Abstract
BACKGROUND: Empiric antibiotic therapy for suspected sepsis is a common practice in neonatal intensive care units (NICU). While standard practice is to discontinue antibiotic therapy when blood cultures remain negative after 48 hours, it remains unclear whether this is the true time range to positivity of blood cultures in the neonatal population across all gestational ages. Gaining an empirical understanding of the time required to detect organisms implicated in neonatal sepsis may allow for a reduction of antibiotic exposure in non-septic neonates. OBJECTIVES: The objectives of this study were to determine (1) the time required for neonatal blood cultures for suspected sepsis to become positive; (2) the proportion of cultures that become positive 24 hours or greater after collection; (3) differences between time to positivity for early versus late onset sepsis and; (4) perinatal and neonatal factors related to time to positivity of blood culture. DESIGN/METHODS: This was a retrospective observational study of blood cultures drawn in neonates within the first 30 days of life with initiation of empiric antibiotic therapy between January 1, 2013 to December 31, 2017 at a level III Canadian NICU. Blood cultures drawn after initiation of antibiotic therapy were excluded. Data was analyzed using descriptive statistics, chi-square and student t-tests to examine differences between characteristics of early versus late onset sepsis, and multivariable logistic regression analyses to examine the relationship between select perinatal and neonatal variables on time to blood culture positivity. This research was approved by the institutional Research Ethics Board. RESULTS: A total of 2213 blood cultures were drawn during the study period. 125 positive blood cultures (5.6%) were identified, out of which 20 were deemed as likely contaminant and hence excluded from the analysis. The median time to positivity of blood cultures was 15 hours (Interquartile range 10–22.3 hours). 22% of positive cultures and 1% of all cultures drawn became positive after 24 hours. Only 8% of the positive cultures were positive after 48 hours. Time to positivity of blood cultures did not differ significantly between suspected early onset sepsis versus late onset sepsis (p=0.75). In the logistic regression analysis that included gestational age, clinical chorioamnionitis, timing of sepsis onset, isolated organism, pre-treatment platelet count, and C-reactive protein, only gestational age was significantly inversely associated with time to positivity (Î(2)=-1.31; p=0.01). CONCLUSION: Results suggest that with the 48-hour empiric antibiotic rule, a large number of neonates potentially receive unnecessary doses of antibiotics as only 1% of all blood cultures become positive after 24 h. A shorter duration of empiric antibiotic therapy may be considered to reduce unnecessary antibiotic exposure in neonates and development of antibiotic resistance in the NICU.