Utility of Cytokines to Predict Neonatal Sepsis.

Qing Ye,Li-zhong Du,Wen-Xia Shao,Shi-qiang Shang
DOI: https://doi.org/10.1038/pr.2016.267
IF: 3.953
2016-01-01
Pediatric Research
Abstract:Background: Sepsis is an important cause of neonatal morbidity and mortality worldwide. Diagnosis and treatment of neonatal sepsis relies on clinical judgment and interpretation of nonspecific laboratory tests. In a prospective cohort, we measured inflammatory cytokines as a potential biomarker for neonatal sepsis. Methods: Serum inflammatory cytokine levels were evaluated in the early stage of neonatal sepsis and after antimicrobial treatment. Receiver operating characteristic curves assessed the diagnostic value of cytokines. We performed multiple logistic regression analysis to characterize the role of each cytokine independently for infants with culture proven sepsis. Results: C-reactive protein, interleukin (IL)-6, IL-10 and IL-6/IL-10 levels were significantly elevated in neonatal sepsis when compared with the control group and there were 1.4 (95% confidence interval (CI): 1.2–1.5), 4.9 (95% CI: 4.6–5.1), 5.1 (95% CI: 4.5–5.6), and 10.2 (95% CI: 9.2–11.1) fold greater odds, respectively, to predict neonatal sepsis when increased. After effective treatment, median IL-6 (pretreatment value: 263.0 pg/ml and post-treatment value: 7.4 pg/ml) and IL-6/IL-10 levels (pretreatment value: 16.6 and post-treatment value: 1.4) significantly decreased. The areas under the curve for IL-6, IL-10, IL-6/IL-10 and C-reactive protein for differential diagnosis were 0.98, 0.82, 0.90, and 0.88, respectively. Conclusion: IL-6 and IL-6/IL-10 outperformed C-reactive protein to diagnose neonatal sepsis. Of the cytokines studied, IL-6 was the most sensitive, whereas IL-6/IL-10 was the most specific predictor of neonatal sepsis.
What problem does this paper attempt to address?