Arrive Health: Simplifying Prior Authorization with AI

Ayush:
Hi everyone, welcome to another episode of HealthTech with Purpose podcast, where we deep dive into the world of healthcare technology and people building innovation. I’m your Ayush Jain, Ayush, and today we have a truly exciting guest — Christie Callahan, CEO of Arrive Health. Christie is a health tech leader with expertise in strategy, innovation, and operations. She has spent over a decade tackling one of healthcare’s biggest pain points: prior authorization. She’s here to help us unpack the challenges, the role of AI and automation, and what the future holds for patient access and transparency. Plus, we’ll learn about her journey and leadership philosophy. Christie, welcome to the show.

Christie:
Thanks for having me. I appreciate it.

Ayush:
We always start with understanding a bit of your background and journey into healthcare. Did you always envision yourself in this space, or how did your career begin?

Christie:
The short answer is no, I didn’t think I’d end up in healthcare. In school, I was an engineer and worked in manufacturing — on the Puffs tissue line in Green Bay, Wisconsin. As a young professional, I wanted more variety, so I moved into consulting at Bain & Company. That wasn’t a great fit either, and eventually CVS recruited me.

At CVS, I joined a sort of “SWAT team” to help operations when things went wrong. My first project involved their Medicare Part D plan, which was under sanctions from CMS. We had to fix technology, processes, and help the operations team get back on track. Later, I worked on coverage determinations and appeals, which are essentially prior authorizations. The team needed automation and better-integrated technology to reduce workload and improve speed.

This was all pre-COVID. During my time at CVS, I saw specialty pharmacies, call centers, and the pain points of prior authorization. But I also realized that while I could make things faster within the PBM, I couldn’t fix the root cause — providers needing better information and workflows.

That’s when I discovered Arrive Health. Their vision matched mine: tackling the problem from the provider side. And during the pandemic, when everything was remote, it was the perfect opportunity to join without relocating.

Ayush:
So you found your calling.

Christie:
Yes, exactly. What’s been fun is getting closer to decision points for providers and patients. The problems in healthcare are big, but solvable. That’s what keeps me motivated — it’s complicated, but not unsolvable.

Ayush:
That’s a great perspective. Let’s talk about prior authorization, since that’s your deep area of expertise. Why does prior authorization exist, and why is it such a persistent challenge?

Christie:
Specifically around medications, prior authorization usually exists for three reasons:

1. Safety – Some medications require monitoring, dosing adjustments, or have addiction risks. For example, sleep medications used daily may trigger a prior auth as a safety check.
2. Off-label use – Drugs sometimes gain popularity for uses not yet clinically validated. Plans typically won’t pay for these until proven. A big example today is GLP-1 drugs like Ozempic, used for weight loss instead of diabetes.
3. Cost – Occasionally, plans use prior auth to ensure patients try lower-cost therapies first before moving to expensive alternatives.

From the plan perspective, they’re balancing patient safety and overall affordability. But in practice, it creates massive inefficiencies.

In just six months, we saw 416,000 unique drug-plan combinations — meaning 416,000 potential prior authorizations. No provider can keep track of that. It feels unpredictable and inconsistent. Small issues, like an outdated lab result, can cause denials, appeals, and weeks of delays even when the patient is eligible. That’s why 80%+ of appeals get overturned — it’s often just procedural.

Ayush:
Exactly. That’s why many even demand prior auth be eliminated. What’s your take?

Christie:
It’s complicated. Some prior auths should no longer exist, and plans could simplify. But with the right technology, we can actually flip the model to work better for providers.

So far, innovation has mostly meant EPA (Electronic Prior Authorization) — which is faster than faxing but still burdensome. Providers still don’t see requirements upfront and still get stuck in denial-appeal loops.

That’s why I’m excited: technology is now capable of truly lightening the load.

Ayush:
That brings us to Arrive Accelerator. Tell us about your approach.

Christie:
Our philosophy is,

“the best prior authorization is no authorization.”

We start by helping doctors avoid it altogether, showing formulary alternatives through real-time benefits enhanced with AI models. Done well, this helps providers avoid up to 48% of prior auths.

Second, we use large language models (Claude Sonnet and others) to extract and summarize coverage criteria from plan documents. This gives providers clarity before they even start, avoiding wasted steps. For example, they’ll know upfront if a lab test needs to be within three months instead of nine.

Third, we aim to prevent denied claims at pharmacies altogether. By resolving authorization within hours, patients avoid confusion and delays at the counter. Our health system partners now see resolution times as fast as four hours.

Ayush:
That’s impressive. How do you see AI performing in this space — where does it shine and where does it fall short?

Christie:
AI is excellent as an assistant — augmenting intelligence, not replacing humans. It helps summarize requirements, validate outputs, and speed decisions.

Where we’re cautious is PHI (Protected Health Information). Today, we don’t feed PHI into our models. We train only on plan documents and policies, keeping PHI separate. This ensures compliance while still gaining insights.

Another area is quality loops. We run multiple validation steps with LLMs to check for hallucinations or errors. As these improve, we’ll reduce manual oversight.

But I don’t see AI making fully autonomous clinical decisions anytime soon. In healthcare, it will remain human-assist, not human-replace.

Ayush:
Agreed. And taking innovation to market is often harder than building it. How did you approach adoption?

Christie:
We learned that providers don’t want more portals. The key is deep EMR integration. Our solutions are embedded natively so that adoption requires little to no extra training. Tight feedback loops with health systems also help refine the workflow.

We also created urgency by holding prescriptions until prior auth was resolved, ensuring providers engage quickly without patient delays.

Lastly, we built on standards. They’re not perfect but provide a starting point for interoperability and scaling without endless custom builds.

Ayush:
That’s smart. Let’s close with leadership. What’s your philosophy?

Christie:
I see my role as twofold:

1. Unlocking performance – Putting people in the right roles, giving them tools, and fostering cross-functional communication.
2. Being a thought partner – Asking questions, offering outside perspective, and sharing context so teams can connect the dots.

I believe information should flow openly across the organization. That context empowers everyone to make better decisions.

Ayush:
Beautifully put. Christie, thank you for sharing your journey and insights. Where can listeners learn more?

Christie:
Follow us on LinkedIn and visit arrivehealth.com for thought leadership, blogs, and resources.

Ayush:
Fantastic. To our listeners, if you enjoyed this episode, don’t forget to subscribe and leave a review. Stay tuned for more conversations. Until next time, take care and keep innovating.