Featured Student Research
Using Machine Learning to Develop Real-Time Labor Predictions
For her Master's Degree in Health Delivery Science research project, Melissa Wong, MD, an assistant professor of obstetrics and gynecology and a fellow in Maternal-Fetal Medicine at Cedars-Sinai, is focused on how machine learning can predict delivery outcomes in real time.
There is a strong nationwide push to reduce the overall cesarean delivery rate, particularly among nulliparous, term, singleton, vertex (NTSV) patients. Part of the difficulty for providers in choosing whether or not to recommend a cesarean delivery is limited predictive utility of the available information. It would be ideal to be able to predict the likelihood of a patient achieving a vaginal delivery. The aim of this project is to develop a Partometer that would utilize intrapartum variables to predict, in real time, the risk of a patient needing a cesarean delivery.
To date, we have been successful in incorporating admission, intrapartum cervical exam, and induction agents into our model, and have generated better prediction models than currently exist in the literature. We anticipate being able to develop and test this further, and, ultimately to operationalize this to aid providers and patients in shared decision making to attempt to achieve a vaginal delivery.
If this project is successful, we should be able to aid in continuing to reduce the cesarean delivery rate while maintaining excellent maternal and neonatal outcomes. If successful, then this could be operationalized for other institutions, and become a guide for patients to safely deliver vaginally.
- Kimberly Gregory, MD, MPH – Vice Chair Women's Healthcare Quality and Performance Improvement
- Tod Davis – Associate Director, Data Sciences
- Matthew Wells – Data Scientist
Enterprise Information Services