Healthcare has an outcomes problem

Rethinking Patient Access in a Conversational World

Healthcare has invested heavily in digital transformation over the past decade.

On paper, it looks impressive: EHRs. Patient portals. Mobile apps. Cloud infrastructure.

But if you’ve ever tried to get a straight answer as a patient, you know how this investment plays out on the ground.

The moment something really matters, whether it’s booking an appointment, disputing a bill, or figuring out the next step to deal with a health issue, you don’t want to click around. You want to talk to someone who understands and can help.

That’s the disconnect.

Despite the digital investment, patients still default to the phone: in a recent study of more than 3,600 U.S. patients, over 70% reported using calls to schedule care—more than any other channel.

Digital healthcare systems were designed with a particular goal in mind: to store and manage information. Patients have entirely different goals: they are trying to accomplish something specific related to their care.

When those goals don’t line up, friction shows up immediately.

That gap between how systems are built and how people behave is where a lot of the frustration in healthcare lives today. It’s also why, despite all the digital investment, the contact center is still the front door to the healthcare system. 

When “modern” systems don’t feel modern

We’ve built incredibly sophisticated systems of record.

But we haven’t built equally sophisticated ways to interact with them.

Even the layers meant to handle interaction (think IVRs, contact centers) were designed to manage volume, not resolve outcomes. So what happens?

→ More transfers → More repetition → More conversations that lead nowhere

If you’ve ever had to explain your situation three times to three different people, you’ve felt this firsthand.

Why voice still wins (even in a digital world)

Patients don’t think in channels.

They move between phone, chat, and digital tools depending on what they need. But when something is urgent or unclear, they default to the most natural interface we have: conversation.

And that’s why voice still plays such a central role in healthcare. As technology progresses, that’s not changing. What is changing is what happens during patients’ conversations.

We’re starting to see AI agents step in, not just to route or respond, but to carry context across interactions and help complete the task the patient is seeking help with.

These AI agents pull context across channels. Meaning a conversation can start on the phone, continue over in-portal messaging, and trigger a workflow in the background. But from the patient’s perspective, it’s seamless. No restarting. No repeating.

That continuity, not the channel, is what makes an experience feel modern.

Where this really matters: high-volume, high-friction workflows

If you spend time with healthcare operators, you know the same types of interactions come up again and again:

• Scheduling and rescheduling

• Eligibility and benefits checks

• Claims and billing questions

• Referral coordination

• Authorization status

These make up the bulk of contact center volume. They’re also the moments that determine whether care moves forward.

And, unfortunately, they’re where things break down most often.

This is where AI is starting to prove its value. Not just because it’s efficient, but because it finally addresses the structural problems that have plagued healthcare systems for eons:

→ It can handle volume.

→ It can maintain context.

→ And, most importantly, it can take action.

The real constraint: systems of record

Here’s the part that often gets overlooked:

Every meaningful outcome in healthcare has to be resolved within a system of record.

Platforms like Epic hold the truth: patient data, authorization status, and workflow state.

So for AI to move beyond surface-level interactions, it has to be able to operate inside those systems. That’s the difference between answering a basic question and actually solving real patient issues.

When AI is connected to systems of record, a conversation can turn into a completed outcome:

→ An appointment gets scheduled.→ A benefit gets verified.→ A claim inquiry leads to a clear next step.

Which, to the patient, was the whole point of their outreach to begin with.

What it takes to make this work in the real world

Making this a reality requires technology, but it also requires orchestration. Healthcare access spans systems, teams, and workflows that were never designed to work together.

That’s never more obvious than in high-friction scenarios.

Take authorization status checks. Providers often have to call repeatedly just to confirm whether something has been approved. It’s manual, it’s time-consuming, and most frustrating of all, it can delay care.

But it’s also exactly the kind of interaction where AI can help, by pulling the right data from the right systems and moving the workflow forward in real time.

What we’re hearing from the front lines

At two recent healthcare conferences, ViVE and HIMSS, this wasn’t a theoretical conversation. We spent days on the show floor speaking directly with payers, providers, and health plan reps. And we kept hearing the same thing.

We heard urgency around implementing AI. But honestly, we also heard a ton of concern.

Some of the things keeping leaders up at night? Contact centers are overwhelmed with high-volume interactions. Staff are stretched to their breaking point. Administrative workflows are fragmented across systems. And whenever they try to automate, they’re stymied by questions about accuracy, compliance, and control.

Teams are feeling the heat. They know they need to improve access, cut operational load, and give patients and members more modern experiences. But, understandably, they’re cautious. They’ve seen amazing pilots that check all the boxes in controlled environments, but then fall flat when they’re out in the wild, exposed to production-level complexity.

Where this is all heading

The loud-and-clear message we walked away with was that the big struggle for organizations isn’t that they don’t get AI. They do. They just haven’t figured out how to trust it in production yet.

Bridging that divide doesn’t mean existing systems should be ripped and replaced. It just means we need to make systems more responsive. More usable. 

In the end, healthcare industry leaders want to be able to create experiences that start with a conversation and end with a real outcome, without the friction, repetition, and dead ends we’ve all experienced.

They understand that in healthcare, the issue isn’t access to technology.

It’s access to outcomes.