Early Differentiation of Gram-Negative and Gram-Positive Bloodstream Infections Using Hematological Indices, Serum Endotoxin Activity, and Procalcitonin in Hospitalized Sepsis Patients

Prompt and accurate differences in reference to Gram-negative and Gram-positive bloodstream infections are critical for appropriate empiric therapy in case of sepsis. The potential of automated hematological indices together with serum endotoxin activity and procalcitonin for early pathogen-class stratification at the time of hospital admission was evaluated in this study. For this study, 120 participants who were at least 18 years of age and were suspected of sepsis and were admitted to the tertiary care facility were involved for a period of 12 months. The sample size was determined given the medium expected effect size (Cohen's d = 0.5) for differences in endotoxin activity for Gram-negative and Gram-positive infections. Given an 80% power and 5% significance level, 102 participants were necessary. Additional subjects were included to account for cultures that were negative or for case exclusions. Pre-antimicrobial modification and admission, blood samples were taken within six hours of admission. An automated hematology analyzer was used to perform a complete blood count with differential, and NLR, PLR, and SII were calculated. Serum endotoxin activity was measured using an endotoxin activity assay which is standardized and is based on the detection of lipopolysaccharide. Procalcitonin serum levels were conducted via an automated immunoassay that is used routinely. For the definitive etiological classification, the standard hospital blood culture results that were obtained in the standard hospital diagnostics were used exclusively. The patients with Gram-negative bacteremia (n ≈ 65) presented with considerably higher levels of endotoxin activity and procalcitonin than did Gram-positive cases (n ≈ 55) (p < 0.01). Infections of Gram-negative bacteria also had a statistically significantly higher NLR and SII values. In the multivariate analysis, the combination of endotoxin activity, procalcitonin, and NLR had improved predictive accuracy over each marker alone. This confirms the clinical value of straightforward, non-genetic biomarkers for the preliminary classification of sepsis.