817. Initial Empiric Carbapenem Use Does Not Improve Mortality in Critically Ill patients with Gram-Negative Bacteremia BUT Increases Multi-Drug Resistant Organisms
Session: Poster Abstract Session: Bacteremia and Endocarditis
Friday, October 9, 2015
Room: Poster Hall
  • CBM GNB IDweek 2015 poster FINAL.pdf (520.9 kB)
  • Background:

    With the increase in antibiotic resistance worldwide, use of empiric broad spectrum antibiotics such as carbapenems in intensive care units (ICU) has been widely recommended. However, the risk of “collateral damage” with subsequent infection or colonization with multi-drug resistant organisms (MDRO) is a concern. Few have studied effect of empiric carbapenem treatment on mortality and subsequent new MDROs.


    A prospective observational study (2007-2011) was performed, including all patients with gram-negative bacteremia (GNB) who stayed for more than 24 hours at medical and surgical ICUs of our 1000 bed hospital. The primary outcome was 7-day mortality. Secondary outcome was the development of subsequent new MDRO (defined as resistance to >=3 antibiotic classes) in the index admission and the ensuing 6 months.


    There were 107 ICU patients with GNB during the study period. Causative organisms were E. coli (50%), Klebsiella pneumoniae (42%), A. baumanii (11%), P.aeruginosa (3%).  There was no significant difference in 7-day mortality regardless of the use of empiric carbapenems in the first 24hours (odds ratio (OR) 1.34 (p=0.61, 95% Confidence interval (CI) 0.48-3.70). Of the 107 GNB, 47(43.9%) were MDRO-GNB. 18(18.8%) patients died within 7 days of GNB onset, 16 of which were MDR-GNB. Among the 16, there was no significant difference in mortality regardless of the use of initial carbapenems (OR 0.9,p=1.0, 95% CI 0.3-3.0).

    24 of 107 (22.4%) patients developed new MDRO in the index admission and ensuing 6 months. Among those who survived the first admission, there were more new MDROs in patients treated with carbapenems in the index admission (OR 7.3,p=0.04, 95%CI 0. 9-60.0).


    Empiric use of carbapenems in critically ill patients with GNB, and even in MDRO-GNB does not improve mortality. Exposure to carbapenems however increases the risk of development of new MDROs. Novel techniques to identify MDRO within 24hours may help reduce the use of broad spectrum antibiotics, allowing targeted therapy to reduce collateral damage without affecting mortality.

    Rui Min Foo, MBBS, MRCP (UK)1, Anupama Vasudevan, BDS, MPH1, Isaac Taoyang Low, BSS2, Amartya Mukhopadhyay, MBBS, MD, MRCP, FRCP, EDIC2 and Paul Tambyah, MBBS, M.D., FSHEA3, (1)Infectious Diseases, National University Health System, Singapore, Singapore, (2)National University Health System, Singapore, Singapore, (3)Division of Infectious Disease, National University Hospital, Singapore, Singapore


    R. M. Foo, None

    A. Vasudevan, None

    I. T. Low, None

    A. Mukhopadhyay, None

    P. Tambyah, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.