Introduction: India has the highest incidence of clinical sepsis i.e.17,000/ 1,00,000 live births. In Neonatal sepsis septicaemia, pneumonia, meningitis, osteomyelitis, arthritis and urinary tract infections can be included. Mortality in the neonatal period each year account for 41% (3.6 million) of all deaths in children under 5 years and most of these deaths occur in low income countries and about one million of these deaths are due to infectious causes including neonatal sepsis, meningitis, and pneumonia. In early onset neonatal sepsis (EOS) Clinical features are non-specific and are inefficient for identifying neonates with early-onset sepsis. Culture results take up to 48 hours and may give false-positive or low-yield results because of the antenatal antibiotic exposure. Reviews of risk factors has been used globally to guide the development of management guidelines for neonatal sepsis, and it is similarly recommended that such evidence be used to inform guideline development for management of neonatal sepsis.
Material and Methods: This study was carried out using institution based cross section study . The total number neonates admitted in the hospital in given study period was 644, of which 234 were diagnosed for neonatal sepsis by the treating pediatrician based on the signs and symptoms during admission. The data was collected: Sociodemographic characteristics; maternal information; and neonatal information for neonatal sepsis like neonatal age on admission, sex, gestational age, birth weight, crying immediately at birth, and resuscitation at birth.
Results: Out of 644 neonates admitted 234 (36.34%) were diagnosed for neonatal sepsis by the paediatrician based on the signs and symptoms during admission. Of the 234 neonates, 189 (80.77%) infants were in the age range of 0 to 7 days (Early onset sepsis) while 45 (19.23%) were aged between 8 and 28 days (Late onset sepsis). Male to female ratio in our study was 53.8% and 46% respectively. Out of total 126 male neonates 91(72.2%) were having early onset sepsis while 35 (27.8%) were late onset type. Out of total 108 female neonates 89(82.4%) were having early onset sepsis while 19 (17.6%) were late onset type. Maternal risk factors were identified in 103(57.2%) of early onset sepsis cases while in late onset sepsis cases were 11(20.4%). Foul smelling liquor in early onset sepsis and in late onset sepsis was 10(5.56%) and 2 (3.70%) respectively. In early onset sepsis cases maternal UTI, Meconium stained amniotic fluid, Multipara and Premature rupture of membrane was seen in 21(11.67%), 19 (10.56%), 20(11.11%) and 33 (18.33%) cases respectively. In late onset sepsis cases maternal UTI, Meconium stained amniotic fluid, Multipara and Premature rupture of membrane was seen in 2 (3.70%), 1(1.85%), 3 (5.56%) and 3 (5.56%) cases respectively.
Conclusion: Maternal risk identification may help in the early identification and empirical antibiotic treatment in neonatal sepsis and thus mortality and morbidity can be reduced.
Deep learning approach on tabular data to predict early-onset neonatal sepsis
