Four questions with Lisa Hornick: Human & AI for Payment Integrity

Lisa Hornick BA, CPC, CPMA, CEDC, CPhT, AAPC Approved Instructor is the Medical Coding Team Lead for Shift clients in the healthcare space. 

In this installment of Four Questions, we talk with Lisa about the career path that brought her to Shift, her thoughts on payment integrity’s evolution throughout the years, the role artificial intelligence and human collaboration has played, and will continue to play, in changing how the industry thinks approaches payment integrity.

A quick glance at your LinkedIn profile shows a truly interesting career from pharmacy technician to leading the Medical Coding team for Shift. Can you tell us a little more about your journey?

When I take a step back and look at my career, how I got here from there makes a lot of sense to me. My time as a certified pharmacy technician and allied health instructor exposed me to not only the clinical side of medicine but also the business side—particularly health insurance management. From there, it felt like a natural progression to apply my experience to high-cost drug auditing. I was responsible for reviewing medical drug claims, across multiple accounts, to identify instances of overpayments. And from auditing I moved into medical coding and began working in the payer space. I handled appeals, medical records review, and worked with providers to resolve any issues to ensure they had a good experience working with our organization. From coding, I moved into ideation where I initiated cost savings initiatives and ensured compliance—all with the goal of saving money for the company and reducing premiums.

Ultimately, I was promoted to coding manager with oversight of coding appeals, the SIU, payment integrity, and ideation. It was there that I began to really understand the impact technology could have on the payment integrity process and I wanted to be a part of that. So, naturally I came to Shift.

How have you seen payment integrity evolve over the course of your career?

I’ve worked in payment integrity for more than a decade now and have witnessed a ton of changes. When the term “payment integrity” was first introduced it was really kind of a buzzword. Payers would say they had a payment integrity division, but the impact was negligible. From there, the industry moved into what many still think of as payment integrity, a basic check of coverage and services provided. Essentially, payment integrity was all about determining if the patient is a plan member, if the service being billed is covered by the plan and is the rate within contracted guidelines. And don’t get me wrong, it’s a good start. But as we’ve seen over the years, it’s a system that can be easily manipulated. Claims that look legitimate on the surface can be hiding any number of inaccuracies, misrepresentations, or other aberrations. If they’re not found before the claim gets paid, now you need to try and get it back.

What we’re beginning to see emerge is a new way of approaching payment integrity that goes beyond the basics and is really charged by AI. Payers are finding savings opportunities by incorporating AI into areas like policy analysis, uncovering unknown patterns in claim activity, optimizing edits for areas of vulnerability and continuing a feedback loop between postpay and prepay activity. They’re looking for what may be honest mistakes or lags from our changing healthcare landscape, but still have the potential to significantly impact the business. It’s truly an exciting time to be in the payment integrity space. I can’t wait to see what’s next!

AI and human collaboration is essential for a modern, trusted strategy. How does that drive your role, and what does that mean for the future of payment integrity? 

My role sits right at that intersection between AI decisioning and my former roles in medical coding for plans. As far as how that collaboration concept impacts my role, I validate updated edits, serve as the final checkpoint before results go to the plan, and bring first-hand coding experience to every decision we make. That context matters. AI can surface patterns and anomalies (at a scale no team I've seen can manage manually), but it takes someone who has actually coded claims to know whether a flagged result reflects a real problem or a nuance the model hasn't fully learned yet.

That feedback loop is where that scale becomes closer to possible. When I flag something; yes, it's for that single result, but it also makes the model smarter over time for future decisions. So the relationship isn't human vs AI; it's human expertise shaping AI accuracy and outputs continuously. For payment integrity broadly, I think that's the future. AI isn't replacing clinical and coding judgment, but amplifying it so plans can act faster, with more confidence, and at scale.

What do you think is next for payment integrity?

I think we're going to see plans and payer organizations thinking differently about payment integrity, and I see that happening in a couple of ways.

We're going to see payment integrity looking a lot more like a closed-loop system where pre-pay and post-pay analysis inform each other to drive real visibility into what's happening with individual claims, providers, and networks. With the growing adoption of AI on the provider side, claims are going to come in faster and with higher complexity - keeping up will call for a tight partnership between AI and human experts in the loop. And that loop only works when human expertise is part of it: interpreting what AI surfaces, validating it, and feeding those decisions back into the system so it keeps getting smarter. It’s a powerful combination.

I also think we're going to see how AI can help break down the silos that limit current payment integrity strategies. Take dental benefits, for example. We often see dental benefits submitted and paid on a medical plan because the provider is both an oral maxillofacial surgeon and a dentist, billing for cleanings, surgeries, Botox for TMJ. There are so many scenarios where a dentist could bill against medical benefits instead of dental. AI can surface those patterns at scale, but it still takes someone with the clinical and coding background to recognize what's actually happening and make the right call. When AI surfaces the big picture and humans act on it, that's where good decisions get made.