sizes in interventions that involved human therapeutic guidance (g = .63) compared with self-help interventions (g = .34), and significantly lower effect sizes for effectiveness trials (g = .30) compared with efficacy trials (g = .59). We found no significant difference in outcomes between smartphone-based apps and computer- and Internet-based interventions and no significant difference between human-guided digital interventions and face-to-face psychotherapy for depression, although the number of studies in both comparisons was low. Findings from the current meta-analysis provide evidence for the efficacy and effectiveness of digital interventions for the treatment of depression for a variety of populations. However, reported effect sizes may be exaggerated because of publication bias, and compliance with digital interventions outside of highly controlled settings remains a significant challenge. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

中文翻译：

---

治疗抑郁症的数字干预：元分析综述。

全球抑郁症的高患病率，以及最近因 COVID-19 大流行而加速的远程护理，促使人们对数字干预治疗抑郁症的功效越来越感兴趣。我们根据迄今为止最大的研究样本，总结了数字干预治疗抑郁症的最新证据基础。一项系统的文献搜索确定了 83 项研究（N = 15,530），这些研究将参与者随机分配到抑郁症的数字干预与主动或非主动控制条件。总体异质性非常高（I2 = 84%）。使用随机效应多级元回归模型，我们发现与所有控制条件 (g = .52) 相比，数字干预的总体效应大小显着中等。亚组分析显示干预措施与不同对照条件之间存在显着差异（WLC：g = .70；注意力：g = .36；TAU：g = .31），涉及人类治疗指导的干预措施的效果显着更高（g = .63 ) 与自助干预 (g = .34) 相比，并且与功效试验 (g = .59) 相比，有效性试验 (g = .30) 的效应量显着降低。我们发现基于智能手机的应用程序与基于计算机和互联网的干预措施之间的结果没有显着差异，人工指导的数字干预措施和面对面的抑郁症心理治疗之间也没有显着差异，尽管这两种比较的研究数量都很少. 当前荟萃分析的结果为数字干预治疗各种人群的抑郁症的有效性和有效性提供了证据。然而，由于发表偏倚，报告的效应量可能被夸大，并且在高度控制的环境之外遵守数字干预仍然是一个重大挑战。（PsycInfo 数据库记录 (c) 2021 APA，保留所有权利）。