Researchers Develop AI Clinical Decision Support Tool for Nutrition

Researchers from the University of Texas at Austin’s Dell Medical School have developed an artificial intelligence (AI) clinical decision support tool designed to help clinicians discuss nutrition and engage in shared decision-making with patients regarding diet improvements.

The tool, known as Nutri, integrates with the EHR and presents information about a patient's diet to clinicians to support personalized goal setting and progress tracking between visits. The tool also drafts potential chart notes to help reduce clinician burden.

“Nutrition can be a difficult subject,” said Marissa Burgermaster, PhD, an assistant professor of nutritional sciences at UT Austin’s College of Natural Sciences and population health at UT Austin’s Dell Medical School, in the press release. “It’s a fast-moving area of study, with changing recommendations. It can be an emotional subject for patients. Add in limited time in each appointment, and it’s a discussion that is easily missed. But nutrition can be one of the most critical tools in addressing conditions such as diabetes and heart disease.”

The press release states that patients whose primary care providers talk with them about nutrition are more likely to lose weight and less likely to develop related diseases such as diabetes, heart disease, and obesity-related cancer.

Nutri will first be deployed at Lone Star Circle of Care in Austin, a federally qualified health center that focuses on providing care to uninsured, underinsured, and at-risk populations.

“This is an exciting new tool in our wheelhouse as providers that will help guide our discussion on nutrition and personalize it to the individual patient with information our patients can use,” said Lola Okunade, MD, medical director of family and adult services at Lone Star Circle of Care, in the press release. “Nutri also allows us to get more comfortable with our talking points on nutrition so we can have a robust conversation to help our patients meet their goals and our goals for them. It gives a starting point for collaboration.”

A pilot program was already launched at Lone Star Circle of Care this summer with 20 participating physicians. It is expected to last through 2022 and include up to 80 patients. A study alongside the program will track whether the tool's use prompts patients to change their eating habits and whether they feel confident enough to try new strategies after discussions with their doctor.

“Currently, there are no tools like this anywhere. It’s one of a kind,” Burgermaster said. “I can’t think of a better place to launch this tool than in a clinic network that serves a community that could benefit from these discussions. Nutrition is an important piece of health equity.”

The tool is one of the latest to result from the growing research interest in personalized and precision nutrition.

In February, the University of North Carolina (UNC) at Chapel Hill received two of 14 awards from the National Institutes of Health Common Fund as part of the Nutrition for Precision Health (NPH) study. The awards will total $170 million over five years. Researchers from UNC’s Gillings School of Global Public Health will direct a $13 million NPH Clinical Center and a $19 million Metabolomics and Clinical Assay Center.

Further, as part of the NPH study, researchers will develop predictive analytics algorithms to determine an individual’s responses to food and dietary patterns. The program will build on recent advances in biomedical science, microbiome research, AI, and areas of the All of Us Research Program.