Miller School, University Join National Effort to Make Health Care Data More Inclusive

Leaders from the University of Miami Institute for Data Science and Computing are working to build a larger, more diverse database for more accurate research on health care disparities.

While kidney disease affects one in seven adults in the United States, it impacts Black patients much more. They are four times more likely to suffer from kidney failure than white U.S. residents, and a 2021 study indicates that computer algorithms that determine eligibility for a kidney transplant often put Black people at a disadvantage, widening the gap for a successful recovery.

This is just one example of how artificial intelligence solutions and algorithms for data collection need to improve, so that other health disparities impacting minority populations—in diabetes, heart disease, and cancer care—do not continue to increase in the United States.

Recognizing this problem, in July the National Institutes of Health (NIH) began the Artificial Intelligence/Machine Learning Consortium to Advance Health Equity and Researcher Diversity program, or AIM-AHEAD, with a goal to reduce these health disparities by creating new algorithms and health care databases that more accurately reflect the diverse U.S. population. They selected the University of Miami’s Institute for Data Science and Computing (IDSC) as one of the institutions to spearhead the program’s infrastructure core, which is one of four pillars of the program. The others are partnerships, research, and data science training.

“Most of our current data is biased, and often the people collecting this data are not representative of all minorities and cultural differences,” said Nick Tsinoremas, Ph.D., the University’s vice provost of data science and computing, professor of biochemistry and molecular biology, as well as the founding director of IDSC and lead investigator on the AIM-AHEAD grant. “We want these biases eliminated, but we also want to create an infrastructure that encourages minority serving institutions to do this research because these are the people who understand biases in data and algorithms and know how to create more equitable or unbiased approaches.”

As a result, the University received a $1.3 million grant to work with historically Black colleges and universities, tribal colleges, and other minority serving institutions to create a computing structure for these institutions to share patient data that is void of personal information and that will improve the quality and breadth of health care research.

Far-reaching Collaboration

Joining Tsinoremas to lead the AIM-AHEAD project are Azizi Seixas, Ph.D., associate professor of psychiatry at the Miller School of Medicine and director of IDSC’s population health informatics program, as well as computer science professor Yelena Yesha, Ph.D., who also serves as the Knight Foundation chair of data science and artificial intelligence and IDSC’s chief innovation officer. The three are now working with minority serving institutions across Florida to improve their access to artificial intelligence tools and help them to do more efficient health equity research. Seixas said he and his researchers at the Miller School’s Media and Innovation Lab have reached out to approximately 15 institutions across the state.

“Many of these institutions have fragmented electronic health record systems, which can lead to waste, inefficiencies, and poor communication in health care delivery, which is critical. We’ll focus on trying to bring these fragmented systems together,” said Dr. Seixas, who is also the associate director of the Center for Translational Sleep and Circadian Sciences. “If we are really serious about tackling the health disparities in Florida that make us unique—like the high prevalence of cardiovascular disease and dementia—we need to build a larger network to really unravel those issues.”

On the AIM-AHEAD infrastructure team, Drs. Tsinoremas, Seixas, and Yesha are also working with the National Alliance for Disparities in Public Health, Harvard University, and Vanderbilt University. Currently, University leaders are working closely with four institutions in Florida to create pilot programs, and recently received additional funding to support two of these pilot programs.

Additional Grant

The team is now part of an additional $500,000 grant to work with Florida Atlantic University’s (FAU) Schmidt College of Medicine and the Caridad Center, Inc.—the largest free clinic in Florida for uninsured and underserved children and families of Palm Beach County—to improve their electronic health records, so that they can be used more often for research.

“In terms of technology infrastructure, security, cloud computing, and supercomputing, we have some of the best faculty and staff working at the University of Miami. So, we are able to do more as far as research productivity and output. But if we look just up the road to many of our minority serving institutions, they don’t have those infrastructures,” said Dr. Seixas. “We are trying to spread our resources around, so these institutions that are often under-resourced can also do cutting-edge research and provide top of the line care, too.”

As a first step, leaders from IDSC are sharing a tool to collect and de-identify patient electronic health records for research that they developed a few years ago for the Miller School. The tool, called University Research Informatics Data Environment, or URIDE, is a web-based platform that collects, sorts, and helps researchers visualize this type of patient data from multiple clinical health systems. This could help health care professionals explore demographics, diagnoses, procedures, vital signs, medications, labs, allergies, co-morbidities, and other information for certain patient populations to pinpoint trends or optimal treatment practices. URIDE is co-funded by the Miami Clinical and Translational Science Institute.

A second award for $362,000 will help IDSC leaders create another pilot program with Florida Memorial University (FMU) and Miami-Dade College to train 40 existing faculty members and students to use artificial intelligence and machine-learning techniques in their clinical practice, research, and curriculum. In this program, IDSC leaders will also build upon another program they created at the University to attract and foster the careers of underrepresented minorities in science and especially in the burgeoning field of data science.

“We are very excited to collaborate with FAU, Caridad, FMU, and Miami-Dade College to expand the URIDE platform and to become an important piece of the infrastructure that supports the entire AIM-AHEAD consortium,” Dr. Tsinoremas said. “We also want to establish an open-source community around this platform to engage data scientists, data engineers, and developers for continuous improvement of this open-source effort, which we call Hi-RiDE—for Health Informatics Research Integrated Data Environment.”

Tags: artificial intelligence, Department of Psychiatry, Department of Psychiatry and Behavioral Sciences, Dr. Azizi Seixas, Dr. Nick Tsinoremas, health disparities, Miller School of Medicine, National Alliance for Disparities in Public Health, National Institutes of Health, University of Miami Institute for Data Science and Computing