Sepsis Upcoding: When a Diagnosis Raises Revenue and Lowers Reported Mortality

Medicare Advantage sepsis upcoding can make patients look sicker for payment and hospitals look better at keeping them alive. That is the allegation in the sepsis complaint against Kaiser Permanente: unsupported diagnoses increased the plan’s payments while swelling the number of cases underneath its reported death rate. The same coding decision allegedly improved the revenue and the reputation.

How Medicare Advantage Sepsis Upcoding Works

Medicare Advantage sepsis upcoding uses unsupported diagnoses to increase a plan’s risk-adjusted payments. In the alleged Kaiser scheme, adding false cases also enlarged the denominator of the hospital’s reported sepsis mortality rate, making that rate fall.

Medicare Advantage plans receive a monthly payment for each enrollee, adjusted for expected health care costs. The diagnoses a plan submits help determine that adjustment, including eligible diagnoses from hospital stays, outpatient visits and physician visits during the prior year. This is how a diagnosis becomes revenue, even when the immediate issue is a hospital’s coding decision.

The 2016 sepsis complaint describes a physician leadership meeting in October 2009. A coding presentation explained the need for clinical documentation. A physician liaison allegedly told those present to disregard it and reminded them that diagnosing sepsis was worth money. The complaint put the risk-score weight at .754 and the additional payment at approximately $8,500 per diagnosed enrollee for the following year.

The second effect is the stranger one. A reported mortality rate has deaths on top and cases underneath. Add false sepsis cases to that denominator without adding sepsis deaths, and the rate falls without a single life being saved. The complaint alleges that Kaiser used broader diagnosis practices to produce that result, giving the risk-adjustment allegation a second purpose.

It says sepsis diagnoses increased by as much as 200% to 300% at some California facilities between 2009 and 2013. By 2013, according to the complaint, Kaiser claimed a reported sepsis mortality rate of approximately 9%.

Sepsis Definitions and the Missing Query

The alleged sepsis coding policy did more than use a broad definition. It stopped coders from asking the question that might have removed the diagnosis.

The complaint describes a standard based on two or more elevated SIRS criteria, signs of systemic inflammation, plus infection. It contrasts that approach with the Sepsis-3 definition published in 2016, which emphasizes a life-threatening response to infection that injures tissues and organs. Its criticism concerns the clinical support for the diagnosis, not simply whether someone typed the word.

The definition problem also appears in the HHS Office of Inspector General’s sepsis billing work plan, announced in March 2024. The Inspector General described the newer, narrower definition and the older, broader definition that the Centers for Medicare & Medicaid Services and the Centers for Disease Control and Prevention recognized at the time. It proposed examining inpatient sepsis billing and the Medicare costs associated with using the broader definition. Here, the additional question is what a diagnosis does to the plan’s risk-adjusted payment.

A coding query asks the physician to clarify the medical record. The whistleblowers alleged that Kaiser required coders to code sepsis whenever the physician wrote it, while prohibiting queries about the supporting documentation. According to the complaint, coders could question other diagnoses without that restriction.

The route for resolving uncertainty stayed open for other conditions and closed for sepsis. The complaint attributes the duty to query unclear documentation to coding guidelines, while attributing the prohibition to Kaiser’s unwritten policies. The person reading the chart allegedly had to accept the diagnosis without asking for the findings behind it.

Short Stays and a Mortality Outlier

The whistleblowers tied the alleged policy to encounters a hospital coder would recognize: brief stays, limited treatment and a sepsis diagnosis the chart did not support.

One coder described patients at Kaiser’s South Bay hospital between 2010 and 2014 who went home after observation stays of three to 48 hours. The complaint says they had no ICU or inpatient admission, had not received the aggressive treatment described in Kaiser’s sepsis protocols, including blood cultures before antibiotics, and lacked sufficient documented clinical findings supporting sepsis. It also describes patients with brief inpatient stays and the same alleged mismatch between diagnosis, treatment and documentation.

The quality-reporting allegation becomes more concrete at Panorama City. According to the complaint, management flagged that hospital in November 2010 for reported sepsis mortality of 25%, even while describing it as one of the system’s best hospitals by other quality measures. The alleged response was to enlarge the case count:

An intensive-care physician allegedly challenged the accuracy of the coding and the credibility of the other hospitals’ reported mortality figures. The complaint says she received no response. The alarm went off, and the alleged answer was more denominator.

When an Audit Found Sepsis Without Support

A separate Kaiser complaint describes a sepsis diagnosis that an internal audit had already challenged. The alleged failure came after the problem was found.

In the separate amended complaint, one whistleblower describes a 2015 audit of a clinic encounter. Sepsis had been added four months after the patient was seen. She alleged that the addition had no supporting lab values, signs, symptoms, treatment or clinical indicators explaining how the diagnosis was determined.

She agreed with the National Compliance Office’s finding that sepsis had been improperly coded. Her director allegedly told her never to agree with that office’s audits, to respond with thanks if the topic arose, and to speak at the audit exit call only when asked.

This account adds something the short-stay allegations cannot: a description of an internal reviewer identifying an unsupported code and then being told how to respond to the audit. It puts the response to a compliance finding at the center of the story. Finding the problem was not the end of the alleged failure.

The Incentive Runs Through the Organization

A plan’s ability to influence the people assigning diagnoses matters to its revenue. Broader research on coding intensity helps explain why the alleged instructions deserve attention.

The 2015 working paper Upcoding or Selection? Evidence from Medicare on Squishy Risk Adjustment, hosted by Duke University, estimated that Medicare Advantage plans generated risk scores averaging 6.4% higher than the same enrollees would have generated under fee-for-service Medicare. It found greater coding intensity with closer integration between insurers and physicians. Its explanation reaches directly into the coding office:

In its October 2024 report on health risk assessments, the Inspector General estimated that diagnoses found only on those assessments and linked chart reviews generated $7.5 billion in Medicare Advantage risk-adjusted payments for 2023. With no other service records for those diagnoses in the data it examined, the Inspector General raised concerns about diagnosis accuracy and whether patients received needed care. Its question is relevant here: what happened in the patient’s care besides the appearance of a payment diagnosis?

What the Government Pursued at Kaiser

The government joined part of the Kaiser case and pursued diagnoses inserted into the record after outpatient visits. The sepsis mortality theory remains an allegation from the whistleblowers’ complaint.

In its amended complaint, the government alleged that Kaiser pressured physicians to add diagnoses that had not required or affected care at the visit. The January 14, 2026 settlement announcement put the resolution at $556 million for those allegations, covering conduct from 2009 through 2018. It described diagnosis targets, financial incentives and ignored internal warnings.

The alleged mechanisms point in the same direction. In the outpatient case, queries pushed physicians to add diagnoses. In the sepsis complaint, queries that might challenge a diagnosis were prohibited. One encouraged the addition. The other allegedly prevented the question.

The same concern about review that works only toward more payment appears in another insurer’s case. Aetna allegedly used chart reviews to add diagnoses while failing to withdraw previously submitted codes those reviews did not substantiate. The Department of Justice’s March 2026 announcement describes that conduct. The parallel is the treatment of contrary evidence: a review can find a problem while the payment input survives.

The Department of Justice announced a $95 million whistleblower share of the Kaiser recovery for the resolved claims. A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers.

When the Measure Moves With the Code

The sepsis allegation exposes a question that travels beyond this case: does the code that earns the payment also decide who enters the quality measure? If it does, a better-looking rate can be a product of the same coding decision that raises revenue.

In the scheme alleged here, adding unsupported diagnoses would make patients look more expensive to insure while making the hospital look more successful at treating sepsis. The measure does not decouple. It inverts.

For someone inside a hospital or plan, the useful connection is between the clinical findings, the coding instructions, the submitted diagnosis and the reported outcome. The alleged query prohibition, the response to the audit and the instruction to find more cases put decisions behind the numbers. That is where this story becomes more than a dispute about a definition.

When You Know Why the Code Stayed

If you work in healthcare and have seen sepsis diagnoses survive contradictory charts, blocked queries or adverse audit findings, speak with our whistleblower experts about what happened. We build your strongest sepsis upcoding case around the clinical evidence and the decisions that kept unsupported codes in place, bringing in Medicare Advantage whistleblower lawyers who understand hospital coding. We stay on your side as partners through the whole journey.