A Systematic Review of Infection Control Interventions for Extended-Spectrum Beta-Lactamase Producing Enterobacteriaceae (ESBL-E)

Background: The global pandemic of ESBL-E is a representative example of the infection control challenges posed by multi-drug resistant gram negative bacteria.  Currently, there are no accepted infection control guidelines for the management of patients colonized or infected with ESBL-E.  This has resulted in widespread variations in practice, particularly in the non-outbreak setting.

Objective: To identify infection control interventions that reduce the incidence of ESBL-E infection and/or colonization in hospitalized patients in the non-outbreak setting.

Methods: We searched Medline, Embase, CINAHL and the Cochrane Library from 1980 to April 2009 for experimental or quasi-experimental studies that examined the ability of infection control interventions to reduce the incidence of ESBL-E colonization or infection among hospitalized patients in the non-outbreak setting.  Observational studies, studies of antimicrobial stewardship, and studies conducted during an outbreak were excluded.  All studies were reviewed independently by both authors.

Results: We reviewed 799 articles of which 10 were considered potentially eligible and were reviewed in detail.  Of these, 6 were excluded as they failed to measure ESBL-E at baseline (n=2), did not describe a specific intervention (n=2), were conducted during an outbreak (n=1) or included an antimicrobial stewardship intervention (n=1). Four studies were included.  Two studies examined interventions designed to improve hand hygiene and other routine practices.  Two studies examined isolation and barrier precautions combined with surveillance cultures for all admissions or for contacts of cases.  No prospective or controlled studies were identified and no study reported seeking the approval of a research ethics board (REB).  The study designs were either pre-test/post-test studies (n=3) or time series analyses (n=1).  Methodological flaws were present in all studies and included the presence of co-interventions (n=1), failure to measure relevant confounders (n=4), failure to exclude the presence of an outbreak (n=2), failure to distinguish nosocomial and non-nosocomial cases (n=3), and consideration of ESBL-E incidence only as a post-hoc analysis (n=1). 

Conclusions: There are no high quality studies evaluating the ability of infection control interventions to reduce the incidence of ESBL-E colonization or infection in the non-outbreak setting.  In fact, there was a complete lack of prospective research in this area with no evidence of prospective, controlled or REB approved studies.  This is concerning, given the importance of this problem.  Prospective, controlled studies of the efficacy of barrier precautions and/or surveillance cultures to reduce the incidence of ESBL-E and other multi-drug resistant gram negative bacteria) are urgently needed.