Searching for the best HCC coding software for providers usually produces feature lists, vendor rankings, and broad claims about AI. Those comparisons often miss the most important question: what does provider-focused HCC software need to do under the CMS rules taking effect for CY 2027?
The answer is changing.
CMS will continue using the 2024 CMS-HCC model, commonly called V28, for CY 2027. However, CMS is also tightening which diagnosis sources count toward risk scores. Diagnoses from most unlinked chart review records and certain audio-only encounters will begin to be excluded during the CY 2027 Midyear model run. That run will affect payments for July through December 2027 and may also trigger retroactive adjustments to payments already made for January through June. The records being created during 2026 therefore matter directly to 2027 payment. CMS’s CY 2027 Risk Adjustment Implementation Information explains the implementation timeline in detail.
For providers, this reinforces a clear direction: valid HCCs need to be identified, assessed, documented, and supported inside qualifying clinical encounters. Software that only searches for codes after the visit will not be enough.
The best HCC coding software for providers captures clinically supported diagnoses during qualifying patient encounters, integrates directly into the EHR, and helps clinicians document the specificity required under V28. It should also validate diagnosis support before submission, reduce dependence on retrospective chart review, and improve risk adjustment accuracy without increasing provider burden.
This is not simply a race to identify more possible codes. The software must help providers capture the right conditions, with the right clinical evidence, through the right encounter.
A complete provider workflow should support:
Provider HCC software should address four major pressures at once: continued operation under V28, the exclusion of unlinked chart review diagnoses, the exclusion of certain audio-only diagnoses, and a significantly expanded RADV audit environment.
In the 2027 Medicare Advantage and Part D Rate Announcement, CMS finalized continued use of the 2024 CMS-HCC model for CY 2027 rather than moving forward with the newly proposed Part C model. The agency said this would give the market more time to adjust following the model’s completed phase-in in 2026.
That does not mean providers can return to older documentation habits. V28 remains fully implemented, and organizations still need software that understands its condition categories, hierarchy logic, specificity requirements, and current ICD-10 mappings.
Provider teams should be monitoring more than final RAF results. They should be evaluating first-pass capture, specificity, unsupported diagnosis rates, recapture performance, and dependence on retrospective cleanup. The V28 Mid-Year Checkpoint explains which operational metrics can reveal whether a workflow is actually performing under the model.
Beginning with the CY 2027 Midyear model run, CMS will exclude diagnoses from unlinked chart review records for most non-PACE organizations. A limited exception applies when a beneficiary moves between Medicare Advantage contracts under different parent organizations.
Providers may not control how plans submit chart review records, but they do control where the underlying documentation originates. If valid chronic conditions are recognized and documented during the patient encounter, plans are less dependent on disconnected retrospective recovery.
This makes prospective and point-of-care capture more valuable. Suspected conditions should be surfaced before or during the visit, reviewed through clinical judgment, and connected to documentation showing what the provider assessed or managed.
For a more detailed explanation of the timing and payment implications, see CMS’s CY 2027 Risk Adjustment Playbook.
CMS will also exclude diagnoses from certain audio-only encounters identified with modifier 93 or FQ beginning with the CY 2027 Midyear model run. The exclusion applies when all risk-adjustment-eligible service lines associated with the record carry one of those modifiers.
Provider software should therefore understand more than diagnosis codes. It should be able to evaluate the encounter type and source associated with the diagnosis.
A clinically relevant condition does not automatically qualify for risk adjustment. The diagnosis must come from an eligible service and be supported through the required documentation.
CMS has announced plans to audit all eligible Medicare Advantage contracts in newly initiated payment-year audits. CMS also said it would increase review samples from 35 records to between 35 and 200 records per contract, depending on plan size, and use technology to identify diagnoses that may lack medical-record support.
That means providers need more than a code-suspecting engine. They need software that can explain:
A platform that only adds diagnoses may increase audit exposure. The best software should also identify conditions that should not be submitted.
Providers need different HCC software because they create the clinical record on which risk adjustment ultimately depends.
A health plan generally receives diagnosis and encounter data after care has occurred. Its workflow focuses on validation, submission, analytics, and oversight. Providers must recognize the condition, assess it, document the clinical reasoning, and code it while managing the patient’s immediate care needs.
That creates distinct requirements.
Provider-focused HCC software must:
The software should translate regulatory and coding complexity into a small number of clinically relevant actions. It should not transfer that complexity to the provider.
The software should use structured and unstructured data to identify conditions that may be relevant to the current encounter.
A historical claim or problem-list entry alone should not automatically produce an alert. The system should evaluate notes, medications, laboratory results, prior diagnoses, external records, and the current clinical context before surfacing a suggestion.
Providers should also be able to see why the condition was suggested. Black-box recommendations weaken trust and encourage providers to dismiss alerts.
For more on closing documentation gaps while the provider still has clinical context, read Closing HCC Coding Gaps at the Point of Care.
The platform should help clinicians recognize when a diagnosis needs additional specificity around severity, staging, complications, or relationships between conditions.
This matters particularly for:
The goal is not to turn clinicians into coders. It is to show them the documentation detail needed to accurately reflect the patient’s condition while that information is still available.
EHR integration should be a threshold requirement, not an optional enhancement.
The software should ingest patient data automatically, surface relevant information inside the existing workflow, and allow accepted conditions or documentation to flow into the appropriate part of the medical record.
Separate portals introduce context switching, additional logins, duplicate work, and adoption risk. The best system should feel like part of the EHR rather than another task layered on top of it.
This is especially important when the objective is to reduce documentation burden while improving coding accuracy.
No single stage of the workflow will catch every condition.
A strong platform should support the full coding lifecycle:
Pre-visit: Identify clinically relevant conditions before the appointment and prepare a focused suspect list.
Concurrent: Help the provider assess and document the condition during the encounter.
Post-visit: Identify remaining gaps while the encounter is still current and connected to the original visit.
Pre-submission validation: Confirm that final diagnoses are supported before they become payment or audit risk.
Retrospective review remains useful as a safety net. It should not remain the primary method by which the organization finds valid conditions.
Finding a possible HCC is only half of the process. The platform should also determine whether the record supports the diagnosis.
The software should identify evidence that the condition was monitored, evaluated, assessed or addressed, or treated. It should connect the diagnosis to the relevant documentation and make that support easy for coding and compliance teams to review.
That traceability is increasingly important as CMS expands the scale and speed of RADV audits. A diagnosis that cannot be defended creates more risk than value.
The CY 2027 Medicare Advantage and Part D final rule removes 11 administrative or low-variation Star Ratings measures and adds a new Depression Screening and Follow-Up measure beginning with the 2027 measurement year for 2029 Stars. CMS said the changes refocus Stars on clinical care, outcomes, and patient experience.
For providers, that means quality and risk adjustment increasingly depend on the same encounter.
The software should help clinicians see relevant HCC opportunities and care gaps through a coordinated workflow rather than separate, competing alert systems. Diabetes documentation, laboratory monitoring, screening status, medication management, and follow-up planning may all need attention during the same visit.
Use these questions during vendor evaluation:
Inferscience connects the major functions provider organizations need as CMS moves risk adjustment closer to the documented encounter.
HCC Assistant analyzes structured and unstructured chart data and surfaces clinically supported HCC opportunities directly in the EHR. Providers can review the underlying evidence, determine whether the diagnosis applies, and add appropriate conditions to the assessment and plan.
AI Chart Assistant synthesizes diagnoses, medications, laboratories, imaging, clinical notes, and external records into one patient view. This reduces the time clinicians spend searching fragmented records before they can assess the full clinical picture.
Quality Assistant brings care-gap information into the provider workflow so clinicians can review and address outstanding quality needs without switching applications.
HCC Validator reviews diagnoses against supporting documentation, identifies MEAT deficiencies, flags high-risk codes, and creates an evidence trail for RADV preparedness.
Together, these capabilities support the full workflow: understand the patient, identify clinically relevant conditions, document them during the encounter, address related quality needs, and validate the resulting diagnosis before submission.
The best HCC coding software should help provider organizations achieve:
The most important result is not the number of suspects the software generates. It is whether accurate, supported HCC capture becomes a natural output of the provider’s existing workflow.
The best HCC coding software for providers captures clinically supported diagnoses during qualifying encounters, integrates directly into the EHR, supports V28 specificity and annual recapture, and validates documentation before submission. It should improve risk adjustment accuracy without adding a separate workflow or increasing provider burden.
Yes. CMS will continue using the 2024 CMS-HCC model, commonly called V28, for CY 2027 rather than implementing the newly proposed Part C model. Providers still need to meet the specificity, hierarchy, and documentation requirements of the fully implemented V28 model.
Beginning with the CY 2027 Midyear model run, diagnoses from most unlinked chart review records will no longer count toward risk scores. Providers therefore need workflows that identify, assess, and document valid conditions within qualifying patient encounters instead of depending on disconnected retrospective capture.
CMS will exclude diagnoses from certain audio-only encounters identified with modifier 93 or FQ beginning with the CY 2027 Midyear model run. Provider organizations need systems that can identify the encounter source and ensure diagnoses are captured through qualifying services.
HCC coding software supports RADV readiness by connecting each diagnosis to its supporting medical-record evidence, identifying unsupported conditions, and validating documentation before submission. The strongest systems also preserve a traceable evidence trail for internal review and audit response.
The best HCC coding software for providers in CY 2027 is not the platform that produces the longest suspect list. It is the one that helps clinicians create accurate, encounter-linked, defensible documentation without slowing care delivery.
CMS is retaining V28, but it is tightening the connection between diagnosis capture and qualifying encounters. Unlinked chart review and audio-only exclusions will affect Midyear risk scores. RADV review is expanding. Star Ratings performance is moving further into clinical workflows.
Providers need software that can operate across all those pressures at once.
That means native EHR integration, evidence-backed suspecting, point-of-care specificity support, coordinated quality workflows, and pre-submission validation. Anything less leaves too much dependent on retrospective recovery at the exact moment CMS is making that strategy less reliable.
Contact Inferscience to see how HCC Assistant, AI Chart Assistant, Quality Assistant, and HCC Validator can support provider risk adjustment workflows for CMS CY 2027.
