Asembia AXS25: How AI Will Reshape Prescription Fulfillment

In an interview with Pharmacy Times®, Harry Travis, president of The Travis Group, discussed how AI can significantly improve pharmacy operations by streamlining prescription fulfillment, resolving prior authorization questions, and assisting with copay support. He highlighted his presentation, “AI and its Impact on Healthcare and Pharmacy - An Introduction,” which he shared at Asembia's AXS25 Summit.

Travis highlighted potential ethical concerns around using AI to predict patient medication adherence, particularly regarding potential biases in algorithmic decision-making. He emphasized that AI should primarily give pharmacy teams more time to deploy their existing skills more effectively. Travis also stressed the need to address regulatory and payer-side challenges to fully realize the potential of AI in pharmacy settings.

Pharmacy Times: Beyond drug interactions, where else in pharmacy could AI have the biggest impact?

Harry Travis: I think in pharmacy operations, AI can have an impact on practically every step in the fulfillment of a prescription. Drug interactions are a good example where the data is there, out there in various clinical databases as to what drug interacts with what drug, but there's also data in the medical record where AI can help pharmacists resolve or answer prior auth questions; that's one area. Copay assistance is another area, using any number of AI tools to quickly, essentially, assess or investigate the vast amount of resources that are available—different sites, different companies, different funds that could be available for a patient to help them pay for their medication. So, prior auth is one, patient assistance is another, just data entry. The entry of the prescription, whether it comes by fax, electronically, or whatever, into the operating system of the pharmacy. I don't think that there is any step along the fulfillment or dispensing of a prescription where AI can have some positive impact.

Pharmacy Times: What are the top ethical concerns pharmacists should have about using AI?

Travis: It depends on how the AI is being used. When it comes to things like drug interactions or prior auth, I'm not too worried about ethical considerations. If this is more hypothetical, as I'm thinking about it, because everyone's trying to figure this out as we go, the answers aren't really obvious. Let's just say you're using AI to predict the adherence of a certain patient class to certain medications. How adherent is this patient? Can I risk stratify patients so that I deploy my most aggressive or most optimal adherence tools to a certain patient class? What are the algorithms that are being used to drive those decisions where this patient is likely to be more adherent than that patient? What's underlying that? Are you really sure of what the social determinants were used to figure that out, and is there a certain bias in there? Because it's all built on some database. When it gets to the personalization of care based on some model, that's where my spidey sense starts going up. Are we sure we've got this really worked out?

Pharmacy Times: How will AI change the pharmacist's job in the future, and what new skills will be needed?

Travis: The basic answer, the very basic answer, is AI should give the pharmacist more time. It should make a pharmacist, or, let's say, more accurately, the pharmacy team, because typically a pharmacist is not operating solo. There are technicians or customer service reps and patient service reps. Depending on the setting, be it a small community pharmacy or a big specialty pharmacy, there's a team. I like to say, how is AI going to improve the team? The simplest way to look at it is time. Your team gets X amount of work done in an 8-hour shift today. They're going to be able to do that in 6 hours or 5 hours. What are you going to do with those extra 2 hours? Now I don't think that the pharmacy team needs to learn anything new, because I think they're trained to do a lot that they can't today because of the administrative burden. Now you've lightened that load. Are they ready to deploy the skills that they have when they couldn't prior to that? Or the next question is, are they going to get paid for that? In some situations, they will. In some situations, they won't. In some states, they may not even be able to do certain things because the regulations are active. We have to work on the regulatory side to lighten up regulations, and then we have to work on the payer side to get them paid. Then they unleash the skills that they already have.