Implementing a Digital Self-Care Cardiometabolic Program in Nigeria: Outcomes From a Virtual Coaching Model

Objective: To assess the impact of a digitally enabled, human- and AI-supported virtual coaching model on blood pressure and self-care engagement among adults with cardiometabolic risk. Design and method: Between January 2024 to December 2025, 200 participants were onboarded onto a mobile application, CompleteHealth™, and enrolled in the virtual cardiometabolic coaching program for self-monitoring of blood pressure. Each participant was assigned a dedicated human health coach who supported goal setting co-creation and provided personalized engagement through weekly one-on-one phone calls and weekly group coaching sessions. In addition, participants had 24/7 access to Kem, an AI virtual health coach, who delivered continuous health education, symptom triage, and referral support to appropriate health facilities. All AI-driven interactions operated under 100% human clinician oversight, with clearly defined escalation pathways and seamless handoffs to human coaches when clinically indicated. Group coaching sessions emphasized lifestyle modification and self-care behaviors relevant to cardiometabolic risk reduction. An average of 11 one on one phone call interactions between human coach and participants with a total of 41 group coaching sessions centered on nutrition, medication adherence, mental health, exercise with 126 participants (63%) having at least one interaction with the AI health coach Kem for nudges, continuous health education and clinical referrals Results: The mean age was 47 years, with approximately 57% being female. Participants’ cardiometabolic profiles at baseline were 73% (147) with hypertension, 66%(133) with obesity and 30% (60) with diabetes with mean BP at 144/90 mmHg. Participants recorded an average of one BP reading weekly on the app. After 24 weeks of the intervention, there was a mean reduction of -15.2 mmHg for systolic blood pressure and -9.4 mmHg for diastolic blood pressure. Conclusions: A digitally enabled cardiometabolic self-care program integrating human and AI health coaching was feasible and well adopted among adults with a high burden of cardiometabolic risk. Participants demonstrated clinically meaningful reductions in blood pressure from a hypertensive baseline, alongside strong engagement with digital self-monitoring, human coaching, and AI-supported education. These findings suggest that integrated virtual coaching models may offer a scalable approach to improving cardiometabolic outcomes in resource-constrained settings