This is the streamlined podcast version of the blog post, “Retrospective Risk Adjustment: How Health Plans Build Defensible Coding Programs in 2026.” We’ve shaped it into a simple, story-driven experience to help you grasp the most important points with ease.
In January 2026, CMS confirmed that federal estimates point to approximately $17 billion in overpayments annually from unsupported diagnosis data submitted by MA organizations [1]. That single figure explains why retrospective risk adjustment has shifted from a revenue function to a compliance discipline. For health plans, coding teams, and healthcare organizations managing risk adjustment programs, defensible accuracy is no longer optional.
This guide explains how retrospective risk adjustment works in practice, why it matters under current regulatory enforcement, and how AI-powered clinical tools are transforming the coding process for Medicare Advantage plans, navigating value-based care.
What Is Retrospective Risk Adjustment?
Retrospective risk adjustment is the systematic review of medical records after patient encounters to identify, validate, and correct diagnosis codes that reflect a member’s true health conditions. Coding teams examine clinical documentation, lab results, specialist consultations, and medication lists to find chronic conditions that were missed or miscoded during initial claims processing.
Every validated HCC code must be supported by MEAT documentation: evidence that the condition was Monitored, Evaluated, Assessed, or Treated during a qualified face-to-face encounter [2]. This process ensures proper reimbursement from the Centers for Medicare and Medicaid Services while maintaining coding accuracy across the patient population.
Unlike prospective risk adjustment, which captures diagnoses during the encounter itself, retrospective reviews analyze completed charts to close gaps in documentation. Both approaches serve the same goal: ensuring that risk adjustment data accurately reflects patient complexity and health conditions. The difference is timing and method.
Autonomous Retrospective Risk Adjustment Solution
One platform. Every HCC validated. Revenue secured.
Why Is Retrospective Risk Adjustment Critical for Health Plans in 2026?
Three forces are converging to make retrospective risk programs a compliance priority rather than a revenue function. The financial impact of getting risk adjustment wrong has never been higher.
- Regulatory enforcement is accelerating: CMS confirmed in its January 2026 RADV memo that Payment Year 2020 audits will begin by February 2026, with variable sample sizes of 35 to 200 enrollees per contract [1]. Completed audits for Payment Years 2011 through 2013 found overpayment rates between five and eight percent, and CMS stated recoveries “will begin soon.” CMS will also introduce an AI-powered medical coder support tool to streamline coding reviews, though all final decisions remain with human certified medical coders [1].
- OIG compliance guidance raises the bar: The February 2026 OIG MA Industry Compliance Program Guidance identified specific risk adjustment practices that have drawn federal investigations [2]. These include using chart reviews solely to add diagnoses without removing unsupported codes, conducting health risk assessments that generate diagnoses not considered in patient care, and prompting physicians through electronic health record platforms to add risk-adjusting diagnoses that patients did not have [2]. Physician engagement in compliance education is now essential to avoid these patterns.
- Financial performance depends on risk adjustment accuracy, not volume: The OIG found that chart reviews accounted for an estimated $6.7 billion in MA payments for 2017, with over 99 percent of those reviews adding diagnoses rather than deleting them [3]. This pattern of one-way code addition is now a regulatory red flag. Health plans that want financial sustainability must prove that their retrospective programs produce accurate documentation, not inflated risk scores. Financial success depends on defensible coding, not volume. Patient outcomes suffer when the focus shifts from care quality to revenue capture.
How Does the Retrospective Risk Adjustment Process Work?
A compliance-first retrospective risk adjustment process follows structured steps that build defensible evidence trails for every HCC code.
Step 1: Strategic Chart Selection
Rather than reviewing all patient records, risk adjustment coders use predictive analytics to identify high-value charts. Priority targets include members with multiple chronic conditions, patients with medications that suggest undocumented patient conditions, and members whose risk adjustment factor scores dropped unexpectedly. This prospective approach to chart selection maximizes efficient review while focusing coding teams on charts with the highest likelihood of missed diagnoses.
Step 2: Clinical Documentation Review and HCC Validation
Coders analyze medical records, lab results, imaging reports, and unstructured clinical data to identify relevant diagnoses. Each suspected condition must map to a valid hierarchical condition category code at the highest ICD-10-CM specificity. Under the V28 model, now in full effect for 2026, the number of HCC codes has changed significantly, requiring updated training for coding teams and medical coders [4]. Accurate HCC capture depends on coders who understand the clinical context behind each code.
Step 3: MEAT Evidence Capture and Two-Way Coding
This is where compliance-first programs diverge from legacy approaches. Every diagnosis must link to MEAT evidence from a qualified encounter. Equally important, the review must identify and flag unsupported codes for deletion, not just find correct codes to add. The OIG specifically noted that programs designed only to add diagnoses create payment integrity concerns [2]. Proper documentation accuracy requires both accurate risk capture (adds) and removal of unsupported HCC codes (deletes).
Step 4: Quality Assurance and Claim Submission
Multi-level quality checks validate that every HCC code has defensible evidence before data reaches CMS. This includes verifying face-to-face encounter compliance, confirming annual documentation requirements, and ensuring proper documentation of patient health status for each chronic condition. When patient care involves multiple providers, the review must reconcile documentation across all patient encounters to confirm that each patient visit generated appropriate clinical evidence. Care quality controls at this stage prevent critical diagnoses from being submitted without sufficient support.
Industry First RADV Audit Solution
AI-powered solution enables health plans to efficiently manage and streamline RADV audits
What Is the Difference Between Prospective and Retrospective Risk Adjustment?
Healthcare organizations use both prospective risk adjustment and retrospective approaches as part of a complete risk adjustment strategy in value-based care. Understanding the difference is essential for healthcare providers and health plans managing financial performance and patient health outcomes.
- Prospective Risk Adjustment happens during or before the patient visit. Pre-visit planning identifies suspected conditions based on patient records, claims history, and care quality data. During the encounter, clinical decision support prompts providers to assess and document relevant codes. Prospective risk adjustment captures diagnoses at the point of care, which CMS considers the safest method for risk capture. A strong prospective coding workflow reduces downstream retrospective needs.
- Retrospective Risk Adjustment occurs after the encounter. Coding teams review completed charts to find conditions that were present but not coded. Retrospective reviews catch what prospective risk adjustment missed and, when done correctly, also remove codes that lack adequate support. Prospective reviews and concurrent reviews complement this work by catching issues earlier in the cycle.
- Concurrent Risk Adjustment sits between the two, reviewing documentation during or shortly after an episode of care. Concurrent risk adjustment allows healthcare organizations to address issues faster than traditional retrospective approaches while capturing encounters that the prospective approach may have missed.
The most effective risk adjustment programs integrate all three approaches. Prospective risk adjustment sets the foundation during visits. Concurrent reviews provide a real-time safety net. Retrospective reviews serve as the final quality layer, ensuring every HCC code submitted to CMS reflects patient health accurately and is properly documented. This integrated model delivers the strongest compliance outcomes and supports long-term financial performance.
How AI-Powered Technology Is Transforming Retrospective Reviews
The limitations of manual chart review are well documented. Complex reviews take 30 to 45 minutes each. Accuracy drops after extended sessions. Coding teams face burnout from repetitive work on unstructured clinical data, and improving accuracy through manual processes alone is no longer sustainable.
AI-powered clinical tools are changing this equation. Neuro-Symbolic AI, which combines deep learning with structured clinical reasoning, addresses three failures that limited traditional technology.
First, it understands clinical context. Neuro-Symbolic AI recognizes that a metformin prescription combined with an elevated A1C indicates active diabetes management. It distinguishes between “family history of diabetes” and “diabetes with peripheral neuropathy,” a distinction that directly affects HCC codes and risk scoring.
Second, it provides explainable results. Every suggested diagnosis links to specific evidence in the clinical record, creating a transparent audit trail. When paired with seamless EHR integration, AI-powered risk adjustment tools connect directly to existing workflows, reducing administrative burden on coding teams and healthcare providers.
Third, it supports two-way coding. AI-powered risk adjustment tools should flag unsupported HCC codes for removal with the same rigor with which they identify missed diagnoses. This reflects the compliance outcomes regulators expect: accurate data, not maximized risk scores.
Organizations deploying this technology report chart review times under 10 minutes, first-pass accuracy above 95 percent, and 60 to 80 percent productivity improvements for coding teams. The financial benefits are significant, but the real value is building programs that produce defensible, audit-ready documentation.
Preparing for RADV: How Retrospective Programs Build Audit Defense
CMS’s accelerated RADV audit strategy means health plans must treat every chart review as potential audit evidence. The January 2026 memo confirmed that CMS plans to initiate audits approximately every three months [1].
To build defensible programs that protect financial performance, healthcare organizations should consider these practices from the OIG’s compliance guidance [2]:
Pair chart review processes with data accuracy controls. Implement data filtering logic to identify anomalies in diagnosis data. Benchmark HCC prevalence rates across years to spot unusual coding patterns. Analyze provider coding intensity and conduct follow-up education where patterns suggest overcoding. Ensure that retrospective reviews generate both adds and deletes. Report unsupported codes to CMS and address overpayments per Medicare overpayment law (42 U.S.C. § 1320a-7k) [2].
The OIG also recommends that health plans review any software used in risk adjustment, including vendor-created tools, to ensure they are not designed primarily to increase risk scores without supporting clinical validity [2]. For complex chronic conditions spanning multiple providers, the review process must ensure accurate risk adjustment across the entire patient population.
Ready to Evaluate a Collaborative Path?
Over a Partnership Evaluation Call.
Frequently Asked Questions
Retrospective risk adjustment is the process of reviewing clinical documentation after care to identify and validate codes that accurately reflect a member’s patient health status. It ensures that health plans receive fair reimbursement while maintaining compliance with CMS documentation requirements for Medicare Advantage plans.
Retrospective HCC coding involves certified coders reviewing completed encounters to identify hierarchical condition category codes that were missed during initial claims processing. Each code must be supported by MEAT evidence from a face-to-face visit.
Prospective coding captures diagnoses during or before the visit in real time as part of prospective risk adjustment. Retrospective chart review analyzes records after encounters to identify missed conditions, correct errors, and close documentation gaps. Both are needed for accurate risk capture and appropriate reimbursement in value-based care.
Concurrent coding happens during or shortly after a care episode, allowing faster correction of issues. Retrospective review occurs after the encounter is completed, serving as a final quality and compliance check on all submitted HCC codes. Many organizations also use prospective risk adjustment at the point of care to reduce the volume of corrections needed downstream.
AI-powered tools analyze medical records faster than manual review, identify clinical evidence with higher documentation accuracy, and create transparent audit trails linking each HCC code to supporting documentation. This improves coding accuracy, reduces administrative burden on coding teams, and strengthens RADV audit defense across the entire risk adjustment program.
CMS is accelerating RADV audits with Payment Year 2020 audits beginning February 2026, using variable sample sizes of 35 to 200 enrollees per contract with a 5-month medical record submission window. Health plans should ensure their retrospective risk adjustment programs produce defensible documentation now.
Codes for risk adjustment must be supported by current-period documentation showing that the condition was actively managed during a face-to-face encounter. Listing a condition in past medical history without current-year evidence of monitoring, evaluation, assessment, or treatment does not meet CMS requirements.
