Clinical Performance of Check-Direct CPE, a Multiplex PCR for Direct Detection of bla(KPC), bla(NDM) and/or bla(VIM), and bla(OXA)-48 from Perirectal Swabs.

We evaluated the clinical performance of Check-Direct CPE for carbapenemase detection directly from 301 perirectal swabs (258 patients) in a nonoutbreak setting. Culture of a PCR-confirmed, carbapenemase-containing organism, or history of colonization with such organism within the previous 2 weeks, was used as the reference standard. Check-Direct CPE demonstrated a sensitivity value, specificity value, positive predictive value (PPV), and negative predictive value (NPV) of 100% (all bla(KPC)), 88%, 21%, and 100%, respectively. False positives accounted for 79% (n = 34) of samples for which a cycle threshold (C(T)) value was reached. Simulated studies to evaluate specimen pooling as an approach to minimize costs showed no difference in C(T) values for pooled groups of three or five that each contained a single specimen spiked with ∼1,500 CFU bla(KPC) Klebsiella pneumoniae; however, the detection rate dropped to 60% at a seeded concentration of ∼150 CFU. When data were pooled, C(T) values for bla(KPC) were higher for heavy-feces-containing than for light-feces-containing liquid-suspended specimens. Furthermore, C(T) values for liquid-suspended specimens were 4 to 5 C(T) values lower (i.e., represented greater sensitivity) than those seen in direct swab analysis. Culture was equivalent to or better than Check-Direct CPE for 13/15 (87%) isolates tested in a limit-of-detection analysis. Detection of a carbapenemase gene at a C(T) cutoff value of ≤35 was culture confirmed in 23/24 (96%) of cases; however, C(T) values of >35 overlapped broadly between culture-positive (n = 21) and culture-negative (n = 36) specimens. Check-Direct CPE will likely prove most useful in high-prevalence areas or in outbreak settings where rapid carbapenemase detection is critical for infection control management.