2376. Attributable Mortality of Carbapenem-Susceptible and Carbapenem-Non-Susceptible Phenotypes among Hospital-Onset Bacteremia due to Klebsiella spp.: A Multicenter Matched Cohort Study in the Veterans Health Administration
Session: Oral Abstract Session: Epidemiology and Outcomes of Resistant Gram-Negative Organisms
Saturday, October 29, 2016: 2:45 PM
Room: 388-390

Background: Antimicrobial resistance to carbapenems among Klebsiella spp. has emerged globally, and accurate estimate of attributable mortality (AM) is crucially important to assess its burden. We aimed to estimate AMs of carbapenem susceptible (CS) and carbapenem non-susceptible (CNS) phenotypes among Klebsiella spp. blood isolates by a case-case-control study design from a large bacteremia cohort within the Veterans Health Administration (VHA).

Methods: A retrospective cohort of all patients who had a positive blood culture for Klebsiella spp. >=48 hours after admission at 130 VHA hospitals between 2003 and 2013 was analyzed. If patients had more than one episode of Klebsiella spp. bacteremia, only the first one was included. Cases were classified as CS and CNS groups according to the susceptibility of isolates. For each case, up to 3 uninfected control patients were selected, matched on admission facility, month of admission, and length of stay. 30-day all-cause mortality after bacteremia onset was the primary outcome. Crude AMs were calculated for both CS and CNS groups by comparing to their uninfected controls. Age, body mass index (BMI), and Charlson comorbidity index were considered as potential patient-level confounders, and Cox proportional hazard models were used for risk-adjustment. Adjusted AMs were calculated by Levin’s formula based on hazard ratios (HRs).

Results: A total of 4,244 patients with Klebsiella spp. bacteremia (CS: 4,023; CNS: 221) and 11,196 uninfected control patients (10,537 matched to CS; 619 matched to CNS) were included in the analysis. Crude 30-day mortality after the initial positive blood culture was 26.2% for CS and 58.4% for CNS. Crude 30-day AMs compared to uninfected controls were 12.5% for CS and 39.0% for CNS. From Cox proportional hazard models, risk-adjusted AM was 21.7% (95% CI: 19.0% to 24.4%) for CS and 42.9% (95% CI: 34.1% to 51.3%) for CNS.

Conclusion: In this large cohort of Klebsiella spp. bacteremia from the VHA, we observed high AMs in both CS and CNS phenotypes. Risk-adjusted AM in CNS phenotypes was >20% higher than adjusted AM in CS phenotypes. Continued efforts towards development of novel treatment and prevention strategies for CNS Klebsiella spp. are needed.

Michihiko Goto, MD, MSCI1,2, Daniel Livorsi, MD, MSc1,2, Makoto Jones, MD, MS3,4, Jennifer Mcdanel, PhD1,5, Bruce Alexander, PharmD1, Brice Beck, MA1, Kelly Richardson, PhD1 and Eli Perencevich, MD, MS, FIDSA, FSHEA1,2, (1)Iowa City VA Health Care System, Iowa City, IA, (2)Department of Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA, (3)Department of Internal Medicine, University of Utah School of Medicine, Salt Lake City, UT, (4)VA Salt Lake City Health Care System, Salt Lake City, UT, (5)University of Iowa Carver College of Medicine, Iowa City, IA


M. Goto, None

D. Livorsi, None

M. Jones, None

J. Mcdanel, None

B. Alexander, None

B. Beck, None

K. Richardson, None

E. Perencevich, Merck & Co, inc.: Grant Investigator , Research grant

Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.