Fellow University of California Irvine Orange, California, United States
Background: Culture negative early onset sepsis (EOS) is a clinical diagnosis that is often treated with broad spectrum antibiotics. Some institutions utilize cultures obtained from fetal umbilical venous blood (FUVB) as a replacement or in conjunction with infant blood cultures. Objective: We aimed to evaluate the utility and sensitivity of FUVB cultures in directing duration and choice of antibiotic for cases that may have otherwise been treated broadly as “culture negative” EOS. Design/Methods: This is a single-center, paired-sampling prospective study carried out at UC Irvine Medical Center. Over the study period (9/1/2017 to 10/31/2022), all patients admitted to our NICU were eligible for inclusion. Neonatal admission blood culture was drawn per standard of care. For every positive FUVB culture, risk factors for EOS were collected. Five criteria were established to define true positive FUVB culture vs contaminants. These included 1) time to positivity < 24h for aerobes and < 36h for anaerobes, 2) the bacterium is among the usual culprits that cause EOS, 3) the culture grew 2 or fewer bacteria, 4) the presence of at least one risk factor for EOS, and 5) at least 1 mL of blood volume recovered. Results: Paired samples were collected on 404 patients out of 8095 deliveries including 2234 NICU admissions. There were 20 true infections and 6 contaminants, leading to a 5% incidence of EOS with culture positivity in this high-risk group. 19 of 20 (95%) of blood cultures were positive in FUVB only. 10 of 20 patients (50%) were premature and 7 of 20 (35%) were delivered via c-section. Gestational age ranged from 23 3/7 to 41 weeks and BW ranged from 633 g to 4670 g. Bacteria recovered from FUVB included 9 E. coli (45%), 3 E. faecalis (16%), 2 GBS (10%), 2 Viridans streptococci (10%), and 1 each for Klebsiella, Proteus and Bacteroides (5% each). There was one case of positive infant blood culture (E. coli) with a negative concurrent FUVB sample out of the 404 paired cultures (1/404 = 0.2%). Based on clinical course, in addition to inflammatory markers and placental histology, 5 patients of 20 were treated with 2-3 days of antibiotics and the rest were treated with a 10-day tailored course. Bacterial contaminants included 3 CONS, one 2 Viridans streptococci and 2 mixed microorganisms, for a contamination rate of 1.5%.
Conclusion(s): Culture-negative EOS remains a challenge in the absence of culture data to drive tailored antibiotic therapy decisions. FUVB culture is a useful additional tool to inform antibiotic decisions for EOS as it’s more sensitive and reveals more pathogens than recovered by infant’s blood culture.
