Hospital for Special Surgery (HSS) has a catchphrase for AI strategy. “Stay in the game, win the game, change the game.”
Dr. Assisballad, Chief Digital and Transformation Director for the New York City-based system, says Ambient Scribe hasn't changed the game Newsweek July 18th. They will be needed to “stay in the game” and are expected from the clinical community.
To win the AI game of healthcare, health systems must use technology to automate and augment more than patient visits. As the competition for patient preferences recovers, there will be people who can completely revolutionize the care experience.
“What we're doing in today's core business needs to be fundamentally different in the future,” Balado said.
That would be a lot of work. Vendors and health systems leaders are hoping to face many regulatory, infrastructure and refund challenges as they surpass the rapid victory of previous iterations of generative AI, sources say Newsweek.
Newsweek Illustration / Getty
What does the next stage of healthcare AI look like?
Historically, healthcare leaders have had to make forward-looking decisions based on retrospective data. AI can be used for prediction and predictive modeling that allows health systems to adjust strategies in real time based on the “center brain,” Balado said.
Some health systems have already used predictive modeling to coordinate staff, reduce admission delays and deploy equipment. Around 300 hospitals around the world are using GE Healthcare's command center technology to improve operational efficiency.
“We don't rely on word of mouth, potent or systems that don't work well,” said Christie Baraz, president and COO of Duke Health Lake Norman Hospital in Mooresville, North Carolina. Newsweek The System Command Center Experience. “There is one source of truth.”
Innovaccer is an AI platform used in over 100 US healthcare systems, including seven of the 10 top-grossing healthcare systems, and used predictive modeling to identify expensive and avoidable patient readmissions and estimate readmission management tools. This solution includes AI agents that can actively reach out to patients at high risk of readmission, preventing them from losing contact with their health care team.
Abhinav Shashank, co-founder and CEO of Innovaccer, said Newsweek The company is deliberately avoiding “point solutions” (individual tools that address individual problems) in favor of maximizing functionality on a single platform.
“Everyone wants a platform that happens in a coordinated way that works together and coordinated with these AI capabilities. This happens in a coordinated way, not this accidental way that happens when you grow point solutions,” he said.
In fact, continuity is a big part of the next play in the health system. Varado and his team are in the midst of an 18-month focused effort at HSS, creating a “Lakehouse” architecture to house data in a single location.
This allows the system to separate data entry from data storage without disrupting employee workflows, Barad said. Through our partnership with HSS's Amazon Web Services and Databricks, we are working to focus data from early entry points, including initial entry points, grand data systems, for a central pool that can provide a more comprehensive view of system operations and patient journeys.
With all that data in one place, it's easier to deploy AI-based analytics to tackle pain points at the enterprise level. This is a way to break down the silos across the brand without having to change your behavior or ask employees to introduce a new data entry system. For health systems like HSS, that seamless data flow is essential to creating a seamless patient experience.
Ultimately, HSS wants to be like Disney, according to Ballard. They are now known for their world-class hospitals on Manhattan's Upper East Side, providing temporary surgical care. These hospitals are like a medical version of Walt Disney World, attracting international visitors and providing an inclusive experience while still inside the wall.
But even after you leave the theme park, Disney continues to be a part of everyday life. The brand is a constant part of public consciousness through films, streaming services, music and products.
“We're in the Disney world, which is so amazing, but we're staying in the Disney world,” Ballad said. “How do we expand and become a part of your high school's physical or annual tests?”
He sees AI as the answer to that question, without deploying a limited number of talent. As wearable devices and fitness trackers become more mainstream, health systems may have more immediate access to more data than ever before. AI agents are particularly useful in understanding its constant flow of information and use it to prevent patients from falling through cracks between episodes of care.
HSS will work with Palantir to build a complete orchestration layer that brings all this data together, providing a horizontal view of the patient experience for the healthcare system. Schedule appointments from the moment you say “it hurts” to the moment your insurance allows you to pay.
“Up to this point, nothing will organize all the assets we have in our ecosystem into a continuous journey for consumers to pass,” Balado said. But doing so “it actually allows for no handoffs or dropoffs, so the results are better.”
Health Systems has also hired new talent to tackle complex challenges ahead, including Chief Data and Technology Director Dr. Michael Apcon.
“This new world has to be built,” Ballad said.
What challenges do healthcare still face when implementing AI?
The excitement of the AI chatbots with the surrounding scribes was getting tired. And healthcare is running out of quick victory as it appears to build this “new world.”
AI promises efficiency, but the health system has to spend a considerable amount of time unlocking its full advantages.
Beyond the note-taking tools, renowned Dr. AI & Health, in the Department of Computer Science at MIT's School of Electrical Engineering and Computer Science, Dr. Regina Barzilay, AI undergraduate lead at MIT's Jameel Clinic, says that beyond note-taking tools, there are “lol success stories.”
She said Newsweek There is a rich story of new AI algorithms that show promising results, but few people have actually proven themselves beyond just promises.
“You need to know how to bring it safely [health care] The system is, “There is a new science, the science of implementation.”
Like all science, implementation is complicated. Each hospital has a different experience when translating AI into practice due to dramatic fluctuations in infrastructure. The tools are still new, but touching on multiple departments of the health system makes it difficult to find a single point person who can lead your strategy.
Furthermore, according to Barzilay, AI is rarely mentioned in clinical guidelines that govern physician decisions. For example, the National Comprehensive Cancer Network (NCCN) maintains practice guidelines for 72 cancers. Only one guideline for prostate cancer recognizes the mention of AI.
“The clinicians who treat you must follow the guidelines and follow what is considered standard,” Baltzilai said. “AI-powered solutions are not part of the guidelines, so clinicians won't use them.”
Next is the issue of refunds. If the AI tool recommends that patients require follow-up appointments, will the insurer consider sufficient evidence to pay for that visit? With each Barzilay, few AI tools have been able to obtain their own diagnostic codes, and the refund status is “a little wild West.”
As she moves forward, she believes clinical trials will be key to solving guidelines and refund issues. AIs must be tested and validated in a controlled environment to prove their effectiveness, including prescription drugs, new treatments, and surgical devices.
However, science is moving slowly and if cuts to federal health agency funding, the process could be further maintained. “These studies aren't cheap,” Barzilay said.
When healthcare looks to the next phase of AI, it will need to rethink old systems, rewrite past guidelines and conduct long-term research. This all takes time.
And unfortunately, the clock is ticking. According to Barzilay, the US is already behind other countries when it comes to healthcare AI. It is easier to deploy new technologies across government-led or nationalized healthcare systems, like in the UK, Israel and the Netherlands. These models are usually more uniform among hospitals in the same country.
However, in the US, hospitals can have very different sources of financial resources and digital infrastructure, even within the same state. Each health system must start from near-scratch when selecting and implementing AI tools.
“I think the federal government really needs to lead this transformation,” she said.
Barad ultimately wants to use AI to improve health equity. They hope to expand the expertise of HSS's world-class surgeons to even the smallest rural hospitals. At this point, codifying AI agents with medical knowledge of eOns is the “easy” part. Hospitals will need to change the game to grow results across state lines and siloed systems.
Want to hear more? Dr. Asis Varado speaks NewsweekWith leaders from the Digital Healthcare Forum in New York City on September 16th, Kaiser Permanente, City of Hope, Sanford Health, Northwestern Medicine and others. Click here to see your tickets in more detail.
