Guest expert and AMCP 2024 session presenter Joseph Honcz, RPh, MBA, explores the current role of AI in managed care, barriers to implementation, and strategies for collaborating with managed care leadership to advance the testing of AI technologies. 

Please share your name, title, and a brief overview of your professional history. 

My name is Joe Honcz. I'm senior vice president at a company called Petauri. It’s about a year-old market access and agency company. We help organizations access patients and coordinate with manufacturers and managed care. 

A little bit about my professional history. I went to the University of Connecticut School of Pharmacy and after school I worked in retail before jumping around to different parts of the health care space. I worked as an academic detailer, in medical communications, and for Pfizer in the research pharmacy.

While pursuing my MBA, I applied for a job with Anthem Blue Cross Blue Shield. It was my first managed care job and sparked the next 22 years of working in managed care, mostly for national health plans and PBMs. I worked at Aetna, CVS, Humana, HealthNet, and in all different parts of the business. While at Aetna, I spent time in an innovation lab and enjoyed learning about digital health. Now I take all those experiences and I use them in a consulting fashion to ensure these drugs and digital innovations get to patients in the best way possible. 

Can you provide a specific example of how AI can transform health care management and delivery within health care organizations?

I think there's a myriad of examples, but I'm going to build off an example from one of my prior companies. In the claims processing format for a cancer patient, we wanted to ensure the smoothest path for access possible. We searched patient histories to see whether they had a cancer diagnosis at any time. And if they had a cancer diagnosis, the claim tended to sail through utilization management. 

Now, this was a long time ago and if we take that example and extrapolate it out, that was probably the most infantile AI possible, but it was still an algorithm. As AI becomes more articulate, I think we can be even more intelligent about how to clinically manage utilization and drug access. We might be able to the prescriber at the beginning and say, “Hey, listen, this patient's drug requires prior authorization and based on their history, it's missing x.” And then that could be an opportunity for the patient to provide additional information not in the chart or for the physician to make a credible argument regarding that access should be provided. 

Having this conversation at the bedside, if you will, really allows for a more patient-centric conversation. It allows the patient to participate in that decision, which they don't get to do today. They usually go to the pharmacy, drop off a prescription, and are then told it needs prior authorization. And then it goes into a black hole. 

I think AI is going to allow us to be more humanistic and take us back to talking to each other and working through the best health care decisions to be made. Eric Topol at the Scripps Research Translational Institute in San Diego was the doctor who discovered Vioxx was killing people. He's a cardiologist and a geneticist and writes books all the time, including a book about AI. He thinks AI is going to allow us more human interactions, that we shouldn’t be afraid of AI but instead use it to become more human again. 

What are the common challenges that health care leaders encounter when trying to implement AI solutions in managed care?

When talking with my peers and trying to encourage the adoption of innovative technology, maybe folks don't have a broad understanding or a base of knowledge. I think we can all recognize we may fear things we don't know. The unknown is a bit scary. And it provides us the natural recourse of pulling back instead of leaning in. So right now, that challenge is trying to explain what AI is and then helping people understand how it can be beneficial and an incremental change. The biggest hurdle is helping people to understand, which is typical of any emerging technology. 

How do you garner consensus and support from organizational leadership for AI initiatives? Can you share any particularly effective strategies?

Since a lack of knowledge is a key barrier, I think one of the first initial strategies is being as transparent as possible and just sharing knowledge so that leaders can appropriately understand what we're talking about. AI is like a brain, and we're training that brain to do different tasks. Just like a child if you raise them poorly then they will misbehave. But if you raise them properly, they can be a great human being at some point. 

Another example is when I was moderating a panel at the Academy of Managed Care Pharmacy, and I heard from a national leader at a health plan in the US that they had deployed their own AI tool so that every employee had access to this AI capability. Their AI was walled off so that it can’t get out into the wild. He noted that about 50% of the employees now use that AI tool daily. This is tens of thousands of employees, a very large plan. And I think that's probably one of the best ways to get people to lean in and use a tool. And the AI can intuitively learn at the same time.  Once the idea is tangible and you have that baseline understanding, you can dream more appropriately. 

Many organizations talk about scale. Although the AI tool was being deployed across all the employees was a pretty large scale, it's not driving the business operations because if you introduce an AI tool and it breaks business operations, that's the last time you're going to get a chance to try it out. But if you can deploy it on an individualized basis and let people get comfortable, there's no greater way to learn. It's like the residency model: see one, do one, teach one. 

In what ways do AI technologies have the potential to enhance patient care and operational efficiencies in managed care settings? 

That's a bit more of a reactive process. I’d love for us to start thinking about proactive care and management. We don't do that very well here in this country, it's all sort of reactive. You get sick, then you get treated. Performing a physical is probably the most proactive thing we do even though we might know, statistically speaking, that patients of a certain age, gender, and ethnicity have a higher risk for x, y, and z diseases. But we don't typically go out and try to find or prevent disease. I would love for AI to take all the data of who you are and what we know to create a lifelong plan of care that could ensure that from infancy to old age you're following a plan of care or at least you're provided with the knowledge to made informed decisions about whether to follow it. 

A simple example of this could be if you have high cholesterol you start on a statin. Usually, a patient will start on the lowest dose and titrate up. It might take a year for you to have incremental visits with the physician and for them to start upping the dose to get your cholesterol level down. That is how our system is set up. But if AI could be involved, it could read sensors, consider lab values, and help prompt an increase in titration. This wouldn’t happen outside the doctor's knowledge but could help that physician recognize an opportunity. I think AI can be a buffer connection point to enable folks to treat diseases more proactively. 

What future advancements and opportunities do you foresee for AI in managed care pharmacy practices?

I think in the near future we're going to stay in a phase of learning. I would love to see curriculum begin to proliferate within schools of pharmacy, maybe crossovers between the school of pharmacy and the school of technology or engineering where we start to mint different types of pharmacists. Pharma informaticists, if you will, who are capable of utilizing AI, before beginning to engage at a managed care population level management.

Let's start using AI as a tool to be much more articulate about patients at risk, identifying them and hopefully productively engaging with them. Even now when we try to identify risks among patients, there are plenty of people who fall through the cracks. 

As we move forward, if I get dreamy and excited, I start to think about truly personalized medicine, meaning a drug that's developed for your disease. I just read an article today that there's an AI out there capable of identifying all the different interactions between receptors within the body. If we can find all the locks in the body, AI can help us define all the keys necessary to either turn something on or turn something off. And that becomes incredibly powerful. Instead of just symptom management or controlling disease progression, we're potentially getting into personal cures. Candidly, I can't think of something more exciting than that. I think pharmacists will be integral to this conversation, especially this new kind of pharmacist I'm envisioning, who can access what's going on and then figure out what therapy is needed. 

This could be the change we need at just the right time. We're running out of cash flow in this country in terms of the cost of care. It’s a big dream but if we can start curing diseases at a reasonable price, I'm hoping we can improve the overall longevity of humans. 

In multiple forums with executives, I have asked them to raise their hands if they had ever tried using ChatGPT. It's a very simple, basic AI that is in the wild. At best, usually, only half of the executives raise their hands. What I'm looking for is increased curiosity instead of fear. I hope anyone reading this will get excited and want to lean in and figure it out. And hopefully, then that allows them to grow and maybe help someone else be less fearful. Once it is out, you can't put the genie back on the bottle, whether you're ready or not.