Usefulness of c- reactive protein and newborn scale sepsis for screening of neonatal sepsis at Muhimbili national hospital in Dar es salaam.

Abstract

Background Significant challenges exist in the diagnosis of neonatal septicaemia due to the non specific nature of clinical presentation as well as problems in laboratory confirmation of suspected cases. In most deprived countries often diagnosis is made on the basis of clinical suspicion and antibiotics are prescribed empirically. Objective To determine prevalence of neonatal sepsis, spectrum of etiological agents and their antibiotic susceptibility pattern and evaluate the efficacy of Rubarth’s newborn of sepsis and C- reactive protein in the diagnosis of neonatal sepsis using blood culture as a gold standard. Methodology: This was a hospital based prospective cross sectional study conducted at the neonatal ward of the Muhimbili National Hospital in Dar es Salaam city from March 2012 to April 2013. Participants were neonates who had clinical features suggestive of neonatal sepsis by WHO criteria.Babies were physically examined and investigations which included; a bloodfor culture and sensitivity, full blood picture andtwo serial measurements of C – reactive protein (CRP), taken 12 hours apart were collected. Rubarth’s newborn scale of sepsis which is based on combining physical findings and FBP results, was used for scoring sepsis. The efficacy of Rubarth’s newborn scale of sepsis and serial CRP was assesses by calculating sensitivity, specificity, negative and positive predictive values as well as likelihood ratios (LHR). Blood culture was used as a gold standard and if the confidence did not include 1 and p value of < 0.05 was considered statistically significant. Results: A total of 208 neonateswere recruited out of whom 40 (19.2%) blood culture had a positive blood culture. The isolated organisms were Klebsiella species (35%), Coagulase Negative viii Staphylococus (30%), Staphylococcus aureus (10%), Escherichiacoli (22.5%), and Pseudomonas species (2.5%). The overall resistance to the WHO recommended first line antibiotics was for cloxacillin 100% , ampicillin 96.4% and gentamicin 39% while for second line drug ceftriaxone was 41%. Resistance to vancomycin and amikacin was 53.6% and 13.9% respectively. Clinical features that were significantly associated with neonatal sepsis were fever (67.3%), fast breathing (79.8%),and low muscle tone (38.9%) and this was statistically significant withp< 0.001.The first CRPhadsensitivity of 87.5% and specificity of 70.8% , the likelihood ratio for positive test was 3, while LHR for negative tests 0.18 both with a p value of <0.001. Combined CRP had sensitivity of 97.4% and specificity of 53.7%. Rubarth’sneonatal score of sepsis had sensitivity of 65% and specificity of 79.7% serial CRP had sensitivity of97.4%. Combining CRP and neonatal scale of sepsis reduced sensitivity to 94.71%as compared to when CRP was used alone. Conclusion: Serial CRP has a high sensitivity of 97.4% to diagnose neonatal sepsis. The etiological agents to neonatal sepsiswereidentified to be Klebsiellaspp, CoNS, S. aureus,E. coli, and Pseudomonas spp. These agents were found to be sensitive to amikacin and resistant to the currently prescribed first line antibiotics and moderate resistant to the second line drugs whose use can be guided by sensitivity results. Recommendation:In the management of suspected neonatal sepsis, serial CRP done twelve hours apart in combination with Rubarth’s Neonatal scale of sepsis can be used for screening of neonatal sepsis. Once the results for the screening tests are available a blood sample for culture can be taken for those neonates with a positive test. There is also a room for follow up studies on usefulness of CRP in combination with RNSOS in monitoring for response to treatment. Because of high resistance observed, there is a need to have follow up studies to assess the efficacy of the current first line antibiotics used in the management of neonatal sepsis.