Biomarkers of Sepsis: A Retrospective Approach

Jose M. Roberts, The University of Akron



Background: Biomarkers are taking the spot light in becoming the norm for early diagnoses. Sepsis is an inflammatory disease that increases metabolic rate in children. The first biomarker is hyponatremia. Hyponatremia is a frequent electrolyte imbalance in clinical practice, often observed in children with inflmmatrroy disesase and infection. .Presence of hyponatremia is associated with electrical signaling imbalances, inflammation and renal dysfunction. The clinical value of hyponatremia in febrile pediatric patients is unknown. The C-reactive protein is the second biomarker. Its presence signifies dying cells and inflammation are present.

Objectives: To evaluate the use of biomarkers in children seen in the Emergency Department with a diagnoses sepsis.

Design/Methods: This is a retrospective study including children between 11 months -18 years old presenting to Akron Children’s Hospital ED with fever and sepsis between January - December 2014. Electrolytes were collected in 731 patients, and 31 patients were excluded due to immune insufficiency or diuretic use. Hyponatremia was defined as serum sodium of < 132 mEq/l, and patients were divided into two groups accordingly. Outcomes of interest included hospital admission and length of hospital stay(LOS). Categorical variables were presented as counts and percentages and were compared using χ2. The Wilcoxon Rank Sum test was applied to compare non-normally distributed variables between groups. Spearman correlation was evaluated between continuous variables. Logistic regression models were developed to evaluate admissions based on potential predictors selected at the univariate analysis. Level of significance was set at 0.05.

Results: Twelve percent (n = 84 vs. n= 616) of patients had hyponatremia. Patients with hyponatremia had higher serum CRP and lower serum Bicarb. Hyponatremia was associated with a higher admission rate and longer LOS (days) (Table 1). Spearman correlations coefficients revealed correlation between hyponatremia, Maximum Temperature in ED (rs = -0.17, P < 0.0001), Bicarb (rs = 0.13, P = 0.0008), WBC (rs = -0.08, P = 0.0449), Anion Gap (rs = 0.13, P = 0.0005), and CRP (rs = -0.23, P = 0.0004). Patient with Hyponatremia are 1.79 (P = 0.0211, 95% CI 1.08 – 2.95) more likely to be admitted.

Conclusion: In febrile children with sepsis, hyponatremia was associated with higher levels of inflammatory marker CRP, and elevated leukocytes, and acidosis. Patients with hyponatremia had higher admission rates and longer LOS. The correlation was low, but the use of hyponatremia and CRP as potential biomarkers in diagnosing sepsis early could be better understood in future studies.