The phenomenon of artificial intelligence (AI), which encompasses machine learning, has been the subject of tremendously hyped speculation and commentary for several years now in healthcare.
But ambient voice, the hands-free AI-driven technology already being used by some physicians—both primary care physicians and a range of specialists—to record and document patient visits—is a reality in the present moment. And ambient voice’s success so far has been encouraging, and provided a counterpoint to what has in some cases been overdone hype in other AI-related areas.
On the broader level, “When we look at the realm of healthcare right now, there’s a lot of excitement around AI, in terms of what it could be and how it could improve care quality while also driving financial outcomes; but not a lot of organizations have dipped their toes in yet,” says Ryan Pretnik, director of strategy and research, at the Orem, Utah-based KLAS Research. “Why is that? Number one,” he says, “because the knowledge is lacking in healthcare organizations. They don’t know the use cases yet.” And, beyond that, there continue to be widespread struggles with data formats (spread across claims, electronic health record, structured, and unstructured realms), and an inability to corral and normalize data. In addition, the lack of highly skilled data scientists in patient care organizations is holding back progress, he emphasizes. For now, most of the use cases are around operational issues, with, perhaps surprisingly, the lowest percentage of use cases, a recent KLAS survey found, in the area of analytics for reimbursement revenue cycle, and waste/cost/fraud avoidance.
KLAS’s recent study of this area “confirms that we are very early on in this journey,” says Tim Zoph, client executive and strategist at the Naperville, Ill.-based Impact Advisors. Some of the most targeted use cases, Zoph, a former CIO, notes, have “focused on the sweet spot of clinical, for which there’s a direct impact on financial” health of patient care organizations, with sepsis and readmissions having been some “narrow-banded and fairly pragmatic applications of this. I would say, too, that it’s going into analytics teams that are already working on this but need a better tool set and data for its effectiveness, but culturally, AI is still a departmental function and not an enterprise strategy” being yet another systemic challenge across U.S. healthcare.
The KLAS report to which Pretnik and Zoph refer revealed the explication of widely held misconceptions, including the idea that “Building the models is the most time-consuming AI task.” Indeed, the report noted, “Don’t underestimate the time and effort it will take to prepare the data needed to test and build the models. Healthcare data is hard to clean and comes from many sources, and your organization may not have the expertise to feed the right variables or features into your models. Vendors and tools can help, but you need to do your own evaluation of the time and effort required to be successful with your models.”
In the broader context of AI implementation, some physicians are very happy with their successful use of ambient voice applications, through a few different vendors, One of those clinicians is William Silver, M.D., medical director of the Triangle Region of the Durham, N.C.-based EmergeOrtho, a 100-plus-orthopedist orthopedics group that serves patients in 21 counties across North Carolina, from 49 offices. Silver, who has practiced orthopedics for 17 years, is the official medical director of one of the organization’s regions, and the de facto medical director of the entire organization.
Silver and his fellow orthopedists have been using products from the Burlington, Mass.-based Nuance, including Dragon Speaking Naturally, for years; but moving to the Nuance Dragon Ambient eXperience (DAX) solution, he notes, means that “I literally don’t have to interact with the computer at all; I start the summary, I talk to the patient just as I did back in the days when I had to dictate; but I’m not interacting with the computer.” And that, he says, is where the future lies—with unencumbering the physicians—practicing in all specialties, and in primary care—in order to remove from them the tasks (including keyboarding for physician documentation) that are continuing to cause physician burnout.
Going forward, what should patient care organization leaders expect to happen in the next few years? “Obviously, AI is at the peak of its hype cycle, so no matter what it delivers, it will not meet expectations, because expectations are so inappropriately high,” says Christopher A. Longhurst, M.D., CIO and associate CMO at the 799-bed UCSD Health integrated health system in San Diego. “But where it will deliver the most value in the near term, will be around imaging. We’ve got a pipeline in place here” to continue to advance AI-facilitated clinical decision support at USCD Health, he says. “I don’t think it will replace radiologists; I think it will make them more efficient. Take mammography; if AI can screen out some of the more redundant films, it will help them. I think we’ll see it in pathology and ophthalmology, as for retinal screenings; and in gastroenterology, for colorectal screenings. I think we’ll see some AI in operations, whether it’s in scheduling or to help improve claims, etc.” Ironically, Longhurst says, “I think a place where there’s a lot of excitement, but the least demonstrated benefit is in the EHR.”
All those interviewed for this article agree: AI will evolve forward significantly in the coming years—but less rapidly than has been predicted until recently, with bright spots, including ambient voice, in which pioneers are leading the way.
