Large language models help usher in a new era of personalized healthcare

An interview with Notable co-founder and chief executive officer Pranay Kapadia on how ChatGPT-like technology is being applied to healthcare

Over the years, trillions of dollars have been spent to move healthcare from a largely analog industry to one that is far more digital. And it has mostly worked. Today, it’s estimated that healthcare accounts for nearly a third of the world’s entire data volume. What’s been far less successful is harnessing healthcare’s vast data to improve the way patients experience care and ultimately drive better outcomes.

But, progress is being made. By now, most people have heard of tools like ChatGPT, which take advantage of large language models (LLMs) and generative pre-trained transformer (GPT) technology, to generate responses to complex queries faster and oftentimes more accurately than humans are able to. According to Pranay Kapadia, Notable’s co-founder and chief executive officer, LLMs “...will revolutionize the way we access and process data across virtually every industry.”

When LLMs are intelligently applied to massive datasets like those that exist in healthcare, that’s when the magic happens. Notable’s Patient AI represents the world’s first use of LLM and GPT to bring personalization at scale to healthcare.

In this interview, Kapadia shares more about Patient AI and explains how it represents a significant step toward ushering in a new era of personalized healthcare.

Q: What is Patient AI and why does it matter to health systems?

Kapadia: Let’s start with why it matters. It matters because despite healthcare being deeply personal for every single one of us, the way most of us experience healthcare in this country is decidedly impersonal. We get asked the same questions multiple times. We have to fill out generic forms and paperwork repeatedly. We have to coordinate our own care. All of this has to change. We have to put healthcare on par with other industries that have been doing personalization in a digital context really well for a long time. 

Patient AI is Notable’s intelligence engine. It powers all of the automated workflows on our platform by comprehensively reviewing millions of data points, both within the medical record and from third-party sources, to provide personalized recommendations to every patient throughout the care continuum. It’s invigorating to do something that’s never been done before – namely deploying large language models and GPT to bring personalized healthcare at scale to the industry. 

Q: What were some of the major considerations when you and the team were building Patient AI to ensure personalized healthcare?

Kapadia: There were many, but I’d specifically call out three. First, we know that patient expectations for personalized healthcare are sky-high, and for good reason. We all experience personalization in other parts of our lives, whether it is when banking, or making a restaurant reservation, or when we turn on Netflix and look for something to watch. So, whatever we built, in terms of the experience, needed to feel personal, and truly magical to patients. When you see how Patient AI proactively surfaces care recommendations to an individual patient, we believe we’ve achieved that.

Second, healthcare is extremely complex. At present, most patients are required to coordinate their own care, or the care of a loved one, despite sometimes having little to no understanding of how the healthcare system actually works. How do they find information about a relevant and available clinical trial? How can they efficiently identify an in-network specialist and schedule that follow-up appointment? We wanted to remove this burden from the patient. 

Finally, patient safety and privacy are paramount. Much of the data we’re talking about analyzing and activating to deliver a personalized experience is sensitive. So, we spent a lot of time thinking through best practices and approaches to ensure we respect and protect that data at all times.  

Q: For physicians, what is the value of Patient AI in delivering personalized healthcare?

Kapadia: Two things immediately jump to mind: increased capacity for providers and staff, and ensuring patients feel heard and cared for. It’s no secret that physicians are busier than ever. In a normal clinic day or hospital shift, clinicians are running from patient to patient with almost no time in between to read up on who they are about to see next or review their charts. This leads to a highly impersonal, transactional care experience for patients, and the common sentiment from patients that it feels like the health system doesn’t know them.

With Patient AI, we can imbue every care encounter with highly personalized, patient-specific context. This not only helps patients feel heard and cared for but can also improve outcomes and administrative efficiency. It’s a significant step forward for the industry.

Q: It feels like every industry is talking about large language models and tools like chatGPT. What is Notable’s unique approach to utilizing some of these underlying technologies and applying them to healthcare?

Kapadia: There is certainly a lot of hype around these technological innovations, and for good reason. The reality is that things like LLMs enable almost anyone to access an incredibly powerful technology capability, so what makes our application any different?

To start, Patient AI is able to analyze a patient’s medical record in seconds, learning their clinical history. From there, Patient AI enhances the clinical data with third-party data like financial history, employment information, pharmacy data, and more. The end result is that in a matter of seconds, we are able to create a comprehensive clinical and social understanding of each patient without any staff intervention. At Notable, we are leveraging our existing distribution such that our Patient AI application can immediately impact more than 20 million patients and supercharge the operations of our health system partners. 

Q: Why is now the time for Patient AI?

Kapadia: There is no doubt; large language models will revolutionize the way we access and process data across virtually every industry. Healthcare, in particular, is ideally suited to leverage large language models because of the sheer volume of latent data that exists across a given health system’s medical records. For years there has been discussion about unlocking the potential of all this data and now technology has evolved to a point where that is possible. When you combine the technology evolution with the current highly transactional, impersonal nature of clinical encounters, you have an environment that is ripe for this type of innovation. 

With Patient AI, we can restore the personal touch to healthcare while also democratizing access to a concierge-like care experience. Patient AI can be used to anticipate and outreach to patients for any number of needs – from rideshare to clinical trial matches to consolidating appointments. Imagine empowering every patient with their own personal care coordinator who knows them best – that’s what we’re doing with Patient AI.

—

Learn more about Patient AI and put your health system on a path to personalization at scale.