Accuracy and reproducibility of the AutoMicrobic System Gram-Negative General Susceptibility-Plus Card for testing selected challenge organisms.

A total of 180 selected strains of gram-negative bacilli were tested for susceptibility to nine antibiotics by the prototype General Susceptibility Card and reference broth microdilution MIC method, and 112 of 180 were tested by the modified Gram-Negative General Susceptibility-Plus Card (Vitek Systems, Inc., Hazelwood, Mo.). When category calls--susceptible, intermediate, and resistant for the prototype card and very susceptible, moderately susceptible, and resistant for the modified card--were identical or compatible (very susceptible by one method and moderately susceptible by the other), the methods were considered in qualitative or categorical agreement. The overall categorical agreement improved from 83% for the prototype card to 91.5% for the modified card, and the frequency of major and very major disagreements was reduced from 7.5 to 1%, respectively. Overall agreement between the modified card and reference method for both category calls and MICs (+/- 2 dilutions) was 90%. Of the results, 67% were identical, 22% were more susceptible with the modified card, and 11% were more resistant. Reproducibility of identical or compatible category calls and MICs obtained from three trials was 94.5%. The poorest accuracy and reproducibility were observed when testing Pseudomonas aeruginosa. The study suggests that the modified card can rapidly and efficiently perform susceptibility tests with an acceptable level of accuracy and reproducibility provided that the system is appropriately used for testing strains of the family Enterobacteriaceae with preset criteria for excluding organism-antibiotic combinations. The evaluation also indicates dramatic improvement in the technical performance of the modified card compared with its earlier prototype.