What Does the OIG Sepsis Billing Report Mean for Hospital Coding Teams?

Jessica Miller, MHA, CPC, CPC-I

13 min read · November 19, 2026

Photograph of two hospital coding and CDI colleagues reviewing closed chart folders at a meeting table, representing what the OIG sepsis billing report means for hospital coding teams.

The OIG sepsis report, data brief OEI-02-24-00230, counted 615,007 FY 2024 Medicare inpatient stays billed with sepsis but not severe sepsis or septic shock. OIG estimates $2.2 billion could have been put to better use under the 2016 definition. That figure comes from a claims model, not chart review.

OIG sepsis data brief, OEI-02-24-00230What the report says
Issued and postedIssued November 16, 2026; posted November 19, 2026
DataPaid Medicare fee-for-service claims, FY 2024 dates of service, IPPS hospitals only
"Sepsis stays"615,007 stays with sepsis but not severe sepsis or septic shock, 9.4% of 6,556,071 inpatient stays
All sepsis-related stays1,079,620, or 16.5% of inpatient stays (Appendix B)
Medicare paid for sepsis staysOver $12 billion, about $20,000 each
Funds that could be put to better useAn estimated $2.2 billion, modeled from 407,389 stays where sepsis directly affected payment
Record reviewNone. The model assumes the information hospitals submitted was accurate
Recommendations to CMSAdjust payments to reduce the financial incentive to bill sepsis stays; use monitoring to take follow-up actions
CMS responseNeither concurred nor nonconcurred, and asked that both be removed. OIG kept them

Sources: the OIG report page and the full data brief (PDF) .

What did OIG find about Medicare sepsis billing in FY 2024?

OIG found that hospitals billed 615,007 Medicare inpatient stays in FY 2024 with a sepsis diagnosis but without severe sepsis or septic shock, under the broad definition Medicare uses. That is 9.4% of the 6,556,071 inpatient stays OIG analyzed. Medicare paid over $12 billion for those stays, about $20,000 each.

OIG calls this group "sepsis stays," and the label matters when you read the numbers. Its headline says hospitals included a sepsis diagnosis in "about 1 in 10" inpatient stays. That count leaves out severe sepsis, septic shock, and stays in MS-DRG 870, which OIG treated as consistent with the updated definition because more than 96 hours of mechanical ventilation indicates organ dysfunction. Appendix B puts all sepsis-related stays at 1,079,620, or 16.5% of inpatient stays. Keep the two figures apart in any summary for leadership.

The scope is narrow. OIG analyzed paid Medicare fee-for-service claims from the National Claims History File with FY 2024 dates of service. It excluded hospitals not paid under the Inpatient Prospective Payment System (IPPS), such as long-term care hospitals, critical access hospitals, and hospitals in Maryland. If you serve those facilities, these counts do not describe them.

OIG also found that billing for sepsis stays "varied widely among hospitals, raising concerns that some may be leveraging the broad definition to maximize their Medicare payments." Read the "may be" as written. OIG named no hospital and made no finding against any hospital.

How does the 1991 broad definition differ from the 2016 definition?

Medicare bills and pays sepsis under a broad definition developed in 1991. Under it, OIG says, a fever and elevated heart rate with an infection can be considered sepsis. Experts updated the definition in 2016. OIG refers to the 2016 version as "Sepsis-3," and it includes organ dysfunction, which the 1991 broad definition does not.

The report's primer describes the 1991 broad definition as Systemic Inflammatory Response Syndrome (SIRS) with a documented infection. It quotes the 2016 definition as "life-threatening organ dysfunction caused by a dysregulated host response to infection."

Medicare has not switched. OIG says the ICD-10 Coordination and Maintenance Committee "has not adopted the updated definition, though it has been adopted by others, such as insurers and some government programs." OIG adds that a CDC sepsis program and CMS's hospital quality program use the updated definition, and that "some insurance payors, including some Medicare Advantage insurers, use the updated definition of sepsis in determining payment." The report says nothing beyond that about how those payors review claims.

CMS's comment letter adds two points for coders. First, "the ICD-10-CM standard does not include diagnostic criteria. The code assignment is based on a clinical provider's documentation of sepsis in the health record." Second, CDC proposed ICD-10-CM changes on sepsis in March 2026, including implementing the Sepsis-3 definition. The report treats that as a proposal only, and an earlier proposal from September 2019 did not go forward. Our post on proposed ICD-10-CM code changes tracks the code-set side.

How did OIG estimate $2.2 billion?

OIG removed the sepsis diagnosis from each sepsis-stay claim and modeled the MS-DRG the claim would have received. The $2.2 billion comes from 407,389 stays where sepsis directly affected payment. OIG calls it funds that could be put to better use. It is not an overpayment finding and not a savings figure.

On average, Medicare paid about $5,000 more for each of those 407,389 stays because of the sepsis diagnosis. OIG split them into two groups:

  • Different base DRG. 355,004 stays likely would have moved to a different base DRG, most often related to pneumonia or metabolic encephalopathy. Medicare would have paid $4,405 less per stay on average, about $1.6 billion in total.
  • Same base DRG, lower severity. 52,385 stays likely would have stayed in the same base DRG at a lower severity level. Medicare would have paid $11,493 less per stay on average, about $602 million. For these stays, sepsis was typically the only secondary diagnosis counted as a major complication.

For the other 207,618 sepsis stays, the diagnosis did not directly affect what Medicare paid.

Two limits belong in any internal summary. OIG did not review records: "We did not conduct medical record reviews to verify the presence or absence of organ dysfunction." And "the model assumes the accuracy of the information submitted by hospitals," meaning a severe sepsis code appears when organ dysfunction was present. CMS objected that the method "does not account for patients diagnosed with sepsis under the current code set that are at risk of developing organ dysfunction and the resources required to treat them."

OIG also notes that MS-DRG changes in annual rulemaking must be budget neutral. The $2.2 billion "would remain within the IPPS and could be put to better use via different hospitals and MS-DRGs." OIG says CMS could look within its existing authority, or seek statutory authority, for the dollars to accrue as savings.

How much did sepsis billing vary between hospitals?

OIG found wide variation. Some hospitals billed more than one in five of their Medicare inpatient stays with sepsis, while others billed none. Stays of 3 or fewer days made up 30% of sepsis stays overall, but 75% at some hospitals and 5% at others.

OIG ran these comparisons on the 2,975 hospitals with 50 or more inpatient stays in FY 2024. Severity varied too. Some hospitals had more than 90% of their sepsis stays at the high severity level, while others had less than half there. Across all hospitals, 78% of sepsis stays were paid at the high severity level. Without the sepsis diagnosis, OIG estimates that share would likely have dropped to 53%.

OIG notes that stays in a given MS-DRG are paid at the same rate regardless of length of stay. OIG says hospitals were paid about $16,600 for each stay assigned to MS-DRG 871, whether the patient stayed 3 days or 2 weeks.

None of these patterns proves that a hospital billed in error. OIG did not review records, so the report cannot say why any one hospital's share is high. Your own record review can answer that for your hospital.

What did OIG recommend, and how did CMS respond?

OIG made two recommendations, both to CMS. CMS should adjust hospital payments to reduce the financial incentive to bill sepsis stays, and use its monitoring of hospitals' sepsis billing to take follow-up actions. CMS neither concurred nor nonconcurred, and asked that both be removed. OIG did not remove them.

For the first recommendation, OIG gave one option: CMS "could require a diagnosis of severe sepsis or septic shock for payment under the MS-DRGs 870-872." OIG also asked CMS to reconsider the role of sepsis diagnoses that are not severe sepsis or septic shock on the major complications and comorbidities list.

CMS said it "cannot propose to modify the MS-DRG GROUPER logic until the diagnosis code changes are made through the established consensus-based process," and that its analysis supports keeping sepsis on that list. OIG replied that its "recommendation is not contingent on updates to the ICD-10 diagnosis codes," because CMS updates MS-DRGs every year through rulemaking.

For the second recommendation, OIG notes that CMS "has resumed its PEPPER reports, which have consistently named sepsis as a target area for billing errors." OIG suggests targeted follow-up, "such as medical record reviews, or refer hospitals to the Recovery Auditors where appropriate." CMS said this "does not appear to require action outside of our existing oversight framework." Our Q2 FY 2026 PEPPER guide covers how to read the Septicemia target area, and our summary of HHS OIG's 2026 top challenges covers a separate OIG document.

What should hospital coding and CDI teams check now?

OIG's recommendations go to CMS, not hospitals, and Medicare payment has not changed. The suggestions below are MediCodio AI's advice, not OIG's or CMS's. Each row starts from a statement in the report and pairs it with an internal check you can run on your own sepsis stays.

Report statement it responds toMediCodio AI advice (our suggestion, not an OIG or CMS requirement)
Sepsis was typically the only secondary diagnosis counted as a major complication for 52,385 staysPull a sample of stays where sepsis is the only MCC and confirm the record documents sepsis
Stays of 3 or fewer days were 30% of sepsis staysTrend your share of short sepsis stays by quarter
78% of sepsis stays were paid at the high severity levelTrack how often sepsis is the diagnosis that moves a stay to high severity
Code assignment is based on the provider's documentation of sepsisCheck that each sepsis code has a supporting documentation passage, and send unclear records to clinical validation and your CDI query process
CMS could use PEPPER or new monitoring for follow-upReview your Septicemia target area in PEPPER alongside this sample

Our guide to clinical validation challenges covers the method behind the fourth row. For sampling and scoring the review, see periodic coding audits and CDI metrics . Keep this OIG estimate separate from CERT results: our CERT improper payment post covers a different program with a different method.

Where do pre-bill coding review and an audit trail fit?

OIG's model assumed hospital claims were accurate and did not test that against records. If a payer or auditor asks why a stay carried sepsis, your answer sits in the documentation. In our view, checking sepsis codes before the claim goes out, and keeping the supporting passage with each code, makes that answer faster to show.

That is our reading, not an OIG finding. No source here says pre-bill review will change a hospital's sepsis billing rate or prevent a review. Our post on what a medical coding audit trail must show covers the evidence to keep, and our guide to revenue integrity in healthcare covers the controls around it. If a payer review leads to a corrected or denied claim, our coding denials playbook covers the follow-up.

MediCodio AI works on the coding side. Complete charts flow through AutoPilot for fully autonomous coding. Charts where the documentation is thin or contradictory stay in CoPilot, where certified coders review and finalize the output. Nothing runs autonomous on day one. Everything starts in review. MediCodio AI keeps a 100% code-level audit trail. Every code ships with its documentation passage and the compliance rule behind it. Clinical judgment about sepsis, and the decisions that follow a review, stay with your coding, CDI, and compliance teams.

Frequently asked questions

What is the OIG sepsis report?

It is an HHS OIG data brief, report number OEI-02-24-00230, issued November 16, 2026 and posted November 19, 2026. It analyzes paid Medicare fee-for-service inpatient claims from FY 2024 at IPPS hospitals and estimates how payments would change if they followed the 2016 definition of sepsis.

Does OIG say Medicare overpaid $2.2 billion for sepsis?

No. OIG says Medicare could have put an estimated $2.2 billion to better use in FY 2024 had payments been consistent with the updated definition. Because MS-DRG changes in annual rulemaking must be budget neutral, OIG says that amount would remain within the IPPS. It is not an overpayment finding.

Did OIG review medical records for this report?

No. OIG analyzed FY 2024 claims data. Its model assumes the information hospitals submitted was accurate, including that a severe sepsis code was used when organ dysfunction was present. OIG states that it did not conduct medical record reviews to verify the presence or absence of organ dysfunction.

Did CMS agree with OIG's recommendations?

CMS neither concurred nor nonconcurred with the two recommendations and asked that they be removed. CMS said it cannot change MS-DRG GROUPER logic until diagnosis code changes go through the consensus-based process. OIG replied that its payment recommendation does not depend on new codes, and it kept both recommendations.

Has Medicare adopted the 2016 sepsis definition for billing?

No. Medicare still uses the 1991 broad definition for billing and payment, and OIG says the ICD-10 Coordination and Maintenance Committee has not adopted the updated definition. CMS's letter says CDC proposed ICD-10-CM changes in March 2026, including implementing the Sepsis-3 definition. The report treats that as a proposal.

Book a Demo to see how MediCodio AI codes your own charts, with the documentation passage and compliance rule behind every code. This article is general information, not legal advice. Book a Demo

Sources

  • HHS Office of Inspector General. Medicare Could Put Billions of Dollars To Better Use by Adjusting Hospital Payments To Be Consistent with the Updated Definition of Sepsis (OEI-02-24-00230), report page. https://oig.hhs.gov/reports/all/2026/medicare-could-put-billions-of-dollars-to-better-use-by-adjusting-hospital-payments-to-be-consistent-with-the-updated-definition-of-sepsis/
  • HHS Office of Inspector General. Data brief OEI-02-24-00230, full report (PDF), November 2026. https://oig.hhs.gov/documents/evaluation/11987/OEI-02-24-00230.pdf

See it in action

Ready to transform your medical coding?

See how MediCodio's AI platform delivers 98%+ accuracy across deployments since 2023, with turnaround under 24 hours at enterprise volume, inclusive of coder review, across 35+ specialties.

About the Author

JM
Jessica MillerMHA, CPC, CPC-I

Founding Member & Product Evangelist

Jessica Miller is the Head of Coding Operations at MediCodio, where she leads our AI-powered medical coding services with over a decade of experience in medical coding, compliance, and healthcare management. She specializes in developing innovative medical coding workflows that combine human expertise with cutting-edge technology to deliver exceptional accuracy and efficiency for healthcare organizations.

LinkedIn

Loading posts...