Evaluation of the Containment of Antimicrobial-Resistant Salmonella Species from a Hazard Analysis and Critical Control Point (HACCP) and a Non-HACCP Pig Slaughterhouses in Northeast Thailand

To evaluate the containment of antimicrobial-resistant (AMR) Salmonella contaminations of a HACCP slaughterhouse (HACCP SH) and a non-HACCP slaughterhouse (non-HACCPSH), 360 paired pig rectal (representing the farm pig status) and carcass samples (representing the contamination) were collected equally from the two slaughterhouses that serviced 6 and 12 farms, respectively, in Northeast Thailand (n = 720). The purified Salmonella isolates were serotype identified, antimicrobial susceptibility tested, and pulsed-field gel electrophoresis (PFGE) assessed. Four evaluations of two slaughterhouses were examined: (1) the means of slaughtering contamination rates (SCR) (to evaluate the contamination level by averaged farm SCRs): the HACCP SH decreased contamination (SCR: −48.89% ± 8.80%, n = 6), whereas the non-HACCP SH increased (SCR: 14.31% ± 9.35%, n = 12). (2) The serotype diversity: the HACCP SH decreased the diversity from the rectal group (110 isolates, 9 serotypes) to carcass group (23 isolates, 3 serotypes), whereas there was no decrease in the non-HACCP SH (rectal group (66 isolates, 14 serotypes) and carcass group (31 isolates, 10 serotypes)). (3) The AMR patterns: the HACCP SH decreased from rectal group (96 isolates, 7 patterns) to carcass group (22 isolates, 1 pattern), whereas there was no decrease from the non-HACCP SH rectal group (22 isolates, 7 patterns) to carcass group (48 isolates, 8 patterns). (4) The estimated indirect contamination rate (by serotype screening and PFGE confirmation): the HACCP SH was 60.87% (14/23), whereas the non-HACCP SH was 98.48% (65/66). This study indicates that both the slaughterhouses keep a high level of indirect contamination; the HACCP SH decreases Salmonella contaminations and reduces the AMR patterns, the non-HACCP SH increases contaminations.


Introduction
Salmonella spp. are dangerous pathogens often found in the food chain. Salmonella (non-typhoidal) are one of the four chief global causes of 550 million annual diarrheal disease cases, announced by the World Health Organization [1]. Salmonella enteritidis was the most common detected agent of

Materials and Methods
Slaughtering methods: The slaughtering procedures of the two slaughterhouse varied. For HACCP slaughterhouse (HACCP SH), they included: stunning (shackling), sticking (bleeding), scalding, dehairing, gambrelling, polishing (singeing or shaving), pre-evisceration wash, bunging, head dropping, brisket opening, pull leaf fat, final trim, grade stand, and final wash. For the non-HACCP slaughterhouse (non-HACCP SH) procedures, they had not been normalized and they did not include a pre-evisceration wash nor a final wash step.
Pig sample collection and Salmonella isolation: For this study, 360 paired rectal and carcass samples were collected from 360 pigs at two slaughterhouses in Northeast Thailand from 2017 to 2018 (n = 720 total samples). Collection of these two samples from an individual pig represent the farm source Salmonella as determined by a positive rectal swab and contaminated product as determined by positive carcass swab of the pig.
One hundred and eighty paired samples were obtained from a single large-scale HACCP slaughterhouse (HACCP SH) that serviced six different farms (farms No. 1-6). The HACCP SH had good manufacturing practices (GMP) and HACCP certificates from the Department of Livestock Development (DLD, Bangkok, Thailand), Thailand. The other 180 paired samples were obtained from a small-scale slaughterhouse (non-HACCP SH) servicing twelve different farms (farms No. [7][8][9][10][11][12][13][14][15][16][17][18]. The non-HACCP SH was registered with the local DLD office, Thailand. The HACCP SH was characterized by (1) restricting access to the slaughterhouse, (2) the assembly line, and 3) the cleaning and disinfection of critical points of the slaughtering process. The non-HACCP SH was characterized by (1) open access to the slaughterhouse, (2) disordered slaughtering processes, and 3) limited cleaning and disinfection steps. To determine whether the contaminated products in both slaughterhouses were produced indirectly from slaughtering, rectal samples were collected after stunning the pigs, and the carcass samples were collected by sterilized swabs covering a 600 cm 2 area of each half pig carcass from the primal cuts after the final wash (from the HACCP SH) or from the primal cuts without a final wash (from the non-HACCP SH). All paired samples were meticulously matched and stored in an icebox for transportation.
Salmonella spp. isolation and detection were performed at the laboratory of Khon Kaen University Veterinary Teaching Hospital. The isolation of Salmonella spp. was performed, following the International Organization for Standardization (ISO, Geneva, Switzerland) standard procedure 6579-1:2002 [18]. The presumptive colonies (n = 3) of each sample were identified on XLD agar plates as Salmonella by their black color. Then the presumptive colonies were biochemically screened by use of triple sugar iron agar (TSI; Becton Dickinson, NJ, USA) and motility indole lysine medium (MIL; Becton Dickinson, NJ, USA) techniques. The identified Salmonella isolates were subsequently serotyped. The O (somatic) and H (flagellar) antigens of all rectum and carcass isolates were characterized by agglutination with hyperimmune sera in the World Health Organization (WHO) National Salmonella and Shigella Center for serotype identification, and the serotypes were assigned according to the Kauffmann-White scheme [19].
The slaughtering contamination rate (SCR) analysis for Salmomella: To evaluate the slaughterhouse efficiency at stopping Salmonella spread, the positive Salmonella rates of carcass and pig rectal isolates by farms were transformed into Salmonella slaughtering contamination rate (SCR). Then, the SCR means of two slaughterhouses were calculated.
The formula of the Salmonella SCR of one farm is: The formula used to determine the contaminated product rate for a farm was determined to be: the number of Salmonella positive carcass samples divided by the total number of individuals collected, e.g., for carcass contamination rate of farm NO.1: 22 Salmonella positive samples divided by 30 collected individuals times 100% = 73.33%. The formula of the pig source rate for a farm was determined to be: the number of Salmonella positive rectum samples divided by the total number of individuals collected, e.g., for rectum (infected animal) Salmonella positive rate of farm NO.1: 29 positive samples divided by 30 times 100% = 96.67%. Then, the SCR of farm NO.1 is 73.33%-96.67% = −23.33%.
This formula reveals the difference of the Salmonella-contaminated product rate to Salmonella infected animal rate on one farm (Table 1). The SCR quantifies the efficiency at stopping Salmonella spread during the slaughtering process. A negative SCR means that the contaminated product rate is lower than the infected animal rate according to the formula, indicating that the contamination rate decreased. Conversely, a positive SCR indicates increasing contamination by the slaughtering process. The lower SCR illustrates a better slaughtering work at blocking pathogen spread. Mean ± standard error of mean 14.31% ± 9.35% Note: (a) The formula of SCR = (contaminated product rate − infected animal rate) × 100% reveals the value difference of the Salmonella-contaminated product rate to the Salmonella infected animal rate of each farm. 'Rectum' stands for the Salmonella infected animal rate. 'Carcass' means the Salmonella-contaminated product rate. (b) Interestingly, although the rising tendency of contaminations in non-HACCP SH was presented, three reduced contamination cases were observed. These results suggest that a single SCR analysis of one-time slaughtering is inadequate to estimate the true SCR for the slaughterhouse. Serotype diversity of the two slaughterhouses: From the serotype diversity result, the serotype diversity including the serotypes and their amounts of carcass group (23 isolates with 3 serotypes) was significantly lower than the pig rectal sample group (110 isolates with 9 serotypes) (p < 0.05) in HACCP SH; the diversity of the carcass group (66 isolates with 14 serotypes) was not significantly different from the pig rectal group (31 isolates with 10 serotypes) (p > 0.05) in the non-HACCP SH. In particular, six serotypes were prevented from entering the food chain by HACCP slaughtering process. In non-HACCP SH, seven extra serotypes were introduced into the food chain, and three serotypes were blocked. The HACCP SH procedures can reduce the serotype diversities from entering into the food chain, but the non-HACCP SH introduced extra serotypes into the food chain (Table 2 and Figure 1).   Note: These results show that among pigs from the 6 farms that were provided to the HACCP slaughterhouse (HACCP SH), high carrier status of Salmonella (61.11%) was found. However, in the HACCP slaughterhouse, only 12.78% of the carcass samples were Salmonella positive, suggesting a reduced contamination level. In contradiction to these observations of the HACCP SH, the presence of Salmonella found among pigs from the 12 farms provided to non-HACCP slaughterhouse (non-HACCP SH) was only 17.22 %, but the contamination proportion was increased to 36.67%.
Evaluation on Salmonella serotype diversities: The serotype diversities of rectal and carcass groups of the two slaughterhouses (including the serotypes and their amounts) were statistically analyzed by SPSS Statistics 17.0 (SPSS Inc, Chicago, IL, USA), using Pearson's chi-squared test. p < 0.05 is considered significant.
The indirect contamination cases analysis: The sources of contaminations could be sorted into (1) intrinsic pathogen presence on the muscle tissue (the muscle tissue infected with pathogens); (2) direct contamination from the pig's own gut flora; (3) indirect contamination from other pigs or the slaughterhouse environment. The intrinsic pathogen presence cases and the direct contamination cases would exhibit isolates common to both the rectal and carcass i.e., have the same Salmonella serotype and genetic type. The indirect contamination cases would consist of Salmonella spp. from the carcass sample that were not derived from the pig itself. Therefore, indirect contamination could be defined as the rest of the contamination cases except those of the same subtype pairs of isolates (assumed to be due to the intrinsic pathogen presence cases and direct contamination).
The same subtype pairs of isolates were attained following these procedures: (1) the serotypes of all pig rectal and carcass Salmonella isolates were screened for matched pairs of isolates by web-based software: calculate and draw custom Venn diagrams [21]; (2) the matched pair isolates were genetically confirmed by pulsed field gel electrophoresis (PFGE).
The PFGE analysis was performed following the CDC standard [22]. The DNA fragments were digested by Xbal and separated by a CHEF-DRIII pulsed-field electrophoresis system. The gels were stained with ethidium bromide and documented by ChemiDocTM XRS+ (Bio-Rad, Hercules, California, USA). The dendrogram was produced using band clustering with the dice coefficient similarity index of 1% optimization and 1% tolerance by the unweighted pair group method with arithmetic means (UPGMA) by BioNumerics version 7.6 (Applied Maths, Sint-Martens-Latem, Belgium).

Results and Discussion
The means of slaughtering contamination rate (SCR) of the two slaughterhouses: The paired samples from the HACCP SH consisted of 61.11% (110/180) and 12.78% (23/180) Salmonella positive rectal and carcass samples, respectively. Whereas, the paired samples from the non-HACCP SH consisted of 17.22% (31/180) and 36.67% (66/180) Salmonella positive rectal and carcass samples, respectively ( Table 2). The SCR of Salmonella utilized here provides evidence on the efficiency at stopping the spread of Salmonella in the slaughterhouse. It is defined as the rate difference of Salmonella-contaminated products to Salmonella infected animals on one farm. The SCRs of each farm and the SCR means of two slaughterhouses were calculated (Table 1). The parameter "n" is the sample size of the investigated farms for one slaughterhouse. It influences the estimation power of the true SCR for the slaughterhouse. The SCR mean of HACCP SH was −48.89% ± 8.80% (n = 6); the SCR mean of non-HACCP SH was 14.31% ± 9.35% (n = 12), indicating this HACCP SH reduced the contamination level of Salmonella, while the non-HACCP SH contaminated more products. The SCR was utilized to evaluate the efficiency at stopping Salmonella contamination of meat products. We suggest it to be used as an important parameter assessed during the evaluation of contamination in slaughterhouses. In addition, the limited function of the HACCP system was exposed, that is some Salmonella-contaminated products were detected. Thus, additional processes are needed to decrease Salmonella contamination further. This study's results are consistent with the other reports of positive Salmonella outcomes [11,12].
Antimicrobial resistance (AMR) of the two slaughterhouses: The susceptibility of Salmonella isolates to 7 antimicrobial agents was examined. The positive rectum source Salmonella spp. from the HACCP SH was resistant to AMP (76.36%), CHL (11.82%), STR (60.00%), SXT (17.27%), and TET (85.45%), while the isolates from carcass were resistant to AMP (95.65%), STR (95.65%), and TET (95.65%) ( Table 3). From the non-HACCP SH, the pig rectum source were AMP (61.29%), CHL (3.23%), STR (29.03%), SXT (38.71%), and TET (64.52%) resistant, while the carcass source were AMP (69.70%), CHL (1.52%), STR (45.45%), SXT (15.15%), and TET (69.70%) resistant ( Table 3). The most common resistance antimicrobials are similar to that reported in Thailand (AMP, STR, and TET resistance) [7,23]. The AMP, STR, and TET of the carcass were significantly higher than the rectum source in HACCP SH. The CHL and SXT of the carcass were significantly lower than the rectal group in HACCP SH. The STR of the carcass was significantly higher than the rectal group in non-HACCP SH. The SXT of the carcass was significantly lower than the rectal group in non-HACCP SH. It is unclear as to the antimicrobial resistance relationship in the two slaughterhouses and sample sources.  From the AMR pattern comparison result, the AMR patterns of the carcass group were significantly different from the pig rectal group (p < 0.01) in HACCP SH, while the AMR patterns of the carcass group were not significantly different from the pig rectal group (p > 0.05) in the non-HACCP SH. In the HACCP SH, the numbers of AMR patterns in carcass group (22 isolates with 1 pattern) were less than pig rectal group (96 isolates with 7 patterns), but the proportion of AMP-STR-TET in carcass group was higher than the pig rectal group. These results demonstrate that the AMR patterns in HACCP SH were reduced to one pattern after slaughtering. Nevertherless, the non-HACCP SH did not reduce the presence and diversity of AMR among the Salmonella-contaminated products, between the rectal group (22 isolates with 7 patterns) and carcass group (48 isolates with 8 patterns) ( Table 4). Table 4. Antimicrobial resistance (AMR) patterns of Salmonella spp. from the two slaughterhouses.

Pattern HACCP (%) (n/N) non-HACCP (%) (n/N) Total (%) (n/N) Rectum
Carcass The indirect contamination cases of the two slaughterhouses: To assess indirect contamination cases (the rest results of contaminations of the same subtype pairs of isolates). All isolated serotypes were screened by web-based software: calculate and draw custom Venn diagrams based on serotyping and genetically confirmed by PFGE. The 12 rectal and 12 carcass isolates were serologically matched in pairs from the total 141 rectal and 89 carcass isolates. Ten of the 12 pairs were from HACCP SH, and two pairs were from non-HACCP SH (Figure 2). Then, the 12 serologically paired isolates were further assessed by PFGE (n = 24). The serotype matched pairs: R19 and C19, R341 and C341 were genetically different (Figure 3). The advanced indirect contamination cases were 14 from the total 23 contaminations in HACCP SH, and 65 from the total 66 contaminations in non-HACCP SH after PFGE confirmation. The rate of indirect contamination was 88.76% for the total contaminated carcasses (from two slaughterhouses) (79/89), revealing most cases were indirectly contaminated. Most of the indirect contaminations from HACCP slaughterhouse happened at one farm slaughtering, rated as 63.64% (14/22), total indirect contamination rate was 60.87% (14/23). Besides, almost all contaminations of each farm from the non-HACCP slaughtering were indirect contaminations, rated as 98.48% (65/66) ( Figure 4). These results illustrate that both the non-HACCP SH and the HACCP SH maintain a high level of indirect contamination. These indirect contaminations could be prevented by the improvement of the slaughtering procedures. Therefore, this research proposed that the management of slaughtering procedures should be paid more attention. Some limitations to the research should be noted. (1) There could be some 'same subtype pairs of isolates' due to indirect contamination. This would lead to a lower estimation rate than the true indirect contamination rate. Despite the imprecise estimation of indirect contamination cases (the true indirect contamination rate is higher than the estimation), the current results still exposed a serious indirect contamination situation. (2) Although the isolation of Salmonella from each sample was repeated, there still might be some Salmonella spp. not discovered. Nevertheless, all data were analyzed based on the same level of the detection limit.
Pathogens 2020, 9, 20 9 of 12 10 true indirect contamination rate. Despite the imprecise estimation of indirect contamination cases (the true indirect contamination rate is higher than the estimation), the current results still exposed a serious indirect contamination situation. (2) Although the isolation of Salmonella from each sample was repeated, there still might be some Salmonella spp. not discovered. Nevertheless, all data were analyzed based on the same level of the detection limit. Note: The 141 pig rectum and 89 carcass Salmonella isolates were screened by serotyping. Twelve paired isolates were matched, including ten pairs from HACCP SH and two pairs from non-HACCP SH. The serotype matched pair isolates need to be confirmed genetically. Only the serotype and genotype different carcass isolates were considered indirect contamination cases. Figure 2. Serotype matched pairs of isolates between two slaughterhouses. Note: The 141 pig rectum and 89 carcass Salmonella isolates were screened by serotyping. Twelve paired isolates were matched, including ten pairs from HACCP SH and two pairs from non-HACCP SH. The serotype matched pair isolates need to be confirmed genetically. Only the serotype and genotype different carcass isolates were considered indirect contamination cases.

Conclusions
According to the results of this study, it is worth considering the slaughtering procedures used in controlling AMR Salmonella contamination of pig products from farms to the dining tables. Firstly, the HACCP slaughterhouse reduced the Salmonella occurrence, serotype numbers, and serotype diversity; whereas the non-HACCP slaughterhouse increased the Salmonella occurrence and introduced extra serotypes of Salmonella spp. into the meat products. Secondly, in both slaughterhouses, most of the AMR Salmonella spp.-contaminated products were associated with indirect contamination during slaughtering. This indicates that the AMR pathogens found at the dining tables may not directly relate to Salmonella from the farms. The results of this work may explain the phenomenon of the increasing Salmonella prevalence from farms to the markets [6,7]. These indirect contaminations may keep accumulating from slaughtering until selling. More work should focus on the individual evaluations of each slaughterhouse to improve their efficiencies at pathogen containment. Further, the food chain includes many steps from slaughtering to the final products. More studies on the other parts of the whole food chain process are needed to limit the risk of food contamination.