OIG Hospital Severity Upcoding Report: Rising Severity, Shorter Stays and a Declined Review

The OIG hospital severity upcoding report found Medicare stays billed at the highest severity level rising while their average length fell. More bills described the sickest patients, while those patients spent less time in the hospital. For a coding specialist or physician who has watched one diagnosis change the payment for an entire admission, the national pattern has a familiar shape. The chart and the decision to keep that diagnosis are where the story becomes specific.

In its February 2021 report on billing before COVID-19, the Department of Health and Human Services Office of Inspector General recommended targeted reviews of stays vulnerable to hospital DRG upcoding and the hospitals billing them. The Centers for Medicare & Medicaid Services did not agree.

Are Medicare Severity DRGs Vulnerable to Inappropriate Billing Practices?

The Office of Inspector General identified a particular vulnerability in Medicare Severity DRGs: a single major complication or comorbidity can set the payment for a whole stay. Hospitals could submit up to 24 secondary diagnoses per stay under the system OIG examined. The most complicated qualifying diagnosis generally determined the severity level.

In Medicare’s inpatient payment system, the hospital receives a rate for the diagnosis-related group rather than payment for each day. Higher severity carries a higher rate because those patients typically need more resources. That makes documentation a decision about health care spending as well as a clinical one. An unsupported diagnosis can raise payment for the whole admission without adding a procedure or another hospital day.

What Did OIG Find About Hospital Severity?

Highest-severity stays increased 19% from fiscal 2014 through fiscal 2019 while total inpatient stays decreased 5%. Stays at each of the lower severity levels declined. The average highest-severity stay shortened from 6.9 to 6.4 days.

The billing and the time in the hospital were moving in opposite directions. The Inspector General concluded that changes in hospital billing practices likely drove the increase, rather than changes in the patient population. Medicare spent $54.6 billion on highest-severity stays in fiscal 2019, nearly half of its inpatient hospital spending.

One MCC Diagnosis Raised the Payment for the Whole Stay

A single MCC diagnosis determined the highest severity level for 54% of top-tier stays in fiscal 2019, nearly two million admissions. Medicare paid $26.8 billion for those stays. The financial consequence of that one diagnosis was substantial:

The report’s pneumonia example makes the mechanism concrete. One major complication put the stay in the highest severity group and produced an $8,500 payment. If that diagnosis had been billed inappropriately, the next lower group would have paid approximately $2,800 less.

More than 80 percent of highest-severity stays for kidney and urinary tract infections, MS-DRG 689, reached the highest severity on one major complication. Nearly 70 percent of highest-severity stays for pneumonia, MS-DRG 193, chronic obstructive pulmonary disease, MS-DRG 190, and renal failure, MS-DRG 682, did too.

Short Stays and Hospitals Far From the Pack

Almost 30% of highest-severity stays were more than 20% shorter than the average for their payment group. Short top-severity hospital stays and a single complication gave reviewers specific places to look. About a third of highest-severity stays for heart failure and shock, MS-DRG 291, pneumonia and renal failure were particularly short.

Medicare paid approximately $14.5 billion for particularly short highest-severity stays. That was $4.9 billion more than it would have paid at the next lower severity level.

Hospitals also differed sharply. At most hospitals, the share of stays billed at the highest severity ranged from 13 to 52 percent. The highest 5% of hospitals billed between 52 and 79 percent of their stays that way. Some hospitals billed 80 to 100 percent of their highest-severity stays on one major complication.

A separate short-stay oversight audit, issued in June 2024, measured how much improper payment routine review caught. For short inpatient stays during 2016 through 2020, Quality Improvement Organizations denied payments equal to just 0.6% of the improper payments the Centers for Medicare & Medicaid Services estimated.

CMS Declined the Targeted Hospital Severity Reviews

The Centers for Medicare & Medicaid Services did not concur with the Inspector General’s hospital severity review recommendation. It said more work was needed to determine conclusively which billing changes came from upcoding, and promised to share the findings with Recovery Audit Contractors, continue monitoring and educate providers.

The Inspector General’s answer was pointed. Finding out which stays involved upcoding was precisely why it had recommended targeted reviews. The agency kept its recommendation and explained:

The Severe Malnutrition Audit Went Into the Charts

A separate severe malnutrition diagnosis-code audit, issued in July 2020, tested the medical records behind claims carrying E41 or E43, both major-complication codes. The Inspector General found 173 of 200 randomly sampled claims incorrectly billed.

Severe malnutrition upcoding accounted for an estimated $1 billion in Medicare overpayments for fiscal 2016 and 2017, according to the audit. A CDI query asking a physician to add severe malnutrition can bring the diagnosis that raises payment into the chart. The question is what supported that diagnosis, what the physician said and why the code stayed when someone challenged it. Clinical documentation improvement reaches the same payment decision the national data can only flag.

The kwashiorkor audit at Promise Hospital of Ascension shows the dispute at one hospital. The Inspector General found all 62 reviewed claims incorrectly coded and recommended a $465,079 refund. The hospital maintained that it had submitted the claims appropriately. The agency said the charts called for another form of malnutrition or no malnutrition diagnosis at all. For an insider dealing with a kwashiorkor diagnosis on an adult inpatient claim, the hospital’s audit response and the underlying chart belong in the same conversation.

Sepsis raises a separate documentation issue. Septicemia or severe sepsis with a major complication, MS-DRG 871, was the most frequently billed payment group in fiscal 2019. The Office of Inspector General’s sepsis review concerns the gap between the narrower clinical consensus definition and the coding definition Medicare uses.

What the Studies Add to the Billing Pattern

A 2024 study in Health Affairs went beyond counting higher-severity claims. Upcoding Linked To Up To Two-Thirds Of Growth In Highest-Intensity Hospital Discharges In 5 States, 2011–19 found that highest-intensity discharges rose 41% in its five-state data. After adjusting for patient characteristics, length of stay and hospital characteristics, it estimated a 13% increase if coding behavior had stayed the same. The study’s comparison puts the payment incentive in sharper focus: coding behavior can change how severe the hospital’s patients appear on the bill.

The gap between a billing flag and a chart has a long history. In Using Software to Detect Upcoding of Hospital Bills, issued in August 1998, the Office of Inspector General tested two software products against blinded medical-record review. Only 10 to 20 percent of cases the products flagged as upcoded were actually upcoded. An employee who knows how a diagnosis entered the record can supply information a billing screen never saw.

Hospital Settlements Reach the Decisions Behind the Codes

Hospital fraud settlements show how those decisions become allegations against a particular business. Prime Healthcare and its chief executive settled admission and upcoding allegations for $65 million in 2018. The Justice Department alleged that Prime falsified diagnosis information, including complications and comorbidities, from 2006 through 2014 to increase Medicare reimbursement. A performance-improvement director brought the whistleblower case, and the government joined it.

That role sits close to the pressure a coding department whistleblower may recognize: case mix index targets, physician education focused on diagnosis wording and comparisons with peers. You may have seen a CDI query produce a higher-paying diagnosis even after a clinician objected. The connection between that pressure and the billing decision is the part a national trend cannot show.

Oroville Hospital‘s $10.25 million settlement in 2024 included allegations of false systemic inflammatory response syndrome codes in claims to Medicare and Medicaid. The SIRS coding allegations sat alongside an alleged admission-bonus system that paid physicians according to how many patients they admitted.

A healthcare whistleblower bringing a qui tam case under the False Claims Act can receive 15% to 30% of what the government recovers.

The Decision a National Report Cannot Show

You may have seen the decision the national report cannot show: an objection overruled to keep a higher-paying code. Speak to our whistleblower experts about why that code stayed. We stay beside you from that first account through the case, connecting the clinical decisions to the billing and bringing in whistleblower lawyers who know hospital coding.