Abstract:
Neonatal sepsis (NS) is the third most common contributor to neonatal deaths worldwide, the majority of which
occur within the first 72 hours of life (Early-onset sepsis [(EOS)]). Diagnosis of EOS is challenging due to limitations
with blood volume, poor sensitivity of culture, delay in culture results, and most importantly, lack of bacterial blood
culture capacity in high burden settings. Current syndromic algorithms for diagnosis of EOS lack validations and are
needed to enable clinical decision making for management. To evaluate the diagnostic performance of a severe
illness syndromic algorithm in distinguishing culture-proven or probable EOS from unlikely sepsis. Neonates with
their mothers with suspected neonatal sepsis that gave a written consent and made the enrollment criteria that
fulfilled the WHO case definition of septicemia within the first 72 hours of life were enrolled from maternity and
newborn units at Kisii and Homa Bay District hospitals in Kenya. Blood samples (1-2 mLs) for culture were collected
and cultured for bacteria. Between April 2015 to Jan 2016, Out of the 256 newborns infants were enrolled. Fourteen
(5.7%) infants had a bacterial pathogen identified on culture, 3 were, 1 Escherichia coli, 1 Klebsiella, 1
Staphylococcus 1 Aureus and 3 Enterobacter spp. number at risk, 14 had sepsis giving an early onset sepsis
prevalence of 5.7%, (81.6%) had a negative culture but had probable sepsis and (13.29%) Of the confirmed sepsis
the majority the neonates had more than one neonatal and maternal factor of which premature rapture of
membranes (PROM) was the most common maternal risk factor and refusal to feed and chest in drawing were the
most common clinical featured. Out of the 223, which were followed up to day 7, we had 18(7.03%) death of
probable sepsis and 0(0.00) of confirmed sepsis.
