Sepsis - Changes In ICD-10

Sepsis – every coder’s best friend! 


Many of us have faced the challenge of reconciling the variations that exist in defining sepsis between ICD-9, ICD-10, providers as a group, and even amongst providers at a single hospital.  Change is in the air.

The most traditional concept of sepsis is a bacterial infection of the blood (septicemia) and the havoc this wreaks on the body.  Indeed ICD-9 and ICD-10 both fit best to this idea as the codes for sepsis are found squarely among the bacterial infectious disease codes.  Of course ICD-10 arguably clarified the situation somewhat by re-titling the codes from A40-A41 as “sepsis” whereas ICD-9 used “septicemia” for 038 codes.  But a mountain of guidelines, references, and Coding Clinics all inferred that sepsis can only be caused by septicemia.  In ICD-9 we couldn’t code 995.91 “sepsis” without 038 first, and in ICD-10 there is no sepsis or SIRS due to infection at all except the infection codes themselves – septicemia or local infection as the case may be.

If you’ve coded a few sepsis charts it becomes evident that many providers’ concepts of sepsis differs from that which is apparent from understanding ICD-9 and ICD-10.  The damage done by sepsis occurs from the patient’s own abnormal immune response to the infection – local or systemic – and not from the infection itself.

The first evolution in the concept of sepsis occurred in 1991 with the First International Sepsis Definitions Conference (“Sepsis-1”), which established the criteria for Systemic Inflammatory Response Syndrome:

  • Body temperature less than 36 °C(96.8 °F) or greater than 38 °C (100.4 °F)
  • Heart rate greater than 90 beats per minute
  • Tachypnea with greater than 20 breaths per minute.
  • White blood cell count less than 4000 cells/mm³ (4 x 109 cells/L) or greater than 12,000 cells/mm³ (12 x 109 cells/L).

The Second International Sepsis Definitions Conference in 2001 (“Sepsis-2”) added some slightly broader criteria but substantially validated the existing criteria.

Pause to note that these criteria make no reference to actual septicemia!  These findings were found after study to represent a certain elevated risk of morbidity and mortality from any infection due to the dysfunctional response of patients’ immune system to that infection.  SIRS criteria are used to predict these adverse outcomes and guide a provider’s medical decision-making in the care of the patient with any number of localized or systemic infection, septicemia included.  So a provider documenting “pneumonia and sepsis” may not mean the patient has septicemia in the ICD sense at all, but that due to their pneumonia the patient is responding in an aberrant and harmful way.

Now with the publication of the Sepsis-3 criteria, drafted by the Third International Sepsis Definitions Conference in 2015 the proposed definition of sepsis has radically changed to focus not on exam and blood findings but on end-organ dysfunction caused by the immune response to the infection.  Looking through the lens of ICD, we are familiar with this concept as being “severe sepsis”.  So by Sepsis-3 criteria, little consideration is given to traditional vital signs, WBC count, or blood cultures but rather to the cozy-sounding “Sequential (or sepsis-related) Organ Failure Assessment” or “SOFA” score.  This provides an aggregate tally of the type and severity of organ system dysfunction including respiratory, coagulation, liver, cardiovascular, CNS, and renal function.  A mercifully shortened version “qSOFA” (quick-SOFA) seems to blend organ dysfunction with SIRS criteria and pulls the sepsis trigger if a patient has 2 or more of clinical criteria:  altered mental status, respiratory rate >22, systolic blood pressure <100.

Whether the new Sepsis-3 criteria will be adopted is a matter of great controversy among clinicians as they offer some benefit in specificity in predicting outcomes of sepsis while sacrificing some sensitivity in revealing patients at risk.

Where does this leave us as coders?  Excepting any forthcoming guidance we are still reliant on the provider crossing that magic line and documenting the word “sepsis” in order to confidently code A40-A41.  You can’t lay your head on “SOFA criteria” and you get no love from “SIRS criteria” when the source is infectious.  If sepsis is documented we will dutifully select the right code from among A40-A41 depending on the organism, if specified.

Some of us might hope that R65 SIRS of non-infectious origin may get a sister code which provides the option of coding SIRS of infectious origin, ANY infection.  Wouldn’t this mesh better with the clinical concept of sepsis as it has evolved?

For more information about Sepsis-3 criteria and definitions see:


Ed O'Beirne



Ed O’Beirne PA, MHS, CCS, CDIP is an ICD-10 educator of providers and coders with a specialization in procedure coding ICD-10-PCS and CPT. Before joining HRG, Ed was a director of physician assistants and patient relations for an 80k visit a year ER, a physician assistant in emergency medicine for over 10 years, coding supervisor, auditor and consultant for 9 years and EMT and respiratory therapist for 5 years.

In need of Ed and HRG's Optimization Teams' expertise?