Appropriate sites for the active surveillance of carbapenemase-producing Enterobacteriaceae

Abstract

Background: Active surveillance tests for carbapenemase-producing Enterobacteriaceae (CPE) are recommended in patients showing risk factors for colonization by these bacteria. There are limited data however on whether surveillance tests for anatomic sites other than the stool would be useful to detect CPE colonization, and we investigated this in our present study.Methods: Retrospective analysis was performed on cases at our tertiary care hospital during a 5-year period. Patients with CPE colonization had been admitted to our surgical intensive care unit (SICU) or sub-ICU for liver transplantation in this period and undergone surveillance tests for both the stool and other sites. Patients were grouped as stool CPE negative (but which included CPE positive cases from initial sputum and other site tests) or positive. Results: Among the total study cohort of 158 patients, 138 (87.3%) were included in the stool CPE positive group and the remaining 20 (12.7%) in the stool CPE negative group. While the sensitivity of CPE surveillance testing of the stool was 87.3% (95% CI 81.1-92.1), the sensitivity when combining stool and sputum samples was 93.7% (88.7-96.9). The transmission rates were similar for patients showing CPE positivity in the stool, sputum and other sites, at 4.8% (27/557), 4.7% (3/64), and 6.7% (1/15), respectively (p = 0.95).Conclusions: The sensitivity of CPE detection in a stool sample is suboptimal for ruling out CPE colonization and the transmission rates are similar between stool-positive or -negative cases. Thus, combining surveillance of the stool with other sites may be needed for detecting CPE.