Background: Neonatal sepsis is a serious systemic infection that might be fatal. Early diagnosis is challenging. The c-reactive protein (CRP)/albumin ratio is a promising biomarker that may help with early diagnosis and timely management of sepsis.
Aim of the work: to study the sensitivity and specificity of the CRP/albumin ratio as a diagnostic marker of neonatal sepsis in preterm newborns.
Patients and Methods: in this case-control study CRP/albumin ratio was estimated in 154 preterm newborns admitted to Neonatal Intensive Care Units (NICUs) of Cairo University Children's Hospitals. They were divided into 2 groups. Group (I), the sepsis group “based on positive blood culture" (n=54), and group (II), a control group, with no clinical signs or laboratory evidence of sepsis (n=100).
Results: The mean age of newborns in the sepsis group was 3 ± 2.5 days vs 4.5 ± 2.5 days for the control group. The sepsis group had a mean ± SD CRP/albumin ratio higher than that of the control group (18.5 ± 20 vs 0.6± 0.7, p= 0.00). 30 (55%) of those in the sepsis group died; they had a mean ± SD CRP/albumin ratio of 20± 21, vs 15 ± 17 for those who survived (p value=0.3). Neonatal Sepsis was best diagnosed by the CRP/albumin ratio above the cutoff value of 1.5 with 100 % sensitivity, 95% specificity, 91.5 % positive predictive value (PPV), 100 % negative predictive value (NPV), (area under the curve (AUC)= 0.99, and 95 % confidence interval (CI) 0.983 - 1), which was not different than that of CRP alone (p=0.800). CRP above the cutoff point of 4.3 showed 96.3 % sensitivity, 96% specificity, 92.9 % PPV, 98 % NPV, (AUC= 0.99 , 95 % CI  0.983 - 1). The I/T ratio (shift to the left) above the cutoff point of 0.2 showed 64.8% sensitivity, 100 % specificity, 100% PPV and 84% NPV (AUC= 0.94 , 95% CI 0.891 - 0.973).
Conclusion: CRP/albumin ratio is a highly sensitive and specific early diagnostic biomarker for neonatal sepsis in premature newborns that has marginal superiority to the CRP in the  diagnosis of neonatal bacterial sepsis. More studies are needed to validate the sensitivity and specificity of a score that combines both the CRP/ albumin ratio above 1.2 and the I/T ratio above 0.2 in predication of true positive cases and exclusion of the true negative cases.