Association between inappropriate empirical antimicrobial therapy and mortality in gram-negative bloodstream infections in patients with febrile neutropenia and hematological malignancy

Ariel Fernando Flórez Riaño,Oscar Julián Rojas Castro,Sigifredo Ospina,Isabel Cristina Ramírez-Sánchez
DOI: https://doi.org/10.1016/j.jiac.2024.10.006
2024-10-11
Abstract:Background and objective: Inappropriate initial antimicrobial therapy has been associated with high mortality in patients with gram-negative bacilli bloodstream infections during febrile neutropenia following chemotherapy for hematological malignancies. The aim of this study is to determine this association in our hospital. Methods: A single center, retrospective, cohort study of bloodstream infection due to gram-negative bacilli and febrile neutropenia was conducted. Clinical characteristics, microbiological etiology, antimicrobial resistance profile, empirical and targeted antibiotic therapy, intensive care unit admission, persistent bacteremia and mortality were analyzed. Results: Of the 171 episodes of bloodstream infection due to gram-negative bacilli, empirical antimicrobial therapy was inappropriate in 43 episodes (25.1 %). There was a significant difference in mortality at 7 and 30 days between patients who received appropriate versus inappropriate empirical treatment (4.6 % versus 13.9 %, p = 0.04; 15.6 % versus 32.5 %, p = 0.016). Inappropriate empirical treatment (RR, 2.97 [95 % CI, 1.01-8.74]), shock at the time of febrile neutropenia diagnosis (RR, 6.5 [95 % CI, 1.83-23.05]) carbapenem-resistant microorganism (RR, 3.73 [95 % CI, 1.14-12.24]) and persistent bacteremia (RR, 84.6 [95 % CI, 11.3-629.4]) were associated with an increased mortality at 7 and 30 days. In the multivariate analysis, shock (RR, 4.85 [95 % CI, 2.10-11.65]) and persistent bacteremia was independently associated with increased 30-day mortality, but inappropriate empirical antimicrobial therapy was not an independent prognostic determinant (RR, 1.66 [0.53-4.82]). Conclusion: Shock at the time of febrile neutropenia diagnosis contributes to mortality in patients with gram-negative bacilli bloodstream infection, in this scenario, appropriate empirical antimicrobial therapy should be encouraged.
What problem does this paper attempt to address?