There’s no denying that artificial intelligence is well on its way to changing the systems that keep the world running. In U.S. hospital systems, AI is particularly revolutionary, with some enlisting the technology to help lighten administrative burdens and improve diagnostic accuracy. Here’s what some Northern Virginia health care professionals have to say about the ways they’re using AI now — and where they might use it in the future.
Time-Saving Technologies
A recent Stanford Medicine study on physician burnout found that nearly half the doctors surveyed experienced at least one symptom of burnout. And with a shortage of physicians in the U.S. — that’s projected to reach about 187,000 doctors by 2037 — it’s no wonder that medical providers are stressed.
“The way that we look at this [issue] is, can we do one of two things: Can we unburden [health care professionals] by automating some portion [of their jobs], or can we make a thing that they do easier?” says Matt Kull, Inova Health System’s chief information and digital strategy officer, about physicians and providers experiencing burnout.
Kull says ambient technologies allow physicians to converse with patients while AI “does all of the charting in the background.” AI ambient listening applications can filter out inconsequential conversations and make sure that the summary, notes, and treatment plan are documented properly while the physician is focused on the patient.
“It makes the notes better, which allows future care to be better and the physicians [to be] unburdened,” Kull says. “Some of the feedback we get has been incredible from our providers, in that, ‘You’ve given me my life back.’”
Dr. Joe Evans, co-chair of Sentara’s AI Committee, also points to changes he’s seeing after the incorporation of ambient technologies in the ambulatory environment. Providers say it has “really changed their [lives]. … They can walk out of the office at the end of the day and all their notes are done, and they’re not having to stay up for two or three hours after they put their kids to bed at night to finish their work,” Evans says. “So, I think it’s had a huge impact in that regard.”
A Game Changer
Dr. Mary Chasko, chief medical information officer and anesthesiologist for Kaiser Permanente of the Mid-Atlantic States, likens ambient listening technology to “an electronic scribe” that can reduce busywork and make patient interactions more meaningful. Chasko says it’s “a total game changer when it comes to how physicians are performing their documentation. A lot of physicians beg us, ‘Please don’t take this away.’ And what we’re seeing over the past year is new physicians that are joining Kaiser Permanente have had experience with ambient listening. And so I expect that it will become the default preference for documentation for new physicians as they complete their training and are looking for medical groups to join.”
Evans even anticipates medical school potentially being shortened, which could be attractive to those contemplating entering the field. “When I think of a pipeline of physicians, I think AI in general has an incredible potential to change the way we do medical education. And really change and potentially even accelerate that process,” Evans says. “I think it may help by reducing potentially some of the training time on some of the basic science, and accelerate the clinical portion of the training while incorporating the utilization of AI in clinical practice.”
Improving Patient Outcomes
Because it can process massive amounts of medical data and spot patterns that a human might miss or recognize too late, AI also has the power to catch some critical conditions early.
For example, AI tools can scan radiology images and detect tiny anomalies that aren’t the primary focus of the radiologist’s review. In one case, Kull says, an AI tool picked up a tiny lung nodule on the chest X-ray of a patient presenting with chest discomfort. This allowed doctors to quickly investigate and remove a potentially cancerous growth before it progressed, dramatically improving the patient’s prognosis.
AI also analyzes patient data over time, identifying subtle changes in walking speed, heart rate, or lab results that may signal early disease. “If we can pick up on that, then we can get really proactive about intervening when you’re starting to have a problem, as opposed to after you have an acute cardiac failure,” Kull says.
Addressing Privacy Concerns
While AI is revolutionizing patient care, privacy and trust remain central concerns for both patients and providers. Hospital systems have processes in place to ensure patient information is protected and treated with care. “First and foremost, we want to keep patients’ information safe. We want to keep their data safe,” says Kull.
Sentara’s AI Oversight Committee was formed two years ago to assess potential risks and benefits of AI solutions, a vital undertaking in a field that’s advancing at breakneck speed. The committee asks questions like, “Are we doing this in an ethical way? Have we taken all approaches to remove bias?” says Evans.
Patients do have a role in the process, providers emphasize. Patients must be comfortable with how AI is used, and consent is built into every step. “We always ask for permission if we are going to be recording or using other technology to facilitate our interactions,” Chasko says. “We are going to be very conscientious about honoring and respecting the patient’s wishes.” Patients can ask for alternatives — like opting out of ambient listening — if the technology makes them uneasy.
Future Progress
The future of AI in health care promises to go far beyond early detection and clinical efficiency. The next wave will build on the foundation of ambient technologies.
AI‑driven “augmented response” tools are being tested to review patient messages, draft replies, and even initiate routine refills, reducing delays and administrative burdens. And voice bots are gaining sophistication. They are already able to route patient calls, reschedule appointments, and respond to routine requests. “It can shorten call times and save patients time,” says Kull, of the bot Inova has in place at the front end of its call center.
One thing is definite: AI is not going away.
“I predict that in probably three or four years, definitely within five years, we’ll see almost every aspect of clinical medicine having some sort of AI-enabled tool to not only assist us with diagnosis and care, but also to help us provide this information to the patients and to assist us with communication as well,” Chasko says. “I know I’m making it sound like it’s revolutionary, but it really is.”
Feature image, LALAKA/stock.adobe.com
This story originally ran in our August issue. For more stories like this, subscribe to Northern Virginia Magazine.
