Multicenter Evaluation of the Modified Carbapenem Inactivation Method and the Carba NP for Detection of Carbapenemase-Producing Pseudomonas aeruginosa and Acinetobacter baumannii.

The purpose of this study was to develop the modified carbapenem inactivation method (mCIM) for the detection of carbapenemase-producing (CP-PA) and carbapenemase-producing (CP-AB) and perform a multicenter evaluation of the mCIM and Carba NP tests for these nonfermenters. Thirty and 30 isolates previously characterized by whole-genome sequencing from the CDC-FDA Antibiotic Resistance Isolate Bank were evaluated, including CP isolates (Ambler class A, B, and D), non-carbapenemase-producing (non-CP) carbapenem-resistant isolates, and carbapenem-susceptible isolates. Initial comparison of a 1-μl versus 10-μl loop inoculum for the mCIM was performed by two testing sites and showed that 10 μl was required for reliable detection of carbapenemase production among and Ten testing sites then evaluated the mCIM using a 10-μl loop inoculum. Overall, the mean sensitivity and specificity of the mCIM for detection of CP-PA across all 10 sites were 98.0% (95% confidence interval [CI], 94.3 to 99.6; range, 86.7 to 100) and 95% (95% CI, 89.8 to 97.7; range, 93.3 to 100), whereas the mean sensitivity and specificity among CP-AB were 79.8% (95% CI, 74.0 to 84.9; range, 36.3 to 95.7) and 52.9% (95% CI, 40.6 to 64.9; range, 28.6 to 100), respectively. At three sites that evaluated the performance of the Carba NP test using the same set of isolates, the mean sensitivity and specificity of the Carba NP test were 97.8% (95% CI, 88.2 to 99.9; range, 93.3 to 100) and 97.8% (95% CI, 88.2 to 99.9; range, 93.3 to 100) for and 18.8% (95% CI, 10.4 to 30.1; range, 8.7 to 26.1) and 100% (95% CI, 83.9 to 100; range, 100) for Overall, we found both the mCIM and the Carba NP test to be accurate for detection of carbapenemase production among isolates and less reliable for use with isolates.