NCSI: Data Driven Analysis of Protective Factors to Prevent Adverse Maternal Outcomes

Status	Completed
Mentor Name	Kelly Gaither
Mentor's XSEDE Affiliation	co-PI and L2 Area lead of CEE
Mentor Has Been in XSEDE Community	4-5 years
Project Title	Data Driven Analysis of Protective Factors to Prevent Adverse Maternal Outcomes
Summary	Today, the U.S. spends five times more per capita on health care than countries with similar life expectancies and costs are still rising. Despite this investment, the U.S. has one of the highest rates of maternal mortality and morbidity. There is a significant disparity in maternal outcomes for women of color, particularly for Black and African American women and Indigenous Alaskan Native women. However, Indigenous Hawaiian Native women have one of the lowest rates of adverse maternal outcomes. In 2017, approximately 810 women died every day from preventable causes related to pregnancy and childbirth. This risk has led to a growing body of research to investigate race as it pertains to maternal risk. Skilled care before, during and after childbirth can save the lives of women and newborns, and barriers that limit access to quality maternal health services must be identified and addressed at both health system and societal levels. This project will use machine learning and advanced computational analytics to examine 10 years of birth and death data in the U.S. provided through the public repository curated by the National Vital Statistics System. The goal of this project is to highlight any possible protective factors present for Native Hawaiian women in an effort to further inform known problems in maternal health.
Job Description	The student will work alongside me and will develop a set of methods/algorithms to do both supervised and unsupervised learning from 10 years of birth and death data provided by NVSS. The student will work in R and Python to have access to the latest in both machine learning and statistical methods for high-dimensional data. The student will also work with me to develop a publication from the work that will be submitted to a journal in maternal fetal medicine. Dr. Radek Bukowski, a practicing maternal fetal medicine physician will also participate and provide oversight on the medical aspects of the project.
Computational Resources	We will use both Stampede2 and Frontera for all computations using both the Vis portals and Jupyter Notebooks. All computations will be done in R and Python. We will also use dashboarding to
Contribution to Community	The work will directly contribute to the state of the art in both data science and maternal health.
Position Type	Intern
Training Plan	The student will be gaining hands-on experience in all aspects of the project and will work directly with me in the research.
Student Prerequisites/Conditions/Qualifications	Computational experience in maternal health data and computational skills in R/RStudio, Python, and modern washboarding technologies.
Duration	Semester
Start Date	01/15/2022
End Date	05/01/2022