Tuba ÇİFTÇİ, Ayşe Hümeyra TAŞKIN KAFA
Sağlık Hizmetleri Araştırma Dergisi - 2026;4(1):74-81
Multidrug-resistant (MDR) Gram-negative bacteria have become a major healthcare problem, particularly in hospital-acquired infections. Colistin remains a last-resort antibiotic against pathogens such as Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. Accurate determination of colistin susceptibility is therefore critical for appropriate patient management. This study compared the BD Phoenix(TM) 100 automated system with the broth microdilution (BMD) method for colistin susceptibility testing. A total of 758 Gram-negative isolates were retrospectively evaluated. According to the BMD method, 546 (72.0 %) isolates were susceptible and 212 (28.0 %) were resistant, whereas the Phoenix(TM) 100 system reported 647 (85.4%) isolates as susceptible and 111 (14.6%) as resistant. Performance analysis of the BD Phoenix(TM) 100 system showed an overall categorical agreement (CA) of 81.4 %, a major error (ME) rate of 3.7 %, and a very major error (VME) rate of 57.1 %. Species-based analysis revealed CA, ME, and VME rates of 81.3%, 2.4%, and 65.0% for A. baumannii; 80.8%, 4.9%, and 50.5% for K. pneumoniae; and 86.4%, 5.7%, and 44.4% for P. aeruginosa, respectively. The high VME rates indicate that a substantial proportion of resistant isolates may be falsely reported as susceptible, posing a serious risk of treatment failure. In conclusion, the broth microdilution method should remain the gold standard for colistin susceptibility testing, and Phoenix(TM) 100 results should be confirmed by BMD, particularly for critical isolates.
