AbstractBackgroundCarbapenem-resistant Enterobacteriaceae (CRE) are an emerging problem in the paediatric population worldwide with high mortality rates in bloodstream infection (BSI).ObjectivesTo evaluate predictors of 30 day mortality in CRE BSI in a paediatric cohort.MethodsA retrospective observational single-centre study (December 2005–August 2018) was conducted. Cases of CRE BSI in children 0 to 16 years were included. Microbiological identification (MALDI Biotyper) and antimicrobial susceptibility testing (Vitek2® and MicroScan panel NBC44) according to EUCAST breakpoints were performed. PCR OXVIKP® was used to confirm carbapenemase genes (OXA-48, VIM, KPC, NDM). Demographic characteristics, underlying diseases, source of bacteraemia, antimicrobial therapy and outcomes were collected from medical records. Survival analysis to establish predictors of 30 day mortality was performed.ResultsThirty-eight cases were included; 76.3% were hospital-acquired infections and 23.7% related to healthcare. All patients had at least one underlying comorbidity and 52.6% were recipients of an organ transplant. VIM carbapenemase was the predominant mechanism (92.1%). Previous CRE colonization or infection rate was 52.6%. Intestinal tract (26.3%) and vascular catheter (21.1%) were the most common sources of infection. Crude mortality within 30 days was 18.4% (7/38); directly related 30 day mortality was 10.5%. Conditions associated with an increment in 30 day mortality were intensive care admission and inadequate empirical therapy (P < 0.05). Combination-antibiotic targeted treatment and a low meropenem MIC were not related to improved survival.ConclusionsCRE BSI mortality rate is high. The most important factor related to 30 day survival in our CRE BSI cohort in children was empirical treatment that included at least one active antibiotic.