10th U6+ Consortium of African Universities International Conference

Maternal mortality stayed a huge public health challenge in rural Kenya, where delayed recognition of pregnancy related complications, played a big role in preventable maternal and neonatal deaths. The World Health Organization noted that Kenya recorded around 379 deaths per 100,000 live births in 2023, and that number was far above the Sustainable Development Goal target which is under 70 deaths per 100,000 live births. In addition, geographic barriers along with shortages of skilled health personnel, and weak referral systems all limited timely access to good quality maternal healthcare, especially for people in underserved rural communities. This work intended to build and test a citizen science driven; Artificial Intelligence (AI)-assisted framework meant for early detection of maternal health risks. It also aimed to see whether this framework could reasonably reinforce Community Health Volunteer (CHV)-led maternal care, and also the referral systems in rural Kenya. A mixed methods design was planned, bringing in CHVs, pregnant women, and healthcare providers from chosen rural counties. Community Health Volunteers were expected to gather maternal health information using a mobile health application. The data were to cover clinical symptoms, antenatal care indicators, plus socio demographic characteristics. In the machine learning part, a Random Forest approach was proposed, to predict pregnancy related complications such as pre-eclampsia, gestational diabetes, severe anaemia, and postpartum haemorrhage. Model quality was supposed to be judged using sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve. At the same time, qualitative interviews were arranged, to look at feasibility, acceptability, and usability, from the perspective of users. The proposed framework was meant to help with earlier identification of high-risk pregnancies, tighten referral routes, boost community engagement, and improve timely access to maternal healthcare services. The study was also expected to generate practical evidence for scaled AI enabled maternal health interventions, in line with Kenya’s Universal Health Coverage direction, Sustainable Development Goal 3, and Africa’s Fifth Industrial Revolution. Ethical clearance was planned to be obtained from the relevant Institutional Ethics Review Committee before the study started.