Modeling Infectious Disease Drivers for Gestational Diabetes Outcomes

Nicki Tiffin and Tsaone Tamuhla of the University of the Western Cape in South Africa will model how infectious and non-communicable diseases interact to affect maternal, neonatal, and child health outcomes, using gestational diabetes as a case study. These interactions include those between multiple chronic conditions and multiple medications in individuals and those due to variable access to health care. They will develop risk factor models for gestational diabetes, harnessing Large Language Models for data harmonization and standardization. The models will be applied to mother and child clinical datasets held by collaborators across the Global South through a federated data analysis approach (joint analysis without sharing the data itself). This collaboration will generate new models and evidence for gestational diabetes outcomes. It will also establish guidelines more broadly for health data modeling to inform policy, while helping build a collaborative Global South data modeling community.