Pediatric Sepsis Defined: Experts Endorse Organ Dysfunction

By Erica E. Remer, MD, FACEP, CCDS, ACPA-C

Those of you who live in the adult world may not even be aware that the sepsis conundrum (i.e., Sepsis-2 vs. Sepsis-3) didn’t really pertain to pediatrics. That matter was tabled whilst the so-called grown-ups squabbled. Their previous criteria hailed from 2005, specifically from the International Pediatric Sepsis Consensus Conference. But it was already clear then that sepsis needed to be updated and better defined for the pediatric population as well, so an international consensus panel was convened in 2019, and the results were published online on Jan. 21, 2024.

The International Consensus Criteria for Pediatric Sepsis and Septic Shock (hereafter referred to as the International Consensus) were derived by 35 clinical experts from diverse pediatric practice in 12 different countries. The first thing that struck me was that they refer to a sentinel article titled Guidance for Modifying the Definition of Diseases: A Checklist, published in the Journal of the American Medical Association (JAMA) Internal Medicine in July 2017. It didn’t occur to me that there is a process to make adjustments in disease definitions. I wonder if it was crafted in response to the mess that arose after Sepsis-3 was introduced. The article explains that there is a fine line to walk between over-diagnosing conditions and capturing them early and often, versus harming patients by not entertaining a diagnosis quickly enough.

Something specific in that article struck a chord with me: “Diseases do not generally have discrete boundaries, and (my inserted word: clinical) judgment is required to determine the thresholds for diagnoses.” It seems as though it is more prudent to have many multidisciplinary representatives with expertise in the condition being addressed, making thoughtful decisions about diagnostic criteria than it being the Wild West, where everyone uses whatever criteria suit them that day – and for their own purposes.

The approach they took for the pediatric sepsis criteria was first to administer a global survey of 2,835 clinicians, published in The Current and Future State of Pediatric Sepsis Definitions: An International Survey, asking what the condition sepsis constituted and what the word should mean. I found this curious until I realized before there was even a Sepsis-1, we used to diagnose sepsis by gestalt. You walked in a room and said, “Uh-oh, this patient is really sick!”

The ultimate conclusion, supported by the majority of respondents, was that sepsis is an “infection with associated organ dysfunction.” The International Consensus authors noted that this was preferable to “infection-associated SIRS (Systemic Inflammatory Response Syndrome),” and that this evolved definition indicates “widespread adoption of the Sepsis-3 conceptual framework.”

Next, they undertook a systematic review of more than 3 million pediatric encounters, letting that inform the design of a derivation and validation study. The original concept was an eight-organ system model, but they ultimately dropped the renal, hepatic, endocrine, and immunological dysfunction criteria, seemingly to reduce “requirements for laboratory investigation and data collection.” I watched a presentation by the Children’s Hospital Association, and their commentary was that isolated hepatic, renal, endocrine, or immunological dysfunction was unusual, and it was far more common to be seen in conjunction with dysfunction of one of the other included systems, thus rendering those systems redundant.

The International Consensus settled on a composite four-organ system model, which was coined the Phoenix Sepsis Score (PSS). It seems as though the PSS was originally intended to be a prognosticating tool for mortality, similar to the Sequential Organ Failure Assessment (SOFA) score, but ultimately, attaining a score of 2 or more on the PSS was defined as identifying sepsis in an “unwell child with suspected infection.” Accruing ≥ 1 point from cardiovascular dysfunction establishes septic shock.

The PSS includes respiratory, cardiovascular (CV) (with age-based mean arterial pressure values), coagulation, and neurological systems. They are variably weighted, with CV potentially offering up to 6 points, respiratory up to 3 points, and the other two only up to 2 points. The score was designed to be able to be utilized even in low-resource areas.

There are a few more points regarding the International Consensus:

• If a child manifests organ dysfunction remote from the site of localized infection, they are recognized as being at higher risk of mortality than if they only have a localized, single-organ system impairment (e.g., respiratory failure in pneumonia).

• The PSS is assessed in the first 24 hours of presentation to the hospital. The article notes that the PSS “is not intended for early screening or recognition of possible sepsis and management before organ dysfunction is overt.”

• Due to the difficulty of defining organ dysfunction in neonates born < 37 weeks’ gestation and the contribution of perinatally acquired infection, term newborns remaining in the hospital after delivery and neonates whose postconceptional age is younger than 37 weeks are excluded.

Here are some of my concerns regarding the International Consensus:

• They use mortality as their only endpoint (morbidity is also a major concern post-sepsis);

• They limited the development of the PSS to data from the first 24 hours of hospitalization (again, with the primary endpoint of predicting risk of mortality). Clearly, children with infections and organ dysfunction discovered later on, or who acquire infections during the encounter, can suffer from sepsis. I asked the corresponding author, and he asserted that they expect the PSS to perform similarly whenever during the hospitalization the condition crops up; and

• If the Phoenix-8 score had comparable performance to the Phoenix-4 score (PSS), shouldn’t we diagnose sepsis if a patient has an infection and organ dysfunction not included in the PSS (e.g., hepatic or renal failure)? I’m going to hope that this is a rare occurrence, but what is the role for clinician judgment? Are payors going to deny claims of sepsis even if the clinician believes there is life-threatening organ dysfunction of a non-Phoenix-4 organ system?

This International Consensus statement certainly demonstrated rigor in development. The most important sentence to me in the paper is “SIRS should no longer be used to diagnose sepsis in children, and because any life-threatening condition is severe, the term severe sepsis is redundant.”

Fortunately, we have guidance permitting the capture of the code for severe sepsis if organ dysfunction is linked to the sepsis, even if the provider doesn’t document the word “severe.” Until the World Health Organization (WHO) and Centers for Disease Control and Prevention (CDC) eliminate the ICD-10/-CM code for severe sepsis without shock (R65.20), all patients with sepsis and sepsis-related organ dysfunction should receive at least four codes:

• A code indicating sepsis (e.g., unspecified or organism-specific sepsis; perinatal, obstetrical, or postprocedural sepsis);

• An R65.2- code indicating severe sepsis without or with septic shock;

• A code specifying the underlying localized infection that is the source of the sepsis; and

• At least one code detailing the organ dysfunction.

Pediatric experts have expressed their agreement: there is no such thing as sepsis without organ dysfunction.

Now, if we can only get the adult practitioners to buy in, too.