Background: Coagulase-negative staphylococci (CoNS) frequently causes late-onset sepsis in preterm infants. Vancomycin is the first-line therapy, but the emergence of reduced vancomycin-susceptibility strains has resulted in linezolid use, of which long-term safety in preterm infants is unknown.
Objective: Evaluate the association between linezolid exposure and neurodevelopmental impairment (NDI) or death at 18-21 months of corrected age, in preterm infants with CoNS sepsis.
Methods: Multicentric retrospective cohort study comparing long-term outcomes of preterm infants exposed to linezolid versus other antistaphylococcal antimicrobials. We included infants ≤28 weeks' gestational age (GA), with CoNS sepsis, admitted between January 2011 and June 2015 in 3 level-3 Canadian NICUs. Primary outcome was a composite of death or significant NDI (sNDI) at 18-21 months of corrected age. Secondary outcomes included NDI and individual components of the primary outcome. We assessed the relationship between linezolid exposure and outcomes using a multivariable logistic regression.
Results: Of 274 infants included, 67 (24.4%) were exposed to linezolid. Median GA was 26 weeks and clinical characteristics were similar between groups. There was no difference in composite outcome of death or sNDI among the infants of both groups, but significantly more death by 18-21 months in the linezolid group (29.9% vs. 17.6%; P = 0.01).
Conclusions: Linezolid exposure was not associated with composite outcome of death or sNDI at 18-21 months. The association between linezolid and death may be due to indication bias. Further studies are warranted.