Group Health Cooperative: The Chart Said Resolved but the Code Said Otherwise

In the Group Health Cooperative Medicare Advantage settlement, the Washington health plan agreed to pay $6,375,000 to resolve allegations that it knowingly submitted unsupported diagnoses to increase Medicare payments. The organization is now Kaiser Foundation Health Plan of Washington. The case reached into the work of its chart-review vendor, DxID, and the objections raised inside the plan. The government’s later complaint alleged that Group Health submitted a major-depression code even though the physician had recorded that the condition had resolved.

The Department of Justice announced the settlement on November 16, 2020. A former Group Health risk-adjustment director brought the qui tam case under the False Claims Act, and the government joined it.

What DxID Sold to Group Health

DxID sold chart reviews that found additional diagnoses for Medicare payment. The vendor was a wholly owned subsidiary of Independent Health, another Medicare Advantage insurer, and served both its owner and other plans. Group Health was a customer of that business.

The financial opportunity came from how a diagnosis changes the plan’s monthly payment. Medicare pays a fixed monthly amount for each member’s health care coverage, adjusted for the member’s health status. Eligible diagnoses feed into hierarchical condition categories, which help determine the risk score. The score changes the payment. A condition added after a chart review can therefore produce additional revenue from care that has already happened.

According to the government’s complaint, DxID’s work caused Group Health to submit 4,946 new diagnosis codes in January 2012 for the 2010 service year. The complaint alleged that those submissions generated about $12.34 million in additional Medicare revenue. DxID’s contract gave it 20% of the added revenue.

DxID’s own sales pitch, quoted in that complaint, described the arrangement plainly:

That percentage-of-revenue fee put a price on the reviewer’s findings. An added diagnosis could pay both the plan and the vendor. A correction that reduced Medicare’s payment worked in the opposite direction. This is the incentive behind one-way chart review: additions earn revenue, while removing unsupported codes gives it up.

What Group Health’s Own Reviewers Found

The whistleblower alleged that a department head and a physician reviewed 117 charts and found 74% of the submitted codes unsupported. She also alleged that leadership’s concern was that the vendor had been overpaid. Taxpayers had paid for diagnoses the charts did not support, according to that review.

The government’s later complaint gives concrete examples from Group Health’s records. They show the difference between finding a condition somewhere in a chart and reading what the clinician actually said about it. A problem list can carry an old condition forward. A visit note can explain that it has resolved, or describe an encounter concerned with something else. In the alleged DxID process, those distinctions were lost on the way to a payment code.

Depression the Physician Said Had Resolved

The government alleged that DxID captured, and Group Health submitted, a major-depression diagnosis from a problem list despite conflicting progress notes. The complaint quoted the physician’s assessment:

The alleged contradiction was inside the same medical record. The problem list supplied the diagnosis. The physician’s assessment said it had resolved. For a coder or reviewer, that is a much more specific concern than a high risk score. It identifies the entry used for payment and the clinical information pointing the other way.

Old History and an Unrelated Visit

In another Group Health review described by the government, DxID captured alcohol dependence from an old medical-history entry. The reviewer reported no updated entry for the relevant service year, and the chart’s snapshot marked the condition resolved. The alleged coding decision revived a condition the record had already closed.

The complaint also described an uncontrolled-diabetes code drawn from a problem list during a dermatology appointment for a skin lesion and a bump behind the patient’s ear. According to the government, the record did not document diabetes as a factor in that encounter, yet DxID captured the condition and Group Health submitted it.

DxID’s written policy, as quoted in the complaint, directed coders to capture continuing chronic conditions from signed and dated problem lists. The examples show why the instructions and the individual chart belong in the same inquiry. The visit note shows what happened when they followed it.

The Heart-Attack Codes Reached Management

The government’s complaint alleged that Group Health executives approved disputed old-heart-attack codes after clinicians questioned whether the patients had the condition. The clinical objection reached the people deciding what to submit.

In January 2013, the risk-adjustment director reported that clinicians had reviewed 47 old-heart-attack diagnoses and found that 33 patients did not have the condition. The government then described executive emails acknowledging the concern and agreeing to proceed within the risk parameters they had already accepted.

That sequence matters to an insider who sees clinical objections overruled inside a risk-adjustment program. According to the complaint, clinicians challenged the diagnoses and management approved sending them to Medicare anyway.

Why the Payment Measure Matters

A study published in Health Management Policy and Innovation in 2019 tested whether the choice of risk measure changed how sick Medicare Advantage members appeared. Comparing Diagnostic Coding Intensity for Medicare Advantage and Fee for Service compared the payment-driving hierarchical condition category measure with a different diagnosis-based measure that was not used to pay the plans.

Using national claims samples from 2010 through 2014, the study found that the measures gave similar results for traditional Medicare. For Medicare Advantage, the alternative measure reduced the apparent risk to approximately the traditional-Medicare level. The study concluded that its results were consistent with selective upcoding to increase payments.

The government’s Group Health examples put that incentive beside the clinical record: did the diagnosis used for payment survive a reading of the clinical notes? A payment score is built from reported diagnoses. Here, the complaint puts those diagnoses beside the physician’s account, where the conflict becomes visible.

Independent Health’s Late Diagnosis Forms

The government’s complaint also alleged that DxID solicited diagnoses added to records after the visit for Independent Health. Providers were prompted to add conditions to records as much as 12 months after the encounter.

The government described a $25 payment for each signed form and alleged that, with Independent Health’s permission, DxID staff visited providers’ offices, checked diagnoses on forms and presented the completed forms for signature. Meeting notes quoted in the complaint reported that physicians were signing the forms without updating the medical-record diagnosis list.

The shared vendor makes the comparison useful. Group Health’s examples concern diagnoses taken from existing records despite contradictory clinical information. The Independent Health allegations concern a later effort to obtain documentation for additional diagnoses. Both accounts require attention to what the original encounter recorded and what the payment submission eventually said.

One Whistleblower Case, Separate Resolutions

Group Health resolved its part of the case in 2020. The government’s later complaint against the remaining defendants described Group Health transactions as part of its account of DxID’s conduct.

On December 20, 2024, the Justice Department announced Independent Health’s separate agreement to pay up to $98 million. The agreement covered Independent Health and DxID, which ceased operations in 2021. DxID’s former chief executive separately agreed to pay $2 million. The announcement expressly referred back to the government’s earlier settlement of the claims involving Group Health’s use of the same vendor.

The Group Health Whistleblower Reward

The Justice Department said the former Group Health employee would receive approximately $1.5 million from this settlement. The same whistleblower would receive at least $8.2 million from the later Independent Health resolution, according to its announcement.

A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers.

When Your Review Found Something Different

If your clinical review challenged diagnoses that your healthcare organization submitted anyway, tell our whistleblower experts what happened after you raised the issue. We help you build your strongest Medicare Advantage whistleblower case around the patient records, vendor instructions and decisions that kept those codes in use, and bring in Medicare Advantage whistleblower lawyers who understand risk-adjustment cases. We stay beside you through the case, with your knowledge of the program at its center.