<![CDATA[mHealth applications are increasingly relevant in South Africa for addressing chronic diseases, unequal access, and resource constraints. This sequential mixed-methods study evaluated user experience and mHealth effectiveness through qualitative exploration, iterative co-design, a 12–16-week quasi-experimental trial, and an implementation evaluation guided by the CFIR and RE-AIM frameworks. Qualitative findings highlighted the need for local language support, offline-first functionality, low data consumption, granular permissions, and light integration into clinical workflows. In the trial, the intervention achieved 72.4% adoption and 90-day retention of 51.8% (versus 41.2% and 29.7% in the comparator), alongside higher usability (SUS 78.6 vs. 65.9) and perceived usefulness (TAM) scores, and greater weekly engagement. Trust and privacy mediated the effect of privacy-by-design features on retention, while digital literacy and device ownership moderated adoption. Secondary clinical indicators showed improved HIV therapy adherence and better blood pressure control. Implementation fidelity reached 84%, and 79% of healthcare providers rated workflow compatibility as good or very good. These results support the view that mHealth success depends on fit-for-purpose, low-data, offline-first, and privacy-by-default design, supported by adaptive onboarding and workflow integration. Key recommendations include continuous engagement monitoring aligned with the IDHEM framework, transparent data governance, and low-bandwidth-friendly interoperability to enable scalability and sustainability.]]>
