Answers are grounded in this paper and cite their sources.
Background Carbapenem-resistant Gram-negative bloodstream infections (CR-GNB BSIs) are associated with substantial mortality, but prognostic factors may not exert the same influence throughout the clinical course. We aimed to determine whether factors associated with early mortality differ from those associated with later mortality in hospitalized adults with CR-GNB BSI. Methods This retrospective, single-center cohort included 75 adults with CR-GNB BSI between January 1, 2024, and January 1, 2025. Demographic, clinical, laboratory, microbiological, and treatment-related variables were evaluated. Multivariable logistic regression and receiver operating characteristic analyses were performed separately for 14-day and 28-day all-cause mortality, and a landmark analysis restricted to patients alive on day 14 was used to examine death between days 15 and 28. Models were internally validated by bootstrap resampling. Results Fourteen-day and 28-day mortality were 42.7% and 68.0%, respectively. Older age (OR 1.073, 95% CI 1.014–1.135) and higher neutrophil-to-lymphocyte ratio (NLR) (OR 1.062, 95% CI 1.003–1.126) were independently associated with higher 14-day mortality, whereas appropriate antimicrobial therapy (OR 0.232, 95% CI 0.076–0.712) and hospitalization within the previous 3 months (OR 0.259, 95% CI 0.077–0.869) were associated with lower early mortality. Chronic obstructive pulmonary disease (OR 6.81, 95% CI 1.97–23.49) and Klebsiella spp. infection (OR 3.82, 95% CI 1.18–12.38) were independently associated with higher 28-day mortality. Among the 43 patients alive on day 14, COPD was the only factor associated with death between days 15 and 28 (OR 6.80, 95% CI 1.70-27.09), whereas Klebsiella spp. infection was associated with mortality mainly within the first 14 days. Serum albumin discriminated both mortality endpoints, whereas NLR discriminated 14-day mortality. Findings remained consistent after exclusion of six polymicrobial episodes and after adjustment for SOFA or APACHE II scores; optimism-corrected AUCs were 0.80 (14-day) and 0.77 (28-day). The association of appropriate therapy was attenuated in sensitivity analyses addressing immortal time bias. Conclusions The determinants of mortality in CR-GNB BSI differed by time point. Early mortality was associated with older age, systemic inflammation, Klebsiella spp. infection, and lack of appropriate antimicrobial therapy, whereas death after day 14 was associated mainly with COPD. Early therapy active against locally prevalent CR-GNB is a plausible modifiable target, although its association with survival was attenuated in analyses addressing immortal time bias, while patients with COPD remain at high risk beyond the second week and warrant continued close monitoring. These exploratory findings require external validation.
No takes yet. Share an insight, caveat, or question.
EKİCİ et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: