Use of Serial C-reactive Protein Measure-ments for Determination of the Length of Empiric Antibiotic Therapy in Suspected Neonatal Sepsis

J. Khashabi, M. Karamiyar, H. Taghinejhad, M. Shirazi


Background: The management of neonatal sepsis especially in developing countries is problematic.  There is no single reliable marker of infection available at the present.  C-reactive protein (CRP) has long been used as a marker of infection. Serial measurements of CRP are recommended as a guide for duration of antibiotic therapy.


Objective: To evaluate the serial CRP measurement as a guideline for diagnosis and monitoring therapy and determining the length of antibiotic treatment in suspected neonatal sepsis.


Methods: The present descriptive study involves newborns suspected of having bacterial sepsis.  CRP levels were measured initially and at least twice at 24-hr intervals until blood culture results were available.  Antibiotic therapy started in neonates with clinical signs and symptoms of suspected sepsis. In neonates with negative blood culture, normal CRP (<6mg/L) was used as a criterion for the length of antibiotic treatment provided that the infants were in good clinical condition.  These neonates were followed up for one month after discharge.


Results:  Antibiotic therapy was discontinued in 91 patients who had negative blood cultures and two consecutive normal CRP levels.  These neonates were followed for one month after discharge.  Only one patient was re-admitted with pneumonia 21 days after discharge, giving negative predictive value of 98.9% (CI95%  96.8%-100%).  The mean±SD duration of treatment was 3.3±1.0 days in the study group and 5.9±1.7 days in neonates prior to conducting the study (p<0.000).


Conclusion: Serial CRP measurement is a good practical guide for discontinuing antibiotic therapy in neonates with suspected sepsis.  These neonates can be discharged from the hospital earlier, with significantly reduced cost, complications of treatment and family anxiety. 

Full Text:

View Counter: Abstract | 576 | and PDF | 34 |


  • There are currently no refbacks.

pISSN: 0253-0716         eISSN: 1735-3688