HealthCare Partners and DaVita: More Diagnoses, Same Uncorrected Codes

The DaVita HealthCare Partners Medicare Advantage fraud settlement required a $270 million payment to resolve allegations of improper spinal coding guidance and one-way chart reviews. The provider group supplied diagnoses to health plans and shared in the Medicare payments those diagnoses helped generate. According to the Justice Department, its reviews sought missed diagnoses that would bring in more money while ignoring inaccurate codes that should have brought payments down. Medicare and taxpayers bore the alleged extra cost. Someone inside that process could see what the settlement headline cannot show: which findings the business acted on and which it left alone.

The Justice Department announced the settlement on October 1, 2018. DaVita had acquired the large California independent physician association in 2012. The resolution combined practices DaVita disclosed to the government with separate allegations brought by a whistleblower who had worked for a health plan doing business with DaVita.

The Provider Group Shared in the Plan’s Payments

HealthCare Partners connected the patient encounter to the insurer’s Medicare revenue. DaVita operated a medical services organization that contracted with Medicare Advantage organizations in states including California, Nevada and Florida. It collected diagnoses from the care it provided and passed them to the plans. The plans paid DaVita a share of what they received from Medicare for those patients.

Medicare Advantage plans receive a fixed monthly payment for each enrollee’s health care, adjusted for health status. Diagnoses associated with greater care needs can raise that payment. The plan submits the diagnosis information to the Centers for Medicare & Medicaid Services, but the information starts with healthcare providers. The provider’s coding work can therefore affect revenue on both sides of its contract with the plan.

The route from a diagnosis to payment runs through the risk score. As the HHS Office of Inspector General explains the payment calculation, eligible diagnoses are grouped into categories representing expected medical costs. Those categories and demographic factors contribute to the patient’s score, which multiplies the plan’s base payment rate.

That explains why this case reaches beyond an insurer’s submission department. The alleged payment inflation began with information the provider group controlled. A coding instruction distributed to physicians could travel through the plan and affect the government’s payment, with some of the money coming back to the group that issued it.

Two Practices Reached the Same Payment Stream

The settlement covered both guidance about which diagnosis to use and decisions about what to do after a chart review. Those practices entered the government’s account through different routes.

Spinal Coding Guidance Came Through DaVita’s Disclosure

DaVita voluntarily disclosed practices instituted by HealthCare Partners that allegedly caused plans to send incorrect diagnosis codes to Medicare. The Justice Department described inflated payments shared by DaVita and HealthCare Partners and gave a specific example: guidance directing physicians to use an improper code for a particular spinal condition that increased reimbursement.

The risk adjustment fraud allegation starts with the code a physician was told to use. For an insider, the significant connection is between the instruction, the diagnosis it encouraged and the information that reached the plan.

One-Way Reviews Came Through the Whistleblower

The whistleblower alleged a different practice. HealthCare Partners searched existing patient records for diagnoses its providers had missed and sent those additional diagnoses to Medicare Advantage organizations for higher payments. At the same time, it allegedly ignored inaccurate codes that should have been deleted.

The Justice Department said those deletions would have reduced Medicare reimbursement or required the plans to return money. A review could produce information pointing in both directions, while the business acted only on the direction that paid it more.

Why the Review’s Instructions Mattered

Review instructions can steer the work toward finding more codes while leaving errors uncorrected. In the HealthCare Partners allegations, additions reached the plans while inaccurate codes stayed in place. A retrospective chart review goes back through existing medical records, which can reveal both missed diagnoses and submitted diagnoses the records do not support.

The business question is what the review was built to produce. An additions queue, instructions focused on finding more codes and a correction process that never sends deletions all bear on that question. The person reading the chart and the person deciding what to submit may see different parts of it. A coder can produce a review result that the next stage of the process uses selectively.

The economics help explain why that distinction matters. The 2015 working paper Upcoding or Selection? Evidence from Medicare on Squishy Risk Adjustment, hosted in Duke University’s seminar archive, estimated that private Medicare enrollment generated diagnosis-based risk scores 6% to 16% higher than those same enrollees would generate under fee-for-service Medicare. It also found that coding intensity increased with closer insurer-provider integration. The paper describes a direct way to align the incentives: insurers can base physicians’ pay on the diagnoses they code instead of the visits and procedures they perform.

That is an industry finding about provider incentives. The HealthCare Partners case supplies the particular commercial connection: the provider group collected diagnoses and received a share of the plans’ Medicare payments. Its alleged coding and review practices operated inside that arrangement.

How Blind Chart Reviews Can Feed One-Way Submissions

The Ninth Circuit’s analysis of one-way chart reviews addressed allegations that organizations deliberately avoided identifying wrongly submitted diagnoses while certifying their payment data as accurate, complete and truthful.

Its explanation reaches into how the review works. In blind coding, reviewers do not see the codes previously submitted. The court explained that this can keep reviewers from looking only for support for an earlier coding decision. The next stage still matters: how the newly identified diagnoses are compared with the diagnoses already sent to Medicare.

The opinion described both withholding identified errors and running a comparison that reveals only missing codes. In the latter arrangement, even a comprehensive review can feed a one-way submission process. The useful evidence can sit in the comparison instructions and the treatment of the results, as well as in the review itself.

The Whistleblower’s Reward Belonged to the Chart-Review Component

The one-way allegations came through a qui tam case under the False Claims Act. The Justice Department announced a whistleblower reward of approximately $10.2 million specifically for settling those allegations. The reward belonged to that component of the resolution, which also included the company’s voluntary disclosures.

The whistleblower had worked for a health plan doing business with DaVita. The plan side of a provider relationship offers a view of how diagnoses arrive, how review results are used and what the payment information contains.

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

The Same Choice Appears in Other Chart-Review Cases

The allegations against Cigna and Aetna describe the same selective use of information. In Cigna’s settlement announcement, the government alleged that chart-review results supported requests for more money while results indicating overpayment went unused. In Aetna’s announcement, it again alleged that the plan used findings to seek additional payments while ignoring findings that it had been overpaid.

The government complaint against Anthem shows how the objective could appear in an instruction to the review vendor: Anthem allegedly pointed Medi-Connect’s reviewers and coders toward every new code that could bring in revenue.

That instruction tells the vendor what success looks like. Elsewhere in the same complaint, the government alleged that Anthem declined to investigate flagged records when the flags did not concern new revenue-generating codes. The connection to HealthCare Partners is the alleged business choice: a review could uncover information that reduced payment, yet its instructions and follow-through favored additions.

The Records Behind the Coding Decision

For a provider-group insider, the HealthCare Partners case connects several kinds of knowledge. A physician may recognize the coding guidance. A reviewer may know how findings were selected for submission. A person working with the plan may know which diagnoses arrived, and someone responsible for the contract may understand how the provider shared in the payment.

We examine those connections with the person who understands the work. The coding instruction shows what physicians were told. The review output and the submission show what changed afterward. The payment-sharing arrangement explains who stood to benefit. The response to correction findings can connect the original information to the later decision to leave it in place.

Someone who knows where that information changed, or why an identified error survived, can explain a decision that the payment total alone will never reveal.

You Know Which Findings Reached the Plan

If you know who directed the coding or why a review’s corrections never reached the plan, talk with our whistleblower experts about the decisions you saw. We develop your case with you and bring in Medicare Advantage fraud whistleblower lawyers who understand provider contracts, coding operations and the payment trail. You have a partner from the first conversation through the pursuit of a recovery.