Machine Learning and AI Use Cases for Maternal Health Users on MomConnect

MomConnect is a flagship digital health programme of the South African National Department of Health, designed to strengthen maternal and child health outcomes through mobile-based communication (SMS, USSD and WhatsApp). Since 2014, MomConnect has reached more than 5 million women across more than 95% of public health facilities and supports nearly 300,000 monthly users. It provides women with comprehensive support services extending from pregnancy until their child’s second birthday. MomConnect has incorporated several AI-enabled features to augment maternal health support, from Ask-A-Question feature to AI-powered Symptom Checker. However, improvements in user data collection remain a challenge. Historically, MomConnect has relied on static, multiple-choice surveys to gather basic information such as demographics, ANC attendance, and childhood vaccination status. These surveys can feel impersonal and repetitive, potentially limiting user engagement and reducing the depth of insights captured. In partnership with OpenAI and the Agency Fund, we tested whether LLM-assisted user journeys -embedded in the MomConnect interface - can improve onboarding participation, survey completion, and engagement with ANC feedback processes, without compromising data quality or user satisfaction. Looking ahead, our experimentation with interruptible voice models and AI promise to enable a voice-first user experience, significantly lowering the literacy barrier for MomConnect mothers.

At Reach Digital Health we believe in a future where uninterrupted healthcare is available to all, not only a chance few. We leverage simple technology, such as SMS and WhatsApp, to empower citizens on their health journey, support health workers in building their resilience and provide critical data and insights to health administrators to drive systems change. Building on 18 years of experience scaling digital health programmes we have reached over 100 million people in 200 countries and territories in 26 languages. Our approach leverages our multi-channel, highly scalable technology stack and incorporates all the other elements critical to implement, scale and sustain a personal, digital health system. These elements include our behaviourally informed content, fit-for-purpose measurement frameworks, implementation and scaling strategies and capacity building of partners. We are transforming how communities access and engage with health information. Our digital solutions integrate voice, messaging, and AI to drive deeply personalised and sustained behavioural change by addressing personal behavioural barriers and significantly enhancing individuals' decision-making abilities. In the age of AI application, our focus on personalisation, proxy measures, and risk assessments is crucial for optimising these behaviour change interventions. Our programmes have been proven, through rigorous evaluation to improve both healthy behaviours and the uptake of services. These include an improved likelihood of exclusive breastfeeding, an increase in uptake of family planning post-birth and the use of a condom at the last time they engaged in sex, amongst others. By blending cutting-edge technology with evidence-based behavioural insights, we empower individuals with the knowledge and tools they need to take action, improving health outcomes at scale.