A cluster of guideline releases has quietly rewired how cardiovascular treatment decisions get justified. The 2025 AHA/ACC blood pressure guideline replaced the Pooled Cohort Equations with the PREVENT risk equations and tied initiation thresholds to them. The 2026 dyslipidemia guideline went further, dropping the Class I statin indication to a 5 percent ten-year PREVENT risk, down from 7.5 percent. The first-ever AHA/ACC/ADA/ASN cardiovascular-kidney-metabolic (CKM) guideline, released June 9, 2026, builds its entire staging and prevention logic on the same equations. PREVENT is no longer one tool among several. It is the shared denominator across blood pressure, lipids, and cardiometabolic risk.
The provider side is beginning to move. Elation Health, a cloud-based EHR, licensed the AHA calculator and went live for clinicians ten days after the dyslipidemia guideline dropped. That speed tells you where this goes. Within a year, a meaningful share of prescribing decisions will be documented with a PREVENT score sitting in the note.
Here is the part that has not been worked through. The payer adjudicating the resulting claim usually cannot compute that score independently. The clinician and the actuary are about to be working from different instruments, and the guidelines just raised the stakes on the difference.
The gap is structural, not an oversight
PREVENT earns its discrimination from measured physiologic inputs. Total cholesterol, HDL, systolic blood pressure, body mass index, and estimated glomerular filtration rate form the base model, with HbA1c and urinary albumin-to-creatinine ratio added in the full version. These are values. Adjudicated claims carry codes and fills. A claim tells you that a patient has a diabetes diagnosis and a metformin prescription. It does not tell you the HbA1c, the eGFR, or the UACR that PREVENT actually consumes.
Strip the physiologic inputs out and you are no longer running PREVENT. You are running something closer to an HCC or DxCG cost model wearing a cardiology badge. The cardiovascular-risk literature already documents this boundary directly. A claims-built model for major adverse cardiovascular events in type 2 diabetes found that HbA1c, urinary albumin-to-creatinine ratio, and non-HDL cholesterol could not be assessed from insurance claims at all, which is why claims-only entities cannot reproduce a model that depends on them.
A claims-degraded PREVENT under the AHA brand would carry real validation liability. That is why one will probably never ship, and why payers should stop waiting for it.
This is not a knock on AHA. They built a point-of-care clinical tool on cohort data with measured inputs, and that is exactly what a treating clinician needs. A version retrofitted to run on claims would inherit the AHA name while losing the calibration that justifies it. No responsible standards body releases that.
The coding layer makes the gap more concrete still. Cardiovascular-kidney-metabolic syndrome, the organizing construct of the June 2026 guideline, is not a documentable ICD-10 diagnosis. The 2026 code set has no entry for it, and the nearest option, E88.810 for metabolic syndrome, is a narrower and distinct condition that does not capture CKM staging or its renal and cardiovascular integration. A plan working from claims cannot identify the CKM population at all, because the syndrome the guideline organizes everything around has no code to carry it.
The denominator is invisible before the risk score is ever in question. The guideline names a population that claims data has no way to count.
Why this matters more now than in 2023
When PREVENT was first published in Circulation in late 2023, it was a better risk equation that most plans could safely ignore in their coverage operations. The guideline cascade changed that. Treatment thresholds across three major guidelines now reference a specific PREVENT cutpoint, and the implementation guidance is explicit about it in the Hypertension companion paper.
The downstream consequence is an asymmetry that lands squarely in utilization management. Prescribers will increasingly justify therapy with a score the payer cannot independently recompute. Any UM or prior authorization criterion that references a PREVENT threshold becomes attestation-based by default. The plan is accepting the prescriber's stated score because it has no way to derive its own. Attestation without verification is the precise condition that breeds friction, inconsistent determinations, and appeals.
The lowered thresholds widen the exposure. Moving the statin indication from 7.5 percent to 5 percent enlarges the treatment-eligible population, which raises claim volume in exactly the category where the payer can least verify the clinical rationale. The guideline that expands the denominator also expands the surface area of unverifiable claims, functionally meaning guideline-based care that has an incomplete bridge to access and coverage.
Not every payer faces the same problem
The gap is not uniform. It depends entirely on what data a plan can reach. The distinction worth drawing is between plans that touch clinical data through provider relationships and plans that live on adjudicated claims alone.
| Payer archetype | Access to PREVENT inputs | Practical position |
|---|---|---|
| Integrated payer-provider system | Direct EHR and lab feeds, HIE participation, point-of-care vitals | Can run PREVENT today on a large share of attributed members. The data already exists inside the system. |
| Plan with value-based provider partnerships | Supplemental clinical data flowing through VBC contracts and quality programs | Can run PREVENT on the partnered population, with coverage that tracks contract penetration rather than full membership. |
| Claims-only commercial or MA book | Diagnosis codes and pharmacy fills, no native lab values or vitals | Cannot compute PREVENT for most members. This is where the structural gap is fully exposed. |
An IDN like Highmark-AHN, UPMC, or Kaiser Permanente can already assemble most PREVENT inputs from lab feeds, health information exchange access, and supplemental data tied to value-based contracts. For those organizations the question is operational sequencing rather than feasibility. The plans that are genuinely stuck are the claims-only books, and they hold a large share of commercial, Medicaid, and Medicare Advantage lives.
The realistic paths, and the catch in each
For claims-only plans, three paths can close part of the gap. None is clean, and the ranking depends on a plan's data partnerships and tolerance for building policy on unvalidated proxies. That calculation must also account for the data cleansing work semantic interoperability requires.
| Path | What it provides | The catch |
|---|---|---|
| License lab data feeds | Quest and Labcorp feeds carry a meaningful share of commercial lab values, supplying the cholesterol, HbA1c, and renal markers PREVENT needs | Coverage is partial and uneven across membership. Vitals such as systolic blood pressure and BMI still have to come from somewhere else. |
| Build an imputation model | Estimates missing PREVENT inputs from available claims, pharmacy, and demographic signal to produce a computable proxy score | An imputed PREVENT needs its own validation study before it can carry coverage policy. No such study has been published. |
| Require score documentation in PA | Shifts the computation to the prescriber and captures the score as a prior authorization data element | Reproduces the attestation problem. The plan records a number it still cannot verify, and adds provider burden in the process. |
The imputation path deserves a specific warning. Building coverage policy on an imputed PREVENT means building on a proxy whose calibration against ground-truth PREVENT is unknown. The closest adjacent evidence is the 2025 JACC evaluation of PREVENT across four academic health systems, which found moderate discrimination but calibration that varied widely across systems, demographics, and comorbidity profiles. If the equation itself drifts in calibration when applied to a real population it was not derived on, a claims-imputed approximation of it carries that risk and then compounds it. Anyone proposing imputed PREVENT for utilization management owes a validation study first. In the end, this rising clinical and financial risk is already in the payer’s member population, but invisible to the actuaries for benefit design improvements.
Interoperability is building the pipe, slowly
The claims-only position is a present-tense problem rather than a permanent one, because the data plumbing is being mandated into existence. The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) requires impacted payers to stand up four FHIR-based APIs by January 1, 2027, including a Provider Access API and a Payer-to-Payer API that move clinical and encounter data rather than claims alone. The regulatory trajectory is pointed directly at the kind of data PREVENT needs.
What rides on those rails is the part that matters here. The United States Core Data for Interoperability (USCDI) specifies the data classes those APIs must carry, and two of them map almost exactly onto PREVENT inputs. The Laboratory class includes the test and its result value, and the Vital Signs class includes systolic blood pressure and body mass index. The lipid panel, the HbA1c, the creatinine behind eGFR, the blood pressure, and the BMI that a plan cannot read off a claim are the same elements the federal interoperability stack is being built to exchange as structured FHIR.
This does not close the gap on its own, and treating the 2027 API surface as a finished answer would be a mistake. Exchanged data arrives at uneven completeness, much of it moves on member or provider request rather than as a standing feed, patient matching and latency remain real constraints, and an API that delivers a lab value still does not compute a PREVENT score or bind it to a coverage rule. The honest read is that FHIR and USCDI convert a structurally impossible task into an operationally hard one that improves over time. A payer strategy that designs for the 2027 API surface now will be positioned to compute PREVENT from exchanged clinical data as coverage thickens, instead of retrofitting under a deadline.
The white space
What does not exist yet is the bridge. There is no published, validated claims-plus-partial-lab adaptation of PREVENT, and there is no rigorous gap analysis of what payers can and cannot operationalize from the new prevention guidelines. That absence sits directly at the intersection of payer strategy and utilization management, and it is timely for a reason beyond the guideline cycle.
Quality measurement will hit the same plumbing. If NCQA embeds PREVENT-based measures into HEDIS or Stars, every plan suddenly needs the physiologic inputs or a defensible proxy to report against them. The same data dependency that complicates a coverage determination today becomes a measurement and revenue problem the moment it enters a quality program. A plan that solves the input problem for utilization management will have already solved it for quality reporting, while a plan that defers the work ends up facing both pressures in the same cycle.
There is a second move available, and it runs against the instinct to wait for a perfect instrument. ACC and AHA will not, and should not, degrade the clinical PREVENT tool to run on claims. They could instead partner with the actuarial community to co-develop a separate, clearly labeled claims-based companion model calibrated to approximate PREVENT at the population level. The goal is not to reproduce the point-of-care score patient by patient. It is to give plans a sanctioned, transparent approximation with published calibration bounds, governed by the same bodies that own the clinical standard, in place of a patchwork of proprietary vendor proxies built behind closed doors.
The value of that partnership is as much about trust as about math. A companion model developed jointly by the guideline authors and the actuaries who price the risk would carry credibility no unilateral payer model can earn. It would also document, in a structured way, exactly which inputs claims cannot supply, which is the specification the interoperability roadmap needs in order to prioritize what to move next. The most basic of those gaps is the missing diagnosis code itself. Advancing an ICD-10 code for CKM syndrome would let claims identify the population the guideline defines, which is the precondition for measuring or pricing it at all, and it is a concrete first step the sponsoring societies could take well ahead of any model work.
What makes the absence more striking is that a formal proposal does not appear to have been submitted. A review of the CDC ICD-10 Coordination and Maintenance Committee’s published meeting agendas going back to the original 2023 CKM guideline finds no listed proposal for a CKM syndrome diagnosis code. The clinical literature has engaged the problem. The sponsoring societies have not yet taken the concrete step of petitioning the CDC or WHO for a formal code assignment. Published commentary on the CKM coding gap exists, but the path from literature to code proposal has not been walked.
That step matters considerably more than it might appear. An ICD-10 code for CKM syndrome would give claims-based systems the ability to count the population the guideline defines, converting an invisible denominator into a measurable one. It is the administrative foundation on which quality measurement, benefit design, actuarial pricing, and value-based contracts have to be built before any of those functions can operate against CKM staging.
The cautionary case is CKD and heart failure. The administrative data ontology for both conditions lagged clinical reality for years, and claims-based measurement infrastructure still has not fully caught up. CKM deserves a cleaner path:
- ✓A formal code proposal submitted to the CDC or WHO
- ✓A population claims can identify
- ✓The actuarial visibility to tie the risk to value-based contract architecture
Capture it, measure it, tie it to VBC, and build the bridge of a specialist-informed, primary care-enabled system for clinical and financial impact.
Prevention and management should rest on the same foundation. The actuary pricing the risk and the clinician treating it should not be reading from different instruments.
Conclusion
The most useful move for a payer strategy team right now is to map its own position honestly. Which members can the plan compute PREVENT for today, through which data assets, and what is the realistic path to the rest. That map is the prerequisite for any UM policy that references a PREVENT threshold without generating avoidable appeals. The guidelines have set the clinical standard. Closing the gap between that standard and what a plan can operationalize is where the opportunity sits.
Frequently Asked Questions · Erik’s Hot Take
What is the PREVENT risk equation?
PREVENT (Predicting Risk of Cardiovascular Disease Events) is the AHA/ACC cardiovascular risk model first published in 2023 and now embedded as the shared threshold across three major 2025–2026 guidelines: the 2025 blood pressure guideline, the 2026 dyslipidemia guideline, and the first-ever CKM guideline. It uses measured physiologic inputs including total cholesterol, HDL, systolic blood pressure, body mass index, and eGFR, with HbA1c and urinary albumin-to-creatinine ratio in the full version. PREVENT is no longer one tool among several. It is the shared risk denominator across cardiovascular prevention.
Why can’t payers compute PREVENT from claims data?
PREVENT requires numeric lab values and vitals that do not exist in adjudicated claims. Total cholesterol, HbA1c, eGFR, systolic blood pressure, and BMI are the inputs the equation actually consumes. Claims carry diagnosis codes and pharmacy fills, not measured numeric values. Strip the physiologic inputs and you are running a cost model, not PREVENT. The gap is structural, not a software problem.
Does CKM syndrome have an ICD-10 diagnosis code?
No. As of the 2026 ICD-10 code set, cardiovascular-kidney-metabolic syndrome has no assigned diagnosis code. The nearest available option, E88.810 for metabolic syndrome, is a narrower and distinct condition that does not capture CKM staging or its renal and cardiovascular integration. A plan working from claims cannot identify the CKM population the June 2026 guideline organizes everything around, because the syndrome itself has no code to carry it.
What can claims-only payers do to operationalize PREVENT now?
Three paths exist, each with a meaningful catch. Licensing lab feeds from Quest or Labcorp supplies most lab values PREVENT needs, but vitals coverage remains incomplete. Building an imputation model estimates missing inputs from available claims and pharmacy signal, but requires a published validation study before it can support coverage policy, and no such study exists. Requiring PREVENT score documentation in prior authorization shifts computation to the prescriber but reproduces the attestation problem. The right first step is to map which members you can compute PREVENT for today and build a realistic path to the rest.
How does the CMS interoperability mandate affect this?
The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) requires impacted payers to stand up FHIR-based APIs by January 1, 2027, including a Provider Access API and a Payer-to-Payer API. The United States Core Data for Interoperability standard specifies Laboratory and Vital Signs data classes those APIs must carry, which map almost exactly onto PREVENT inputs. FHIR and USCDI convert a structurally impossible task into an operationally hard one that improves over time. Design for the 2027 API surface now, not under deadline.
What happens if NCQA embeds PREVENT in HEDIS or Stars?
Plans that have not solved the PREVENT data infrastructure problem for utilization management will face the same gap as a measurement and revenue problem simultaneously. The data dependency that complicates a coverage determination today becomes a quality program problem the moment it enters HEDIS or Stars. Plans that solve the input problem for UM will have already solved it for quality reporting, while plans that defer the work end up facing both pressures in the same performance cycle.
Working this problem inside a plan or a vendor?
OneAnother Health works at the seam between clinical guidelines and payer operations, mapping what the new prevention standards mean for utilization management, quality measurement, and the data assets required to run them.
Schedule a call