To evaluate the mortality of bloodstream infections (BSI) caused by Enterobacter cloacae complex (ECC) according to the resistance profile prioritized by the World Health Organization (WHO).

Multicenter, retrospective cohort study. Adults with BSI caused by ECC were included. Duplicates, missing data, and hospital stays ≤ 2 days were excluded. Data were obtained from medical records. The outcome was 30-day mortality. Antimicrobials initiated within 7 days of BSI onset and maintained for ≥ 2 days were evaluated for adequacy (defined according to in vitro susceptibility). ECC isolates were divided into three groups according to WHO priority resistance profile: susceptible to third-generation cephalosporins and carbapenems (WHO 1); resistant to third-generation cephalosporins and susceptible to carbapenems (WHO 2); and resistant to carbapenems (WHO 3). Two univariate analyses were performed (#1 between WHO groups and #2 between deaths and survivors). Variables with P < 0.20 were included in a hierarchical Cox model.

From January 2012 to December 2023, 858 patients were screened; 241 were excluded due to duplicate isolates, 173 due to hospitalization < 2 days, 4 for being under 18 years old, and 11 due to missing data, resulting in 429 analyzed cases: WHO 1: 300 (70%), WHO 2: 103 (24%), WHO 3: 26 (6%). Mortality was 21%, 25%, and 35% among patients with ECC BSI WHO 1, 2, and 3, respectively (P = 0.20). In univariate analysis #1, the groups differed in high-risk primary site and ICU admission (more frequent in WHO 3), and in adequate empirical and definitive therapy (less frequent in WHO 3). In univariate analysis #2, age, Charlson score, Pitt score, glomerular filtration rate (GFR), vasoactive drug use (VAD), high-risk site, and polymicrobial infection were associated with mortality. Deaths received less adequate empirical and definitive therapy. In the hierarchical multivariate analysis, adjustment for age, comorbidities (Charlson, GFR), and infection severity (high-risk site, VAD, polymicrobial infection) equalized the risk of death among patients with ECC BSI WHO 1, 2, and 3. After including definitive adequate therapy, WHO 3 was associated with a lower risk of mortality (hazard ratio 0.47; 95% CI 0.20–0.93; P = 0.03).

The hierarchical multivariate analysis of our study suggests that the higher crude mortality of ECC BSI observed in the WHO 3 group was associated with greater clinical severity and lower frequency of adequate antimicrobial therapy.