Role of serial C-reactive protein measurements for determination of the duration of empirical antibiotic therapy in suspected neonatal sepsis

Abstract

Background: Neonatal sepsis (NS) is a clinical syndrome characterized by signs and symptoms of infection with or without bacteremia in neonatal age. At present, there are a variety of tests available for testing sepsis in neonates including C-reactive protein (CRP). Objective: The objective of the study was to decrease the antibiotic duration in a case of NS and to decrease the hospital stay and financial burden on the patient by measuring serial CRP levels. Materials and Methods: This case–control study was conducted in a tertiary care teaching hospital at a tertiary hospital of central India. A total of 103 neonates were considered as the study population. Estimation of serial CRP levels among neonates was assessed and correlated within two study groups. Results: CRP level at admission was able to successfully identify NS in 66.67% of the cases (sensitivity), while in 12.07% it was able to negate NS when it was absent on blood culture also (specificity). The positive predictive value (PPV) was low (37.04%) with a negative predictive value of 31.82%. The diagnostic accuracy of the CRP level at admission was 35.92% suggesting its accuracy to correctly identify the patient with NS in only 35.92% cases. The results show that CRP is having a good sensitivity, with a poor specificity, and overall low diagnostic accuracy. Conclusion: Our study showed a significant difference in the use of antibiotics in the two groups. The sensitivity of CRP was high with the low specificity and low PPV and high NPV, which makes CRP a vital screening tool, especially if used in conjunction with other screening parameters improving the overall sensitivity and specificity.