Abstract
The increasing incidence of carbapenemase-producing Gram-negative bacilli (C-PGNB) represents a major public health challenge. Rapid detection of digestive colonization with C-PGNB is fundamental to control their spread. We performed the validation of a rapid protocol for C-PGNB detection directly on rectal swabs. We developed a protocol combining enrichment by a rapid selective subculture of the rectal swab medium and realization of a Resist-4 O.K.N.V. K-SeT test on the bacterial pellet obtained. The limit of detection and performances of this protocol were validated in vitro on 52 C-PGNB strains spiked on a calibrated sample suspension and confirmed in clinical settings on 144 rectal swabs sampled from patients with C-PGNB digestive colonization (n = 48) and controls (patients with extended-spectrum beta-lactamase [ESBL] colonization [n = 48] and without carbapenemase/ESBL [n = 48]). The protocol detected, with 100% sensitivity, the presence of the 15 OXA-48-, 14 KPC-, 13 NDM-, and 10 VIM-producing GNB from 10(3) CFU/ml. The limit of detection was 2 × 10(2) CFU/ml. Among the 48 C-PGNB-containing rectal swabs of the validation cohort, 46 were accurately detected. False negative were observed for 1 NDM-producing Acinetobacter baumannii strain and 1 OXA-48-producing Escherichia coli strain. The 96 control swabs were negative. Sensitivity and specificity for C-PGNB detection were 97.7% (95% confidence interval [CI], 87.7 to 100) and 100% (95% CI, 96.2 to 100). The negative likelihood ratio was 0.04 (95% CI, 0.01 to 0.16). Considering a C-PGNB digestive colonization prevalence between 0.01% and 0.1%, positive and negative predictive values were 100%. Our protocol is a rapid and low-cost method detecting accurately the digestive colonization with carbapenemase-producing Enterobacteriaceae in 4 h without any requirement for specific equipment.