Central-line-associated bloodstream infection burden among Dutch neonatal intensive care units

BACKGROUND

The establishment of an epidemiological overview provides valuable insights needed for the (future) dissemination of infection-prevention initiatives.

AIM

To describe the nationwide epidemiology of central-line-associated bloodstream infections (CLABSI) among Dutch Neonatal Intensive Care Units (NICUs).

METHODS

Data from 2935 neonates born at <32 weeks' gestation and/or with a birth weight <1500 g admitted to all nine Dutch NICUs over a two-year surveillance period (2019-2020) were analysed. Variations in baseline characteristics, CLABSI incidence per 1000 central-line days, pathogen distribution and CLABSI care bundles were evaluated. Multi-variable logistic mixed-modelling was used to identify significant predictors for CLABSI.

RESULTS

A total of 1699 (58%) neonates received a central line, in which 160 CLABSI episodes were recorded. Coagulase-negative staphylococci were the most common infecting organisms of all CLABSI episodes (N=100, 63%). An almost six-fold difference in the CLABSI incidence between participating units was found (2.91-16.14 per 1000 line-days). Logistic mixed-modelling revealed longer central line dwell-time (adjusted odds ratio (aOR):1.08, P<0.001), umbilical lines (aOR:1.85, P=0.03) and single rooms (aOR:3.63, P=0.02) to be significant predictors of CLABSI. Variations in bundle elements included intravenous tubing care and antibiotic prophylaxis.

CONCLUSIONS

CLABSI remains a common problem in preterm infants in The Netherlands, with substantial variation in incidence between centres. Being the largest collection of data on the burden of neonatal CLABSI in The Netherlands, this epidemiological overview provides a solid foundation for the development of a collaborative platform for continuous surveillance, ideally leading to refinement of national evidence-based guidelines. Future efforts should focus on ensuring availability and extraction of routine patient data in aggregated formats.