Ceftazidime–avibactam tied to better early ICU survival in resistant Klebsiella, but benefit fades after adjustment

Patients in the intensive care unit treated with ceftazidime–avibactam for carbapenem-resistant Klebsiella pneumoniae infections showed significantly lower mortality at days 7, 14, and 28 than patients who received alternative antibiotic regimens, according to a retrospective cohort study published in BMC Infectious Diseases. Yet when the researchers adjusted for baseline clinical differences between the groups, the drug itself was no longer independently associated with survival, a result that shifts attention toward patient-related risk factors as the dominant drivers of death in this population. The study, conducted by Zuhal Kalaycı Çekin and Mustafa Deniz, included 249 patients diagnosed with CRKP infections between November 2022 and January 2025.

Of these, 115 received ceftazidime–avibactam and 134 were assigned to a comparator group treated with alternative combination antibiotic therapies. The analysis covered all patients treated with the drug in the intensive care unit as well as clinical cultures in which carbapenem-resistant isolates were detected. The two groups were broadly comparable in age, with no statistically significant difference reported, and in comorbidities.

Despite this apparent balance, the raw mortality gap was striking. All-cause mortality across the entire cohort reached 22.5 percent at day 7, 44.6 percent at day 14, and 57.8 percent at day 28. In the ceftazidime–avibactam group, the corresponding figures were 14.8 percent, 31.3 percent, and 44.3 percent, and the differences between the groups were statistically significant at each time point, with p values of 0.011, less than 0.001, and less than 0.001 respectively.

When the investigators moved from simple comparison to multivariable analysis, however, the picture changed. Ceftazidime–avibactam treatment was not independently associated with mortality at day 7, day 14, or day 28, with p values of 0.201, 0.105, and 0.175 respectively. The authors interpret this as evidence that the apparent survival benefit may be partly attributable to variations in baseline clinical characteristics between the groups rather than an independent treatment effect, a caution that matters for clinicians weighing the drug against other options for critically ill patients.

What did emerge from the adjusted models were several patient-level predictors of death. Advanced age was independently associated with both 14-day and 28-day mortality, with p values of 0.016 and 0.018. Diabetes mellitus, a low lymphocyte count, and hypoalbuminemia were each independently associated with 28-day mortality, with p values of 0.038, 0.002, and 0.029.

These findings underline that mortality in critically ill patients with carbapenem-resistant K. pneumoniae reflects a constellation of disease-related risk factors rather than antibiotic choice alone. Other clinical variables carried weight at specific time points. Malignancy was associated only with 7-day mortality, with significantly higher early deaths observed among patients with cancer.

Among markers of clinical status, the need for vasopressor support emerged as a significant risk factor for 14-day mortality. The microbiology of the cohort also offers practical context. Respiratory tract specimens were the most frequently identified source of positive cultures, accounting for 50.6 percent of all cases.

Carbapenem-resistant Klebsiella pneumoniae remains a significant global threat, and the study’s authors frame their results as an argument for rational, evidence-based antibiotic management combined with effective and continuous surveillance strategies. The high overall mortality observed even among patients receiving active therapy illustrates how limited the therapeutic margin remains once these infections take hold in intensive care populations. The study’s design imposes important limits on interpretation.

As a retrospective cohort, it can identify associations but cannot establish causation, and the loss of statistical significance for ceftazidime–avibactam after multivariable adjustment suggests that residual confounding by illness severity or other baseline differences may explain part of the unadjusted benefit. The authors’ conclusion is measured: early survival was substantially better in the treated group, but the correlation was insignificant in multivariate analyses. For clinicians and infection-control teams, the practical takeaway is that decisions about ceftazidime–avibactam should rest on evidence that accounts for these confounders, since the raw survival gap observed in this cohort did not survive statistical adjustment.

Subject of Research: Mortality predictors and clinical outcomes of ceftazidime–avibactam treatment for carbapenem-resistant Klebsiella pneumoniae infections in intensive care unit patients

Article Title: Ceftazidime–avibactam for carbapenem-resistant Klebsiella pneumoniae infections: determining mortality predictors in the intensive care unit

Article References: Çekin, Z. K., & Deniz, M. (2026). Ceftazidime–avibactam for carbapenem-resistant Klebsiella pneumoniae infections: determining mortality predictors in the intensive care unit. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14447-4

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14447-4

Keywords: carbapenem-resistant Klebsiella pneumoniae, ceftazidime-avibactam, intensive care unit, mortality predictors, retrospective cohort study, antimicrobial resistance, multivariable analysis, hypoalbuminemia, lymphopenia, diabetes mellitus, vasopressor requirement, Ceftazidime