Preterm birth risk in pregnant women: developing context-adapted models for India

Preterm birth is any live birth that occurs before 37 weeks of pregnancy. With the highest number of preterm births, India contributes to a quarter of the global preterm-related deaths. The surviving preterm babies have an increased risk for developing severe disabilities and chronic diseases. Unlike other conditions, preterm birth is not based on a specific symptom or pathogenic features but is a time-based complex syndrome with multiple aetiologies. Therefore, due to heterogeneity of the syndrome and complex non-linear temporal relationships among the predictors, predicting preterm birth risk is a tough challenge. GARBH-Ini program, coordinated by THSTI Faridabad, is a DBT India-funded observational cohort to study preterm birth in India. This study has enrolled more than 8000 pregnant participants early in their pregnancy and collected anthropometric, clinical, obstetric, socioeconomic, and ultrasonographic data, generating about 10 million data points. This talk will discuss two primary data science outcomes from this cohort data. The first outcome is to develop models to estimate gestational age accurately. An accurate gestational age will improve the timing of antenatal care and give better preterm birth estimates. The second is developing preterm birth risk prediction models to enable risk stratification of pregnant women to help strategize antenatal and perinatal care ensuring a favourable outcome. Bill and Melinda Gates Foundation funded this data science project through Grand Challenges India, BIRAC.