Assessing Colistin Resistance Among Clinical Isolates: A Comparative Study of BD Phoenix System versus Broth Microdilution Method

Evaluating Colistin Resistance: BD Phoenix System versus Broth Microdilution

Abstract

Background: This study aims to compare the colistin susceptibility detected by the BD Phoenix automated system among multidrug-resistant bacteria with the results of the broth microdilution (BMD) method.

Materials and Methods: The study included 236 isolates (92 A. baumannii, 68 K. pneumoniae, 55 P. aeruginosa, and 21 E. coli) which were isolated from clinical samples submitted to the bacteriology laboratory of XXXX   Hospital. The BD Phoenix 100 automated system was evaluated using BMD as a reference method. Categorical agreement (CA, the number of concordant results), major errors (ME, results falsely categorized as resistant by the automated system despite being sensitive by the reference method), and very major errors (VME, results falsely categorized as sensitive by the automated system despite being resistant by the reference method) were calculated.

Results: Of the 236 isolates, colistin resistance was found in 34 (14.4%) isolates, with A. baumannii, K. pneumoniae, P. aeruginosa, and E. coli exhibiting resistance rates of 15.2%, 20.6%, 1.8%, and 23.8%, respectively. A. baumannii showed 89.1% CA between the two methods, with 10.8% VME. K. pneumoniae had an 83.8% CA with a 14.7% VME rate. P. aeruginosa exhibited a 98.1% CA with a 1.8% VME rate, while E. coli showed an 80.9% CA with a 19% VME rate. The BD Phoenix 100 automated system had a 10.5% VME rate compared to BMD. The automated system identified 9 out of 34 isolates as susceptible, contrary to the reference method which reported them as resistant.

Conclusions: Categorical agreement rates of the automated system varied according to bacterial isolates.The high VME rate of the BD Phoenix 100 system demonstrated the need to confirm colistin results reported as susceptible with SMD.