Department of Psychiatry and Behavioral Sciences, University of Washington School of Medicine, Seattle, WA, United States.
Merck, Upper Gwynedd, PA, United States.
JMIR Hum Factors. 2024 Aug 7;11:e57082. doi: 10.2196/57082.
Digital Mental Health (DMH) tools are an effective, readily accessible, and affordable form of mental health support. However, sustained engagement with DMH is suboptimal, with limited research on DMH engagement. The Health Action Process Approach (HAPA) is an empirically supported theory of health behavior adoption and maintenance. Whether this model also explains DMH tool engagement remains unknown.
This study examined whether an adapted HAPA model predicted engagement with DMH via a self-guided website.
Visitors to the Mental Health America (MHA) website were invited to complete a brief survey measuring HAPA constructs. This cross-sectional study tested the adapted HAPA model with data collected using voluntary response sampling from 16,078 sessions (15,619 unique IP addresses from United States residents) on the MHA website from October 2021 through February 2022. Model fit was examined via structural equation modeling in predicting two engagement outcomes: (1) choice to engage with DMH (ie, spending 3 or more seconds on an MHA page, excluding screening pages) and (2) level of engagement (ie, time spent on MHA pages and number of pages visited, both excluding screening pages).
Participants chose to engage with the MHA website in 94.3% (15,161/16,078) of the sessions. Perceived need (β=.66; P<.001), outcome expectancies (β=.49; P<.001), self-efficacy (β=.44; P<.001), and perceived risk (β=.17-.18; P<.001) significantly predicted intention, and intention (β=.77; P<.001) significantly predicted planning. Planning was not significantly associated with choice to engage (β=.03; P=.18). Within participants who chose to engage, the association between planning with level of engagement was statistically significant (β=.12; P<.001). Model fit indices for both engagement outcomes were poor, with the adapted HAPA model accounting for only 0.1% and 1.4% of the variance in choice to engage and level of engagement, respectively.
Our data suggest that the HAPA model did not predict engagement with DMH via a self-guided website. More research is needed to identify appropriate theoretical frameworks and practical strategies (eg, digital design) to optimize DMH tool engagement.
数字心理健康（DMH）工具是一种有效的、易于获得的、经济实惠的心理健康支持形式。然而，人们对 DMH 的持续参与程度并不理想，对 DMH 参与度的研究也很有限。健康行动过程方法（HAPA）是一种经过实证支持的健康行为采用和维持理论。该模型是否也能解释 DMH 工具的参与度尚不清楚。
本研究通过一个自我指导的网站，检验了一个经过修正的 HAPA 模型是否能预测 DMH 的参与度。
邀请访问心理健康美国（MHA）网站的访客完成一项简短的调查，以测量 HAPA 结构。本横断面研究使用自愿反应抽样，于 2021 年 10 月至 2022 年 2 月期间从 MHA 网站上收集了 16078 个会话的数据（来自美国居民的 15619 个唯一 IP 地址），检验了修正后的 HAPA 模型。通过结构方程建模，模型拟合度在预测两个参与度结果方面进行了检验：（1）选择参与 DMH（即，在 MHA 页面上花费 3 秒或更长时间，不包括筛选页面）和（2）参与度水平（即，在 MHA 页面上花费的时间和访问的页面数量，均不包括筛选页面）。
参与者在 94.3%（15161/16078）的会话中选择参与 MHA 网站。感知需求（β=0.66；P<.001）、结果预期（β=0.49；P<.001）、自我效能感（β=0.44；P<.001）和感知风险（β=0.17-0.18；P<.001）显著预测意向，意向（β=0.77；P<.001）显著预测计划。计划与参与选择之间没有显著关联（β=0.03；P=.18）。在选择参与的参与者中，计划与参与度之间的关联具有统计学意义（β=0.12；P<.001）。两个参与度结果的模型拟合指数都很差，修正后的 HAPA 模型分别仅能解释参与选择和参与度的 0.1%和 1.4%的方差。
我们的数据表明，HAPA 模型不能通过自我指导的网站来预测对 DMH 的参与度。需要进一步研究以确定适当的理论框架和实用策略（例如，数字设计），以优化 DMH 工具的参与度。
