Diagnostic Reliability of Salivary C-Reactive Protein as an Alternative Noninvasive Biomarker of Neonatal Sepsis.

To assess if salivary C-reactive protein (CRP) can be detected in neonatal sepsis and correlate the levels of salivary and serum CRP. This analytical cross-sectional study included all neonates £28 days of life with suspected sepsis or with perinatal risk factors for sepsis. Saliva was collected using an absorbent swab and analyzed by enzyme-linked immunosorbent assay, along with serum CRP. Salivary CRP was detectable in 135 subjects (99%). An increase was seen in median (IQR) levels from 0.25 (0.13,0.3) ng/mL in clinical sepsis group to 0.6 (0.3,1.4) ng/mL in screen positive/blood culture negative group, and to 1.98 (0.54, 2.95) ng/mL in blood culture positive group. There was a moderate positive correlation between salivary and serum CRP (r=0.63, P value 0.01). On receiver-operator characteristics curve, the area under the curve of salivary CRP for predicting serum CRP ≥10 mg/L was 0.861 (95% CI, 0.78 to 0.94; P <0.001), with the optimal salivary CRP cut-off being 0.6 ng/mL. Salivary CRP could be used as an alternative biomarker of neonatal sepsis.