Abstract
BACKGROUND: Maternal and newborn mortality remains a critical public health challenge, particularly in resource-limited settings. Despite global efforts, Kenya continues to report high maternal mortality rates of over 350 deaths per 100,000 live births and a neonatal mortality rate of 21 per 1,000 live births. Artificial Intelligence (AI)-enabled maternal healthcare interventions, such as Obstetric Point-of-Care Ultrasound (OPOCUS) and AI-driven SMS intervention on Promoting Mothers through Pregnancy and Postpartum (PROMPTS), offer innovative solutions to improve early detection, diagnosis, and maternal health-seeking behaviors. However, there is limited evidence on their usability, feasibility, and impact on maternal and neonatal outcomes. OBJECTIVE: This study aims to assess the implementation, user experiences, and impact of OPOCUS and PROMPTS on maternal and neonatal health outcomes in Kenya. Specifically, it evaluates their effectiveness in reducing maternal complications, improving antenatal and postnatal care utilization, and enhancing clinical decision-making while identifying potential barriers to adoption and scalability. METHODS: This mixed-methods, cross-sectional study will be conducted in ten counties in Kenya that have integrated AI-based maternal healthcare interventions. Quantitative data will be collected from health facility records, national health databases (KHIS), and structured surveys, while qualitative data will be gathered through key informant interviews (KIIs) with healthcare providers and policymakers, as well as focus group discussions (FGDs) with maternal health service users. Statistical analyses will include comparative pre- and post-AI implementation assessments, with thematic analysis for qualitative insights. EXPECTED OUTCOMES: The study will generate empirical evidence on the feasibility, effectiveness, and barriers to AI integration in maternal health services. Findings will inform policy recommendations, enhance AI-assisted maternal healthcare design, and support the scaling of AI-driven interventions to improve maternal and neonatal health outcomes in Kenya and other low-resource settings. CONCLUSION: AI-based maternal health interventions hold promise for reducing maternal mortality, improving diagnostic accuracy, and enhancing health-seeking behaviors. However, their success depends on user experiences, healthcare system readiness, and policy alignment. This study will provide critical insights for evidence-based scaling and policy integration of AI in maternal healthcare.