What AHIP 2026 Revealed About the Biggest Unsolved Problem In Healthcare Engagement

by Abraham Cruz-Peña

I just spent three days at AHIP 2026 in Las Vegas sitting in sessions, talking to plan executives between presentations and listening to some of the most thoughtful leaders in Medicare Advantage and Medicaid managed care wrestle with the same problem from every possible angle.  

I came out with a clear picture of exactly where the industry is stuck, and why the solutions being offered aren’t unsticking it.  

Here’s what the room told me.

Risk Stratification Is Solved. Everything After That Is Broken.  

Session after session, the language was the same. Health plans know who their high-risk members are. The data is there.  The 30% of members driving 80% of cost are identified.

What nobody could answer--and what a full room of care management executives openly discussed in the chronic care session--is what should happen next.

Once you know who needs to be reached, how do you actually get them to respond?

One of the most striking moments of the conference came from a senior clinical leader at one of the largest plans in the country, who stated candidly that current engagement solution vendors are not delivering promised cost savings. Nobody argued. The bar is high, and current tools are not clearing it.  

This is not a data problem. It is not a targeting problem. What plans lack is a reliable answer to why a member who should engage – considering clinical demographic measures – simply doesn’t.  

And the stakes of getting it wrong compound quickly. Nearly one in five Medicare Advantage members switch plans every year. Health plans are not just racing against the measurement window – they are racing against member churn. Every month of trial-and-error outreach is a month a high-need member could leave the plan before a single care gap closes.  

The Question Nobody Asked – And Why It Matters

Across 17 sessions and dozens of panel conversations, I heard a lot about engagement. I heard almost nothing about motivation.

These are not the same thing.  

Engagement is the outcome. Motivation is what produces it. And the data that explains what motivates a specific person to act -- to call back, to show up to an appointment, to take the medication, to close the care gap -- does not live in an Electronic Health Record today. Nor does it live in claims data or SDOH records.

Motivation is a separate layer entirely; one that is currently invisible to most health plans.

The gaps being discussed at AHIP aren't niche. 70% of eligible adults miss colorectal cancer screenings. 61% of diabetes patients skipped recommended kidney tests. These aren’t edge cases, but the majority. The engagement wall is the rule, not the exception.

One data point at the conference stood out starkly: one of the most well-documented in-home care models in the industry offers the highest-touch intervention available: at home clinician visits. Still, their data showed a 43% completion rate for colorectal cancer screening kits. That means 57% of members who were handed a screening kit by a clinician in their own living rooms still did not complete it.

If maximum intervention intensity does not move more than half of members to act, the problem is not access. It is not the channel.  It is not the clinical profile. It is not social determinants. It is something none of those datasets can surface: the motivational why behind individual member behavior. That is the layer the industry has not yet named, and the gap nobody at AHIP had an answer for.  

“Read, Deleted, Ignored”: The AI Problem Nobody Is Talking About Honestly

Health plans are investing millions in AI-powered engagement. The capability is real. The outputs -- personalized messages, optimized channels, automated outreach sequences -- are technically impressive. But the irony of AI-powered engagement at scale is that it has made the problem worse in one important way. As every plan optimizes their outreach – more messages, more channels, more frequency – members begin ignoring all of it equally. “Read, Deleted, Ignored” was not just a session title but an accurate description of what most members do with the number of health plan communications hitting them every week.  

More sophisticated messages that don’t resonate are not a solution. They are a faster path to the same wall.

The AI solutions that dominated conversations at AHIP mostly solve for delivery – how a message reaches a member, when it arrives, and through which channel. But delivery optimization assumes you already know what will motivate that specific member to act. And that assumption is where most engagement strategies quietly break down.  

Optimizing delivery without understanding motivation is like building a faster postal service for letters nobody wants to open. Two members. Same diagnosis. Same risk score. Same AI-optimized message, delivered at the same time through the same channel. One responds. One disappears. The delivery was identical. What was different was the motivation -- and no delivery platform can explain why.

Care managers are not struggling because they lack outreach tools. They are struggling because none of those tools tell them what to say, or how to say it to move a specific person before the first call is made. That is not a delivery problem. It is a decision problem. And it is the one worth pointing AI at.

The First Interaction Is Everything….And Nobody Is Designing For It

This theme surfaced repeatedly and offers clues to how to address the engagement gap.  

In a session titled "Trust-Based Strategies for Hard-to-Reach Members," care management leaders from human-first outreach organizations and large managed care plans described a reality that everyone in the room recognized. For Medicaid members, dual-eligible populations, BIPOC individuals, and people navigating housing instability, behavioral health challenges, and economic hardship, trust is not built through a protocol. It is built or lost in a single moment: often the very first interaction.

Plans are not designing for that moment. They are reactive to it. The outreach goes out, the coordinator makes the call, and the relationship either forms, or it doesn't, with little ability to predict which will happen and even less ability to explain why.

What would it look like if a care coordinator knew before that first call what kind of connection a specific member needed? For example, whether they respond to autonomy and data, to warmth and reassurance, to step-by-step guidance, or to community and belonging?

That is not a generic segmentation question. It is a precision question. And the answer is not in any clinical database.

A Few Things Worth Carrying Forward

The Engagement wall is real. Every session confirmed it. The data, the vendors, the plans – all pointing at the same problem, all still searching for the same answer.  

What motivates a specific member to act is not in any database we have today. It is not in the EHR. It is not in claims. It is not in SDOH records. It is a layer the industry has not yet named – and the plans that find it first will not just improve Star Ratings. They will redefine what member engagement can look like.  

That is the gap we are building toward at Rosemark. If you want to see how we are approaching it, you can learn more here. If you were in the room and you want to compare notes, contact us to connect. We’ll also be at RIZE in San Diego in September – coffee is always on.

‍