Managing Risk Adjustment Coding

Documents Required

• Complete medical record for encounters within the data collection period
• Patient problem list and medication list
• Prior year's HCC profile and RAF score
• CMS-HCC model coefficient tables (current payment year version)
• ICD-10-CM to HCC crosswalk (current model version)
• MEAT documentation criteria reference
• Provider attestation templates (if applicable for chart reviews)

Step 1 — Understand the CMS-HCC Model Structure

Map the relationship between diagnosis codes, HCCs, and RAF scores.

• ICD-10-CM codes map to HCCs through the CMS crosswalk. Not all ICD-10-CM codes map to HCCs — only those indicating chronic, significant, or predictive conditions.
• CMS-HCC V28 (phased in starting 2024): Reduced from 86 to 115 HCCs with recalibrated coefficients. New model increases specificity requirements for many conditions.
• CMS-HCC V24 (being phased out): The legacy model used through 2023. During transition, CMS blends V24 and V28 scores (67% V24 / 33% V28 for 2024, shifting annually).
• HCCs are hierarchical — when a patient has multiple related conditions, only the highest-severity HCC in each hierarchy counts. Example: HCC 17 (Diabetes with Acute Complications) supersedes HCC 18 (Diabetes with Chronic Complications) which supersedes HCC 19 (Diabetes without Complications).
• Disease interaction factors: Certain HCC combinations generate additional RAF value (e.g., diabetes + CHF, HCC 85 + HCC 18).
• Demographics (age, sex, dual-eligible status, institutional/community status) generate base RAF components independent of diagnosis codes.

Step 2 — Validate Documentation with MEAT Criteria

Every HCC-eligible condition must be supported by MEAT in the encounter note.

• M — Monitor: The provider documents monitoring activities — reviewing labs, assessing symptoms, tracking progression. "Patient's HbA1c reviewed at 7.8%, up from 7.2% last visit."
• E — Evaluate: The provider documents assessment of the condition's status. "CHF is currently compensated, no peripheral edema, weight stable."
• A — Assess: The provider documents clinical judgment about the condition. "COPD with acute exacerbation, likely triggered by viral URI."
• T — Treat: The provider documents treatment or management. "Continue metformin 1000mg BID, add glipizide 5mg daily for worsening glucose control."

MEAT validation rules:

• At least ONE MEAT element must be present in the encounter for the condition to be coded.
• A problem list entry alone does NOT satisfy MEAT — the condition must be actively addressed in the encounter note.
• Historical conditions (e.g., "history of stroke") should be coded only if the residual condition is still being managed.
• Each HCC-eligible condition must be recaptured annually — HCCs do not carry forward from year to year.

Step 3 — Map ICD-10-CM Codes to HCCs

Select the most specific code that maps to the highest appropriate HCC.

• Use the current year's ICD-10-CM to HCC crosswalk — mappings change with each model year.
• Specificity matters: E11.9 (Type 2 diabetes without complications) maps to HCC 19 (RAF ~0.105), while E11.65 (Type 2 diabetes with hyperglycemia) maps to HCC 18 (RAF ~0.302). Ensure documentation supports the more specific code.
• Common high-value HCCs: Major depressive disorder (HCC 59), Vascular disease (HCC 108), COPD (HCC 111), CHF (HCC 85), CKD Stage 4 (HCC 138), Morbid obesity (HCC 48 in V28).
• V28 changes to watch: Dementia (reclassified), Substance use disorders (new HCCs), Obesity (now mapped), Depression (recalibrated).
• Never assign a more specific code than documentation supports — "coding to the highest HCC" without documentation is fraud.
• Conditions that are resolved, inactive, or not being managed should not be coded.

Step 4 — Analyze RAF Score Impact

Calculate the financial impact of HCC capture decisions.

• RAF score = sum of all applicable HCC coefficients + demographic factors + disease interaction factors.
• The payment is: (RAF score × county base rate × plan-specific adjustment) = monthly capitation.
• Each 0.100 RAF increment represents approximately $1,000–$1,200 in annual revenue per member (varies by county).
• Calculate the marginal RAF impact of capturing vs. not capturing each HCC — this drives prioritization for chart reviews.
• Track suspecting vs. confirming: "suspected" HCC gaps (conditions from prior year not yet recaptured) should be prioritized for provider education.
• Monitor recapture rates by HCC category — target >85% annual recapture for chronic conditions.

Step 5 — Conduct Retrospective Chart Reviews

Audit coded encounters for missed HCC opportunities and overcoded conditions.

• Pull all encounter records for the member during the data collection period.
• Compare coded diagnoses against the medical record documentation.
• Identify missed HCC-eligible conditions that are documented and meet MEAT but were not coded.
• Identify overcoded conditions where HCCs are assigned without adequate MEAT documentation.
• For missed conditions, determine whether an addendum or retrospective chart review attestation is compliant under the plan's policies.
• For overcoded conditions, submit deletions to avoid RADV audit findings.
• Document all chart review findings with specific record citations.

Step 6 — Prepare for RADV Audit Defense

Ensure documentation withstands CMS audit scrutiny.

• RADV auditors require the original medical record to support every HCC submitted for the member.
• Each diagnosis must be supported by a face-to-face encounter with an eligible provider type (MD, DO, NP, PA — not all provider types qualify).
• Acceptable documentation sources: physician office notes, hospital records, diagnostic reports interpreted by a physician. Lab results alone without physician interpretation do not support HCC coding.
• The encounter must occur within the data collection period for the payment year.
• Provider signature (electronic or physical) and credentials must be present on the encounter note.
• Ensure all chart review addenda follow CMS guidelines — addenda must be dated, signed, and reference the original encounter.

Checkpoint B — Review

• Every HCC-eligible code is supported by MEAT documentation in the encounter note
• ICD-10-CM codes are assigned at the highest specificity supported by documentation
• HCC hierarchy rules are correctly applied — only the highest HCC per hierarchy is counted
• Disease interaction factors are identified for qualifying HCC combinations
• Annual recapture is verified for all chronic conditions from the prior payment year
• RAF score calculation is accurate with current model coefficients
• No conditions are coded based solely on problem list entries, medication lists, or lab results without provider assessment
• Chart review findings distinguish between missed HCCs and overcoded HCCs

Quality Audit

• MEAT criteria validation is documented for every HCC code submitted
• Recapture rates for chronic HCCs are tracked by provider and condition category
• RAF score trends per member are monitored for unexpected spikes (potential overcoding) or drops (undercoding)
• Chart review findings are reconciled with corrective action — deletions submitted for unsupported codes, additions submitted for documented but uncaptured conditions
• Provider education is targeted based on HCC capture gaps by condition type
• RADV-readiness assessments are conducted on a sample of members annually
• V24 to V28 transition impact is tracked — identify conditions that gain or lose HCC status under the new model

Guidelines

• Follow CMS Risk Adjustment Data Technical Assistance for Medicare Advantage Organizations document (current year)
• Apply ICD-10-CM Official Guidelines for Coding and Reporting — risk adjustment coding must follow the same rules as all other coding
• Reference the CMS-HCC Risk Adjustment Model software and crosswalk files for current mappings
• Follow AHIMA Practice Brief on Risk Adjustment Coding for documentation and compliance standards
• Never code a condition for HCC capture that is not documented and managed in the encounter — this is a False Claims Act violation
• Never prompt or lead providers to document conditions that do not exist — CDI queries in risk adjustment must be non-leading
• Mark with [VERIFY] any HCC assignment where MEAT documentation is marginal
• Include disclaimer that risk adjustment coding must be documentation-supported and does not constitute guidance to maximize revenue beyond what clinical evidence supports