RESUMO
BACKGROUND: Carbapenemase-producing Enterobacterales (CPE) can colonize the gut and are of major clinical concern. Identification of CPE colonization is problematic; there is no gold-standard detection method, and the effects of antibiotic exposure and microbiota dysbiosis on detection are unknown. AIM: Based on a national survey we selected four CPE screening assays in common use. We used a clinically reflective in vitro model of human gut microbiota to investigate the performance of each test to detect three different CPE strains under different, clinically relevant antibiotic exposures. METHODS: Twelve gut models were seeded with a pooled faecal slurry and exposed to CPE either before, after, concomitant with, or in the absence of piperacillin-tazobactam (358 mg/L, 3 × daily, seven days). Total Enterobacterales and CPE populations were enumerated daily. Regular screening for CPE was performed using Cepheid Xpert® Carba-R molecular test, and with Brilliance™ CRE, Colorex™ mSuperCARBA and CHROMID® CARBA SMART agars. FINDINGS: Detection of CPE when the microbiota are intact is problematic. Antibiotic exposure disrupts microbiota populations and allows CPE proliferation, increasing detection. The performances of assays varied, particularly with respect to different CPE strains. The Cepheid assay performed better than the three agar methods for detecting a low level of CPE within an intact microbiota, although performance of all screening methods was comparable when CPE populations increased in a disrupted microbiota. CONCLUSION: CPE strains differed in their dynamics of colonization in an in vitro gut model and in their subsequent response to antibiotic exposure. This affected detection by molecular and screening methods, which has implications for the sensitivity of CPE screening in healthcare settings.