Clinical factors associated with carbapenem resistance and in-hospital mortality in Klebsiella pneumoniae bloodstream infection: a 7-year retrospective cohort study

Abstract

Background Carbapenem-resistant Klebsiella pneumoniae bloodstream infection (CRKP BSI) is associated with substantial mortality and limited therapeutic options, yet institutional epidemiology and risk profiles vary. We evaluated antimicrobial susceptibility, non-blood-site detection, factors associated with carbapenem resistance, and in-hospital all-cause mortality among patients with K. pneumoniae BSI at a secondary-care hospital over seven years. Methods We retrospectively analyzed 141 adults with K. pneumoniae BSI hospitalized at Shidong Hospital, Shanghai, China, from January 2019 through December 2025, including 49 patients with CRKP BSI and 92 with carbapenem-susceptible K. pneumoniae BSI (CSKP BSI). Clinical and microbiological data were analyzed using univariable comparisons and multivariable logistic regression. Results Crude in-hospital mortality was higher among patients with CRKP BSI than among those with CSKP BSI (40.8% vs. 16.3%; p = 0.002). Among the tested CRKP isolates, high in vitro susceptibility to ceftazidime/avibactam and tigecycline was observed. In multivariable analysis, pulmonary disease (aOR, 9.441; 95% CI, 2.958–30.136; p < 0.001), urinary catheterization before BSI onset (aOR, 13.821; 95% CI, 3.987–47.904; p < 0.001), prior carbapenem exposure (aOR, 3.982; 95% CI, 1.274–12.448; p = 0.018), and prior cephalosporin exposure (aOR, 11.341; 95% CI, 3.411–37.713; p < 0.001) were associated with carbapenem resistance. Non-blood-site detection of K. pneumoniae was more frequent in CRKP BSI than in CSKP BSI (73.5% vs. 35.9%; p < 0.001). In the overall mortality model, sepsis or septic shock assessed at BSI onset or within 48 h thereafter was associated with higher odds of in-hospital mortality (aOR, 4.738; 95% CI, 1.916–11.720; p < 0.001), whereas the adjusted association of CRKP BSI status with mortality did not reach statistical significance. Conclusion In multivariable analysis, pulmonary disease, urinary catheterization before BSI onset, prior carbapenem exposure, and prior cephalosporin exposure were associated with carbapenem resistance. Although crude mortality was higher among patients with CRKP BSI, sepsis or septic shock was associated with higher odds of mortality after adjustment for age and CRKP BSI status, whereas the adjusted association of CRKP BSI status with mortality did not reach statistical significance. These findings support targeted infection prevention, antimicrobial stewardship, appropriate urinary catheter use, and early recognition and management of sepsis.