EFFICACY OF THE MICROTALLY® MITT FOR SAMPLING OF BEEF CARCASSES

Abstract

ABSTRACT EFFICACY OF THE MICROTALLY® MITT FOR SAMPLING OF BEEF CARCASSES Food safety is crucial in preventing foodborne illnesses and potential product loss due to pathogen contamination. Seven Shiga toxin-producing Escherichia coli (STEC) have been classified as adulterants in ground beef and any beef cuts intended for nonintact use. Therefore, testing carcasses and trimmings is a crucial step for identifying pathogenically contaminated products. Several sampling methods, including N60 excision, N60 PlusTM, and continuous and manual sampling devices using the Microtally® Swab or Mitt, are widely used in industry for verification testing. Among these, the Microtally® Mitt (MT-Mitt) is a sterile cloth mitt approved for sampling beef trim. However, the efficacy of the MT-Mitt for sampling beef carcasses has not been validated. Therefore, the objectives of Phase 1 of this Study were to identify an appropriate lot size for carcass sampling (Study 1) and compare the MT-Mitt’s effectiveness in bacterial recovery with that of carcass surface excision sampling (Study 2). The objective of Phase 2 was to compare the detection of pathogen index targets from carcass surface MT-Mitt samples with those from trim surface MT-Mitt samples from the same carcass. For Study 1 of Phase 1, nine sampling lot sizes were evaluated, where each lot size was represented by the number of carcasses sampled using the same MT-Mitt. The lot sizes were 1, 2, 3, 4, 6, 8, 10, 12, and 14 carcasses, with each carcass comprising two sides. For example, a lot size of 1 was 1 carcass (2 sides), a lot size of 2 was 2 carcasses (4 sides), and so on, were sampled with the same MT-Mitt. Additionally, for all lot sizes, sampling was conducted at two sites per carcass side using separate MT-Mitts: an upper sampling area that included the inside and outside round, and rectal area, and a lower sampling area that consisted of the chuck and brisket. Lot sizes 1, 2, 3, and 4 were replicated 5 times (n = 5) in plant A. Separately, lot sizes 4, 6, 8, 10, 12, and 14 were replicated 10 times (n = 10) in plant B. For Study 2 in Phase 1, carcass surface excision and MT-Mitt samples (n = 10) were obtained from two different plants on a different set of carcass sides than those utilized in Study 1. These samples were collected right after the harvest floor and in the chill cooler, and a lot size of 8 carcasses were used. The MT-Mitt samples were collected from the upper and lower areas of the sides, as previously described. Excision samples were collected using a sterile scalpel or a boning knife, depending on the facility, which was used to excise approximately 3 cm x 10 cm of carcass surface tissue from three locations in each sampling area (upper and lower). Excisions from the upper sampling area were from the hock, round, and rump, whereas excisions from the lower sampling area were from the foreshank, brisket, and short plate. All samples except for half of the samples collected in Study 2 were analyzed for indicator bacteria counts, using appropriate Petrifilm plates, to enumerate aerobic plate counts (APC), Enterobacteriaceae counts (EBC), total coliform counts (TCC), and E. coli counts (ECC). The remaining half of the Study 2 samples were analyzed for Shiga toxin-producing Escherichia coli (STEC) pathogen index targets. Data were analyzed using R using a linear regression model comparing lot sizes for Study 1, and excisions vs. MT-Mitt samples for Study 2. Phase 2 was conducted over 13 consecutive production days in a commercial beef processing plant using the same upper and lower sampling areas as used in Phase 1 for sampling carcasses with the MT-Mitt. Beef trim samples were collected by swabbing either a box or combo bin with the MT-Mitt for 90 seconds. A lot size of 8 carcasses was used based on the results from Phase 1, and a lot size of 5 boxes per pallet or a single trim combo bin was utilized for trim samples. Trim samples were correlated to carcass samples taken 48 hours previously. Both carcass and trim samples were evaluated for 8 pathogen index targets using polymerase chain reaction (PCR). One hundred and thirty carcass swab samples (a total of 1,040 carcasses) and 166 trim swab samples were collected and analyzed over the 13 days. Regardless of lot size in Study 1, the mean APC of the upper and lower sampling areas ranged from 1.2 to 2.9 and 1.6 to 2.2 log CFU/sample at plant A, respectively, and 2.1 to 2.7 and 1.8 to 3.0 log CFU/g at plant B, respectively. For the upper sampling area, there was a difference (P < 0.05) between the APC of lot sizes 1, 2, and 3, at plant A, and lot sizes 8 and 10 at plant B, whereas for the lower sampling area, there was a difference (P < 0.05) between the APC of lot sizes 12 and 14 at plant B, but no difference between lot sizes 1-4 evaluated at plant A. Out of the 120 samples collected from all lot sizes and both sampling areas at plant B, 73%, 72%, and 3% had quantifiable (greater than the 0.7 log CFU/g detection limit) EBC (range: <0.7 to 4.3 log CFU/g), TCC (range: <0.7 to 4.1 log CFU/g), and ECC (range: <0.7 to 1.2 log CFU/g), respectively. In Study 2, the overall recovery of APC from MT-Mitt samples was greater (P < 0.05) than that obtained from the excision samples. Specifically, quantifiable APC (greater than the 1.1 log CFU/g detection limit) were recovered from 80% of the MT-Mitt samples collected from the upper sampling site and from 100% of the samples from the lower sampling site. In contrast, quantifiable APC (greater than the 0.5 log CFU/g detection limit) was recovered from 0% and 70% of the upper and lower sampling area excision samples, respectively. Overall, the mean APC of the MT-Mitt samples from the upper and lower sampling locations were <1.9 and 1.7 log CFU/g, respectively, and <0.5 and <0.9 log CFU/g, respectively, for the excision samples. For mitt and excision samples evaluated for pathogen index targets, there were no targets identified for any samples. The results of the study indicated that a lot size of 8 carcasses (16 sides) was an appropriate practical sampling size for sampling of beef carcass surfaces using the MT-Mitt. Moreover, under the experimental conditions of the study, the MT-Mitt was more effective in recovering indicator organisms compared to the excision sampling method. For Phase 2, across all 8 indicators (STEC, O157, O111, O103, O26, O145, O121, and O45) for all carcass samples collected in Round 1, only two samples had detectable values: one for STEC and the other for O45. Of the 166 beef trim samples collected, 76 total detectable values were identified across the 8 indicators. In Round 2, more similar results were seen between carcass and trim samples for index target prevalence, with an average of 4% difference between carcass and trim samples on days 1 and 2. Carcasses on day 3 outperformed trim samples by having a higher target prevalence. Overall, there were inconsistencies between the carcass and trim results, suggesting the need for additional testing to fully validate the MT-Mitt for carcass sampling.