Skip to main content

SILS Professor, CHIP Director Javed Mostafa Awarded Grants to Lead Research Projects on Improving Maternal Health and Veteran Cancer Treatment

The UNC School of Information and Library Science (SILS) Professor Javed Mostafa has been awarded grants from National Institute of Health (NIH) and The Duke Endowment to conduct two research projects centered on improving maternal health informatics. Mostafa will serve as the principal investigator (PI) for both research projects, which are anticipated to be funded for four years.

The projects support research from the Analytics and Machine-learning for Maternal-health Interventions (AMMI) initiative, which is a collaborative effort between researchers at UNC-Chapel Hill, Duke, and Wake Forrest University and aims to address racial disparities in maternal mortality and morbidity through community-engaged development and implementation of machine learning technology.

Mostafa was also awarded funding through the United States (U.S.) Department of Defense to launch a prostate cancer support system program through the interactive Prostate Cancer Information, Communications and Support program (iPICS). Mostafa will serve as site PI on the project which also involves researchers from SILS, The UNC School of Nursing and The UNC School of Medicine.

The AMMI and iPICS projects will involve current Carolina Health Informatics Program (CHIP), a program housed at SILS, students as they receive hands-on training and experience in machine-learning healthcare research.

“One of the translational benefits of this work is that we have the opportunity to train the next generation of healthcare researchers,” Mostafa, who also serves as CHIP director, said. “These scholars will complete research at the intersection of healthcare, informatics, and technology.”

Learn more about each of the projects below.

AMMI and NIH: Improving communication between providers and expectant African American mothers for enhanced treatment 

The NIH grant provides funding for AMMI researchers to develop machine-learning based technology that study maternal morbidity and mortality in expectant African American mothers, who experience higher rates of maternal mortality than white expectant mothers in the U.S.

The researchers will develop a data mart by linking biological, clinical and social determinants of health data. The information in the data mart will be used to identify risk factors for conditions antecedent to maternal mortality and translate this understanding of risk into user-facing apps designed to support individual and shared decision-making of patients, providers, and community support personnel.

The goal of the study is to develop an app that can be incorporated through Epic, the hospital software program that stores patient records. By developing and incorporating the app, researchers hope to improve communications between patients and providers, improve patient care from providers and nurses, and improve knowledge on potential risks in expectant African American mothers.

AMMI and The Duke Endowment: Bridging the gap between clinical data and social determinants of health data 

With higher maternal mortality rates in the U.S. than other high-income nations and maternal mortality rates in North Carolina exceeding the national rate, AMMI researchers aim to develop a tool that fills information gaps on electronic health records (EHR) within the UNC Health hospital system.

Through grants awarded by The Duke Endowment, researchers will work closely with UNC Health to integrate a validated risk-screening tool to expectant mothers who seek care at UNC hospitals. The tool will incorporate social determinant of health data (SDoH) of expectant mothers that can increase provider understanding of how to care for their patient and patient decision-making. SDoH data can include factors such as education, access to healthy food, secure housing, and more. Currently, SDoH data is not included in patients’ electronic health records, but can provide

The study will focus on identifying how the tool can be implemented into the patient screening process in a stakeholder-friendly manner for electronic use.

“This project will identify patient and provider challenges to integrating this screening and implement and electronic workflow that will broadly improve healthcare for expectant mothers that come to UNC Health,” Mostafa said.

By the end of the project, AMMI researchers expect between 80-100 maternity health providers at UNC Health at Chapel Hill to use this tool as part of their workflow with every prenatal patient to improve patient understanding, support and care. Additionally, the project will also serve as a model to incorporate this technology into electronic screening systems at other hospitals and health systems.

IPICS: Using machine learning to create personalized treatment and support for veterans with prostate cancer 

Mostafa will also serve as site PI for research conducted through the interactive Prostate Cancer Information, Communication and Support (iPICS) program, which is a multidimensional, scalable and innovative eHealth supportive intervention supplementing existing decision aid and survivorship programs.

The iPICS project was awarded funding through the U.S. Department of Defense to provide U.S. veterans diagnosed with prostate cancer a theory-based, integrated information and supportive care system that provides tailored and personalized information about localized prostate cancer, treatment options, and their short- and long-term effects, before, during, and after treatment.

“The existing programs used to inform patients with localized prostate cancer about their treatments are limited,” Mostafa said. “Which is why we felt the need to create a better support system for these patients, who are dealing with a long-term disease, to help them and their caregivers make more informed decisions regarding their treatment and quality of life.”

iPICS is designed to positively impact quality of life for patients diagnosed with localized prostate cancer, regardless of their health literacy or educational background, by tailoring and personalizing information and support to the individual patient.

Unlike prior programs that only include generic information, iPICS will utilize machine learning techniques to identify and integrate scientifically credible consumer health information from the public information system MedlinePlus. In turn, patients and caregivers will receive access to accurate, current research evidence based on their healthy literacy, treatment needs or preferences. Thus, patients will be provided with information to support their personal treatment decision-making and care based on data that is unique to them.

The goal of the study is to develop a system and process that can be widely adopted in clinical settings, potentially benefiting many prostate cancer patients, caregivers, and health care providers beyond those participating in the study.

Veterans and military members and Black men with prostate cancer will be especially sought out for participation in the study, to ensure that iPICS’ final design will help improve the quality of life for these patients and their family caregivers.