601 - Revisiting Time to Blood Culture Positivity: Can We Decrease Antibiotic Exposure in the NICU?

Rachel Graf, UH Rainbow Babies & Children's Hospital, Cleveland, OH, United States; Amy Edwards, UH Rainbow Babies & Children's Hospital, Cleveland, OH, United States; Devashis Mukherjee, Case Western Reserve University School of Medicine, Cleveland, OH, United States; Moira Crowley, UH Rainbow Babies & Children's Hospital, Cleveland, OH, United States; Rita M.. Ryan, UH Rainbow Babies & Children's Hospital, Cleveland, OH, United States

Background: Sepsis is significantly associated with neonatal morbidity and mortality. Thus, empiric antibiotics are routinely started pending blood culture (BCx) results. Prolonged antibiotic durations are associated with neonatal morbidity. At Rainbow Babies and Children's Hospital (RBC) NICU we discontinue antibiotics after 36 hours (h) of no BCx growth. We investigated if we could shorten empiric antibiotic exposure duration.

Objective: 1. To identify all positive BCx time to positivity (TTP).
2. To identify factors associated with a higher likelihood of earlier vs. later TTP.

Design/Methods: All positive BCx isolated from infants admitted to RBC NICU from July 1, 2018-July 1, 2023 were included. Maternal and infant demographic, clinical, and laboratory data were obtained from the medical record. Stata 15.0 was used for statistical analysis, Wilcoxon rank sum test for continuous variables, and Fisher's exact test for categorical variables. Kruskal-Wallis test with post-hoc Dunn test was used for more than one group continuous variable analysis and linear regression for two continuous variables.

Results: Over the study period, 264 positive BCx were isolated from 125 infants with 79 (63%) having only one positive BCx. The remainder had 2 to 11 positive BCx. The 125 infants (64% male; 61% Black) had a median gestational age (GA) of 26.6 weeks (w) (IQR: 24.5, 31.3) and median birth weight of 920 grams (650, 1650) (Table 1). Their median age at first positive BCx was 14 days (IQR: 2, 33). Of all positive BCx, only 14% were drawn at < 72h of age, representing early-onset sepsis (EOS). The remaining 86% represent late-onset sepsis. The median TTP for all positive BCx was 21h (IQR: 16, 33), with 60% positive by 24h and 84% positive by 36h. There were significant differences in TTP among microorganism categories (P < 0.0001) (Table 2). Maternal fever was associated with shorter TTP for all positive first BCx (P < 0.002) (Table 1). GA significantly impacted TTP for all BCx (Table 1): For every 0.4w lower GA at birth, BCx TTP increased by 1h (P < 0.02).

Conclusion(s): Our data do not support discontinuing empiric antibiotics for neonatal sepsis at 24h since only 60% of BCx were positive at 24h. While more BCx are drawn on admission, the burden of positive NICU BCx falls in infants over two weeks of age. A clinical conundrum arises as to why lower GA contributes to a longer TTP. At lower GA, lower blood volume may be taken resulting in longer TTP. Overall, these results underline the importance of looking at an individual NICU’s data before implementing a practice change.

Table 1. Clinical and demographic characteristics of NICU infants with positive blood cultures


Table 2. Microbiological characteristics of positive NICU blood cultures


Table 1. Clinical and demographic characteristics of NICU infants with positive blood cultures


Table 2. Microbiological characteristics of positive NICU blood cultures