A.I. takes center stage at HIMSS as method to reduce paperwork

Debates over artificial intelligence and its role in health care took center stage on the HIMSS Global Health Conference in Chicago this week, where greater than 35,000 physicians, other health-care employees, executives and engineers convened to debate the newest advancements in health and technology. 

Corporations equivalent to Microsoft, Google and Amazon prominently advertised recent health applications for AI on booths across a sprawling exhibition floor, and panels of experts answered questions on how the technology might be used to handle industrywide challenges equivalent to staffing shortages and physician burnout.

Many health-care organizations and corporations have been using AI in various capacities for years, but a subset referred to as generative AI exploded into public consciousness late last 12 months when Microsoft-backed OpenAI launched its viral recent chatbot called ChatGPT. Generative AI refers to programs that may use fairly complicated prompts from end users to generate text or images.

Just as generative AI has captured the eye of most people, it has also captivated the medical community.

AI was the main focus of the HIMSS conference’s opening keynote, and HIMSS CEO Hal Wolf prefaced the discussion by revealing that he had asked ChatGPT how one can solve global health-care challenges. The Healthcare Information and Management Systems Society, or HIMSS, holds the conference every year. 

Wolf posed the query to ChatGPT in jest, but David Rhew, global chief medical officer at Microsoft, told CNBC in an interview that generative AI could really be “transformative” for solving big problems within the health-care industry. 

“The chance to use these large language models and the factitious intelligence in clinical workflows is tremendous, and we now have to do it responsibly,” he said. 

For Rhew, which means starting with “high-impact, low-risk” uses for the technology, equivalent to streamlining administrative tasks.

Developing diagnostic or directly patient-facing generative AI applications are higher risk since they pose significant regulatory questions for corporations, academics and federal agencies equivalent to the Food and Drug Administration to work through. Rhew said to think about AI as if the health-care industry has just been introduced to a automotive, while not one of the stop signs, traffic lights or roads have been created yet.

“We still must determine how one can do that together,” he said. 

But within the meantime, administrative or “back office” tasks require less regulatory oversight, and there may be an actual need for efficient solutions, since clerical work is usually burdensome for clinicians.

A study funded by the American Medical Association in 2016 found that for each hour a physician spent with a patient, they spent a further two hours on administrative work. The study said that physicians also are likely to spend a further one to 2 hours doing clerical work outside of working hours. 

Similarly, in 2017, the Journal of the Association of American Medical Colleges published a survey where respondents said around 24% of their working hours are spent on administrative tasks. Greater than two-thirds of the physicians surveyed reported that administrative responsibilities “negatively affect their ability to deliver high-quality care.”  

HIMSS attendees told CNBC they consider generative AI may help with these tasks.

On Monday, Microsoft announced an expanded partnership with Epic Systems, a health-care software company that helps hospitals and other health systems store, share and access electronic health records. Greater than 160 million people use Epic’s MyChart software, which provides patients with direct access to their health information and care team. 

Epic’s first application of the AI technology routinely generates draft responses to the messages that physicians receive from patients through MyChart. The physicians haven’t got to make use of the suggested draft in any respect, but it surely saves them time in the event that they decide to edit or send it. 

Seth Hain, senior vice chairman of R&D at Epic, told CNBC in an interview that AI could function an impactful hypothesis generation tool for physicians in the longer term. He said they may give you the chance to ask patient-specific questions equivalent to: What do you’re thinking that I should have a look at next in regard to this problem?

Peter Lee, corporate vice chairman of research and incubations at Microsoft, told CNBC that an early have a look at Epic’s AI developments brought tears to his eyes. 

“It just blew me away,” he said. 

Microsoft’s speech recognition subsidiary Nuance Communications also announced a clinical notes application called DAX Express ahead of HIMSS in March. DAX Express goals to assist reduce clinicians’ administrative burdens by routinely drafting a clinical note inside seconds after a patient visit. 

In a live demo at HIMSS, Nuance previewed future projects and showcased DAX Express’ capabilities, which were met with gasps and joyful exclamations from among the physicians, nurses and health-care employees within the room. 

Other corporations are also working to make use of generative AI to cut back administrative burdens.

Amazon Web Services on Monday announced an expanded partnership with Philips, a Netherlands-based health technology company. AWS has already been supporting a lot of Philips’ existing cloud-based and AI initiatives, equivalent to those who help radiologists analyze scans and medical images more quickly — even from their homes.  

But Monday’s announcement means Philips may also use AWS’ generative AI technology to simplify its clinical workflows and advance its imaging capabilities even further. 

“What’s most enjoyable is the undeniable fact that we’re approaching a precipice where we now have this tipping point, where we make the precise thing the straightforward thing,” Shez Partovi, Philips’ chief innovation and strategy officer, told CNBC in an interview. “And straight away, in most technology, the precise thing is loads of clicks away.” 

Partovi said all of the small tasks that physicians have to finish are like “death by 1,000 cuts,” so using AI to tease out administrative challenges could make an actual impact on the standard of physicians’ lives.

On Tuesday, 3M Health Information Systems also announced that additionally it is working with Amazon Web Services’ machine learning and generative AI to assist reduce physicians’ administrative workload. 3M HIS supports a conversational AI platform utilized by greater than 300,000 physicians, and the corporate said in a release that the AWS technology will make it easier for doctors to automate and complete accurate clinical notes within the electronic health record. 

Similarly, Google Cloud announced a Claims Acceleration Suite last week that uses AI to streamline medical health insurance claims processing and prior authorization.

In response to the Centers for Medicare & Medicaid Services, the present prior authorization process takes a mean of 10 days. Google’s AI will help alleviate a few of that administrative burden for providers by converting the unstructured data that appears in images, PDFs or other health records right into a more easily digestible, structured format. 

“They really require a human being to go in there and to take that data and rekey it into the system for review,” Amy Waldron, director of worldwide health plans strategy and solutions at Google Cloud, said during a media briefing with reporters at HIMSS. “Which, to me, makes absolutely no sense provided that someone has to take time to place all that wealthy data there, and we now have AI that may unlock that value.”

Generative AI has “tremendous” potential to enhance administrative efficiency in health care, said Microsoft’s Rhew. But as health-care and technology corporations proceed to make more sophisticated advancements, industry leaders, regulators and academics in the neighborhood may have to make sure that generative AI is equitable and doesn’t cause harm to communities. 

The technology is vulnerable to bias and discrimination whether it is trained on health-care data that doesn’t properly represent a patient population, which could ultimately result in inadequate decision-making or treatment plans. 

In consequence, Rhew said, there may be a collective responsibility to determine how one can deploy AI with care. 

“It’s a transformative technology,” he said, “but we now have to determine how one can do it responsibly.”