JMIR Publications

ABSTRACT

Background:

Pregnant women are considered a “high-risk” group with limited access to health facilities in urban slums in India. Barriers to utilizing the health services appropriately may lead to maternal and child mortality, morbidity, low birth weight, and children with stunted growth. With the insurgence of the application of artificial intelligence (AI) and machine learning (ML) in the health sector, we plan to develop a predictive model that can enable substantial uptake of maternal health services and improvements in healthcare adverse pregnancy outcomes from early diagnostics to treatment in urban slum settings.

Objective:

The objective of our study is to develop and evaluate the AI-guided citizen-centric platform that will support the uptake of maternal health services among pregnant women seeking antenatal care living in urban slum settings.

Methods:

We will conduct a cross-sectional study using a mixed method approach to enroll 225 pregnant women living in the urban slums of Delhi for more than 6 months, aged 18-44 years, seeking antenatal care and having smartphones. Quantitative and qualitative data will be collected using an Open Data Kit (ODK) android-based tool. Variables gathered will include socio-demographics, clinical history, pregnancy history, dietary history, COVID history, healthcare facility data, socioeconomic status, and pregnancy outcomes. All data gathered will be aggregated into a common database. We will employ AI to predict the early at-risk pregnancy outcomes (in terms of the type of delivery method, term, and related complications) depending on the needs of the beneficiaries translating into effective service-delivery improvements in enhancing the utilization of maternal health services amongst pregnant women seeking antenatal care. The proposed research will help policymakers to prioritize resource planning, resource allocation, and the development of programmes and policies to enhance maternal health outcomes. The academic research study has received ethical approval from the University Research Ethics Committee (UREC) of Dehradun Institute of Technology (DIT) University, Dehradun, India.

Results:

The proposed AI-guided citizen-centric tool will be designed, developed, implemented, and evaluated using principles of human-centered design that will help to predict early at-risk pregnancy outcomes.

Conclusions:

The proposed internet-enabled AI-guided prediction model will help identify potential risk associated with pregnancies and enhance the uptake of maternal health services among those seeking antenatal care, for safer deliveries. We will explore the scalability of the proposed platform up to different geographic locations for adoption for similar and other health conditions.