Interleukin-6 as a sepsis biomarker in neonatal and pediatric care: A retrospective real world cohort study.

UNLABELLED: This study aimed to assess the diagnostic performance of Interleukin-6 (IL-6) in comparison to other biomarkers (C-reactive protein [CRP], procalcitonin [PCT], neutrophil-lymphocyte ratio [NLR]) in a pediatric and neonatal setting. This retrospective cohort study in 2 Dublin hospitals (one pediatric and one maternity/neonatal) included all pediatric and neonatal patients who had IL-6 testing for clinically suspected sepsis in a one-year period. Cases were categorized according to clinical, microbiological and radiological criteria. Paediatric cases were described along two axes -physiological (from normal to septic shock) and etiological (bacterial, viral or no infection), while neonatal sepsis was defined from a composite of bacteremia, septic shock and clinical sepsis. Biomarker performance was assessed by area under the receiver operating characteristic curves (AUROC) and measures of diagnostic accuracy at optimal cut-offs. Overall, 180 cases were included (111 pediatric cases and 69 neonates). Amongst pediatric cases 34.2% were classified as sepsis or septic shock, and 55.8% had bacterial infections. Among neonates, 49.3% met sepsis criteria. In the pediatric analysis all biomarkers had statistically significant differences between physiological and etiological categories, although only IL-6 and PCT showed stepwise progression along both axes. IL-6 had statistically higher AUROC values than all other biomarkers for the primary physiological and etiological axes outcomes (0.85 [Interquartile range 0.78, 0.93] and 0.91 [IQR 0.85, 0.96], with higher sensitivity than other biomarkers. Over time, IL-6 showed significantly greater negative rates of change. In neonates, all biomarkers other than CRP showed statistically significant differences between categories, although IL-6 alone showed clinically significant differences. For neonatal sepsis the IL-6 AUROC was 0.86 (IQR 0.77, 0.95), statistically higher than CRP and NLR. However, sensitivity was lower than other analyses (67.6%), although this increased to 81.2% when the sepsis category was restricted to culture positive or shock cases only. CONCLUSION: IL-6 had superior diagnostic performance to most other sepsis biomarkers across axes, with high sensitivities and AUROC, in a broad general pediatric and neonatal population. Combined with favorable biomarker kinetics over serial sampling, these findings encourage further prospective evaluation and evaluation of the impact on patient outcomes. WHAT IS KNOWN: • Interleukin-6 (IL-6) is a promising sepsis diagnostic biomarker, due to its kinetics it more closely temporally mirrors infectious or inflammatory insults. • Evaluation of IL-6 in pediatric sepsis has been limited to specific populations, with limited evaluation in a broad general pediatric population of children with suspected sepsis. WHAT IS NEW: • IL-6 is a discerning biomarker, categorizing children into physiological and etiological categories with higher diagnostic accuracy than other biomarkers in common use. • Sensitivity for both bacterial infection and physiological compromise was high for pediatric cases, but lower in neonates.