What $117.7 Million Buys in Lessons
In March 2026, Aetna agreed to pay $117.7 million to resolve False Claims Act allegations tied to its risk adjustment practices. The settlement broke into two parts. $106.2 million covered an add-only chart review program from payment year 2015 that the government alleged submitted diagnoses without adequate clinical support. $11.5 million addressed false morbid obesity diagnosis codes submitted from 2018 through 2023. The case was brought by Mary Melette Thomas, a former Aetna risk adjustment coding auditor, who received $2.01 million as the whistleblower.
This settlement, combined with Kaiser’s $556 million resolution, provides a specific, documented blueprint of what goes wrong in risk adjustment programs and what technology should be designed to prevent. The failures weren’t exotic. They were structural. And they’re repeated across the industry in programs that haven’t updated their design.
Failure Point One: Add-Only Architecture
The Aetna program that generated the $106.2 million portion of the settlement was an add-only chart review operation. Coders reviewed historical charts to find diagnoses. Found diagnoses were submitted. No systematic process existed to evaluate whether previously submitted codes still had adequate documentation support or whether they should be removed.
This is the most common design pattern in the industry, and it’s the one OIG explicitly flagged as high-risk in its February 2026 guidance. The technical fix is two-way coding capability: AI and workflow that evaluates documentation for both adds and deletes, treating removal of unsupported codes as equal in priority to identification of missed ones.
The organizational fix is harder. It requires redefining performance metrics so that deleting a code counts as a compliance win, not a productivity failure. It requires leadership to accept that a cleaned, accurate submission set is more valuable than a larger, unvalidated one.
Failure Point Two: Documentation Validation Gaps
The morbid obesity codes that generated the $11.5 million portion of the settlement were submitted without adequate clinical support. The diagnoses appeared in charts, but the documentation didn’t meet the evidentiary standard CMS applies during audits. The coding process found the codes. It didn’t validate whether the documentation could defend them.
This is the MEAT validation gap. Systems that identify diagnosis mentions without evaluating whether the clinical note shows active Monitoring, Evaluation, Assessment, or Treatment of the condition produce codes that look supported on the surface but collapse under scrutiny. The fix is technology that evaluates evidence quality as its primary function, mapping every coding recommendation to specific clinical language and specific MEAT elements before the code is ever submitted.
Failure Point Three: Insufficient Human Oversight
The settlement revealed a program where the volume of coding activity outpaced the quality of human review. Coders processed charts under throughput pressure. Quality assurance checks were insufficient to catch documentation gaps at scale. The system produced volume. The oversight structure couldn’t keep up.
The technology lesson: AI should function as decision support, not as an automation engine. The system recommends. The coder validates with full visibility into the evidence. The quality layer audits both the AI’s recommendations and the coder’s decisions. Each step creates documentation. Each step is auditable.
Building What the Settlement Says You Need
The Aetna settlement is a specification document for what risk adjustment technology should do. It should code in two directions. It should validate MEAT evidence before submission. It should produce explainable reasoning that humans can review and auditors can follow. It should support human oversight without creating throughput pressure that undermines validation quality.
Any Risk Adjustment Solution that meets these specifications was designed for the enforcement environment that produced the settlement. Any solution that doesn’t is carrying the same structural risks that cost Aetna $117.7 million. The lessons are public, the dollar amounts are documented, and the regulatory posture is only tightening.
