When it comes to augmented intelligence (AI)—often referred to as artificial intelligence—physicians say they want to give and provide feedback, be assured of privacy and security, have AI-enabled tools that fit into their workflow, and they want adequate training so they can use these tools effectively.
Ochsner Health in New Orleans provides its physicians with all of these.
The AMA surveyed almost 1,200 physicians (PDF) last fall and found that:
- 88% of physicians say there should be a dedicated channel for feedback should issues arise.
- 85% want privacy assurances.
- 84% want to get proper training.
- 84% want AI to be integrated into the EHR workflow.
Ochsner Health is part of the AMA Health System Member Program, which provides enterprise solutions to equip leadership, physicians and care teams with resources to help drive the future of medicine.
AI projects get a clinical champion
“This is always about what problem are we solving and how do we know if we've solved it?” said Denise Basow, MD, Ochsner’s first chief digital officer.
“We don't have enough resources to just do cool stuff for the sake of doing cool stuff—and those types of projects tend to fail anyway,” Dr. Basow added. “We take the mindset of, we're not looking to deploy AI, we're looking to solve problems—and it just so happens that there's a lot more AI now to solve those problems than there was 18 months ago.”
She noted that Ochsner Health considers virtually every AI project to be in what they call the “big project category.” Such projects get a dedicated clinical champion who is accountable for driving at least some of the initial feedback.
“We also have a pretty large team that we call Epic Academy,” she added. “That's about 70 people in our organization who are accountable for training and receiving feedback on a lot of things that we deploy at Ochsner, and so they really sit shoulder to shoulder with our physicians and answer questions about how things work and get their feedback.”
The AMA is committed to ensuring that AI can meet its full potential to advance clinical care and improve clinician well-being. As the number of AI-enabled health care tools continue to grow, it is critical they are designed, developed and deployed in a manner that is ethical, equitable and responsible. The use of AI in health care must be transparent to both physicians and patients.
From AI implementation to EHR adoption and usability, the AMA is fighting to make technology work for physicians, ensuring that it is an asset to doctors—not a burden.
Securing the future
Regarding privacy and security, Dr. Basow noted that while AI technologies are developing at a rapid pace, Ochsner Health continues to “lean in and modify” its security and trust framework.
Ochsner Health’s data scientists are accountable for ensuring that any AI models that the health system deploys “are doing what they're supposed to do and for ongoing maintenance of those,” Dr. Basow said.
The system has had a data governance committee in place for a while now, and about two years ago it also developed a separate AI steering committee that focuses on issues that are more unique to AI and that go beyond just routine data governance.
The committees provide “another set of eyes” and include physicians and nonphysician clinicians, lawyers, technology operators, compliance officers and others with a variety of perspectives. They then make decisions about whether whatever potential gains a new AI application offers is worth the risk that it may pose.
“We always have really cool ideas that we want to do, and this is just an extra check to make sure that our reason for doing it—and the outcomes we're going to drive are worth the risk,” Dr. Basow said.
“It doesn't matter how good you are at security, at testing vendors—all of these things have risk,” she added. “It's never about ‘Can we make it zero risk?’ Because nothing is, and we wouldn't do anything if that was the goal.”
AI technology holds the promise to radically transform health care for both physicians and patients. For AI to meet its potential to improve care delivery and health, the AMA has called for a whole-of-government oversight approach that engages the physician community to ensure necessary safeguards and protections are in place.
The AMA STEPS Forward® “Governance for Augmented Intelligence” toolkit, developed in collaboration with Manatt Health, offers a comprehensive eight-step guide for physicians and health care organizations to establish a governance framework to implement, manage and scale AI solutions.
All-of-the-above training
Up until recently, most of Ochsner Health’s physician AI training has been voluntary and offered both online and in person with special sessions for those who have struggled.
But as clinical use cases become more complex, more training sessions have become mandatory.
AI-enabled chart summarization is one example of this that Dr. Basow cited.
“It's really important for the physicians to understand what the AI is doing well and where the potential risks and pitfalls are,” she said.
“We're developing mandatory courses, and they can't begin to utilize those tools until they've taken a course on it,” Dr. Basow added. “That's been the minority right now, but that number will grow.”
Ochsner not following this trend
The physicians responding to the AMA survey said that they thought their influence in their organization’s tech decisions was declining.
This was especially true compared to the 2023 survey, and was revealed in physician responses to the following statements in the AMA survey:
- Only 17% agreed with the statement “I have all or most of the influence,” compared with 20% in 2023.
- Only 18% agreed with the statement “I share influence equally with others,” compared with 28% in 2023.
- Agreement with the statement “My opinion is considered but I do not have any direct influence” grew to 36% from 32% in 2023.
- Agreement with the statement “I have no influence” grew to 29% from 20%.
At Ochsner Health, physicians do offer input on technology purchasing and integration, Dr. Basow said, but sometimes they have to buttress that input with research.
“The problem that we have—and I suspect is the challenge for a lot of health systems—is there's a lot more demand for things to be implemented than there is capacity to do the work,” she explained. “So where we've really spent some time over the last year is in developing a pretty robust process for submission of requests, and how we process those requests.”
If a physician requests the purchase of a digital health tool, they have to state its purpose, why it is needed, what outcomes it may drive and take accountability for those things happening.
“So I don't know that our doctors would say they have less input, but I think they would say that the prioritization process is more rigorous,” Dr. Basow said. “On the flip side, I think we've also realized that we can't just deploy solutions that we don't have physician sponsorship for … and once we decide to do something, we're going to be their partners in it.”
Physicians are 100% involved in developing the steps taken to ensure new technology enhances workflow instead of disrupting it, Dr. Basow said, noting that “we are never developing these solutions or these workflows without our clinicians’ involvement.”
“That's been true even before everybody was crazy about AI,” she added.
“With most products, it's pretty clear what the end users are supposed to do,” Dr. Basow said. “With AI, it's often much trickier. When can they trust it? When do they need to intervene? When do they need to verify?”
Encouragement for physicians to use AI-powered tools comes from the very top of Ochsner Health leadership.
When the system’s CEO Pete November took the helm in September 2022, he asked Dr. Basow what she needed to be successful. Her response was to make being a leader in digital health one of the organization’s strategic priorities—which he did.
“Setting that strategy from a leadership perspective is really important,” Dr. Basow said.
One way Ochsner Health has used AI to make the lives of physicians simpler and better is to help them clear their EHR inboxes.
“We avoided almost 700,000 messages going to our doctors last year—so just a pretty enormous impact,” Dr. Basow said.
The combination of inbox management and using ambient AI listening to aid in documentation combined to reduce physician “pajama time” by 15% to 20% in 2024.
“For physicians who were spending more than 90 minutes in the EHR after clinic hours, we saved them on average 22 minutes a day,” Dr. Basow explained, adding that those uses of AI were “absolutely” reducing physician burnout.
She said patients appreciate how the ambient-listening tool allows their physicians to focus on them and not the computer.
“They definitely notice ambient listening—we get a lot of feedback on that, and the patients love it,” Dr. Basow said. “My favorite quote on that is, ‘I got my doctor back.’ They definitely notice.”
Learn more with the AMA about the emerging landscape of health care AI. Also, explore how to apply AI to transform health care with the “AMA ChangeMedEd® Artificial Intelligence in Health Care Series.”