VITEK MS Collaborative Study Outline - eBook

The VITEK MS was evaluated for robustness by examining the performance of the assay when small  1  changes in the operating conditions were made using a screening factorial experimental design (Table  2  15) for each group of organisms (Table 16). The following parameters were evaluated: (1) 1 µL loops  3  from two different suppliers, (2) the age of the culture on an agar plate before analysis, (3) three  4  suppliers of the agar media.  5  The organisms (Table 16) were inoculated onto different media according to their growth requirements  6  and incubated in appropriate conditions for 24 to 72 hours according to the experimental design. The  7  experimental design was developed for each group of organisms and contained 18 different  8  experiments.  9  Cross‐validation performances were calculated for each incubation time (all 9 media and 3 suppliers  10  taken together).  This allowed for an assessment of how the results will generalize to an independent  11  data set.  One round of cross‐validation involved partitioning a sample of data into complementary  12  subsets, then performing the analysis on the subset (called the training set) and then validating the  13  analysis on the other subset (called the validation set).   14  The rules for performance calculations:  15  1. One choice correct species  One Choice Correct  16  2. All choices have the same genus as the reference genus  Correct ID  17  3. One choice wrong species  Mis ID  18  4. All choices have the same genus but do not match reference genus  Mis ID  19  5. No ID  No ID  20  6. Choices have multiple genera  No ID  21  22  The acceptance criterion is 90% of overall Correct ID (Once Choice correct + Correct ID) for each  23  incubation time.  In each case, the 95% exact binomial two‐sided confidence interval of the Overall  24  Correct ID was calculated.  25  Results  26  After evaluating the following parameters: (1) 1 µL loops from two different suppliers, (2) the age of the  27  culture on an agar plate before analysis, (3) various suppliers of the agar media, it was determined that  28  these changes still provide acceptable results with the VITEK MS at all time points.  For the bacteria  29  evaluation, 93.02% of the results were correctly identified after 24 hr, 94.29% after 48 hr and 92.81%  30  after 72 hr. For the yeast evaluation, 96.54% of the results were correctly identified after 24 hr, 95.34%  31  after 48 hr and 94.35% after 72 hr. The acceptance criterion is 90%. See Appendix 5, Tables 16‐19 for  32  study details and results. 

33  34 

35

Product Consistency (lot‐to‐lot) and Stability Studies

Page 9 of 58