Clinical sensitivity and specificity of the Check-Points Check-Direct ESBL Screen for BD MAX, a real-time PCR for direct ESBL detection from rectal swabs.

Abstract

To determine the diagnostic accuracy of the Check-Direct ESBL Screen for BD MAX (ESBL qPCR) and an ESBL culture method to identify ESBLs directly from rectal swabs. Rectal swabs were obtained from clinical patients by performing cross-sectional (point)prevalence measurements in three regional hospitals. Rectal swabs were analysed by direct culture (ChromID ESBL agar) and with the ESBL qPCR. Suspected ESBL-producing isolates were confirmed with the combination disc method and analysed by WGS. Out of 354 rectal swabs and 351 patients, 21 rectal swabs and 20 patients were positive for ESBL-producing isolates, resulting in a regional ESBL colonization prevalence of 5.7%. One rectal swab was false negative with the ESBL qPCR (blaTEM-12) and not covered by the ESBL qPCR. Eight ESBL qPCR-positive rectal swabs could not be confirmed by culture and were classified as false ESBL qPCR positive. The sensitivity and specificity of the ESBL qPCR were 95.2% (n = 20) and 97.6% (n = 323), respectively. When an optimal cycle threshold cut-off value of 37 was used, the ESBL qPCR displayed a sensitivity and specificity of 95.2% (n = 20) and 98.8% (n = 327), respectively (AUC = 0.975, 95% CI = 0.922-1). This ESBL qPCR offers rapid direct detection of the most prevalent ESBL types (blaCTX-M group and blaSHV group) from rectal swabs. The relatively high false-positive rate renders this test the most suitable as a screening test in high-prevalence regions or in an outbreak setting where a fast result is essential.