Optimizing Engagement With a Smartphone App to Prevent Violence Against Adolescents: Results From a Cluster Randomized Factorial Trial in Tanzania.

Background: Violence and abuse exert extensive health, social, and economic burdens on adolescents in low- and middle-income countries. Digital parenting interventions are promising for mitigating risks at scale. However, their potential for public health impact hinges on meaningful engagement with the digital platform.

Objective: The objective of this study was to evaluate the impact of 3 intervention design and implementation factors aimed at increasing engagement with a noncommercialized, offline-first smartphone app for caregivers of adolescents in Tanzania, in partnership with the United Nations Children's Fund, the World Health Organization, and the Tanzanian national government.

Methods: Following Multiphase Optimization Strategy (MOST) principles, we conducted a 2×2×2 cluster randomized factorial trial involving caregivers of adolescents aged 10 to 17 years. Caregivers were recruited by community representatives from 16 urban and periurban communities (ie, clusters) in the Mwanza region of Tanzania. Each cluster was randomized to 1 of 2 levels of each factor: guidance (self-guided or guided via facilitator-moderated WhatsApp groups), app design (structured or unstructured), and preprogram digital support (basic or enhanced). Primary outcomes were automatically tracked measures of engagement (app launches, modules completed, and home practice activities reviewed), with secondary outcomes including modules started, time spent in the app, and positive behaviors logged. Generalized linear mixed-effects models assessed the impact of experimental factors on engagement.

Results: Automatically tracked engagement data from 614 caregivers were analyzed, of which 205 (33.4%) were men. Compared to self-guided participants, receiving guidance alongside the app led to significantly more app launches (mean ratio [MR] 2.93, 95% CI 1.84-4.68; P<.001), modules completed (MR 1.29, 95% CI 1.05-1.58; P=.02), modules started (MR 1.20, 95% CI 1.02-1.42; P=.03), time spent in the app (MR 1.45, 95% CI 1.39-1.51; P<.001), and positive behavior logs (MR 2.73, 95% CI 2.07-3.60; P<.001). Compared to the structured design, unstructured design use resulted in significantly more modules completed (MR 1.49, 95% CI 1.26-1.76; P<.001), home practice activity reviews (MR 7.49, 95% CI 5.19-10.82; P<.001), modules started (MR 1.27, 95% CI 1.06-1.52; P=.01), time spent in the app (MR 1.84, 95% CI 1.70-1.99; P<.001), and positive behavior logs (MR 55.68, 95% CI 16.48-188.14; P<.001). While analyses did not detect an effect of enhanced digital support on directly observed engagement, the combination of enhanced digital support and guidance positively influenced engagement across a range of outcomes.

Conclusions: This study is the first to systematically optimize engagement with a digital parenting intervention in a low- and middle-income country. Our findings offer important learnings for developing evidence-based, scalable digital interventions in resource-constrained settings.

Trial Registration: Pan-African Clinical Trial Registry PACTR202210657553944; https://pactr.samrc.ac.za/TrialDisplay.aspx?TrialID=24051.

International Registered Report Identifier (irrid): RR2-10.1186/s12889-023-15989-x.