Detection and epidemiology of extended-spectrum beta-lactamases

Abstract

The rising prevalence of extended-spectrum beta-lactamase (ESBL-)producing Enterobacteriaceae threatens public health, as treatment options in case of infections are reduced. Rapid detection of carriage or infection of ESBL-producing bacteria is crucial for appropriate infection control measures and decision making regarding (empirical) therapy in case of infections. This thesis aims to provide deeper insight in the possibilities to detect patients carrying ESBL-producing bacteria quickly and accurately.

ESBL-detection in the Netherlands is reliable, as in 88% of cases the correct ESBL-status was assigned to isolates in daily practice, using the Dutch guideline for phenotypic detection of ESBLs in Enterobacteriaceae. The positive predictive value of the screentest was on average 70%, but was dependent on the method, the species and the minimum inhibitory concentration (MIC) for the third-generation cephalosporins. The ESBL-Etest was less specific as compared to the combination disk (59% versus 92%) for ESBL confirmation. Another tool to detect ESBL-producing Enterobacteriaceae is the Check-KPC ESBL microarray, which had a sensitivity of 97%, specificity of 98%, positive-predictive value of 99% and negative-predictive value of 92%.

Resistance for third-generation cephalosporins in Enterobacteriaceae in the Netherlands results mainly from CTX-M-15 ESBL genes in E. coli, frequently belonging to sequence type (ST)131. CTX-M-15 isolates were –on average– susceptible to less antibiotics than isolates harboring TEM-52, CTX-M-1, or CTX-M-14. The high prevalence of CTX-M-15 is comparable to most other countries.

The hypothesis that ESBL-producing Enterobacteriaceae circulate in two separate compartments, one in the community fuelled by food contamination and one in hospitals fuelled by cross-transmission, could not be confirmed. Resistance patterns differed between the poultry-associated and non-poultry associated isolates with higher susceptibility for aminoglycosides and fluoroquinolones in the CTX-M-1 and TEM-52 positive isolates from humans.

At hospital-admission, the prevalence of ESBL carriage was 8.2%, and was comparable among patients admitted from long-term care-facilities and home settings. Documented ESBL-carriage within 1 year before admission, hospital admission in the previous 6 months and male gender were associated with ESBL-carriage. It was not possible to develop a clinically useful prediction rule for ESBL carriage at hospital admission. ESBL-carriers frequently remained colonized for longer than a year after hospital-discharge, often with the same strain. Carriage of ESBL-producing bacteria positive for CTX-M-15 was associated with a longer duration of ESBL-carriage. At identification of ESBL-carriage, 32% of household contacts were colonized with ESBL-producing bacteria, as this was still >20% after 18 months. As screening all patients at hospital-admission for ESBL-carriage is not feasible, it could be considered to extend the screening for previous identified ESBL-carriers from 12 to 18 months, and additionally, to screen household contacts of previously identified ESBL-carriers at hospital admission.