Comparison of workflow and accuracy of identification and antimicrobial susceptibility testing of clinical isolates of Enterobacteriaceae, Pseudomonas aeruginosa and enterococci by Vitek 2 and routine methods

Abstract

Three hundred and fifty-three consecutive urine cultures growing Enterobacteriaceae, Pseudomonas aeruginosa or enterococci were subjected to parallel identification (ID) and antimicrobial susceptibility testing (AST) by Vitek 2 and routine methods, including simple screening tests or API 20 E for ID and standardized disc diffusion for AST. Accuracy of results, technician hands-on time required by both methods and time to results were compared. Vitek 2 correctly identified 322 (94.7%) of the 340 gram-negative isolates and 17 (81%) of the 21 Enterococcus faecalis strains. AST by Vitek 2 and disc diffusion gave category agreement for 4,058 (95.5%) of 4,248 organism-antimicrobial agent combinations. With MIC determination by E-test as reference, AST by Vitek 2 and disc diffusion produced 15 and 3 very major errors, respectively. Six (40%) of the fifteen very major errors by Vitek 2 were associated with trimethoprim-sulfamethoxazole. With an average of 22 specimens processed per day, use of Vitek 2 saved 80 min per day of technician hands-on time as compared to routine methods. Regarding the cost of hands-on worktime and consumables, use of Vitek 2 for identification of Escherichia coli-screened Enterobacteriaceae saved 0.70 p per sample in comparison to API 20 E. More than 80% of Enterobacteriaceae introduced to Vitek 2 in the morning could be reported by 16:00. © 2006 The Authors.