Design and Early Implementation Outcomes of a Mobile Health-Enabled Social Support Intervention for Adults Living With HIV and Type 2 Diabetes in Uganda

Abstract

Background: Adults living with HIV and type 2 diabetes experience complex treatment demands that affect medication-taking, clinic attendance, privacy, social support, and engagement with care. Mobile health (mHealth)-enabled social support may help reduce treatment burden, but such interventions need to fit patients’ everyday realities, phone access, trusted support networks, and clinic workflow. Objective: This study aimed to describe the user-centered design (UCD) mechanisms of an mHealth-enabled social support intervention for adults living with HIV and type 2 diabetes in Uganda and assess perceived and self-reported early implementation outcomes, including acceptability, feasibility, and appropriateness, after one month of use. Methods: This mixed-methods formative study was conducted among 64 purposively selected participants from the HIV and diabetes clinics at Mbarara Regional Referral Hospital (MRRH), including 39 adults living with HIV and type 2 diabetes, 20 nominated social supporters, and five healthcare workers. Participants were oriented to an Mhealth-enabled social support intervention composed of short message service (SMS) reminders, educational messages, social supporter SMS prompts, discreet messaging, server-supported message routing, and clinic linkage. Patients used the intervention for one month. Post-trial interviews, focus group discussions (FGDs), brief feedback forms, observation notes, and intervention-use records were used to assess acceptability, feasibility, appropriateness, concerns, social support activation, and design refinements. Interview and focus group questions were informed by the Unified Theory of Acceptance and Use of Technology (UTAUT), and refinements followed a UCD approach. Qualitative data were analyzed thematically using NVivo, while quantitative intervention-use records and feedback responses were analyzed in Stata and summarized using frequencies, percentages, and counts. Results: Participants perceived the intervention as useful for supporting medication-taking, clinic appointment preparation, communication with nominated social supporters, and a sense of being supported beyond the clinic. During the one-month trial, 1,092 SMS medication reminders, 74 appointment reminders, 312 educational or supportive messages, and 280 social supporter SMS prompts were delivered. Privacy sensitive wording was strongly preferred because of the fear of unintended HIV status disclosure. The intervention was considered feasible because it used familiar channels, including SMS, voice calls, and simple app functions, although shared phone use, airtime costs, charging problems, network interruptions, language barriers, and possible clinic workload remained concerns. Participants considered the intervention appropriate because it addressed both HIV and diabetes, enabled patient-controlled social support, protected privacy, and fit existing social and clinical support practices. Conclusion: A privacy-sensitive, low-burden intervention that combines reminders, patient-controlled supporter prompts, short educational messages, and clinic linkage could support medication-taking and care coordination for patients managing both conditions. Further implementation research is needed to refine the intervention and assess longer-term adoption, fidelity, sustainability, and preliminary impact.