1210. Defining Multidrug Resistance (MDR) for Gram-Negative (GN) Infections: Who Gets the Gown?
Session: Oral Abstract Session: Identifying and Overcoming Challenges in Preventing Transmission of MDRO GNR Bacterial Infections in Healthcare Settings
Saturday, October 5, 2013: 9:15 AM
Room: The Moscone Center: 250-262

Background:  Isolation of patients with infections due to MDR-GN organisms is recommended, but a standard definition for MDR does not exist.  Our objective was to assess how the definition of MDR impacts the proportion of inpatients meeting criteria for isolation for MDR-GN infections.

Methods:   We retrospectively analyzed all inpatient admissions at Duke University Hospital from 1/1/2011 to 12/31/2011 during which clinical cultures grew E. coli, E. cloacae, K. pneumoniae, P. mirabilis, or P. aeruginosa.  We applied 3 different definitions of MDR: 1) not-susceptible (NS) to ≥ 1 drug in ≥ 3 antibiotic classes to which the organism is not intrinsically resistant; 2) NS to ≥ 1 drug from ≥ 2 of the following classes: fluoroquinolones (FQ), aminoglycosides (AG), carbapenems (CP), or either piperacillin-tazobactam or 3rd/4th generation cephalosporins (BL);  3) NS to ≥ 1 drug from ≥ 3 of the following classes: FQ, AG, CPM, BL.  We calculated the proportion of MDR isolates by organism using the 3 definitions.   Duplicate isolates of an organism in the same admission were excluded.  We then calculated the proportion of all admissions requiring isolation based on the 3 MDR definitions.  If multiple GN organisms were isolated in the same admission, the organism with the greatest number of NS antibiotic classes was included in this overall analysis.

Results:   2269 inpatient encounters associated with the 5 GN organisms were analyzed.  E. coli was the most commonly identified GN pathogen (Figure 1). A total of 679 (30%) patient admissions would require isolation using definition 1; 418 (18%) patients admissions would require isolation using definition 2; and 181 (8%) patient admissions would require isolation using definition 3.  Definition 1 was the most inclusive for Enterobacteriaciae whereas definition 2 was the most inclusive for P. aeruginosa.

Conclusion:   Our data illustrate that different definitions of MDR greatly impact the number of patients who require contact isolation. Furthermore, the above MDR definitions rely on an understanding of antibiotic classes and intrinsic drug resistance, making them impractical to apply in real-world settings.  Our findings and these limitations underscore the need for a national consensus about how to uniformly and simply define MDR.  


Sarah S. Lewis, MD1,2, Rebekah W. Moehring, MD, MPH2,3, Luke F. Chen, MBBS, MPH, CIC, FRACP1,2, Daniel J. Sexton, MD, FIDSA1,2 and Deverick J. Anderson, MD, MPH2,3, (1)Duke University Medical Center, Durham, NC, (2)Duke Infection Control Outreach Network, Durham, NC, (3)Division of Infectious Diseases, Duke University Medical Center, Durham, NC


S. S. Lewis, None

R. W. Moehring, None

L. F. Chen, None

D. J. Sexton, None

D. J. Anderson, None

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.