VITEK MS Collaborative Study Outline - eBook

unknown organism and every organism or organism groups of the knowledge base. Then thresholds  1  were applied in order to retain only the more significant organisms or organism groups.  2  (a) A single identification is displayed with confidence value between 60 to 99.9 when one significant  3  organism or organism group are retained.   4  (b) Low Discrimination identifications are displayed when more than one significant organism or  5  organism group are retained, but not more than 4. In this case, the sum of confidence values is  6  equal to 100.   7  (c) When more than 4 organisms or organism groups are found, the organism is considered as non‐ 8  identified. In this case, a list of possible organisms is displayed and the sum of confidence values  9  is less than 100.   10  (d) When no match is found, the organism is considered as non‐identified  13  This validation study was conducted under the AOAC Research Institute Performance Tested Method SM 14  program and the AOAC INTERNATIONAL Methods Committee Guidelines for Validation of  15  Microbiological Methods for Food and Environmental Surfaces  (1). Two studies were conducted to  16  comprise the Internal Validation portion of the PTM study:  a Food Drug and Administration (FDA)  17  Clinical Trial (2, 3) and a study at bioMerieux. University of Guelph Laboratory Services Division  18  performed the Independent laboratory study. The VITEK MS system identified blind‐coded organisms,  19  either unknown clinical isolates or previously characterized organisms by comparing the characteristics  20  of the obtained spectrum to the VITEK MS knowledge base. A percent agreement calculation was  21  performed.  24  The majority of the claimed organism data included in this study was collected in a FDA Clinical Trial to  25  evaluate the ability of the VITEK MS to identify microorganisms. There were 359 different organisms  26  tested with multiple strains of each organism resulting in a total of 6344 VITEK MS results. The clinical  27  trial data was augmented with 193 blind coded organisms tested at bioMérieux in St. Louis, MO. These  28  two internal studies generated 6413 VITEK MS results.  31  An FDA‐approved protocol was provided to five independent clinical laboratories. Each laboratory  32  tested any unknown clinical isolates using the VITEK MS and DNA sequencing. If there was a shortage of  33  strains for a specific claimed organism, bioMérieux provided stock isolates to each laboratory for  34  analysis.   35  Unknown clinical isolates stored on an agar plate or slant, never frozen. Frozen stock isolates, provided  36  by bioMérieux , were subcultured twice prior to testing.  For this validation study report, organisms from  37  the FDA Clinical Trial that will be claimed for industry use are presented.  There were others tested but  38  only claimed for clinical use.  39 11  12  Validation Study   22  23  Internal Validation Study 29  30  FDA Clinical Trial  

Page 6 of 58