Sepsis is still confusing


The most current authoritative diagnostic criteria for sepsis from 2001 were published in 2003 (1) and should be well known. Yet disinformation and confusion still abound. Some clinicians rely on the original outdated criteria from 1991 (2); others are unaware that specific diagnostic criteria exist, relying on a subjective impression learned long ago. Even a prestigious peer-reviewed medical journal recently published 2 articles on sepsis that contained errors, one using the outdated 1991 criteria and the other citing the 2001 definition of sepsis incorrectly.

Sepsis is such an important condition to recognize and manage correctly, document properly, and code accurately that revisiting the subject is worthwhile, particularly given the persistent confusion and misapplication of criteria. In this month's column, we will explore the current diagnostic standards for sepsis, severe sepsis, and septic shock as well as coding considerations. Next month, in Part 2, we'll delve into the conflicts and contradictions that exist within current sepsis literature, research, and other professional diagnostic standards.

Photo by Thinkstock
Photo by Thinkstock

Let's begin at the beginning. In 1991, the first sepsis definition conference was convened under the chairmanship of Roger C. Bone, MD, FACP, to establish a formal definition of sepsis; the findings of the 1991 conference were published in 1992. The conference defined sepsis as SIRS (systemic inflammatory response syndrome) due to infection.

SIRS was defined as the presence of more than 1 of 4 findings:

  • Body temperature >38.0 °C or <36.0 °C
  • Heart rate >90 beats/min
  • Tachypnea >20 breaths/min or hyperventilation with PaCO2 <32 mm Hg
  • White blood cell (WBC) count >12,000 cells/mm3 or <4,000 cells/mm3

“Severe” sepsis was defined as sepsis associated with organ dysfunction, hypoperfusion, or hypotension, and “septic shock” was defined as sepsis with arterial hypotension despite adequate fluid resuscitation.

Over the following 10 years, research data began to show a need to modify the 1991 definitions to better reflect the evolving understanding of the pathophysiology of SIRS and severe sepsis. A second International Sepsis Definitions Conference was convened in 2001, and the results were published in 2003.

The 2001 consensus retained the definitions of sepsis as SIRS due to infection (presumed or confirmed) and severe sepsis as sepsis associated with acute organ dysfunction. However, the new criteria defining SIRS (Table) were greatly expanded from the 1991 original. Organ dysfunction variables indicative of severe sepsis were also defined and hypotension criteria for septic shock were specified under “hemodynamic variables.”

This new set of sepsis criteria also changed the diagnostic requirement from “more than 1” of the original 4 to “some” of the expanded list. “Some” is still taken to mean more than 1 but allows broad clinical flexibility in applying the criteria. The 2001 consensus also indicates that a patient should appear sick enough in the clinician's opinion to warrant the diagnosis, although not necessarily “toxic.”

In 2001, the 1991 criterion for tachypnea >20 breaths/min was changed to simply “tachypnea” and hyperventilation with PaCO2 <32 mm Hg was eliminated. Tachypnea is still usually defined as >20 breaths/min based on the 1991 recommendation and medical practice standards. While doing so is not specifically prohibited by the 2001 criteria, using only a pulse >90 beats/min and respiratory rate >20 breaths/min in the absence of other criteria for a diagnosis of sepsis would likely be clinically unjustified since so many other situations may be associated with these 2 findings.

Much to the surprise of many clinicians, a positive blood culture is not considered a diagnostic criterion for sepsis but is strongly suggestive or confirmatory in the clinical setting of sepsis. Alternatively, a negative blood culture by no means rules out a diagnosis of sepsis based on the criteria shown in the Table.

Finally, the 2001 consensus requires that if, in the physician's opinion, any criterion is “easily explained” by a coexisting condition other than the infection, that criterion should not be used in determining whether the patient has sepsis. For example, in a patient admitted with pneumonia, acute exacerbation of chronic obstructive pulmonary disease (COPD), and atrial fibrillation, a pulse of 95 beats/min and a respiratory rate of 22 breaths/min might easily be explained by atrial fibrillation and COPD, respectively, and therefore would be excluded as criteria for sepsis.

The 2001 criteria, which replaced the 1991 criteria, remain the current authoritative consensus diagnostic standard for sepsis, applicable in both clinical practice and clinical trials. The Surviving Sepsis Campaign (SSC) literature, including the 2012 SSC guidelines (3), has likewise consistently reaffirmed the 2001 criteria, and no new definition conference has ever been convened.

Coding criteria are an important consideration. The recognition and documentation of sepsis are crucial for correct coding to reflect severity of illness, because in most cases sepsis is considered a more serious condition than the underlying infection causing it. A diagnosis of severe sepsis demonstrates even greater severity, and septic shock demonstrates extreme severity.

“Urosepsis” is no longer recognized as a codable diagnosis and must be clarified as simply a urinary tract infection (UTI) or sepsis. “Bacteremia” means nothing more than a positive blood culture, which, as explained above, is not a sepsis diagnostic criterion. “SIRS” documented as due to a particular infection is no longer codable as sepsis and must be specified as “sepsis.” Noninfectious SIRS, which by definition is not sepsis, remains an important condition requiring documentation.

In summary, look for current SIRS criteria indicative of sepsis in all patients admitted with infection, such as a UTI, pneumonia, cellulitis, diverticulitis, or cholecystitis. Consider the diagnosis of sepsis when 2 or more of the SIRS criteria are met. Look for organ dysfunction indicative of severe sepsis and for signs of septic shock. Use good clinical judgment in applying the SIRS/sepsis criteria, considering whether any of the findings should be excluded as diagnostic criteria because they are “easily explained” by another coexisting condition.