Introduction

Neonatal sepsis has a high mortality and morbidity¹. Timely detection and treatment will help in decreasing mortality and morbidity. Diagnosis of neonatal sepsis is suspected in high risk neonates or in those with clinical features of sepsis. Confirmation of diagnosis is done by growing organism from blood. The main drawback of blood culture is that it takes 24-48 hours for the results to come, which has prompted the evaluation of surrogate markers of inflammation as possible tools for diagnosis of bacterial sepsis. Cytokines and C - reactive protein (CRP) levels have been evaluated in this respect ^2-8. Although several studies confirm that CRP levels are useful, there are reports to the contrary^9-12. It has been suggested that serial levels may be more useful diagnostically^15-18.

We designed this study with the aim of evaluating the role of CRP in early onset sepsis (EOS) of neonates.

Methods

This prospective study was done in the department of neonatology SKIMS Soura from January 2006 to October 2007. Neonates with birth weight >1000 grams, gestational age >28 weeks and born within 48 hours were included in the study if they had Risk factors for EOS &/or Clinical features of sepsis.

At the time of admission a complete septic screen was done which included total and differential leukocyte count, absolute neutrophils count, platelet count, CRP, chest X ray and blood culture. CRP was measured at the time of admission and 8-24 hours later, on the next morning, using a semi-quantitative technique with a detection limit of 8 mg/L (Biocientifica S.A. iturri, Buenous Aires, Argentina).

Newborn babies were classified as having sepsis if they had signs suggestive of sepsis and a positive blood culture. Probable sepsis was diagnosed if they had a negative blood culture with signs suggestive of sepsis and no sepsis if there were no clinical features of sepsis with a negative blood culture. Babies with sepsis or probable sepsis received antibiotics for about 14 days. The remaining newborns were classified as at risk of infection and received antibiotics for an average of 5 days.

Positive blood culture was considered the “gold standard” against which the performance of CRP was compared.

Results

A total of 225 neonates were enrolled for the study. The demographic characteristics of these neonates are shown in table1.

Figure 1

TABLE 1: DEMOGRAPHIC CHARACTERISTICS

Klebsiella pneumoniea was the commonest organism isolated. Majority of the babies were term deliveries. Perinatal asphyxia was present in a significant no. of cases. The commonest risk factor was premature rupture of membranes. Refusal to feed and temperature instability were the commonest presentations (table2).

Figure 2

TABLE 3: CLINICAL CHARACTERISTICS

Performance of screening test

Discussion

Serum concentrations of CRP increase several hundredfold in response to bacterial infection, making it an attractive diagnostic test for neonatal sepsis. Several hours are needed for CRP levels to increase in serum (including activation of neutrophils, elaboration of interleukin-6, and induction of hepatic synthesis of CRP) therefore limiting the sensitivity of this test in diagnosing NNS. CRP levels are consistently elevated 24 to 48 hours after the onset of infection²⁰; therefore serial normal levels may be useful for identification of infants who do not have bacterial infection.

Our findings demonstrates that a single CRP level at the beginning of evaluation lacks sensitivity (40%), which was improved by delaying testing for 24 hours, attesting to the fact that serial levels are better than a single level. Similar findings have been reported by Nuntnarumit P¹⁵ and others who, in a series of 76 neonates with late onset sepsis (LOS), reported a high sensitivity (100%) and NPV (100%), for serial CRP levels done initially and 24-48 hours later. Jankovic B¹⁶ also pointed out the diagnostic accuracy of serial measurements. In the largest series of 1136 neonates, William E benitz¹⁷ showed serial levels to be diagnostically superior. Elizabeth Mathai¹⁹ and others in 250 neonates with risk factors for EOS did not find serial levels to be diagnostically useful. This conclusion could be due to variations in the population studied with respect to age (cord blood vs. neonatal blood) different cut off point (6 mg vs. 8 mg) and selecting only babies with risk factors etc.

Various factors apart from sepsis can cause a rise in CRP levels esp. Meconium aspiration syndrome, perinatal asphyxia, surgery, etc. therefore limiting its Positive Predictive Value. But a negative CRP rules out infection with high certainty i.e. it has a high Negative Predictive Value. Our study also shows a high NPV for CRP of > 90% at all times. This is in agreement with studies of Ehl S¹³ and others who in a group of 176 neonates with suspected sepsis showed that CRP at a cut off value of 10 mg/l had a negative predictive value of 99%. Similar results have also been shown by Jankovic B¹⁴ and others.

Therefore, in conclusion, a single CRP value done at the time of admission may not identify all neonates with sepsis; serial levels done 24 hours apart increase the sensitivity of CRP determinations. A negative CRP is more important than a positive CRP in that it excludes infection with a high degree of certainty.