Adopting Emerging AI Technology for Healthcare

AI technologies offer vast promise for healthcare, a field awash in more data than humans could hope to crunch without the help of advanced machines, and where the stakes for decision-making are dizzying. Already, AI-enabled tools are quietly transforming hospitals by improving such vital functions as scheduling and physician note-taking.

But AI and healthcare experts who attended the most recent program of the SMU Cox Executive Education Healthcare Initiative said AI in the healthcare industry comes with caveats. Although they’re thrilled to be on the forefront of a new wave of technology, they also want to learn its limits and potential drawbacks.

“AI is one of the largest technological leaps I’ve seen in my career,” says Chris Akeroyd, the executive vice president and chief information officer for Children’s Health, a North Texas healthcare system that serves more than 900,000 patients annually. “There’s some urgency not to fall behind. Now is the time to adopt, and being educated about how you adopt is important.

“A lot of us are still trying to figure this out. How do we bring it into our organizations? How do we oversee it? Govern it? That’s why this course was nice. It was well-suited for that.”

The SMU Cox AI For Healthcare Leaders Program

Since 2022, the SMU Cox Executive Education Leadership Initiative has brought together area industry leaders for conferences, roundtable discussions and Executive Education programs. Naturally, the explosion of AI and the rapid changes those new technologies have delivered (or at least promised) have become a topic of interest for executives and managers who hope to keep up with the curve. With consumer needs and patient care as their focus, healthcare organizations are grappling with topics such as leadership, innovation, equity, access, cost and AI—all topics covered in the Cox School’s healthcare programming.

The most recent event in April brought healthcare leaders to SMU’s campus to focus on strategies for implementing AI. The Executive Education program, titled AI for Healthcare Leaders, was developed with input from McKinsey & Company and Perkins Coie LLP in response to feedback from a fall 2023 conference where participants asked for hands-on AI training and real-world use cases.

Instructors led a program that applied theoretical knowledge to practical exercises, case studies and discussions with C-suite industry leaders. Participants left with a better understanding of AI concepts and techniques—and how to successfully apply those solutions in healthcare settings.

“I really encourage hospitals and executives to think about trust and being transparent,” says Vishal Ahuja, an SMU Cox associate professor of information technology and operations management who also serves as an adjunct faculty member at UT Southwestern Medical Center. His research focuses on developing decision analytic tools that can be easily implemented by healthcare professionals and policymakers to improve patient health, advance the quality of care and enhance the efficiency of care delivery.

“Stakeholder alignment and collaboration really are key, and incorporating the human perspective in decision-making is critical,” Ahuja says. “This also means educating patients on the output of AI—for example, explaining what it means when an algorithm tells us there’s a 65% chance of successful treatment.”

AI’s most immediate promise, Ahuja says, is that by outsourcing routine or administrative tasks, doctors and nurses will be free to “practice at the top of their license.” He references the work of Eric Topol, a renowned cardiologist at Scripps Research who predicts that AI’s greatest effects on healthcare will be to free up physicians’ time, enabling them to build stronger connections with patients.

The ramifications for patients’ medical outcomes—as well as for their customer satisfaction—are enormously important to hospitals.

Early in the program, Ahuja recalls, the opening roundtable offered a bold prediction: Businesses that use AI will eventually replace businesses that don’t. As ominous as that might sound, the question remains of what, exactly, the best uses are for AI and when to implement it. And that question is far more nuanced than simply which solutions and software to buy.

“The technology is important and has been around for a while,” Ahuja says. “But we’re often too focused on the tool itself. What’s important—and something I want to emphasize to business leaders—is identifying the ‘right’ problems before looking for the ‘right’ tools to solve them. This will prevent leaders from finding an elegant technological solution to the wrong problem.”

As the two-day program continued, the conversations aimed to demystify the technologies and centered on figuring out just when and how to use them. What specific value do they bring? And, most relevant for executives, what are the ethical considerations of putting a particular tech into practice?

Establishing AI’s True Potential

During the sessions, colleagues leaned on one another to help decide whether the problems they were trying to solve even required AI. With so many startups and vendors clamoring to make sales and establish themselves in the sector, the executives were keen to cut through the noise of the new, burgeoning marketplace.

“It was an aha moment,” Akeroyd says. “The technology, in some cases, is the easy part. You can turn things on, but are they getting the outcome you want? How do you measure that? What structures do you have in your organization to onboard and approve these technologies as you bring them in? They’re not a set-it-and-forget-it type of technology.”

Ahuja and Akeroyd agreed that AI solutions have immense potential to enhance the patient experience in hospitals and other healthcare settings, in ways both conspicuous and subtle. Executives, administrators and caregivers will find applications and solutions that make them more efficient, offering them the chance to improve patient outcomes if they harness the technology effectively.

As providers further adopt AI, those technologies have the potential to lower costs, improve scheduling, and allow patients and doctors to spend more time together developing trust.

Take, for example, the complex problem of scheduling operating rooms, the most valuable piece of real estate in a hospital. When a facility can reliably predict how long an operation will take, it can optimize the use of that expensive resource.

Similarly, AI can help predict the need for beds, so a hospital can allocate them efficiently. AI can also help schedule doctors and nurses using factors such as how many patients they can handle, the practitioners’ experience levels, their personal preferences and the likely needs of the hospital during a given shift.

“Appointments were traditionally scheduled with pen and paper — or, more recently, Excel,” Ahuja says. “The technology has advanced to include tools like AI, machine learning and predictive analytics.” AI tools are already helping hospitals schedule appointments, power chatbots and automate voice message systems. None of those is unique to medical settings, but they do help patients and their families handle the complexity of a hospital.

A more intriguing application of AI that goes straight to patient experience involves clinical documentation. Patients may pay little attention to how their care provider takes notes, and perhaps even less to how that provider revises, types up, disseminates and stores those notes after an appointment.

But doctors spend a great deal of time capturing and processing the information patients tell them. That accuracy and thoroughness of that labor-intensive undertaking has huge implications for patient outcomes, and it’s a job that machines are already helping to complete speedily and reliably, via ambient listening technology.

“A physician and patient can have a very natural conversation and not have to type out notes during a clinic visit,” Akeroyd says. “It adds time back to the physician or the provider in that room. And it gives a more human, connected touch when we visit the doctor.”

Along with technologies such as virtual nursing, which allows other caregivers to attend a meeting virtually, an ambient listening tool can help physicians focus more intently on what their patients are telling them, and to bring a second colleague along to consult with or to validate their findings.

Keeping AI For Healthcare Human

The term of art regarding AI in hospitals these days is “human in the loop.” A physician or clinician may be receiving a boost from an AI tool, but they’re making the final decision on a patient’s care. The technology accompanies those decisions as an advisor or as an assistant—not as a final arbiter.

“This goes back to being more present,” Akeroyd says, “and getting back to why a lot of our clinicians went into medicine.”

At Children’s Health, where Akeroyd works, another huge question centers on the value of the data that AI systems use to train their models. Most medical data comes from adults, after all. Medically, children aren’t simply miniature adults, nor are they a homogenous group. As clinicians treat kids, then, they need to be aware that AI models are likely to be less accurate—perhaps overconfident—in diagnosing illnesses and suggesting treatments.

As hospitals continue to adopt AI to help navigate these new oceans of patient data, the quality, fidelity and security of that data will continue to be priorities.

“HIPAA still applies, and we still have to keep up with our patient data and manage that well,” Akeroyd says. “You can’t just copy protected health information into the public version of ChatGPT and ask a question.”

Registration Now Open

The next 2024 SMU Cox Healthcare Conference, which will focus on improving the consumer experience, will take place Oct. 10–11 at the Cox School of Business. Confirmed speakers include Brian Tyler, the CEO of McKesson and executive co-chair of the SMU Cox Healthcare Initiative; Barclay Berdan, the CEO of Texas Health Resources and executive co-chair of the SMU Cox Healthcare Initiative; and Stephen Gillett, the CEO of Verily.

Registration for the October event is open now. Although the event is geared toward North Texas healthcare professionals at the director level or above, it’s open to anyone who wants to collaborate and network with like-minded business leaders and researchers. Future SMU Cox Executive Education Leadership Initiative conferences will explore topics such as innovation and disruption, health equity, clinical research and leadership.

Sponsorship opportunities are also available. If you’re interested in sponsorship opportunities, please contact Lisa Tran at [email protected].