An Evaluation of Vitek2™ Direct Susceptibility Testing (DST) of 
Gram-Negative Bacteria (GNB) from Positive Blood Cultures (+BC).

BACKGROUND: Delayed reporting of antibiotic susceptibility results impacts the treatment of patients with bacteremia. We evaluated the sensitivity and specificity of DST of GNB from pellets used for rapid identification (ID) directly from +BC. METHODS: Pellets from 1.5 mL of +BC in our lab have been used for the direct speciation of bacteria. For 109 samples identified as aerobic GNB on these pellets, the remainder of the sample was suspended in 0.45% NaCl to a standard density of 0.5 McFarland. Specimens were then processed using Vitek2(TM) N216 cards according to the manufacturer’s instructions. Results were compared with those from susceptibility testing from plate subculture 18 – 24 hours later. Antibiotics not reported on blood isolates from this card including were not included in the analysis. To increase the number of antibiotic-resistant organisms (ARO), 35 highly resistant GNB were seeded into BC and processed as above. RESULTS: Clinical specimens tested included 80 E. coli (EC), 14 K. pneumoniae, 5 K. oxytoca, 3 E. cloacae, and 2 P. aeruginosa, 2 P. mirabilis, and one each of S. marscesens, S. typhi, P. vulgaris, and R. planticola. DST results were available on average 24.8 (95% CI 23.4 - 26.2) hours earlier. Of 235 antibiotic resistant results, there were 2 very major errors (VME) (0.85%) with DST - both in EC against cefepime (CPM). These isolates flagged as possible ESBL-producers based on the ceftazidime (CTAZ) or ceftriaxone (CTRX) DST. Two major errors (MAE) occurred with trimpethoprim-sulfamethoxazole (0.85%). Minor errors occurred in 17 cases all within one dilution of the standard MIC. All of the other tests resulted in categorical or essential agreement. On the ARO tested, there were 405 antibiotic resistant results on standard testing. In this set, there were 10 VME (0.25%); including 1 amikacin, 2 CPM, 1 CTRX, 6 meropenem (MERO). Mechanisms of resistance in these VME included organisms with VIM (6), OXA (3), and KPC (1) genes. All VME with CPM, MERO, and CTRX were flagged as potential ESBL-producers base on the CTAZ result. One MAE with cefoxitin. CONCLUSION: DST from BC produces reliable results 24 hour sooner than standard testing. Isolates flagged for potential resistance to thirdgeneration cephalosporin and/or carbapenem on DST need further testing. DISCLOSURES: All authors: No reported disclosures.