SOS: Sepsis! Let’s Fix ICD-10-CM

Let’s admit it: the International Classification of Diseases, 10^th Edition, Clinical Modification (ICD-10-CM) official conventions, guidelines, and advice involving sepsis and its consequences have been a mess since the introduction of the Sepsis-3 definition in 2016.

However, there’s hope that the Centers for Disease Control and Prevention (CDC) will soon restructure ICD-10-CM in time for the Inpatient Prospective Payment System (IPPS) Final Rule for the 2028 fiscal year (FY), which requires our diligent surveillance and advocacy.

As background, the CDC modified ICD-9-CM in 2003 to mirror how sepsis had been clinically classified since 1992 (known as Sepsis-1 or Sepsis-2) as a systemic inflammatory response syndrome (SIRS) due to infection without associated acute organ dysfunctions (e.g., A40-, A41-, etc. in ICD-10-CM) with additional codes for sepsis with associated organ dysfunctions, also known as “severe sepsis” (R65.2- in ICD-10-CM) to be reported alongside the A40- and A41- codes.

In 2013, the Centers for Medicare & Medicaid Services (CMS) implemented its “SEP-1 – Severe Sepsis and Septic Shock Management Bundle” based on Sepsis-1/Sepsis-2 terminology and maintained Yale University’s Center for Outcomes Research and Evaluation (CORE) methodologies for its pneumonia mortality and readmission measures, to include index encounters for sepsis due to pneumonia while excluding index admissions coded as having severe sepsis due to pneumonia that was present on admission (POA).

Sepsis terminology was disrupted in 2016 when the Society of Critical Care Medicine (SCCM) and the European Society of Intensive Care Medicine (ESICM) completely revamped worldwide sepsis terminology to require an associated organ dysfunction due to an infection to define sepsis which rendered the term “severe sepsis” obsolete. 

Known as Sepsis-3, SCCM and ESICM asserted that for clinical operationalization, organ dysfunction can be represented by an increase in the Sequential [Sepsis-related] Organ Failure Assessment (SOFA) score of 2 points or more and that septic shock can be clinically identified by a vasopressor requirement to maintain a mean arterial pressure of 65 mm Hg or greater and serum lactate level greater than 2 mmol/L.

Despite this, the CDC and CMS continue to maintain the Sepsis-1/Sepsis-2 concepts of sepsis without organ dysfunction and severe sepsis (with associated organ dysfunction) in ICD-10-CM and their applications to payment and quality measures, to this very day.

Not only did CMS stand by the older Sepsis-1/Sepsis-2 terminology for its SEP-1 and pneumonia mortality and readmission measures, it maintained Sepsis-1/Sepsis-2 terminology within Medicare-Severity Diagnosis Related Groups (MS-DRGs) and prohibited its accountability agents from denying traditional Medicare reimbursement for sepsis-related MS-DRGs without an associated organ dysfunction – even though they allowed Medicare Advantage (MA) and managed care Medicaid plans to deny sepsis-related DRG payments if Sepsis-3 criteria were not met.

Accordingly, hospitals have resisted embracing Sepsis-3 for MS-DRG clinical validation since doing so costs multi-million-dollar revenue losses for traditional Medicare sepsis DRGs without an acute associated organ dysfunction.   

Sepsis-related payment denials for encounters not meeting payers’ interpretations of Sepsis-3 criteria remain contentious and controversial since the Coding Clinic editions published for the fourth quarter of 2016, pages 147-149, and the fourth quarter of 2017, page 110, require coding of sepsis when diagnosed, documented, and affirmed by a treating physician using whatever criteria the physician deems appropriate.

Pinson and Tang addressed this in their conversation with Sepsis-3’s lead author, Marvyn Singer, MD, available at https://www.pinsonandtang.com/resources/sepsis-denial-payer-sepsis-3-author/

While the CDC proposed amending ICD-10-CM to conform to Sepsis-3 in 2019, pediatric sepsis was a stumbling block, since the 2005 International Pediatric Sepsis Consensus Conference definitions using Sepsis-1/Sepsis-2 terminology had not yet been revised to embrace Sepsis-3’s sepsis constructs. 

To remedy this, the SCCM Pediatric Sepsis Definition Task Force released its Phoenix criteria for pediatric sepsis in January 2024, which, like Sepsis-3, required an acute organ dysfunction due to infection to define sepsis, promulgating criteria for each organ dysfunction category and rendering the 2005 pediatric sepsis criteria and its concept of “severe sepsis” obsolete.  Consider Dr. Denise Goodman’s discussion of pediatric sepsis at https://acdis.org/sites/acdis/files/PhoenixSepsisPresentation3_7_25.pdf. 

Now, in 2026, the Surviving Sepsis Campaign advocates both the Sepsis-3 for sepsis in adults (over 18) and the Phoenix criteria for sepsis in children (38 weeks of gestation to age 18) that define sepsis as a “life-threatening organ dysfunction due to infection,” which will both likely endure though the criteria defining each organ dysfunction will likely evolve.

To remedy these challenges as a result of public and regulatory pressure (e.g., the U.S. Department of Health and Human Services Office of Inspector General), and with input and insights from the Sepsis Alliance, a leading patient advocacy organization invested in sepsis early detection and intervention, the ICD-10-CM/PCS Coordination and Maintenance Committee requested public input to its March 17, 2026 initial proposal to amend the ICD-10-CM table to address Sepsis-3 terminology that:

1. Eliminates code R65.20, Severe Sepsis;
2. Adds new Chapter 18 codes for other specific organ dysfunction associated with sepsis, such as:
   • R65.220 Respiratory dysfunction associated with sepsis;
   • R65.221 Coagulation dysfunction associated with sepsis;
   • R65.222 Liver dysfunction associated with sepsis;
   • R65.223 Cardiovascular dysfunction associated with sepsis (Excludes1: septic shock (R65.21));
   • R65.224 Central nervous system dysfunction associated with sepsis; and
3. Introduces the concept of “Impending Sepsis” within Chapter 1 sepsis codes that can be trigger simply with a positive sepsis diagnostic aid, such as: A41.92 Impending sepsis, unspecified organism; Infection with SIRS; and Infection with positive sepsis diagnostic aid.

Access their official documents and reasoning at:

• Their official proposal: https://www.cdc.gov/nchs/data/icd/topic-packet-March-2026.pdf  (Page 89)
• A recording of what was said at the March 17, 2026 meeting:
  • Click this link – https://tinyurl.com/Mar26ICD10CMrecording
  • Enter this password (you will likely need to copy and paste it): X.2fZ*B#
  • Advance to the 2 hours, 39 minutes mark to hear the CDC’s comments and those by various speakers, such as Dr. Jeff Linzer of the American Academy of Pediatrics, Dr. James S. Kennedy of CDIMD, and others.

The CDC emphasized that this March 2026 proposal is preliminary, and that a more complete proposal will be presented at their September 2026 meeting after receiving public feedback, which optimally should be sent by electronic mail to NCHSICD10CM@cdc.gov no later than Aug. 15, 2026.

Potential issues involving the CDC’s proposal to amend the sepsis codes include (in no particular order):

• Since the CDC did not offer any proposals to amend the ICD-10-CM Alphabetic Index or their Official Guidelines during their March 2026 meeting, the public must address these in their comments.

• To conform with Sepsis-3/Phoenix concepts, ICD-10-CM must remove the term “severe sepsis” as proposed and, ideally, amend the Alphabetic Index to require the physician to link an associated organ dysfunction to sepsis as to report a sepsis code (e.g., A41.9, Sepsis unspecified). Without this requirement, coders will assign sepsis based on provider documentation (as instructed by Coding Clinic) even if there is no associated organ dysfunction linked to the sepsis diagnosis.

• The CMS SEP-1 core measure currently stipulates that any lactate elevation over 2 mmol/L in the setting of its SIRS criteria defines severe sepsis unless the physician explicitly links the lactate elevation to another cause (e.g., metformin use, thiamine deficiency, end-stage liver disease, malignancies, etc.).
  
  ICD-10-CM currently classifies “excessive lacticemia,” also known as “elevated lactate,” as unspecified acidosis, even if criteria for metabolic acidosis are not met, and does not classify elevated lactate as a cardiovascular disorder as suggested by the pediatric Phoenix criteria or SEP-1.
  
  ICD-10-CM must address these challenges, perhaps by adding a new Chapter 18 code for elevated lactate levels without acidosis or shock, such as R79.84, Abnormal findings of blood lactate, with an Excludes1 note for metabolic acidosis or shock.
  
  Physician engagement of the differential diagnosis of lactate elevation as to differentiate Type A and Type B etiologies and diagnostic approaches is essential. Physicians must also determine if early septic shock exists with evidence of systemic hypoperfusion (e.g., prolonged capillary refill, elevated lactate) if vasopressors are not used as to survive Sepsis-3-defined clinical validation for septic shock.

• While Sepsis-3/Phoenix each has its own promulgated criteria for sepsis and septic shock, there is no universally accepted terminology or criteria for “impending sepsis,” also known as “pre-sepsis,” that describes the infected patient likely to evolve into Sepsis-3/Phoenix-defined sepsis without preemptive intervention. Most clinical literature exploring sepsis-related early-warning systems currently categorize these patients as “risk for sepsis,” not “impending sepsis” or “pre-sepsis.”  
  
  To create a new code for infected patients at high risk to develop Sepsis-3 or Phoenix-defined sepsis, a clinical label (e.g., “pre-sepsis” or “impending sepsis”) and its associated criteria must be clinically constructed by experts in sepsis care (e.g., SCCM) with an emphasis of how various sepsis-related diagnostics (e.g., IntelliSep, ImmunoScore, TriVerity, Epic Sepsis Model, etc.) should factor in establishing this diagnosis. Gerald and colleagues offered a potential option for defining pre-sepsis in July 2025.
  
  Technology companies (e.g., Cytovale, Pronosis, Inflammatix), including those that create electronic health records (e.g., Epic, Oracle), and other entities involved with various pre-sepsis and sepsis-related early warning systems must be approached to standardize their criteria and approach with the yet-to-be determined clinical concept of “pre-sepsis.” 

• Since the CDC proposed that a positive sepsis screen test can qualify as impending sepsis, they must clarify if a coder can report an impending sepsis code solely from the positive test itself (like coders did for CoVID test positivity during the Public Health Emergency) or if the physician must document the positive test in his or her notes.

• If the CDC approves the concept of “pre-sepsis” or “impending sepsis,” the CDC must decide whether this should classified in Chapter 1 (as proposed), which can be sequenced as a principal diagnosis (driving MS-DRGs or APR-DRGs) when appropriate, or as a Chapter 18 “symptom code,” which is unlikely to be a principal diagnosis as to prefer the underlying infection, and how pre-sepsis or impending sepsis relates to obstetric and neonatology chapters.

• The ICD-10-CM Official Guidelines must be extensively revised to address various sepsis and non-infectious SIRS scenarios, such as sepsis and pre-sepsis sequencing, as well as present-on-admission (POA) issues (e.g., if a patient presents with pre-sepsis or impending sepsis at the time of the inpatient order which then evolves into Sepsis-3/Phoenix-defined sepsis afterwards, should the POA status for sepsis be “Y” or “N”) and the coding of coexistent noninfectious SIRS (e.g., due to acute pancreatitis, trauma, or recent surgery) with sepsis since both can contribute to acute organ dysfunctions.
  
  Another POA issue for the CDC to address involves septic shock. Currently if a patient presents with severe sepsis without septic shock at the time of the inpatient order and then develops septic shock afterwards, the ICD-10-CM Official Guidelines require that only one code for severe sepsis with septic shock be reported (R65.21), and that its POA indicator should be “N.  

• The American Hospital Association (AHA) Central Office for ICD-10-CM/PCS, CMS’s and CDC’s publisher of official ICD-10-CM/PCS advice, must offer official coding guidance interpreting these revisions and their day-to-day applications in their publication issued for the fourth quarter of 2027.

• CMS must determine how it will reconfigure MS-DRGs if sepsis must be linked to an associated organ dysfunction and the CDC creates a new code for impending sepsis or pre-sepsis. For example, CMS must decide how to address “pre-sepsis” as a principal or secondary diagnosis and whether it should have its own MS-DRG or be grouped to an existing DRG if sequenced as a principal diagnosis or serve as a complication or comorbidity (CC) or major CC (MCC) if sequenced as a secondary diagnosis.

• Solventum must determine how it will reconfigure APR-DRGs in the same manner as CMS reconfigures MS-DRGs (e.g., should a code for “pre-sepsis” be SOI 1, 2, or 3?) and its other measures (e.g., Enhanced Ambulatory Patient Groups (EAPGs); Potentially Preventable Complications (PPCs or “MHACs” in the state of Maryland); Potentially Preventable Readmissions (PPRs), etc.) for with the CDC’s reconfigured sepsis with an associated organ dysfunction codes and a new code for pre-sepsis.

• Yale University’s Center for Outcomes Research and Evaluation (CORE) must determine how its CMS mortality, readmission, and complication measures (outlined on CMS’s QualityNet website) will be reconfigured with the revised sepsis and new pre-sepsis criteria and coding conventions.

• The Agency for Healthcare Research and Quality (AHRQ) must determine how its Patient Safety Indicators (PSIs), especially PSI 4 (Death Rate among Surgical Inpatients with Serious Treatable Complications) and PSI 13 (Postoperative Sepsis), and other measures that it controls will be reconfigured with the revised sepsis and new pre-sepsis criteria and coding conventions.

• Other eCQMs (e.g., the “failure to rescue model”) or quality algorithms (e.g., Vizient, Premier) must be reconfigured with the revised sepsis and new pre-sepsis and coding conventions.

• CMS must address how it will restructure the SEP-1 core measure with the revised sepsis and new pre-sepsis criteria and coding conventions.

• High-impact physician literature (e.g., JAMA, the Journal of the American Medical Association, Critical Care Medicine, Intensive Care Medicine, etc.) must socialize these revised terminologies, especially those involving pre-sepsis, once they are agreed upon.  For example, they must develop a consensus whether to use the older Sequential Organ Failure Assessment (SOFA-1) or newer (SOFA-2) clinical criteria released in 2025 to define sepsis in adults and how newer pre-sepsis diagnostics are to be interpreted.

• All specialties, especially hospitalists, intensivists, pediatricians and obstetricians, must be engaged to understand and support the new terminologies and their clinical indicators as to appropriately document their critical thinking.

• ICD-10-CM coding and applicable clinical documentation integrity (CDI) personnel must be addressed to embrace the revised terminologies, code changes, and physician engagement strategies.

• Patient and public engagement to socialize early identification of pre-sepsis and sepsis requires deployment, to promote early detection and intervention.

• Last but not least, there must be some way for payers to stop denying sepsis codes and their related DRGs or risk-adjustments once the CDC updates ICD-10-CM with more current definitions.

The project needs everyone’s support. Please, if possible, consider these when addressing your role in advocating for sanity with ICD-10-CM sepsis codes:

• Support the Sepsis Alliance (https://www.sepsis.org), which is taking a leadership role in negotiating these issues with the CDC. Consider attending their “Unite for Sepsis” meeting in Washington, D.C. on June 11-12, 2026, to learn more of their work. Register for this at https://www.uniteforsepsis.org. To support or provide feedback to the Sepsis Alliance, reach out to them at https://www.sepsis.org/contact/ or contact a member of their board, who would most likely be glad to hear from you. 

• Consider sending your proposals for how MS-DRGs or CMS quality measures should be configured for FY 2028 through their Mearis website.
  
  Their deadline is Oct. 10, 2026.

• Network with your professional organizations to submit comments on how sepsis and pre-sepsis should be classified in ICD-10-CM. Options include the following:
  • Physician groups:
    • American Medical Association – https://www.ama-assn.org/health-care-advocacy/advocating-public-health
    • American College of Physicians – https://www.acponline.org/advocacy
    • American College of Surgeons – https://www.facs.org/advocacy/regulatory-issues/
    • American Academy of Pediatrics – https://www.aap.org/en/advocacy/
    • American College of Obstetrics and Gynecology – https://www.acog.org/advocacy/policy-priorities
  • Hospital groups:
    • American Hospital Association – https://www.aha.org/advocacy-public-policy/advocacy
    • Federation of American Hospitals – https://fah.org/issues-advocacy/medicare/
  • Coding/CDI Groups
    • American Health Information Management Association – https://www.ahima.org/advocacy/advocacy-agenda/
    • Association of Clinical Documentation Integrity Specialists – https://acdis.org/acdis-leadership-council
  • Others
    • American College of Physician Advisors – https://www.acpadvisors.org/government-affairs-committee
    • EndSepsis – https://www.endsepsis.org/board/

• Engage your own medical staff to address ICD-10-CM’s approach to sepsis, to advocate for this issue within their own spheres of influence.

• Send your own letter to the CDC as to how you believe sepsis should be addressed in ICD-10-CM at NCHSICD10CM@cdc.gov. 

Thank you for your professional advocacy of great patient care and your dedication to accurate diagnoses and terminologies. Stay tuned to Monitor Mondays and Talk Ten Tuesdays for further information as it becomes available.