We characterised the turn-around-times for pathogen identification with antibiotic susceptibility, and outcomes in newborn infants with early-onset bacterial sepsis (EOS).
Methods Eighty infants with EOS were retrospectively reviewed. EOS was defined by isolation of a pathogen from blood culture drawn within 72 h of birth and antibiotic treatment for ≥5 days.
Results Thirty-seven of the 80 infants were deemed to have true EOS, and 43 were deemed contaminants. The organisms grown in true EOS cases were: E. Coli in 16, Group B Streptococcus in 10, Alpha hemolytic Streptococci in 6, and others in 5.
The median (25%-75% IQR) time noted from blood culture positivity to identification of the organisms with susceptibility testing was almost 4 times longer compared to the time from collection of blood culture specimens to blood culture positivity (79 h, IQR 52 h–101 h, versus 19 h, IQR 16 h–21 h, p < 0.0001) in true cases of EOS. The contaminants took longer to identify compared to true cases (p < 0.05).
Four infants died of gram negative sepsis. Two of these infants with ampicillin resistant E. Coli died from delayed implementation of appropriate organism-specific antibiotic treatment as the susceptibility results took too long to become available.
Conclusions Definitive identification of the pathogen with the currently used laboratory methods take too long affecting outcome of infants with EOS. Empiric antibiotics were continued too long unnecessarily because of delayed identification of the contaminants. Rapid identification of an organism to a species level utilising newer technologies needs to be developed.