LAUSR

Dear Editor, Acute kidney injury (AKI) is common in the pediatric intensive-care unit (PICU) [1]. Similarly, sepsis—the most common cause of AKI in critically ill patients—is frequently encountered [2, 3]. While both are independently associated with poor outcomes in children [1, 3], few data exist examining the attributable mortality from sepsis-associated AKI (SA-AKI) [2]. Assessment of this association is challenging due to confounding by other factors that can contribute to sepsis mortality. While illness severity scores can be used to help adjust for this issue, they are non-specific for pediatric sepsis-related mortality. In this study, we used PERSEVERE-II—a validated tool for estimating baseline mortality risk in pediatric septic shock that incorporates five biomarkers and platelet count measured in the first 24 h [4]—to test for an independent association between severe SA-AKI and mortality. The results of this study have been previously submitted as an abstract [5]. We performed a secondary analysis of a prospective, observational study of 461 children with septic shock admitted to 14 PICUs from 2015 to 2018. A total of 379 patients were included after inclusion and exclusion criteria were applied (Supplement 1). Each subject was assigned a PERSEVERE-II baseline mortality probability as outlined above [4]. We defined severe SA-AKI as Kidney Diseases Improving Global Outcomes stage 2 or greater by serum creatinine (SCr) at day 3 of septic shock (severe D3 SA-AKI). Our primary analysis utilized multivariable logistic regression to test the association between severe D3 SA-AKI and 28-day mortality, after correction for age, illness severity (PRISM III), PERSEVERE-II, the presence of comorbidities, and immune status. A secondary analysis utilized multivariable linear regression to examine the association between severe D3 SA-AKI and PICU-free days. Further details of our methods are provided in Supplement 1 and have been documented previously [4]. Forty-two patients (11%) did not survive to 28 days and 65 (17%) had severe D3 SA-AKI. PERSEVERE-II had an AUC of 0.83 (95% CI: 0.77–0.90) for differentiating between survivors and non-survivors. After adjustment for the aforementioned confounders, severe D3 SAAKI was independently associated with increased odds of mortality (OR 5.6, 95% CI: 2.6–12.5, p < 0.001) and decreased PICU-free days (β = − 7.2, p < 0.001). Figure 1a shows the 28-day survival curves according to AKI stage on day 3 of septic shock (p < 0.001, log-rank survival), and Fig. 1b illustrates how median PICU-free days decreased proportionally with increasing severity of AKI (p < 0.001, Kruskal–Wallis ANOVA on Ranks). Complete results are provided in Supplement 1. In summary, after accounting for PERSEVERE-II and other confounders, severe D3 SA-AKI is independently associated with increased risk of mortality and fewer PICU-free days in a large cohort of children with septic shock. Noted limitations to consider include the reliance on estimated baseline SCr and the lack of urine output *Correspondence: [email protected] 1 Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, MLC 2005, Cincinnati, OH 45229, USA Full author information is available at the end of the article