The American Medical Association (AMA) has announced the recipients of $12 million in grant funding aimed at accelerating precision education.
In all, 11 teams representing more than 80 institutions secured funding from the Transforming Lifelong Learning Through Precision Education Grant Program. Each team will get $1.1 million over four years, with grantees spanning med schools, residency programs, health systems and specialty societies.
This is the third iteration of grant funding from the AMA, and the grants span the continuum of learning, from med students to residents to practicing physicians. The grant teams are:
● Georgia Academy of Family Physicians
● Louisiana State University Health Sciences Center
● Meritus School of Osteopathic Medicine
● Mount Sinai Morningside/West
● Perelman School of Medicine at the University of Pennsylvania
● Stanford University
● University of Cincinnati College of Medicine
● University of Hawaii - John A. Burns School of Medicine
● University of Illinois College of Medicine
● University of Michigan
● University of Wisconsin School of Medicine and Public Health
Several projects will use artificial-intelligence-driven tools to help providers strengthen their communication skills, clinical reasoning and ability to respond to patients’ needs in real time. Other projects will implement mobile sensor technologies to track mastery of clinical skills, build tools that help trainees transition smoothly into practice and enhance coaching and feedback models.
Precision education, as defined by the AMA, uses data and technology to tailor learning to each individual. The idea is to deliver the right education at the right time. For instance, AI can be used to identify a provider that hasn’t seen a certain kind of case in a while and to intervene with that example sooner. This type of targeting is enabled not only by AI but by the massive amounts of data the AMA collects on learners.
Supporting innovation is a major component of the AMA’s work, according to Sanjay Desai, M.D., AMA chief academic officer, and the current model of education is still burdened by “significant inefficiencies.” Using data and AI can “really transform the way we educate physicians,” Desai told Fierce Healthcare.
“The opportunity is to leverage this data that exists, but in the current systems is unable to be aggregated and analyzed in a way that’s effective for learning,” Desai said. “This is what we consider to be the future of lifelong learning.”
The grantees will participate in a learning collaborative to share best practices around implementation. The AMA’s goal is to create interoperability standards now to reduce barriers when projects get repurposed and scaled to other organizations.
The AMA also has a national advisory panel spanning experts in tech and education to think through responsible use, governance, ethics, privacy and technical questions related to AI. The trade group is also using MedBiquitous, a program designed by the Association of American Medical Colleges, that promotes interoperability standards in healthcare education.
The grants align with the mission of the AMA’s newly launched Center for Digital Health and AI, which aims to ensure doctors participate in how relevant tools are used in clinical settings. The center focuses on integrating tech into daily workflows, expanding provider training and encouraging collaboration between the healthcare and tech sectors.
Stanford University, one of the grantees, plans to implement a scalable and sensor-driven program to objectively assess surgical skills. The idea is to shorten learning curves through personalized feedback and targeted training recommendations.
Carla Pugh, M.D., Ph.D., a surgeon scientist and professor at Stanford, is the principal investigator on the project. Pugh’s research involves the use of sensors to revolutionize how competency and mastery in surgical skills get assessed. Pugh, who was the first surgeon in the U.S. to receive a Ph.D. in education, holds six patents on the use of sensor and data acquisition tech to measure hands-on clinical skills, with more than 200 medical and nursing schools using one of these tools.
Across several coaching pilots at Stanford, wearable EEG sensors and video cameras capture movements and audio in an operating room. The sensors can analyze brain waves that drive motor actions during surgery. These data are then used to create a highlight reel recapping what went well and not so well in a particular surgery. While some analysis is already possible with video and AI at other health systems, using biomarker data is novel and adds a new layer of detail, per Pugh.
“We think this is going to make a major headway from a quality perspective in how we share data, how we self assess, how we do peer-to-peer collaborations,” Pugh said.
The standard way of assessing surgical outcomes has historically been based on how a patient fares after a procedure, in terms of 30-day morbidity, mortality or readmissions rates. “It’s a very far distant assessment based on patient outcomes, which is different than procedural outcomes,” Pugh said.
Why not add procedural outcomes to the mix, like what types of stitches were used or a surgeon’s visual assessment of tissue? “Every expert has a routine way of doing certain things, but we’ve never looked at those different ways to see which might be better approaches than others,” Pugh said.
“There’s a lot of procedural and technical decisions that people make based on what we see and none of that gets documented,” Pugh added. “It’s all in the surgeon’s head.”
This system at Stanford is primarily used on practicing physicians rather than on trainees. “The goal is actually to model what the experts do,” Pugh explained. “Otherwise, it’s useless to put on trainees, because you don’t have a benchmark.” The data, once collected and digitized, can indicate competency versus mastery. The reels can be shown to peers and students for learning.
The AMA grant provides Stanford funding for its collaboration with the American Board of American Specialties with the goal of developing a national standard for using digital data and metrics to track medical and surgical skills. Stanford hopes to build a secure data warehouse of “digital performance biomarkers” that can be used for comparative feedback and performance benchmarking, per Pugh.
10 Newtons, a Stanford spinoff of which Pugh is founder and chief scientific officer, will eventually distribute the wearable tech and dashboard software analytics to specialty boards and academic medical centers.
Though it has several pilots that have shown positive results, Stanford has not yet conducted longitudinal studies to track long-term skills improvement with the biomarker data approach. The AMA grant will also support longer-term evaluation of these technologies.