Neonatal Fellow Children's Mercy Hospitals and Clinics Children's Mercy Hospitals and Clinics Kansas City, Missouri, United States
Background: Sepsis is the most common indirect cause of acute lung injury (ALI) in adult and pediatric patients, yet no studies have described this phenomenon in preterm infants. We hypothesize that sepsis-induced ALI occurs in neonates and increases the risk of bronchopulmonary dysplasia (BPD) in preterm infants.
Objective: 1) To investigate whether late onset sepsis (LOS) and other systemic inflammatory diseases are temporally correlated with ALI as quantified by pulmonary severity score (PSS) trajectory. 2) To determine the sepsis subtypes associated with the greatest severity of ALI.
Design/Methods: We included infants < 31 weeks gestational age and < 1500 grams with LOS in this retrospective case control study. The PSS was calculated at 72, 48, and 24 hours (hr) prior to, at the time of, and 24, 48, 72, and 168hr after sepsis diagnosis. We further defined rule out (RO) sepsis (blood culture negative; antibiotics continued only 48-72hr), blood culture positive (Cx+) sepsis, necrotizing enterocolitis (NEC), urinary tract infection (UTI), and culture negative (Cx-) clinical sepsis (blood culture negative; antibiotics continued > 6 days; specific lab criteria met). PSS trajectories for confirmed sepsis events (n=211) and RO sepsis events (n=123) were analyzed for signs of ALI using linear mixed-effects models.
Results: Of those admitted, 168/506 had at least 1 episode of sepsis. There were 211 total episodes of confirmed sepsis analyzed for ALI. The incidence of Cx+ sepsis was 15.6% (Figure 1). Between the baseline window of -72hr to -24hr, infants with RO sepsis had similar PSS scores as infants with confirmed sepsis. There was a significant increase in PSS scores from baseline to 0hr, +24hr, +48hr and subsequent time points with a peak at 0hr. Infants with confirmed sepsis had higher mean PSS scores compared to RO sepsis at +24, +48, +72, and +168hr demarking a distinction in ALI between RO and confirmed sepsis (Figure 2). When analyzed by sepsis subtypes, we observed similar trends with increasing PSS scores across all sepsis phenotypes and with diminishing severity from Cx+ > NEC > UTI > Cx- (Figure 3). Conclusion(s): Our study indicates that LOS in neonates is temporally associated with an increase in PSS after sepsis diagnosis, implying ALI. This is the first study to examine and report ALI in relation to postnatal sepsis subtypes in preterm infants. Future studies will examine whether infants who develop sepsis-induced ALI are at increased risk of moderate or severe BPD.
Authors/Institutions: Megan H. Tucker, Children's Mercy Hospitals and Clinics, Kansas City, Missouri, United States; Hung-Wen Yeh, Children's Mercy Hospitals and Clinics, Kansas City, Missouri, United States; Daniel K. Oh, University of Missouri Kansas City, Kansas City, Missouri, United States; Nicole Shaw, Hurley Medical Center, Flint, Michigan, United States; Navin Kumar, Hurley Medical Center, Flint, Michigan, United States; Venkatesh Sampath, Children's Mercy Hospitals and Clinics, Kansas City, Missouri, United States