Background: Self-guided online interventions offer users the ability to participate in an intervention at their own pace and address some traditional service barriers (eg, attending in-person appointments, cost). However, these interventions suffer from high dropout rates, and current literature provides little guidance for defining and measuring online intervention adherence as it relates to clinical outcomes. Objective: This study aims to develop and test multiple measures of adherence to a specific self-guided online intervention, as guided by best practices from the literature. Methods: We conducted secondary analyses on data from a randomized controlled trial of an 8-week online cognitive behavioral program that targets depression and anxiety in college students. We defined multiple behavioral and attitudinal adherence measures at varying levels of effort (ie, low, moderate, and high). Linear regressions were run with adherence terms predicting improvement in the primary outcome measure, the 21-item Depression, Anxiety, and Stress Scale (DASS-21). Results: Of the 947 participants, 747 initiated any activity and 449 provided posttest data. Results from the intent-to-treat sample indicated that high level of effort for behavioral adherence significantly predicted symptom change (F4,746=17.18, P<.001; and β=–.26, P=.04). Moderate level of effort for attitudinal adherence also significantly predicted symptom change (F4,746=17.25, P<.001; and β=–.36, P=.03). Results differed in the initiators-only sample, such that none of the adherence measures significantly predicted symptom change (P=.09-.27). Conclusions: Our findings highlight the differential results of dose-response models testing adherence measures in predicting clinical outcomes. We summarize recommendations that might provide helpful guidance to future researchers and intervention developers aiming to investigate online intervention adherence. Trial Registration: ClinicalTrials.gov NCT04361045; https://clinicaltrials.gov/ct2/show/NCT04361045

中文翻译：

---

在抑郁和焦虑的自我指导在线干预中测量依从性：随机对照试验的二次分析

背景：自助式在线干预使用户能够以自己的节奏参与干预并解决一些传统的服务障碍（例如，参加面对面的约会、费用）。然而，这些干预措施的辍学率很高，目前的文献很少为定义和衡量与临床结果相关的在线干预依从性提供指导。目的：本研究旨在根据文献中的最佳实践，开发和测试对特定自我引导在线干预的依从性的多种测量方法。方法：我们对一项针对大学生抑郁和焦虑的为期 8 周的在线认知行为计划的随机对照试验的数据进行了二次分析。我们定义了不同努力水平（即低、中和高）的多种行为和态度依从性测量。使用预测主要结果测量、21 项抑郁、焦虑和压力量表 (DASS-21) 改善的依从性术语进行线性回归。结果：在 947 名参与者中，747 人发起了任何活动，449 人提供了后测数据。意向治疗样本的结果表明，行为依从性的高水平努力可显着预测症状变化（F4,746=17.18，P <.001；β=–.26，P=.04)。态度依从性的中等程度的努力也显着预测了症状变化（F4,746=17.25，P <.001；β=–.36，P =.03）。结果在仅发起者样本中有所不同，因此没有一项依从性测量显着预测症状变化（P =.09-.27）。结论：我们的研究结果强调了剂量反应模型测试依从性测量预测临床结果的不同结果。我们总结了一些建议，这些建议可能为未来旨在调查在线干预依从性的研究人员和干预开发人员提供有用的指导。试验注册： ClinicalTrials.gov NCT04361045；https://clinicaltrials.gov/ct2/show/NCT04361045