Time for Sepsis-3 in Children?*
In the recent Third International Consensus statement (1), the Sepsis Definitions Task Force defined sepsis as a life-threatening organ dysfunction due to a dysregulated host response to infection, which has triggered a stimulating debate in the critical care community (2). The Sepsis-3 task force acknowledged that the new criteria were not designed for children, and that future developments should consider age-specific physiology and reflect on the age-specific risk stratification. In this issue of Pediatric Critical Care Medicine, Leclerc et al (3) now report on one of the first studies evaluating a promising approach to embrace Sepsis-3 for pediatric age groups. The study (3) is based on the previously described French-Belgian cohort which was used for the development and validation of the Pediatric Logistic Organ Dysfunction (PELOD) score-2 (4), restricted to 862 children admitted in 2006/2007 to PICU with suspected infection defined by physician diagnosis. The authors tested the performance of PELOD-2 obtained during the first day of admission (d1) to identify ICU deaths in infected children. Using the most abnormal value of each scoring component observed during d1, the resulting area under the curve (AUC) was 0.91 (95% CI, 0.86–0.96), and infection/sepsis mortality increased to 10% in presence of PELOD-2 score on day 1 (d1PELOD-2) of more than or equal to 8, demonstrating the importance and validity of using a measure of organ dysfunction to discriminate children with infection at substantially higher mortality. PELOD-2 shows considerable conceptual similarities to the Sequential Organ Failure Assessment (SOFA) score, and this cohort thus ideally lends itself to the question studied. Partially aligned with Sepsis-3, the authors tested septic shock defined a priori as arterial hypotension and serum lactate more than 2 mmol/L and observed a median difference in d1PELOD-2 of 9 points between survivors and decedents, but did not integrate vasopressor treatment as a shock criterium (5). The reasons for choosing this approach are not clear, in particular, given that children with sepsis may suffer from impaired organ perfusion prior to the onset of systemic hypotension (6, 7). Finally, the authors investigated the predictive performance of a quick PELOD-2 (Glasgow Coma Scale < 11, tachycardia, and systemic hypotension) defined a priori to approximate the quick SOFA (qSOFA) (altered mentation, tachypnea, and hypotension), which compared favorably with studies validating qSOFA in adults (AUC, 0.82; 95% CI, 0.76–0.87) (8).
In view of these results, the authors are to be congratulated to shed light onto some of the challenges ahead for the endeavor of translating Sepsis-3 to critically ill children. Although their findings are supported by the previously described importance of new or progressive multiorgan dysfunction in children with sepsis (9, 10), a number of further questions arise.
First, how should organ dysfunction best be captured, both as a severity marker as well as an important morbidity endpoint given the strong association with mortality and long-term impairment? Severe sepsis as per the 2005 consensus definition (6) was defined as sepsis with organ dysfunction, with more weight given to cardiovascular and respiratory dysfunction, listing 17 potential variables. Yet, in the Sepsis Prevalence, Outcomes, and Therapies Study, the risk of death observed with neurologic, gastrointestinal, hepatic, or renal dysfunction was similar or exceeding the risk increase associated with cardiovascular dysfunction (9), supporting the use of a scoring system allocating similar weight to different organs, with higher subscores reflecting increasing severity of individual organ dysfunction.
Second, a key aspect that has received little attention in the debate around Sepsis-3 relates to the intricate relationship between timing, severity, and diagnosis of “sepsis.