Integrating AI and Pathway Flexibility to Advance Value-Based Cancer Care: Clayton Irvine, PharmD, MBA, MS

At a recent Institute for Value-Based Medicine^® event, Clayton Irvine, PharmD, MBA, MS, highlighted operational and billing challenges in implementing value-based initiatives, including payer-specific biosimilar mandates that complicate inventory management and create inefficiencies in academic and infusion centers. Manual prior authorization processes can delay care and contribute to financial toxicity for both patients and institutions, underscoring the need for integrated digital solutions, such as artificial intelligence (AI)–enabled platforms and CoverMyMeds to streamline workflows.

Additionally, Irvine, senior manager of oncology cancer care at Mayo Clinic in Rochester, Minnesota, emphasizes the importance of pathway flexibility—such as defined off-label arms with documented exceptions—regular stakeholder review to keep pace with evolving guidelines, and the use of molecular tumor boards and AI-driven clinical decision support to balance standardized pathways with personalized, biomarker-driven care, particularly in rare or late-line cancer settings.

This is the second video in a series with Irvine. Please revisit part 1, where Irvine discusses the importance of academic and community oncology collaboration.

This transcript was lightly edited; captions were auto-generated.

Transcript

What operational or billing challenges have you faced when implementing value-based initiatives, and how have you addressed them?

We see lots of challenges and inconsistencies for biosimilar formularies across pairs, and it's created a huge challenge in academic centers and infusion centers, primarily because there's a lot of specific payers that want a certain biosimilar. It creates inventory issues for pharmacies, and pharmacies having to keep all the drugs on formulary, and then at the time when the patient's there for treatment, how do you know which drug is best for the patient? It ties into many institutions have a manual prior authorization process, and so that creates delays and inefficiencies and also sometimes leads to financial toxicity, not just for institutions, but for patients. I think it's a conversation that we don't talk about enough; it's just some of those financial implications then get passed downstream to patients.

I think there is opportunity when we look at the landscape, can we use integrated platforms like CoverMyMeds or other AI-based solutions? There's prior authorization software out there, and I think when we are able to look at those and evaluate how we can integrate them into our workflows, I think that's really where we see the value and benefit. I know that that's a lot easier said than done, and there's a lot of complexities that come with bringing some of those systems in house, but knowing what's out there and knowing the tools that are available is a step in the right direction.

If you aren't able to leverage those technologies, if you aren't able to incorporate those platforms, looking at ways to streamline those properties can really help the downstream bill waste that we see, especially when we're talking about prior authorizations and other decisions that we clinically would like to make without that in place.

As pathway adherence becomes more central to value-based care, how do we balance standardization with the need for personalized, biomarker-driven treatment?

I think pathways really should have an off-label component or an off-label arm to the system that you're using, or however you have pathways setup, but because there’s a need to tie the off-label arm to some type of documented exception and have those clearly defined, because it then helps better streamline those approaches when they happen—and they do absolutely happen I think having stakeholders, and who those stakeholders are should be representative, but having stakeholders review those pathways in a frequent and regular cadence is really important.

Guidelines are changing, and practices are changing. What is happening in practice and how that translates to data in clinical trials or new and emerging trials and information—we need a better way to review those pathways and ensure with new data, even if it's not changing the pathway immediately, those conversations and discussions are taking place.

Additionally, we really need to look at how we can use molecular tumor boards or clinical decision support tools. There's a lot of work being done right now with AI helping to augment and support this approach, and that's a really exciting field. I think that in 5 years from now, we might have the answer to that, maybe faster than the rate that AI and clinical decision support tools are going, but it's an exciting field, and I think that it provides a big opportunity for where we're going in the future.