The Department of Justice Believes United Healthcare Is Defrauding Medicare

Related articles

United Healthcare, the largest provider of Medicare Advantage (MA plans) services, is being sued by the Department of Justice (DOJ) for fraud. To give you a sense of United Healthcare’s size consider that just their third quarter revenue was $46.3 billion, up 11.6 per cent from the same period a year earlier, and the bulk of the growth was billing services and data analysis for healthcare groups, which ws up 34 percent since the beginning of 2016.

And to get a sense of the concern about the DOJ joining a whistle-blower lawsuit, consider this

The Dow Jones Industrial Average was in the red by lunchtime in New York, with UnitedHealth proving the biggest drag a day after a law firm announced that the US had joined a lawsuit alleging that the health insurer and others bilked the Medicare managed care program … the US Justice Department will take a lead role in litigating the claims.”

The Players

United Healthcare – their MA plans have 6.8 million beneficiaries including plans offered by AARP Ingenix is their service subsidiary providing coding services to United Healthcare and other MA plans

MedAssurant – a data analytics company similar to Ingenix

MA Plans that used Ingenix or MedAssurant services [1]

The Game

MA plans or part C of Medicare provide the same or more benefits to beneficiaries by pursuing a risk and gainsharing proposal with the Center for Medicare and Medicaid (CMS). MA plans based on their location, number, gender, and health status of their beneficiaries submit a bid to CMS to provide all services (including administrative costs and profit) in exchange for payment from CMS. CMS compares their benchmarks for these variables to the bids. Bids greater than the benchmark are reduced to the benchmark and beneficiaries pay the difference. Bids less than the benchmark are accepted, and the difference is split, CMS takes 25%; the remaining 75% is to be used by the MA plans for additional benefits, like dental care or paying Medicare co-payments.

Patient’s age and health status vary, risk adjustment accounts for these variations. CMS has defined various categories of chronic disease and assigned them scores related to their cost. The higher the score, the greater CMS's expense. The average risk profile is 1, a 2 costs CMS twice as much, etc.

The ‘catch’

Medicare does not pay the MA plans their bid, their payment is based upon the actual patient’s risk profile they experience during the year. Including additional codes can result in additional payments in the range of $3,000 per beneficiary, so accurate risk scores are important both to CMS and the MA plans. Because of concern about fraud, the coding of risk assessment has strict guidelines:

  • Patient must be treated face-to-face with a qualified physician or hospital.
  • The diagnostic information must be in a particular written form.
  • Insurers have a responsibility for assuring that diagnostic codes come from these face-to-face visits and procedures.
  • If MA plans are made aware of erroneous coding they must delete the claim and file an amended one.
  • MA plans are required to have their CEO or CFO (or their designees) attest to the veracity of the risk adjustment three separate times, in their monthly reports, in annual reports and when their bids are submitted.

CMS monitors the process by a sampling random charts for coding errors. But, unlike similar physician reviews where errors identified in their sample are applied to the whole practice,(e.g. a 5% error rate in our sample results in taking back 5% of all monies paid) identified errors in MA plans lead to repayment for the particular mistake. Without applying reimbursement to the whole plan, there is no financial incentive for MA plans to play fairly. And per the lawsuit, they did not; they cheated.

The alleged fraud

Coding is difficult and is frequently not well understood by providers. After all, we went to school to practice not code medicine. More than once a reviewer has asked whether the patient’s diabetes is related to their kidney disease. As a vascular surgeon, the answer is most times “yes, of course. What a dumb question.” In reality, diabetes without complication has a much lower risk adjustment than diabetes with complications, so this means $3,000 difference to the MA plan. Who knew my answers were so valuable? United Healthcare did. And again according to the lawsuit that took measures to increase those risk scores. The allegations are that United Healthcare upcoded risk adjustments, made indirect payments to providers to upcode data, created fraudulent documentation for their clients to submit, and refused to correct previously filed claims they knew to be false. Specifically,

  • Medical charts, the primary source of diagnostic data were reviewed looking for incremental increases but not for errors that would decrease the risk adjustment.  New diagnostic codes were submitted, but an erroneous diagnosis was not deleted. On a two-way street, “they do not look both ways.” And these reviews were profitable, a $30 cost for chart review resulted in an average of $450 in additional payments.
  • Ingenix sought to find additional diagnosis by using data analytics to identify practices, where chronic conditions were underreported based on population prevalences or where a new diagnosis was suggested by medications patients were receiving. Here the physician was contacted with patient assessment forms (PAFs) to consider these diagnoses for the patient. Physicians receive a small fee for reviewing these PAFs. Ingenix also developed initiatives designed to get patients to visit their doctors each year. But wait a minute, these programs encourage physicians to screen patient’s more carefully and to encourage annual checkups. These are good goals, but as the lawsuit states:

“The program may have certain clinical benefits if and to the extent it helps ensure that members with chronic disease receive treatment for their condition. However, that clinical concerns is not driving this program is demonstrated by which patients are targeted … those with risk adjustment payments from CMS. If United were using the PAFs to improve clinical outcomes, they would include all their members, including their non-Medicare members, in the PAF program. Furthermore, as with the chart review program … Ingenix chooses the conditions it targets … based on revenue impact, not clinical impact, and ignores conditions that are frequently over coded.”

  • When components of the medical chart documenting incremental improvement were missing, Ingenix having identified the error asked the provider to re-attest and corrected the documentation insufficiencies. When the provider re-attested, Ingenix updated the claim; when the provider did not re-attest, they did not delete the claim. And this service was not provided for claims where Ingenix could not see an incremental benefit.
  • Finally, United Healthcare knew that submissions were possibly fraudulent

“The best evidence of both United’s knowledge that the underlying claim data requires verification and United’s fraudulent refusal to correct false claims, is the disparity between its efforts to find ‘incremental’ (new) codes and ‘delete’ (previously submitted, but false) codes.”

In 2009, United reviewed 1.4 million charts for additional codes and 3-4,000 for delete codes.

MedAssurant, the other data-analytic firm and defendant, was more audacious, their coders were instructed to recode diagnostic categories and link conditions without consultation with providers. This is simply illegal. I received letters from MedAssurant while in practice, and I became concerned that my coding was incorrect and that the MA plans wanted to ‘claw’ money back – not an unreasonable expectation given how often insurance companies denied care after it was provided. But I was wrong. Here, again from the suit, is how a medical society newsletter to physicians stated MedAssurant’s role:

“In this audit, MedAssurant is only trying to get more money from Medicare, which it can do if it can ‘jack up’ the intensity of the diagnosis code it finds in your charts. So, in this type of audit, MedAssurant is not looking for money from you.”

These practices generated $100 million or more to United Healthcare annually. And as the saying goes, "A billion here, a billion there, and pretty soon you're talking about real money."


[1] HealthNet, Arcadian Management Services, Tufts Associated Health Plans, Aetna, BC/BS of Florida and Michigan, Bravo Health, Emblemhealth, Healthfirst New York, Humana, Medica Holding Company and Wellcare Health Plans