The Evidence Utilization Gap: Why the Chart You Retrieved May Be More Valuable Than You Think

Two operations, one chart

A Medicare Advantage member — 68 years old, Type 2 diabetes, hypertension — has an annual wellness visit with their primary care physician. One visit. One provider. One chart.

Inside that chart: blood pressure readings, lab results, medication management notes, chronic condition assessments, problem list updates, physician documentation of how each condition was evaluated and managed at this encounter.

A HEDIS team reviewing this chart sees quality evidence — blood pressure documentation relevant to BPC-E, HbA1c results relevant to Comprehensive Diabetes Care, measure closure opportunities.

A risk adjustment team reviewing the same chart sees something different — face-to-face documentation of chronic conditions, diagnosis specificity, evidence relevant to HCC accuracy under the current risk adjustment model.

Same chart. Same encounter. Same provider. Two different lenses, looking for two different things — both present in the same document.

In most health plans, these are two different operations. Separate retrieval requests. Separate vendors. Separate abstraction workflows. Sometimes separate timelines, months apart, for the same underlying documentation.

The chart isn't the asset

The instinct in most quality operations is to think of chart retrieval as the core activity — request the chart, receive the chart, review the chart, move to the next one. Retrieval completion becomes the measure of progress.

But the chart itself is not what has value. The chart is a container. The evidence inside it — the documented clinical encounter, the physician's assessment, the dated and signed record of what was evaluated and how — is the asset.

Once that evidence has entered an organization, in principle every program that can legitimately use it should be able to draw on it: quality measurement, risk adjustment, audit defense, population health analysis. The acquisition cost — the provider's time, the retrieval request, the abstraction review — has already been paid, once.

In practice, most organizations are not structured this way. Each function operates as though it owns the chart it requested. Quality retrieves for quality. Risk adjustment retrieves for coding. Audit preparation retrieves for defense. Each extracts what it specifically needs and the rest of the document's value goes unused — until, sometimes, another team retrieves the same chart again to look for what they need.

This is the Evidence Utilization Gap: the distance between the evidence an organization has acquired and the evidence it has actually used.

Why this is a 2026 problem specifically

The gap between acquisition and utilization is not new. What has changed is the cost of leaving it unaddressed.

CMS-HCC Version 28 is now fully implemented for Medicare Advantage risk adjustment, having phased in across 2024 through 2026. V28 expanded the model from 86 to 115 HCC categories while removing more than 2,000 diagnosis codes that previously contributed to risk scores under the prior model. Conditions that generated risk-adjusted payment under V24 may carry no weight under V28, depending entirely on how — and how specifically — they are documented.

The practical effect: the clinical specificity inside a chart now matters more for risk adjustment accuracy than it did under the prior model. A chart that was adequate evidence for risk adjustment under V24 may not be sufficient under V28, even though nothing about the patient's underlying condition has changed. The documentation has to do more work.

At the same time, RADV audit activity is expanding. CMS has been working through a backlog of audits spanning Payment Years 2018 through 2024 and has scaled its audit infrastructure accordingly. Plans are under increasing pressure to demonstrate that every submitted condition is backed by documentation that meets current standards — not administrative inference, but compliant medical record evidence.

Put together: the documentation inside a retrieved chart now carries more weight for risk adjustment than it used to, audit scrutiny of that documentation is increasing, and — separately — HEDIS audit requirements continue to demand the same standard of source validation and traceability they always have.

The evidence inside every retrieved chart has become more valuable across more programs simultaneously. Organizations that have not adjusted how they extract value from that evidence are leaving more on the table than they were two years ago — not because they are doing anything differently, but because the value of what they are not doing has gone up.

The infrastructure question

HEDIS and risk adjustment are different programs with different objectives, different evidence standards, and different review logic. That is appropriate — they answer different questions, and conflating them would be a mistake.

But the infrastructure underneath both programs is substantially the same. Both require:

• Clinical documentation obtained through a compliant, traceable retrieval process
• A documented chain of custody from source to record
• Evidence that can be validated against current-year requirements
• An audit trail that can reconstruct how a conclusion was reached

This is the case for treating chart retrieval and abstraction infrastructure as an enterprise capability rather than a departmental one. The discipline that produces a defensible HEDIS closure — documented source, validated record, traceable decision — is largely the same discipline that supports RADV readiness. Not because the two programs are the same, but because both depend on the same underlying standard of evidence handling.

When that infrastructure is built and governed separately by each department, the result is duplication: the same discipline, built twice, applied to overlapping sets of charts, often with neither team aware of what the other has already retrieved.

The cost that is easiest to see: provider abrasion

There is a version of this problem that providers experience directly, even if health plans do not always see it from the provider's side.

A physician's office documents an encounter once. One chart is created. But that one chart may generate multiple separate requests from the same health plan — one from a quality program, one from a risk adjustment program, possibly others from vendors supporting still other initiatives, each arriving on its own timeline.

The provider did not create multiple charts. The plan created multiple workflows, each requesting access to the same underlying documentation independently.

Every additional request has a cost that compounds over time: staff time at the practice to locate and transmit records, administrative effort on both sides, and — less measurable but real — the accumulated impression a practice forms of how burdensome it is to work with a given plan. In an environment where provider administrative burden is already a persistent concern, requests that could have been consolidated but were not add friction without adding value.

A test for whether the current approach makes sense

One way to evaluate whether an organization's current evidence workflows make sense is to describe them plainly, the way you would to a CFO, and see whether the description holds up.

Said this way, the natural follow-up questions are straightforward: why are we paying twice to extract value from the same source? Why are two teams operating independently against an overlapping evidence base? Are we organizing around departmental workflows, or around the outcomes the organization actually needs?

These are not questions about whether retrieval happened. Retrieval, in this scenario, worked fine — twice. The question is whether the evidence, once acquired, was used as completely as it could have been the first time.

Retrieval is an activity. Utilization is an outcome. An organization can have a well-functioning retrieval operation — high completion rates, responsive vendors, satisfied internal stakeholders — and still have a significant evidence utilization gap, because the two are not the same measurement.

What closing the gap looks like operationally

Closing the Evidence Utilization Gap does not mean merging HEDIS and risk adjustment into a single program, or asking one team to make decisions that belong to another. It means restructuring how evidence flows once it has been acquired, so that the work already done by one program can inform — without dictating — the work of another.

In practice, this looks like three shifts:

The throughline across all three: the evidence acquired through one process becomes visible to other processes that could use it, without any process losing ownership of the decisions that are properly its own.

The reframe

Most organizations, when asked whether they have an evidence problem, think about it in terms of acquisition: do we have enough charts, enough data, enough documentation to support our programs?

That is the wrong question — or at least, it is usually not the most useful one. For most plans, the more consequential question is whether they are extracting full value from the evidence they have already acquired. One question leads toward more retrieval. The other leads toward better utilization of what retrieval has already produced.

And in 2026, with more programs depending on the specificity and defensibility of that evidence than at any point in recent years, how completely that evidence gets used — not just how much of it gets acquired — is increasingly where performance is determined.