Speaker
Description
Authors: David Stern (IDEMS International), Max Klapow (University of Oxford) FRANCES GARDNER (OXFORD UNIVERSITY), Francisco Calderon (University of Oxford), Jamie Lachman (University of Oxford, Department of Social Policy and Intervention, United Kingdom), Laurie Markle (Parenting for Lifelong Health), Lily Clements (IDEMS International), Moa Schafer (University of Oxford), Qing Han (University of Oxford), Rumaya Binti Juhari
Background: Digital behavior change interventions (DBCI) offer scalable pathways for prevention and support in low-resource settings, but maintaining user engagement remains a significant challenge. Particularly during scale-up, prevention scientists face limited resources and tight timelines, requiring pragmatic strategies for optimizing digital tools. This study demonstrates a rapid optimization process to improve engagement with ParentText, a digital parenting chatbot.
Methods: An explanatory sequential mixed-methods design was used. Engagement data from three pilot studies—Jamaica (n = 1,114), Malaysia (n = 82), and the Philippines (n = 97)—were quantitatively analyzed. Subsequently, an expert panel applied a modified Nominal Group Technique (NGT), integrating these data with qualitative inputs from semi-structured interviews (n = 18) and participant debriefings. Recommendations were prioritized based on feasibility within the technological and implementation constraints.
Findings: Low engagement was observed across sites (average 1.51 modules completed; 3.45–11.18 days of participation over a 38-day period). Experts identified five key barriers: structural constraints, usability issues, limited multimedia, absence of re-engagement mechanisms, and insufficient personalization. The panel proposed 55 modifications including shorter modules, reduced message frequency, flexible delivery timing, and options for managing multiple children. All were considered feasible for rapid implementation.
Conclusion: This study presents a practical, rapid approach to optimizing digital interventions using integrated quantitative and qualitative evidence. Such a method supports timely adaptations to improve engagement and maintain effectiveness during implementation in diverse, low-resource contexts. The proposed model offers a replicable framework for enhancing scalability and retention in digital prevention science.
| Conflict of interest | Dr. Jamie Lachman is affiliated with Parenting for Lifelong Health, which developed ParentText. No financial benefits were derived from this study. Funders did not influence the study design or dissemination. |
|---|
