High 28-day mortality in resistant gram-negative bloodstream infections in the ICU: retrospective cohort study

Hospital-acquired bloodstream infections (BSI) caused by carbapenem-resistant and difficult-to-treat resistant (DTR) Gram-negative bacteria are a global threat associated with high mortality in (ICUs). The aim of this study was to evaluate 28-day mortality, independent risk factors, and the effect of empirical treatment on survival in ICU patients who developed BSI due to carbapenem-resistant Klebsiella pneumoniae , Acinetobacter baumannii , and DTR Pseudomonas aeruginosa. This retrospective observational study was conducted in the anesthesia ICUs of a 700-bed tertiary-care hospital between January 2020 and August 2025. Adult patients with hospital-acquired bloodstream infections caused by carbapenem-resistant K. pneumoniae , carbapenem-resistant A. baumannii , or DTR P. aeruginosa were included. The primary outcome was 28-day all-cause mortality. Demographic, clinical, microbiological, and treatment-related factors associated with mortality were evaluated. Of the patients included in the study, 87.4% ( n = 585) died within the 28-day follow-up period. The most frequently isolated pathogen was A. baumannii (58.3%), with 80.9% of isolates identified as extensively drug-resistant (XDR) and 13.1% as pandrug-resistant (PDR). In multivariate logistic regression analysis, advanced age (OR 1.03, p = 0.009), higher SOFA score on the day of bacteremia (OR 1.13, p = 0.009), presence of COVID-19 co-infection (OR 3.96, p = 0.034), and aminoglycoside resistance (OR 3.68, p = 0.014) were identified as independent predictors of 28-day mortality. Initiation of appropriate empirical antimicrobial therapy within the first 24 h did not show an independent effect on survival in multivariate analysis (OR 1.24, p = 0.576). Carbapenem-resistant and DTR Gram-negative BSI developing in ICUs located in endemic regions are associated with a high mortality rate. The main determinants of mortality are the clinical severity at the time of infection (high SOFA score), COVID-19 co-infection, and accompanying antibiotic resistance. In this critically ill patient population, the effect of early appropriate empirical therapy on survival may have been influenced by factors such as disease severity and organ dysfunction. Not applicable.