As many health systems and organizations work on advancing value-based care to prioritize patient outcomes and cost efficiency, a new tool has emerged as a partner: augmented intelligence (AI), often called artificial intelligence. The integration of AI within value-based care aims to enhance diagnostic accuracy, optimize treatment plans and streamline operations, ultimately improving the quality of care while reducing expenses. But despite its potential, the adoption of AI within health care faces some hurdles that can delay use within value-based care.
Baptist Health Medical Group has been among the top quartile in the country for quality metrics and value-based care. For example, the quality metrics include breast and colon cancer screening, advanced care planning, diabetes A1c testing and control. Each metric aligns with the health system’s value-based care partners.
While it is important to align value-based care initiatives with other health system efforts, “where you can sometimes struggle is if you’ve got value-based care initiatives and you’re going to try to treat certain patients differently because they’re in an ACO [accountable care organization] or something,” said Brett Oliver, MD. He is a family physician and chief medical information office for Baptist Health Medical Group, which is a member of the AMA Health System Program that provides enterprise solutions to equip leadership, physicians and care teams with resources to help drive the future of medicine.
“And—at least from a primary care physician’s perspective—that doesn’t work. If it’s a high-quality, most efficient way to treat patients, that’s what we should do for everybody,” said Dr. Oliver. “We’ve had an aggressive primary-c1are push in terms of recruitment as well as just developing what we would term ‘right place, right time, right cost’ approach to access and care. With that has come some technical pieces such as 24/7 on-demand video visits as well as asynchronous e-visits.”
That includes embracing AI and finding ways to use this technology to improve value-based care offerings while reducing administrative burden for physicians at Baptist Health.
Dr. Oliver spoke at an AMA Insight Network Meeting highlighting how Baptist Health uses AI to improve value-based care and how to overcome challenges to adoption.
Among other benefits, members of the AMA Health System Program have access to the AMA Insight Network’s Quality, Safety and Equity community. This virtual forum provides an opportunity for like-minded leaders from across the country to hear more examples of how leading systems are finding innovative ways to address AI and value-based care in their communities.
Get buy-in from physicians
One of the bigger challenges of implementing health AI is the buy-in from physicians and other health professionals. Physicians are concerned about how it will affect patient care and whether they are at risk when using—or not using—health AI.
There are a lot of important and valid questions circulating about using AI, especially when it comes to value-based care.
The AMA has developed new advocacy principles that build on current AI policy. These new principles (PDF) address the development, deployment and use of health care AI, with particular emphasis on:
- Health care AI oversight.
- When and what to disclose to advance AI transparency.
- Generative AI policies and governance.
- Physician liability for use of AI-enabled technologies.
- AI data privacy and cybersecurity.
- Payer use of AI and automated decision-making systems.
“Being transparent is critical in anything that you do,” Dr. Oliver said. That means “being transparent about this model, this algorithm, how much it can help.”
Let the data do the talking, he said. That is a lesson learned with earlier innovations, such as remote patient monitoring (RPM).
“We’ve had this product for five years now—prior to COVID—and I did a lot of, look what I’ve got, do you want to use it? Look at this. This is really pretty. I did the bright-and-shiny tour,” Dr. Oliver said. “What I needed was data from our own organization.”
“That first clinician that got that vision was our medical director for our heart-failure clinics,” Dr. Oliver said, noting “they had a really good readmission rate of about 18% at 30 days for these high-risk patients with heart failure prior to RPM. After the continuous RPM addition, the readmission rate fell to zero over 90 days. That got everyone’s attention.”
Learn more with the AMA about the emerging landscape of augmented intelligence in health care (PDF).
Integrate AI into existing workflows
For organizations that integrate AI “somewhere outside of the normal workflows, it’s not going to work,” Dr. Oliver said. “I’ve had plenty of meetings with startups or someone with a new application that’s got a lot of potential and they walk me through their workflow. Often, they mention, ‘you just click here, and it takes you to a new screen outside of the original workflow.”
In such a circumstance, the physician’s likely response will be: “You lost me. We’re done,” he noted. The new workflow simply will not succeed “if you understand clinicians’ workflow, the burden that they have, the busyness,” he said, emphasizing “it’s really important to integrate these things into existing workflows.”
For example, Baptist Health’s value-based care efforts initially involved a dashboard that was external to the “EHR and there were all kinds of incentives to try to get us to log into these dashboards and it just didn’t work. It’s out of your workflow,” Dr. Oliver said. “Then it’s not actionable. It’s not part of the EHR. So, I’m looking at this, I’m seeing patients who have a care gap or something along those lines of how I’m doing blood pressure control, but then I can’t do anything with it.”
That is why it is important to get AI “integrated into your EHR, making it easy to access, making it actionable,” he said. “These new technologies and efficiencies need to be in these workflows.”
Address the cost
“Costs have to be talked about. Same thing goes with these ambient [AI] solutions. There are a lot of great things out there, and competition has been healthy and it’s probably helped drive down the price, but there’s still a cost,” Dr. Oliver said. But sometimes it’s hard to find a return on investment.
For example, using AI to draft messages to patients. AI drafts first responses to patients and the physician or team member reviews before sending. To date, there has been a roughly 20% acceptance rate for those messages without editing. While that is low, what Baptist Health has seen is physicians are keeping the empathetic opening that AI provides. The physician is just updating whether a patient needs to come in for a visit and other details of the message.
There is a cost to this, “so as we’re piloting this and figuring out how we can utilize this, one of the conversations is going to be: Who is able to use this? Can we just turn it on for everybody or do we have to be selective with that?” Dr. Oliver said. “I don’t know the answer and that’s probably an individual organization answer as well, but costs have to be talked about.”
Learn with the AMA about ways to improve value-based care data sharing and advance value-based care with alternative payment models in Medicare.
