AbstractAimsAdequately and accurately identifying carbapenemase-producing Enterobacterales (CPE) is vital for selecting appropriate antimicrobial therapy and implementing effective infection control measures. This study aims to optimize the phenotypic detection method of carbapenemase for routine diagnostics in clinical microbiology laboratories.Methods and ResultsCarbapenemase genes in 2665 non-duplicate CRE clinical strains collected from various regions of China were confirmed through whole-genome sequencing (WGS). The carbapenemase inhibition test (CIT) was conducted and interpreted using different methods and breakpoints, then compared with the NG-Test CARBA 5 for carbapenemase detection. The diagnostic performance of the CIT method was optimal when the carbapenemase types were determined by comparing the inhibition zone diameters of the imipenem disc with 3-aminophenylboronic acid (APB) plus ethylenediaminetetraacetic acid (EDTA) to those of the imipenem disc with either APB or EDTA alone, with a breakpoint of 4 mm. The overall sensitivities of the current CIT, the modified CIT, and NG-Test CARBA 5 were 91.4%, 94.9%, and 99.9%, respectively. For detecting isolates co-producing Klebsiella pneumoniae carbapenemase (KPC) and metallo-β-lactamases (MBLs), the modified CIT method had higher sensitivity than the current method (70.0% vs. 53.3%), though this difference was not statistically significant (P = 0.063). The NG-Test CARBA 5 showed excellent performance for multi-carbapenemases diagnosis, with sensitivity and specificity of 97.1% and 100%, respectively.ConclusionsOptimizing and standardizing the CIT method for clinical use is necessary. It has certain advantages in diagnosing multi-carbapenemase and rare carbapenemase production. However, for identifying common carbapenemase types, the NG-Test CARBA 5 demonstrated superior performance.