Risk Factors and Outcomes of Extensively Drug-Resistant Gram-Negative Bacilli in Neonates with Late-Onset Sepsis

Background/Objective: To identify the risks and outcomes of extensively drug-resistant Gram-negative bacilli (XDR-GNB) in neonates. Methods: This retrospective case-control study (1995-2024) included neonates with late-onset sepsis (n = 132) and XDR-GNB bacteremia (n = 26) compared with those without XDR-GNB (n = 106). Results: Median gestational age was 31 weeks and birth weight 1540 g. The prevalence of XDR-GNB was 19.7%. The most common XDR-GNB and non-XDR-GNB pathogens were Acinetobacter baumannii and Klebsiella pneumoniae. Sepsis onset occurred earlier in the XDR-GNB group than in the non-XDR-GNB group (7.0 vs. 12.5 days, p = 0.005). In multivariable analysis using Firth's penalized likelihood method, the XDR-GNB group was more likely to have gastrointestinal anomalies (adjusted odds ratio 3.81, 95% confidence interval 1.24-12.01, p = 0.02) and history of umbilical arterial catheterization (adjusted odds ratio 3.04, 95% confidence interval 1.21-7.95, p = 0.02) compared to the non-XDR-GNB group. The XDR-GNB group had higher rates of septic shock (50.0% vs. 18.9%, p = 0.002) and inadequate empiric antimicrobial therapy (34.6% vs. 13.2%, p = 0.02). The non-susceptibility rates to third-generation cephalosporins, gentamicin, carbapenems, amikacin, and colistin were 83.3%, 58.3%, 48.1%, 30.4%, and 4.4%, respectively. Conclusions: Empirical colistin treatment is warranted for neonates in high-XDR environments who exhibit septic shock and have specific risk factors, such as gastrointestinal anomalies or the presence of an umbilical arterial catheter. Multimodal interventions, including antimicrobial stewardship programs, have been used to prevent or reduce the incidence of neonatal XDR-GNB sepsis.