AOAC 2018 Methods

B ird et al . : J ournal of AOAC I nternational V ol . 102, N o . 1, 2019  109

Collaborative Study

Test Portion Distribution

All samples were labeled with a randomized, blind-coded three-digit number affixed to the sample container. Test portions were shipped on a Wednesday via overnight delivery according to the Category B Dangerous Goods shipment regulations set forth by the International Air Transport Association. Upon receipt, samples were held by the collaborating laboratory at ambient temperature (20–25°C) until the following Monday when analysis was initiated. In addition to each of the test portions and a separate aerobic plate count sample, collaborators received a data logger. Data loggers were programmed to monitor the temperature of the shipment to ensure it did not fall outside the range of 18–25°C. Participants were instructed to notify the study sponsor if the red alarm light was illuminated on the data logger upon receipt of the sample, which would indicate that the temperature fell outside the target temperature range (18–25°C). The information retrieved from the units were recorded on the “Sample Receipt Confirmation” form provided and faxed or emailed back to the study director. No laboratories reported temperatures outside the target range. Collaborators were instructed to follow the appropriate preparation and analysis as outlined in the study protocol. For this paired study design, each collaborator received 36 test portions (12 high, 12 low, and 12 uninoculated controls) preweighed to 10 g in sterile filter stomacher bags. Participants were instructed to add 90 mL of buffered peptone water (BPW) International Organization for Standardization (ISO) formulation (8; BPW-ISO), homogenize by stomaching or hand massaging for 2 min, and incubate at 37 ± 1°C for 18–20 h. Following enrichment, samples were assayed by the 3MMDA 2 – Cronobacter method and, regardless of presumptive result, confirmed following the ISO 22964:2017 reference method, beginning with a transfer to secondary enrichment [ Cronobacter Screening Broth (CSB)]. CSB tubes were incubated at 41.5 ± 2°C for 24 ± 2 h. After incubation, all tubes were streaked on Cronobacter Chromogenic Isolation Agar (CCI). CCI plates were incubated at 41.5 ± 2°C for 24 ± 2 h. Plates were observed for typical colonies (small to medium-sized and blue to blue-green in color), and if present, a well-isolated colony was streaked on TSA and incubated at 37 ± 2°C for 18–24 h. Isolates were confirmed positive by having a negative oxidase test and biochemical tests using API 20 E or VITEK 2 GN biochemical identification test (AOAC Official Method 2011.17 ; 9). Each collaborating laboratory reported results on the data sheets provided. The data sheets were submitted to the study director at the end of testing for statistical analysis. Data for each contamination level was analyzed using the probability of detection (POD) statistical model (10) and conducted using the AOAC Binary Data Interlaboratory Study Workbook, Version 2.3 (11). For each laboratory, the POD was calculated for the candidate presumptive results (POD CP ), the candidate confirmatory results [including false negative (FN) results; Test Portion Analysis Statistical Analysis

Study Design In this collaborative study, one matrix, powdered infant formula (milk-based with iron and docosahexaenoic acid) containing probiotics ( Lactobacillus reuteri ), was evaluated. The matrix was obtained from a local retailer and screened negative for the presence of Cronobacter by the ISO 22964:2017 reference method and by the 3MMDA2 – Cronobacter method. The matrix was artificially contaminated with a lyophilized culture of Cronobacter sakazakii , Q Laboratories isolate 17031.4 (origin: powdered infant formula), at two inoculation levels, a high inoculation level of approximately 2–5 colony-forming units (CFU)/test portion and a low inoculation level of approximately 0.2–2 CFU/test portion. A set of uninoculated control test portions (0 CFU/test portion) was also included. The 3M MDA 2 – Cronobacter and ISO 22964:2017 share a common pre-enrichment; therefore, the study was designed using paired samples. A total of 36 samples were evaluated per collaborator. Twelve replicate samples from each of the three inoculation levels were evaluated by 14 analysts from 11 locations. Collaborators were also sent a test portion for determining the total aerobic plate count (APC) using the ISO 4833-1:2013 (6) reference method on the day samples were received. A detailed collaborative study packet outlining all necessary information related to the study including media preparation, test portion preparation, and documentation of results was sent to each collaborating laboratory prior to the initiation of the study. The C. sakazakii isolate used in this evaluation was lyophilized prior to inoculation. The culture was propagated onto Trypticase Soy Agar with 5% sheep blood (TSA-SB) from a Q Laboratories frozen stock culture stored at –70°C. To prepare the culture for lyophilization, a single, well-isolated colony from TSA-SB was transferred into brain heart infusion broth and incubated at 37 ± 2°C for 18–24 h. The culture was diluted in a sterile cryoprotectant, reconstituted 10% nonfat dry milk, and freeze dried for 48–72 h. A bulk lot of the test matrix was inoculated with the culture at a high level expected to yield all positive results. The bulk lot was placed into a large stainless-steel container and mixed with sterile spatula for 30 ± 1 min. An aliquot of the high-level inoculum was further mixed in the same manner with uninoculated powdered infant formula to produce the low-level inoculum. After inoculation, the matrix was held for a minimum of 2 weeks at ambient temperature (20–25°C). The inoculated test product was packaged into separate 10 g samples in sterile Whirl-Pak ® bags and shipped to the collaborators. To determine the level of Cronobacter in the matrixes, a 5-tube most probable number (MPN) was conducted by the coordinating laboratory on the day of the initiation of analysis using the ISO 22964:2017 reference method. The MPN was determined by analyzing 5 × 20 g test portions, the reference method test portions from the collaborating laboratories 12 × 10 and 5 × 5 g test portions. The MPN and 95% confidence intervals were calculated using the Least Cost Formulations MPN Calculator, Version 1.6, provided by AOAC Research Institute (7). Preparation of Inocula and Test Portions