Extended-Spectrum-β-Lactamase-Producing Klebsiella pneumoniae Bacteremia: risk factors for mortality and clinical outcome with emphasis on antimicrobial therapy

Abstract
Objective:
To evaluate the clinical outcome of bacteremia caused by extended-spectrum β-lactamase producing Klebsiella pneumoniae under different antibiotic regimens and to identify other risk factors for mortality.
Design:
A single-center, retrospective, chart-review study
Setting:
National Taiwan University Hospital (NTUH)─a tertiary-care university hospital in northern Taiwan
Patients:
All adult patients with bacteremia due to extended-spectrum β-lactamase producing Klebsiella pneumoniae between April 1, 2002 and December 31, 2006 were included.
Methods:
Medical records were reviewed for patient’s demographic data and clinical information. Clinical data such as underlying diseases/conditions, bacterial colonization/infection, medical interventions, clinical manifestations, comorbidities, antibiotic therapy and clinical outcome were collected according to the time before, after and at the onset of ESBL-KP bacteremia. The primary endpoint was all-cause 30 -day mortality. Univariate analysis and multivariate stepwise logistic regression analysis were used to identify the risk factors for mortality and to compare the prognosis of patients received different treatment regimens.
Results:
During the study period, 104 patients met criteria for inclusion and 99 (95.2%) of them were hospital-acquired bacteremia. The number of patients with monomicrobial bacteremia was 67 (64.4%). The overall 30-day mortality rate was 28.8%. Analyses of different antibiotic treatment outcomes showed that there was marginal significant difference between patients received adequate and inadequate empirical therapy (P=0.055). Whether the definite therapy was adequate or not, there was no significant influence on the mortality (P=0.342). Comparison between carbapenem and non-carbapenem treatment groups also showed no significant difference (P=0.668). There was a significant difference of the severity of illness between the mortality group and survival group (APACHEⅡscore 26.6±8.3 vs. 19.8±8.5, P<0.001; Pitt bacteremia score 4.8±2.8 vs. 2.9±2.6, P=0.001). Univariate analysis revealed that the risk factors for mortality included: neutropenia (odds ratio 4.73, P=0.043), previous immunosuppressive therapy (especially steroids, odds ratio 9.20, P=0.001), APACHEⅡ score≧17 (odds ratio 4.19, P=0.010), Pitt bacteremia score≧4 (odds ratio 3.01, P=0.013), associated site of infection as pneumonia (odds ratio11.38, P=0.002), presentation of septic shock (odds ratio 6.79, P=0.001) and longer length of hospital stay before bacteremia onset (P=0.024). Multivariate logistic regression analysis showed that septic shock (odds ratio 11.89, P<0.001), associated site of infection as pneumonia (odds ratio 9.96, P=0.035), previous immunosuppressive therapy (odds ratio 9.65, P<0.001), and adequate empirical therapy (odds ratio 0.20, P=0.014) were independent factors for all-cause 30-day mortality. Similar results were obtained from univariate and multivariate analyses of patients with monomicrobial bacteremia.
Conclusions:
Our study showed that adequate empirical antibiotic therapy was associated with a lower 30-day mortality for ESBL-KP bacteremia patients. Regimen of definite therapy did not show significant influence on 30-day mortality, but severity of illness at the onset of bacteremia and previous immunosuppressive therapy had a significant impact on 30-day mortality.