Weingarden Hilary, Garriga Calleja Roger, Greenberg Jennifer L, Snorrason Ivar, Matic Aleksandar, Quist Rachel, Harrison Oliver, Hoeppner Susanne S, Wilhelm Sabine
Department of Psychiatry, Massachusetts General Hospital/Harvard Medical School, Boston, MA, USA.
Koa Health, London, UK.
Internet Interv. 2023 Mar 16;32:100615. doi: 10.1016/j.invent.2023.100615. eCollection 2023 Apr.
Smartphone psychotherapies are growing in popularity, yet little is understood about (1) how people prefer to engage with psychotherapy apps, or (2) which engagement patterns constitute engagement. The present study uses secondary data from a 12-week randomized waitlist-controlled trial of smartphone-delivered cognitive behavioral therapy (CBT) for body dysmorphic disorder (BDD) ( = 77) to address these aims. Additionally, using the present study as a use-case, we seek to provide a roadmap for how researchers may improve upon methodological limitations of existing smartphone psychotherapy engagement research. We measured behavioral engagement via 19 objective variables derived from phone analytics data, which we reduced via factor analysis into two factors: 1) use volume and frequency, and 2) session duration. Cluster analysis based on engagement factors yielded three engager types, which mapped onto "deep" users, "samplers," and "light" users. The clusters did not differ significantly in improvement in BDD severity across treatment, although deep users improved more than light users at a marginally significant level. Results suggest that varying patterns of preferred engagement may be efficacious. Moreover, the study's methods provide an example of how researchers can measure and study behavioral engagement comprehensively and objectively. Trial Registration: ClinicalTrials.gov Identifier: NCT04034693.
智能手机心理疗法越来越受欢迎，但对于（1）人们更喜欢如何使用心理治疗应用程序，或者（2）哪些参与模式构成参与，我们了解得很少。本研究使用了一项为期12周的随机等待列表对照试验的二次数据，该试验针对身体变形障碍（BDD）（n = 77）提供智能手机认知行为疗法（CBT），以实现这些目标。此外，以本研究为案例，我们试图为研究人员如何改进现有智能手机心理治疗参与研究的方法局限性提供一个路线图。我们通过从手机分析数据中得出的19个客观变量来衡量行为参与度，通过因子分析将其简化为两个因子：1）使用量和频率，以及2）会话时长。基于参与因子的聚类分析产生了三种参与类型，分别对应“深度”用户、“取样者”和“轻度”用户。尽管深度用户在BDD严重程度改善方面比轻度用户有略微显著的提高，但各聚类在治疗期间的BDD严重程度改善方面没有显著差异。结果表明，不同的偏好参与模式可能是有效的。此外，该研究的方法为研究人员如何全面、客观地测量和研究行为参与提供了一个范例。试验注册：ClinicalTrials.gov标识符：NCT04034693。
