Government-funded program wants better AI for managing heart failure in the US
The U.S. government's Advanced Research Projects Agency for Health (ARPA-H) recently launched a competition to build patient-facing artificial intelligence (AI), a supervisory agentic AI and a scalable deployment plan. The agency's stated goal is to bring 24-7 cardiac care to every American, even those living in U.S. counties without any cardiologists.
This is part of the Trump administration's wider initiative to push for use of AI in healthcare to try and tackle known cost drivers.
The teams receiving contract awards from its Agentic AI-EnableD CardioVascular CAre TransfOrmation (ADVOCATE) program include Kaiser Permanente, Duke University, Stanford University, Atman Health, Updoc and Tempus AI.
The agency commitment is up to $33.7 million in the first year of the four-year, $62.7 million program. If the technologies can be developed, the agency said the ADVOCATE program will save lives and reduce preventable hospitalizations.
The agency projects this application of AI could generate $28 billion in annual cost savings just in heart failure care alone. Heart failure care and hospitalizations are currently one of the largest drivers in Medicare costs each year and the government wants to see if AI technology can help improve care to keep these patients out of hospitals.
The U.S. spends nearly $500 billion on heart disease annually and still experiences far worse cardiovascular outcomes than other western nations, the agency said. In addition, cardiovascular disease is the leading cause of mortality in the U.S.with more than 200,000 deaths each year. ARPA-H said the root cause is a flawed healthcare model where where in-person cardiology specialist visits are limited or access is difficult in rural areas, and physicians are being asked to manage a chronic condition based on a couple patient visits per year, resulting in missed health status changes, delayed medication titration, avoidable hospitalizations, poor medication adherence, and workforce burnout that hits hardest in rural and underserved communities.
“Every day, Americans are dying from cardiovascular disease that we know how to prevent and treat. Often the barrier isn’t the medicine; it’s access to a clinician who can guide the care,” ADVOCATE Program Manager Haider Warraich, MD, who is also a heart failure cardiologist at Boston Medical Center, said in a statement. “What we are trying to achieve with ADVOCATE is a technology that can serve as a clinician-extender: an autonomous agent smart enough to understand a patient’s treatment needs, capable of providing certain care autonomously, and equally capable of engaging the clinical team when needed. This work will change how we practice medicine—and it would not happen without ARPA-H driving the vision for the future.”
The ADVOCATE program will work with the U.S. Food and Drug Administration (FDA) to develop a regulatory framework for this new class of patient-facing clinical AI. This may include creation of evaluation standards, interoperability requirements and help with reimbursement pathways.
“Patient-facing agentic AI is one of the most consequential frontiers in digital health. Getting the regulatory framework right requires close, iterative collaboration between developers, clinicians, and the FDA. said Rick Abramson, MD, director of the FDA Digital Health Center of Excellence. "ARPA-H’s ADVOCATE program is the kind of ambitious effort that will be critical as we think about continuous ongoing monitoring of autonomous clinical AI for both safety and effectiveness."
Teams involved in the ADVOCATE program
ARPA-H named three teams that will work on various aspects of the program.
The Patient-Facing Clinical AI Agent team (TA1) will develop an AI agentic system capable of autonomously supporting patients with heart failure between regular doctor visits. The AI should be able to escalate issues to the human care team and act as an autonomous member of the clinical team.
Atman Health's team will develop a clinical AI agentic system the will enable large language models (LLMs) with a voice interface that allows the AI to listen as patients talk. The system will be able to adjusts on the fly to ask each patient relevant diagnostic questions most likely to be useful.
Tempus AI's team will develop a clinical AI agentic system that extends the company's existing Olivia patient health app. It will add continuous monitoring with deep clinical analysis to look for changes in a patient’s health.
Updoc's team is developing a clinical AI agentic system with conversational intelligence based on clinician-built rules to validate every proposed action against approved protocols before execution.
The TA1 team is required to submit a first-of-its-kind FDA-authorization package for this product within the next 24 months. All three of the TA1 teams will work together to develop the necessary technical interfaces, safety benchmarks and evaluation protocols.
The Supervisory AI Agent team (TA2) will build a disease-agnostic supervisory AI system to continuously monitor the TA1 clinical agents after successful deployment. This includes detecting unsafe recommendations and out-of-distribution behavior in real time. The goal is to provide a continuous quality assurance layer that the FDA, health systems, patients and clinicians can trust.
The Stanford University team will develop supervisory AI using a three-stage, computer-escalating pipeline that uses outlier filtering, rule-based screening and a deep-research auditing agent.
The third group will work on Scalable Implementation in Health Systems (TA3). These teams based at major health systems will create the real-world implementation strategy that brings ADVOCATE technology to patients and clinicians across various care settings. They will have a specific focus on rural and underserved communities.
The Duke University team will create a multisite validation platform to test ADVOCATE agents across five health systems and rural sites. This will include operations on both Epic and Cerner/Oracle electronic health records.
Kaiser Permanente's team will provide an enterprise-scale deployment across 21 medical centers and more than 260 clinics. They will embed ADVOCATE AI agents into Epic workflows for heart failure patients. They also will create reusable deployment blueprints via shadow-mode deployments and development of randomized clinical trials.
ARPA-H recently announced funding to create autonomous AI combined with robotics to perform emergency thrombectomy procedures in ischemic stroke patients.
Dave Fornell has covered healthcare for more than 17 years, with a focus in cardiology and radiology. Fornell is a 5-time winner of a Jesse H. Neal Award, the most prestigious editorial honors in the field of specialized journalism. The wins included best technical content, best use of social media and best COVID-19 coverage. Fornell was also a three-time Neal finalist for best range of work by a single author. He produces more than 100 editorial videos each year, most of them interviews with key opinion leaders in medicine. He also writes technical articles, covers key trends, conducts video hospital site visits, and is very involved with social media. E-mail: [email protected]