Failure of classical surveillance strategy to detect a large hospital-wide outbreak of a multidrug-resistant Enterobacter cloacae

Abstract number: 10.1111/j.1198-743X.2004.902_o190.x

Leverstein-van Hall M.A.


To identify the determinants causing the failure of a classical surveillance strategy (CSS) to detect a large outbreak of a multidrug resistant (MDR) Enterobacter cloacae (EC). CSS refers to a strategy based on the recognition of an increased incidence of a species with a particular antibiogram at certain wards in a limited time period.


A hospital-wide increase in the number of MDR clinical EC isolates was detected in December 2002. Records were reviewed for the period from January 2001 to Dec 2002 (period I) to identify patients with cultures positive for EC during their stay in the hospital. From January to September 2003 (period II) EC isolates were prospectively stored. Genotyping by PFGE was done on all available EC isolates.


A total of 466 patients were identified for being EC culture positive. From period I, 105 tobramycin-resistant (TR)EC (60 patients) and 27 tobramycin susceptible (TS)EC (24 patients), and from period II, 58 TREC (48 patients) and 49 TSEC (47 patients) were available. PFGE showed that 53 patients in period I and 34 patients in period II carried a TREC that belonged to one clone that was subject to evolution. The susceptibility patterns for eight non-beta-lactam antimicrobials were analysed for all 239 genotyped isolates. The clonal strain expressed 37 different antibiograms (including 2 tobramycin S variants) of which 1/4 were shared by ECs of other genotypes. The clonal strain was isolated from a large variety of clinical sites, including 30% normally sterile sites (eg BC, CSF). During period I cases were detected on nine different divisions (three ICUs and 18 wards). Patients were frequently transferred between wards (median 3) and divisions (median 2) and long laps in time were seen between cases (mean 2 weeks; range 0–18).


The determinants causing the failure of a CSS to detect this outbreak were: (1) the low incidence, (2) the long time interval between cases, (3) the hospital-wide occurrence of new cases, (4) the large variety of clinical sites from which the clone was isolated, and (5) the high variability of the antibiogram. These results illustrate the limitations of the CSS and stress the need of molecular typing facilities and (data mining) surveillance systems that integrate laboratory and hospital information systems to identify patterns indicative for the occurrence of hospital infections.


Session Details

Date: 01/08/2007
Time: 00:00-00:00
Session name: XXIst ISTH Congress
Location: Oxford, UK
Presentation type:
Back to top