OUWB awarded $200,000 grant to study how AI can improve maternal health

Researchers from Oakland University William Beaumont School of Medicine have received a $200,000 grant to study one possible use of artificial intelligence to help improve maternal and infant health outcomes.

The grant was received from the Michigan Health Endowment Fund, a philanthropic foundation that works to improve the health and wellness of Michigan residents while reducing health care costs.

The one-year project is a collaboration between OUWB, Corewell Health in southeast Michigan, and a small stakeholder advisory group consisting of representatives from community-based organizations, patient population, community health workers, and the team leaders.

The project aims to enhance the effectiveness of community health workers (CHWs) by using AI to identify social determinants of health that are associated with poor outcomes.

Project lead is Ramin Homayouni, Ph.D., professor, Department of Foundational Medical Studies and founding director, Population Health Informatics, OUWB.

"This is a pivotal point where our four years of work around AI is actually coming into practicality," says Homayouni. "Until now, it's been theoretical and engineering work.'

"Now we have an opportunity to implement this and show that implementation makes an impact on population health," he adds

Project origins and AI

Homayouni has been with OUWB since 2018 and works closely with colleagues from Corewell Health.

According to Homayouni, he was recently invited to serve on the Corewell Health Equity Council in the Detroit area, which consists of researchers, physicians, and nurses.

He says one of the entity's focus areas is maternal health and maternal health outcomes. Maternal health generally is the health of women during pregnancy, childbirth, and the postpartum period. The Centers for Disease Control and Prevention reports that 50,000 women in the U.S. suffer from pregnancy complications annually, but that Black women are at least three times more likely to die due to pregnancy-related complications when compared with White women.

Ultimately - and in large part because of his expertise in the field - it was decided Homayouni would be the lead on a project that was proposed to the Michigan Health Endowment Fund in November 2023.

The project relies on AI and existing information that doctors already documenting in electronic health records (EHRs). Simply put, the AI can scan the clinical notes in the EHRs to detect social and behavioral factors like tobacco use, depression, food insecurity, stress, and more.

The proprietary AI was developed by Homayouni and a team of data scientists at Corewell Health.

"We have demonstrated that using AI-enabled technology, we can more comprehensively identify the social and behavioral needs in our specific patient population beyond what is being captured in the screening questionnaires," says Homayouni.

The idea is that once risk factors are generally identified for patients, community health workers (CHWs) can tailor individualized care and programming for their patients in ways that address specific patient needs.

The role of CHWs

Hurse

Deidre Hurse, Ph.D., is an assistant professor in OUWB's Department of Foundational Medical Studies. She's also part of the project team.

Her previous experience includes serving on the board of the Michigan Community Health Worker Alliance (MiCHWA). Further, she was interim executive director of MiCHWA when it transitioned to become its own nonprofit.

In short, she says, CHWs are public health workers who connect communities with their health care systems, state health departments, and other resources. CHWs typically have experience related to the area on which they focus.

For example, CHWs who help people navigate poverty are likely to have experience applying for Medicaid or for food assistance programs.

Other CHWs focus on health care. Though they aren't trained physicians or nurses, evidence shows that they can be effective in improving health outcomes, reducing costs, and promoting health equity. Further, numerous studies have demonstrated that CHWs can significantly contribute to improving maternal health outcomes, especially in low-income and rural areas.

"CHWs are so effective because social determinants are those factors that happen outside the exam room and truly impact health the most," says Hurse. "Doctors treat and stabilize patients, provide medicines…but if patients don't have environments where basic needs can be met, it causes problems, and the likelihood of recovery is reduced."

The OUWB project seeks to identify when people require help meeting those basic needs and provide the information to CHWs who can reach out and provide the help patients need

'Excited to use'

Leslie Meyer is senior director of Health Equity and Community Health, Corewell Health.

She calls the project a "continuation of the rich relationship" between the health system and OUWB, and a "very innovative approach to addressing maternal health."

According to Meyer, Corewell Health now handles an estimated 30% of all births in Michigan. That means the project offers the possibility of having a significant impact on maternal health, she says.

"When we talk about addressing maternal health, one of the gaps historically has been with regard to the social determinants of health," she says. "We're very excited to use this innovative tool to help us target not just the pregnancy health of the individual, but also those social conditions that are driving health outcomes for the baby and the parent."

One recent study found that a strong relationship with health care systems and insurers is among the major factors that affect CHW programs that focus on maternal health. Specifically, the study highlights the importance of having access to clinical data from EHRs to identify information that could make CHWs more effective.

The OUWB/Corewell relationship allows for the grant project to establish a pathway to extract relevant information from EHRs without giving CHWs direct access to private records.

Homayouni says this is a much more effective way of identifying the social determinants of health than has been done in the past. That's because traditionally such information has been obtained via surveys, which are hit and miss at best, or not collected at all. More than 90% of patients who receive treatment aren't surveyed, he says.

"The bottom line is if you rely only on surveys, you automatically create a disparity," he says. "There's a huge chunk of people with real needs who aren't even being surveyed."

"That's why AI is so important," he adds. "The information is already there in the medical records, which can complement what is being gathered in the surveys."

For the next year, says Homayouni, the pilot grant will determine the feasibility of the process. Essentially, the team will study the accuracy of the data, track how it's delivered to CHWs, and determine if workflow improves.

The hope is that a second grant will allow the team to take the project further and look at the long-term impact on maternal health.

"We hope to show that with limited resources, you can still have a big impact," says Homayouni.

For more information, contact Andrew Dietderich, senior marketing specialist, OUWB, at [email protected].

To request an interview, visit the OUWB Communications & Marketing webpage.

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Public link

Please use the above public link if you want to share this noodl on another website.