CMS proposes to exclude unlinked diagnoses from Medicare Advantage risk adjustment: An opportunity for plan-provider value-based collaboration?

The Centers for Medicare & Medicaid Services (CMS) published its “Advance Notice of Methodological Changes for Calendar Year (CY) 2027 for Medicare Advantage (MA) Capitation Rates and Part C and Part D Payment Policies” on January 26, 2026, which included proposed changes related to capitation rates and payment policies.

If finalized, one such change would exclude diagnoses from “unlinked” chart review records (CRRs) from risk score calculations. This change may result in the elimination of the only source of diagnoses for certain serious and chronic medical conditions for Medicare beneficiaries enrolled in MA plans. Beyond adversely impacting MA organizations’ risk-adjusted payments from CMS, the change could have unintended consequences for health care providers and other stakeholders providing care management and care coordination support and related technology to this population.

Below, we provide background about MA risk-adjusted payments and address the potential impact of excluding unlinked CRRs when calculating those payments.

What are MA “risk-adjusted” payments?

MA organizations receive monthly capitated payments from CMS for each Medicare beneficiary that enrolls in a plan offered by the applicable MA organization. Certain enrollees may have conditions that would result in higher costs than the average enrollee. CMS will review enrollees’ demographic data and diagnoses from existing records to determine any additional risk and adjust the capitated payment to a higher amount to capture that risk. This risk adjustment process means that an MA organization will receive higher capitated payments for enrollees who have diagnoses suggesting that they are less healthy than the average risk profile.

How does CMS receive data to make risk adjustments?

Pursuant to 42 CFR 422.310(d), CMS requires MA organizations to submit data for risk adjustment purposes. Data is provided to CMS in two formats: (1) encounter data and (2) an abbreviated format called Risk Adjustment Processing System (RAPS) data. Each format is submitted via different CMS systems.

Regarding the former, MA organizations submit encounter data records (EDRs) and chart review records (CRRs) through the Encounter Data System (EDS). EDRs are populated and structured in accordance with strict format and content rules, which include that the only diagnosis codes that may be included in an EDR are those codes that are documented in a health care provider’s medical record as a result of a face-to-face visit during the data collection period for the payment year.

A CCR gives MA organizations an opportunity to review medical records to add additional diagnosis codes that are eligible for risk adjustment (or to delete codes). Added codes must be derived from a face-to-face visit and supported by a medical record. There are two types of CRRs: linked and unlinked. As its name suggests, a “linked” CRR is connected to a previously submitted and accepted EDR. An “unlinked” CRR does not identify a previously submitted EDR.

Both RAPS data and encounter data are processed by CMS to calculate risk scores. CMS uses a blend of the data with encounter data accounting for 25 percent of risk scores and RAPS data making up the remaining 75 percent.

CMS will retroactively audit risk-adjusted payments via risk adjustment data validation (RADV) audits. The results of an RADV audit could result in an MA organization owing repayments to CMS.

Why is CMS proposing to exclude unlinked CCRs from risk adjustment calculations?

Unlinked CRRs were the subject of a report from the United States Department of Health and Human Services, Office of Inspector General (OIG) in 2019. The OIG raised concerns about the use of unlinked CRRs based on its findings that (1) diagnoses reported only on unlinked CRRs corresponded to “some serious and chronic health conditions,” despite no service records containing those diagnoses, which could document what services were provided for the medical conditions indicated by the diagnoses, and (2) almost a third of MA organizations has estimated risk-adjusted payments “driven only by” unlinked CRRs. The OIG also has a number of ongoing audits of MA organizations with respect to unlinked CCR diagnosis codes with a completion date targeting fiscal year 2026.

In its response to the report, CMS challenged certain aspects of the OIG’s methodology. For example, the OIG identified that certain diagnoses were reported on CRRs but not in EDRs. Among the explanations for the discrepancy, CMS pointed out that the EDR might be incomplete (e.g., due to all diagnoses in the medical record not being reported). Likewise, CMS pointed out significant errors in how the OIG calculated the potential overpayments to MA organizations. According to CMS’s analysis, the OIG failed to apply the blending methodology used by CMS for RAPS and encounter data. As a result, CMS notes that the OIG’s assumption regarding the estimated $6.7 billion in risk-adjusted payments may be significantly overstated. By way of example, whereas the OIG calculated $2.7 billion in payments based on unlinked CCRs, CMS estimates the actual value of the payments as approximately $675 million.

Despite disagreement regarding the OIG’s conclusions and methodology, CMS concurred with each OIG recommendation resulting from the report, one of which was for CMS to reassess the risks and benefits of allowing unlinked CRRs to be used as sources of diagnoses for risk adjustment.

What change is CMS proposing with respect to unlinked CRRs and why?

CMS is proposing to exclude all diagnoses submitted on unlinked CRRs from the risk score calculation. If finalized, CMS would still permit MA organizations to submit unlinked CRRs, but it would not account for them in the calculation.

CMS offers two reasons for this change: (1) it would address “program integrity concerns” since unlinked CCRs are not directly linked to an EDR and, with the change, the calculation would only account for reports that link diagnoses to specific EDRs, and (2) it may reduce differences in risk adjustment payment resulting from differences in differential coding across MA organizations.

The change would be effective for risk score calculations starting in CY 2027.

How could the exclusion of unlinked CRRs adversely impact health care providers while presenting an opportunity for value-based cooperation among plans and providers?

While this proposal may appear to only impact MA organizations, it may adversely impact health care providers as well. Under the current model, MA organizations can identify and report diagnosis codes that are not captured in a medical record, an EDR, or otherwise tied to a specific encounter. Instead, the additional codes may reflect an assessment of an enrollee across medical records to which no single billing provider had access. As noted by the OIG, these codes tended to reflect “serious and chronic” conditions. While the OIG and CMS both raised program integrity concerns, neither suggested that the codes were necessarily improper. Excluding unlinked CCRs does not mean that these same enrollees do not in fact have those conditions. Rather, it means that the EDRs and linked CRRs submitted to CMS for risk adjustment might fail to reflect those conditions because the diagnoses were not documented in the medical record or able to be linked to a particular EDR.

As a result, MA organizations will still face the financial risk presented by these enrollees, but without any risk adjustment to their capitated payments. That financial pressure can adversely impact providers as MA organizations implement cost containment strategies to manage the unfunded risk, such as through utilization management, narrower networks, and adjustments to benefit designs.

Under such circumstances, both MA organizations and providers may have a mutual interest to implement value-based arrangements (e.g., care management and care coordination efforts) to continue identifying and managing this higher-cost and at-risk population. From the perspective of the MA organization, these arrangements can help surface and record otherwise undocumented serious and chronic conditions in the medical records while actively engaging providers and the enrollees in initiatives designed to manage those conditions. Once documented in the medical record, the diagnoses would appear in the EDR itself and no additional CRRs would be required to capture them.

From the provider perspective, improving medical record documentation in conjunction with a value-based arrangement can produce higher-quality care and improved patient outcomes while offsetting the provider’s costs in participating in the value-based arrangement. Overall, plan provider collaboration in identifying and documenting diagnoses of serious and chronic conditions in the medical record through value-based arrangements may lead to better risk adjustment reporting by the MA organization, allow for support for provider efforts to manage patient care, and improve health outcomes for the patient.

Any effort to improve the identification and documentation of diagnoses must be carefully structured to comply with applicable law and to mitigate risk in light of the significant governmental scrutiny and various enforcement actions against MA organizations and their vendors, including, in some cases, with respect to participating health care providers.

Do Medicare’s remote chronic care management (CCM) services offer any solution for identifying gaps in documented diagnoses?

Not directly. For CCM services billed under CPT codes 99490/99439/99491, including complex CPT codes 99487/99489, the services are by definition non-face-to-face and do not include audio-video telehealth services. So, documentation in medical records related to these services would not generate information for inclusion in EDRs. However, to the extent that a Medicare beneficiary has serious or chronic conditions that are not already known and subject to the CCM services being provided, these services may help identify information that can be used during in-person or audio-video telehealth visits to surface whether a beneficiary has additional serious or chronic conditions that should be documented as part of such visits.

Conclusion

Whether the proposal will be finalized remains to be seen. As discussed above, the OIG has been critical about unlinked CRRs in the past and views them as a risk to the integrity of the Medicare program. However, CMS has acknowledged not only the benefits that unlinked CRRs can provide in identifying undocumented diagnoses and thereby improving risk score calculations. It has also acknowledged the operational and administrative difficulties for MA organizations related to linking every CRR to an EDR. Feedback from stakeholders over the next month may impact whether CMS finalizes this proposal. CMS is accepting comments on this proposal and the others outlined in the Advance Notice through 11:59 PM EST on Wednesday, February 25, 2026.

For more information, please contact the authors.