ABSTRACT

The Food Safety and Inspection Service (FSIS) conducts microbiological baseline studies to determine national prevalence of select foodborne pathogens in federally inspected meat and poultry products and to obtain data for risk assessments. The FSIS conducted a baseline study from 1 June 2017 through 31 May 2018 to characterize and determine the prevalence of Salmonella and assess the occurrence of Shiga toxin–producing Escherichia coli (STEC) in a variety of raw pork products. In total, 4,014 samples from slaughter and processing establishments were analyzed for Salmonella; a subset of these samples (1,395) from slaughter establishments were also analyzed for STEC. Analyses determined that the national prevalence of Salmonella in raw pork products was highest in comminuted products (28.9%), followed by intact cuts (5.3%) and nonintact cuts (3.9%). Less than 1% of samples analyzed were positive for the top seven STEC. Our findings indicate there is a need for additional pathogen reduction strategies for raw pork products.

As part of the Pathogen Reduction and Hazard Analysis Critical Control Points (HACCP) System implemented in 1996, the U.S. Department of Agriculture's Food Safety and Inspection Service (FSIS) established policies aimed at reducing pathogenic microorganisms on meat and poultry products with the goal of reducing the incidence of foodborne illnesses associated with consuming these products. These policies included performance standards for Salmonella. FSIS also mandated that establishments develop and implement HACCP plans to control and prevent hazards likely to occur during slaughter and processing (28). Foodborne pathogens are among the hazards that HACCP plans are expected to address.

Each year, there are an estimated 48 million illnesses in the United States caused by foodborne pathogens, resulting in approximately 128,000 hospitalizations and 3,000 deaths (6, 24). Approximately 9.6 million of these foodborne illnesses are attributed to 31 known pathogens. Of these, Salmonella ranks among the top five pathogens responsible for ∼90% of illnesses with known etiology. Shiga toxin–producing Escherichia coli (STEC) and Salmonella are among those for which infection results in the most hospitalizations or deaths (6, 24).

Evidence suggests pork is an important source of Salmonella infection and potentially an underrecognized source of STEC infections. Swine are known reservoirs (asymptomatic carriers and shedders) of Salmonella (9), and there are documented cases of Salmonella foodborne illnesses attributable to pork, including those associated with outbreaks (4, 9, 15, 22, 25). Other studies show that swine harbor various serotypes of STEC, including O157:H7 (11, 16, 39) and that STEC O157:H7 organisms are highly pathogenic in swine (8). Foodborne outbreaks caused by consumption of pork products contaminated with STEC O157:H7 occurred in Canada in 2014 and 2018 (10, 13); however, conclusive evidence of widespread pork-associated STEC infections in the United States is limited.

To proactively guide policy decisions, the FSIS reviews sampling data generated through its verification testing (28). Starting in 1996, with the Pathogen Reduction and HACCP rule, market hog carcasses were sampled and assessed under Salmonella performance standards. The very low frequency of positive results called into question whether sampling was warranted in this commodity. The FSIS completed the Market Hog Baseline Survey in 2011 to better ascertain Salmonella prevalence on market hog carcasses. The results of the study found that only an estimated 1.66% of market hog carcasses were positive for Salmonella at postchill (29). This finding, in addition to years of verification testing data, supported the decision of FSIS to suspend hog carcass sampling in 2012. However, one limitation of the Market Hog Baseline Survey was that no data on pathogen contamination in raw pork products was collected.

In 2017, the FSIS conducted a raw pork baseline study to obtain insight into pathogen contamination in raw pork products. The main objectives of the study were to characterize and determine prevalence of Salmonella and occurrence of the top seven STEC (serogroups O157, O111, O103, O26, O121, O45, and O145) in a variety of raw pork products. Other objectives included assessing the relationship between pathogen presence and other factors, such as indicator organisms and seasonality.

EXPERIMENTAL DESIGN AND PROCEDURES

Exploratory sampling (phase I) was conducted to obtain data to inform the design of the baseline study. In total, 1,200 samples were randomly collected from federal slaughter and processing establishments that produced the categories of raw pork products described in Table 1, in addition to two minor “other” categories. Phase I samples were analyzed for Salmonella and indicator organisms (E. coli, coliforms, Enterobacteriaceae, and aerobic counts [AC]), and a subset (200 total) were analyzed for the presence of the top seven STEC. Drawing upon what was learned in phase I (35), the Raw Pork Baseline Study was then initiated to obtain nationally representative data on raw pork products.

TABLE 1

Eligible raw pork productsa

Eligible raw pork productsa
Eligible raw pork productsa

The baseline study was designed to reflect the entire federally inspected supply of raw pork products within the United States. The study identified federally inspected establishments that slaughter or produce pork products and their individual average daily production volume by using data recorded in the FSIS Public Health Information System at the beginning of the study, 1 June 2017. The status of a specific establishment is dynamic and was evaluated throughout the project, with adjustments made to reflect changes in an establishment's status (Supplemental Tables 1S, 2S, and 3S). Establishments included in the sampling frame included all federal establishments in the contiguous United States that produced eligible pork products and those that produced a daily average of greater than 1,000 lb/day (453.59 kg/day) for at least 21 days in a calendar month.

Eligible pork products in the baseline study included three product groups (comminuted, intact, and nonintact cuts) sampled by using a stratified approach and was based on establishment production volume for each product group. Table 2 describes the frequency of scheduled samples for each stratum. Stratum boundaries were determined by applying the visual clustering method (18) to establishment production volume data. Sample allocation within each stratum was based on the percentage of samples positive for Salmonella in phase I and was designed to provide appropriate representation across all production volumes and to provide adequate samples per establishment for risk assessment purposes. The standard deviation of the percent positive values for Salmonella for each raw pork product group was calculated from the phase I data and used to determine the estimated number of samples needed for the baseline study. The number of samples needed was calculated for each raw pork product group; the calculation for comminuted products indicated that 1,680 samples would suffice to yield an 80% chance of detecting the pathogen. The number of samples required for intact and nonintact raw pork product groups were calculated in the same manner, resulting in 1,260 for intact and 1,248 for nonintact cuts, respectively. Therefore, the estimated number of samples for the survey was 4,188. Establishments within the sampling frame were sampled for 12 months.

TABLE 2

Sample allocation for raw pork product groups used

Sample allocation for raw pork product groups used
Sample allocation for raw pork product groups used

As standard agency policy, all samples received in acceptable condition were analyzed for Salmonella and indicator organisms (E. coli and AC). On a subset of samples, STEC testing was conducted; these samples were collected from establishments that performed on-site slaughter and processing. Culture and characterization followed the procedures outlined in the FSIS Microbiology Laboratory Guidebook (3034). Antibiotic susceptibility was done according to procedures developed by the National Antimicrobial Resistance Monitoring System (19). Isolates were sequenced on the Illumina MiSeq platform by using standard laboratory procedures for PulseNet Nextera XT library for the Illumina sequencer (5), and antibiotic resistance genetic markers responsible for the antibiotic resistance phenotype were determined by using the U.S. National Library of Medicine, National Center for Biotechnology Information Database Pathogen Browser in conjunction with the Antibiotic Resistance Isolate Bank hosted by the Centers for Disease Control and Prevention and U.S. Food and Drug Administration (FDA) (7, 20).

Microbiological and other findings for each product group were summarized by using descriptive statistics. The national prevalence or weighted average was calculated for Salmonella to provide a national average of expected values for Salmonella associated with raw pork products (36). The design of the study used class or “strata” by production volume. This design ensured that small establishments were accurately represented, despite their low production volume (36). To counterbalance strata sampling bias, establishments were weighted by assigning each corresponding establishment's fractional contribution to the total national production (36). Once weighted, the individual sample weight was calculated by taking shares in direct proportion to the number of samples taken from the establishment weight. The statistical software WesVar version 5.1 (Westat, Rockville, MD) was used to calculate the national prevalence with confidence limits (36).

Levels of E. coli and AC were estimated by using the most-probable-number (MPN) procedure, and the results were reported as CFU per gram. The MPN estimates the population density of viable microorganisms in a test sample (34). The range of detectable indicator organisms for the samples was quantified. To assess if there was a correlation between indicator organisms and Salmonella, the point biserial correlation coefficient (r) was calculated; this value measures the strength of co-occurrence between indicator organisms and Salmonella as positive or negative. Ranges of r values are between +1 and −1, where −1 indicates a negative correlation, +1 indicates a positive correlation, and a value of 0 indicates no correlation.

RESULTS

During the 12-month period of the survey, 4,232 samples were collected from 285 establishments. Of 4,232 samples collected, 218 samples were discarded because the sample integrity had been compromised, resulting in 4,014 tests samples analyzed. Some of the reasons samples were discarded include temperature of sample was too warm, incomplete sample forms, samples held too long before shipping due to courier error, packaging damaged during shipment, and package seal improperly placed.

Salmonella

Table 3 shows the breakdown of samples by product group, percent positive for Salmonella, and national prevalence. Comminuted products showed the highest rate of Salmonella contamination at 21.2% (380 isolates and 1,796 comminuted samples analyzed); this translated to a national prevalence of 28.9% (95% confidence interval: 24.1 to 33.8%). Of the 545 isolates recovered from the 4,014 samples analyzed, 52 distinct Salmonella serotypes were represented. The top five serotypes were Anatum at 13.8% (75 strains and 545 isolates), Infantis at 13.0% (71 strains and 545 isolates), Johannesburg at 9.0% (49 strains and 545 isolates), Derby at 8.6% (47 strains and 545 isolates), and I 4,[5],12:i:− at 6.0% (33 strains and 545 isolates). Recovery of Salmonella isolates did not show any discernible seasonal trend.

TABLE 3

Baseline study results for Salmonella

Baseline study results for Salmonella
Baseline study results for Salmonella

Antibiotic resistance among Salmonella isolates

Of the 545 isolates recovered, 371 (68.1%) were pansusceptible, 86 (15.8%) were resistant to one or two antibiotics, and 88 (16.1%) were resistant to three or more antibiotic classes. Table 4 shows the number of isolates associated with various resistance profiles by serotype. Twenty-one Salmonella isolates exhibited resistance to ceftriaxone, and three were resistant to azithromycin. Azithromycin and ceftriaxone antibiotics are critically important for human medicine. Eleven isolates exhibited resistance to trimethoprim-sulfamethoxazole, which are highly important antimicrobial drugs used to treat various types of gastrointestinal bacterial infections in humans. Importantly, 5 of 11 of these same isolates were also resistant to either azithromycin or ceftriaxone.

TABLE 4

Antibiotic resistance phenotypes of Salmonella isolates recovered in phase IIa

Antibiotic resistance phenotypes of Salmonella isolates recovered in phase IIa
Antibiotic resistance phenotypes of Salmonella isolates recovered in phase IIa

All 33 isolates of Salmonella I 4,[5],12:i:− carried antibiotic resistance markers, and 82% of these strains had multiple drug resistance phenotypes. Further genomic analysis identified ciprofloxacin resistance qnrB2 and qnrB19 genes in Salmonella I 4,[5],12:i:−. Multiple Salmonella serotypes, including I 4,[5],12:i:−, exhibited an intermediate resistance phenotype to ciprofloxacin (19).

STEC

Of the 1,395 samples analyzed for STEC, 3 (0.2%) were positive, 2 were positive for E. coli O103, and 1 was positive for E. coli O157. All were recovered from comminuted pork products. In addition, 309 samples screened positive for stx and eae genes but did not confirm positive for any of the top seven STEC serotypes currently analyzed by the FSIS. None of the STEC isolates were resistant to any of the antibiotics against which the strains were tested.

Source of raw pork product and Salmonella presence

Market hogs were the predominant swine class used to produce all three of the raw pork products. Market hogs were the largest percentage of source material in comminuted raw pork product samples at 48.6% (872 of 1,796 samples analyzed), and sows at 21.8% (392 of 1,796 samples). In addition, 29.6% came from other classes of swine (532 of 1,796 samples; other swine classes included roaster pig, feral hog, boar, and stag). Comminuted product made from sow material resulted in a proportionately higher number of Salmonella isolates: 170 isolates of the total 380 Salmonella isolates were recovered from comminuted products (44.7%), compared with comminuted product prepared from market hogs at 35.3% (134 strains of 380 Salmonella isolates). Comminuted raw pork products sourced from other swine classes resulted in 20.0% of samples positive for Salmonella (76 strains of 380 Salmonella isolates).

A large portion of the intact and nonintact raw pork samples came from market hogs, 834 (71.3%) of 1,170 intact cut samples and 728 (69.4%) of 1,048 nonintact cut samples; a smaller portion, 1.3% of intact and 2.9% of nonintact cuts, were from sows. The remainder of the intact and nonintact cut samples were made from other swine classes, 27.4% (321 intact cut samples) and 27.7% (290 nonintact cut samples), respectively.

The number of Salmonella isolates recovered from slaughter and processing and processing-only establishments was compared. There were 2,561 samples from processing-only establishments analyzed and 251 (10%) of these samples contained Salmonella. Of 251 samples positive for Salmonella, 152 (60.6%) of these isolates were recovered from comminuted product derived from processing-only establishments. The intact samples were 56 (22.3%) of the 251 isolates, and the nonintact samples were 43 (17.1%) of the 251 isolates. A higher percentage of samples that contained Salmonella was recovered (20.2%) from samples collected from slaughter and processing establishments; specifically, 294 isolates were recovered from the 1,453 samples analyzed. The comminuted samples from slaughter and processing establishments were 228 (77.6%) of the 294 Salmonella isolates recovered. The intact samples included 41 (13.9%) of the 294 isolates, and the nonintact samples consisted of 25 (8.5%) of the 294 isolates.

Indicator organisms and correlation with pathogen presence

Table 5 shows the geometric mean for E. coli and AC for each product group. For comminuted products, the relationship between Salmonella and E. coli was determined by calculation of the biserial correlation coefficient, which resulted in an r value = 0.072 (95% confidence interval: 0.024 to 0.12). On the basis of these results, there is insufficient evidence to show a correlation between E. coli and Salmonella. Likewise, there is insufficient evidence of a correlation between AC and Salmonella in comminuted products: r value = 0.0082 (95% confidence interval: −0.04 to 0.056). There is insufficient evidence of a correlation between the presence of E. coli or AC indicators and Salmonella in intact and nonintact cuts.

TABLE 5

Estimate of indicator organisms associated with raw pork products

Estimate of indicator organisms associated with raw pork products
Estimate of indicator organisms associated with raw pork products

DISCUSSION

Pork is primarily consumed in the United States as minimally processed whole-muscle cuts and as processed raw products that are ground, cured, smoked, or seasoned prior to distribution (12). In this study, comminuted products contained the highest level of Salmonella, which is important given that comminuted products are used to make sausage and that sausage is among the pork products most commonly consumed in the United States (12). The results suggest that there are steps in the processing of raw pork into comminuted products that allow pathogens to survive or grow despite current pathogen reduction strategies. Fabrication and grinding of meat used in the production of comminuted products could allow contamination present in the source materials to spread throughout the product. Studies suggest that comminuted pork might also contain lymphatic material (2, 21) and synovial fluid derived from stripping of meat from bones (26), which could increase Salmonella contamination. Understanding the pathogen risks associated with procedures used in slaughter and processing is essential for mitigation strategies.

Aerobic counts have been suggested as reliable indicators of process control because the magnitude of aerobic bacterial counts is highest on meat products and the estimation of the MPN per gram of sample is reliable (38). However, the data from the baseline study indicate that the presence of AC and E. coli are not correlated with the presence of pathogens, so these indicators may not be good predictors of the presence of foodborne pathogens such as Salmonella. Typically, E. coli occurs within the gastrointestinal tract of animals at low levels; consequently, if present on meat products, the levels would be low and difficult to detect (38). In this study, AC and E. coli were present in all the raw pork product groups.

The results indicate that pork-associated Salmonella isolates had characteristics in common with Salmonella isolated from ill humans. Specifically, serotypes, such as Salmonella Typhimurium associated with foodborne outbreaks caused by a variety of foods, were among those isolated from raw pork samples collected in this study. In addition, all Salmonella isolates recovered during the baseline study belong to serotypes documented as significant or sporadic cause of human illness. Serotypes Anatum, Infantis, and I 4,[5],12:i:−, which together accounted for 32.8% of all Salmonella isolates in the study, are in the top 20 most frequently reported Salmonella serotypes causing illness in the United States (4).

Of the 545 Salmonella strains recovered, 32% carried antibiotic resistance markers. Importantly, all the I 4,[5],12:i:− isolates recovered carried antibiotic resistance genes, with 82% of them carrying multiple antibiotic-resistant genes. Most of the antibiotic resistance genes carried by I 4,[5],12:i:− were antibiotics used in the treatment of human infections. This finding is important because I 4,[5],12:i:− ranks fifth on the list of most frequently reported serotypes linked to human illness (1, 4, 23, 27). Treatment of persons at risk of invasive Salmonella infections require antibiotics, such as trimethoprim-sulfamethoxazole (17), and when these are not effective, a second-line antibiotic may include ceftriaxone (14). In addition, antibiotic resistance genes on mobile genetic elements can transmit to susceptible bacterial species. The FSIS, in collaboration with the FDA and Centers for Disease Control and Prevention, tracks antibiotic resistance in food animal isolates recovered from the FSIS regulatory products as part of the National Antimicrobial Resistance Monitoring System. Ongoing work to study the genetic composition of pathogens recovered from food products and the genetic relatedness of these isolates to those recovered from foodborne outbreaks will increase understanding of how these pathogens persist in food animals and how they survive intervention procedures during slaughter and processing.

Detectable levels of STEC in raw pork suggests that pork consumption provides a potential avenue for human infection. However, the low frequency at which the top seven STEC was recovered, combined with limited epidemiologic evidence that pork-associated infections are a common occurrence, supports the view that this pathogen may not be a major food safety concern in this commodity. Studies by other investigators suggests E. coli that carry stx genes and associated with human illness are infrequently harbored by swine (11, 16, 39). In this study, several of the raw pork samples screened positive for stx and eae genes but did not confirm positive for any of the top seven STEC currently analyzed by the FSIS. It is possible that a screen positive for stx and eae indicated the presence of 1 or more of the 50 other serotypes of E. coli that carry the stx and eae genes. These STEC are known causes of human disease (3), although infections occur less often than the top seven serotypes that the FSIS classifies as adulterants; the top seven serotypes cause the majority (∼90%) of all STEC illnesses and outbreaks. The E. coli O157:H7 outbreak that occurred in Canada in 2018, which was linked to comminuted and intact cuts of pork (10), underscores the need for the continued monitoring of public health trends because Canada is a major exporter of pork products and of feeder pigs into the United States (12). Additional research is needed to understand the significance of STEC serotypes in raw pork products (37).

CONCLUSIONS

Findings from the Raw Pork Baseline Study show that Salmonella in raw pork merits policy consideration. The high rates of Salmonella found in raw pork products indicate that in contrast to what was observed in the past with hog carcasses, performance standards focused on comminuted pork might provide a meaningful measure of process control. Furthermore, the serotypes and antibiotic resistance profiles seen in pork-associated Salmonella is relevant to public health. With respect to STEC, additional research is needed to understand the full range of STEC serotypes present in raw pork products and pork's contribution to human STEC infections (36).

ACKNOWLEDGMENTS

The authors recognize the many Food Safety and Inspection Service individuals responsible for making a baseline study, such as the Raw Pork Baseline Study successful. By working together, daunting tasks become achievable. Special recognition goes to the following: Office of Field Operations and Inspection Program Personnel and those at headquarters; Office of Public Health Science and Field Service Laboratories and headquarters staff; Office of Policy and Program Development; the Office of Planning, Analysis, and Risk Management; and others who helped steer the study in the right direction.

SUPPLEMENTAL MATERIAL

Supplemental material associated with this article can be found online at: https://doi.org/10.4315/0362-028X.JFP-19-360.s1

REFERENCES

REFERENCES
1.
Arruda,
B. L.,
Burrough
E. R.,
and
Schwartz
K. J.
2019
.
Salmonella enterica I 4,[5],12:i:− associated with lesions typical of swine enteric salmonellosis
.
Emerg. Infect. Dis
.
25
:
1377
1379
.
2.
Bellido-Carreras,
N.,
Arguello
H.,
Zaldivar-Lopez
S.,
Jimenez-Marin
A.,
R.,
.
Martins
P.,
Arce
C.,
Morera
L.,
Carvajal
A.,
and
Garrido
J. J.
2019
.
Salmonella Typhimurium infection along the porcine gastrointestinal tract and associated lymphoid tissues
.
Vet. Pathol
.
56
:
681
690
.
3.
Centers for Disease Control and Prevention (CDC)
.
2012
.
National Shiga toxin–producing Escherichia coli (STEC) surveillance
.
4.
Centers for Disease Control and Prevention (CDC)
.
2016
.
National Salmonella surveillance annual report, 2016
.
5.
Centers for Disease Control and Prevention
.
2016
.
Whole genome sequencing using laboratory standard operating procedure for PulseNet Nextera XT library preparation and setup for Illumina MiSeq
.
Available at: https://www.cdc.gov/pulsenet/pdf/PNL32-MiSeq-Nextera-XT.pdf. Accessed 8 January 2020.
6.
Centers for Disease Control and Prevention (CDC)
.
2018
.
Estimates of foodborne illnesses in the United States
.
7.
Centers for Disease Control and Prevention and the U.S. Food and Drug Administration
.
2015.
Antibiotic resistance isolate bank
.
AR gene glossary
.
Available at: https://wwwn.cdc.gov/ARIsolateBank/GeneGlossary. Accessed 8 January 2020.
8.
Dean-Nystrom,
E. A.,
Melton-Celsa
A. R.,
Pohlenz
J. F. L.,
Moon
H. W.,
and
O'Brien
A. D.
2003
.
Comparative pathogenicity of Escherichia coli O157 and intimin-negative non-O157 Shiga toxin-producing E. coli strains in neonatal pigs
.
Infect. Immun
.
71
:
6526
6533
9.
Fabrega,
A.,
and
Vila
J.
2013
.
Salmonella enterica serovar Typhimurium skills to succeed in the host: virulence and regulation
.
Clin. Microbiol. Rev
.
26
:
308
341
.
10.
Food Safety News
.
2018
.
Deadly E. coli outbreak traced to meat shop; pork recalled
.
11.
Fratamico,
P. M.,
Bagi
L. K.,
Bush
E. J.,
and
Solow
B. T.
2004
.
Prevalence and characterization of Shiga toxin-producing Escherichia coli in swine feces recovered in the National Animal Health Monitoring System's Swine 2000 study
.
Appl. Environ. Microbiol
.
70
:
7173
7178
.
12.
Giamalva,
J.
2014
.
Pork and swine: industry & trade summary. Publication ITS-011
.
U.S. International Trade Commission
,
Washington, DC
.
13.
Honish,
L.,
Punja
N.,
Nunn
S.,
Nelson
D.,
Hislop
N.,
Gosselin
G.,
Stashko
N.,
and
Dittrich
D.
2017
.
Escherichia coli O157:H7 infections associated with contaminated pork products—Alberta, Canada, July–October 2014
.
Morb. Mortal. Wkly. Rep
.
65
:
1477
1481
.
14.
Iwamoto,
M.,
Reynolds
J.,
Karp
B. E.,
Tate
H.,
Fedorka-Cray
P. J.,
Plumblee
J. R.,
Hoekstra
R. M.,
Whichard
J. M.,
and
Mahon
B. E.
2017
.
Ceftriaxone-resistant nontyphoidal Salmonella form humans, retail meats, and food animals in the United States, 1996–2013
.
Foodborne Pathog. Dis
.
14
:
74
83
.
15.
Jackson,
B. R.,
Griffin
P. M.,
Cole
D.,
Walsh
K. A.,
and
Chai
S. J.
2013
.
Outbreak-associated Salmonella enterica serotypes and food commodities, United States, 1998–2008
.
Emerg. Infect. Dis
.
19
:
1239
1244
.
16.
Kaufmann,
M.,
Zweifel
C.,
Blanco
M.,
Blanco
J. E.,
Blanco
J.,
Beutin
L.,
and
Stephan
R.
2006
.
Escherichia coli O157 and non-O157 Shiga toxin–producing Escherichia coli in fecal samples of finished pigs at slaughter in Switzerland
.
J. Food Prot
.
69
:
260
266
.
17.
Lombardo,
M. N.,
G-Dayanandan
N.,
Wright
D. L.,
and
Anderson
A. C.
2016
.
Crystal structures of trimethoprim-resistant DfrA1 rationalize potent inhibition by propargyl-linked antifolates
.
ACS Infect. Dis
.
12
:
149
156
.
18.
Montgomery,
D. C.
2005
.
Design and analysis of experiments, 6th ed.
John Wiley & Sons
,
New York
.
19.
National Antimicrobial Resistance Monitoring System
.
Manual of laboratory methods
.
20.
National Institutes of Health, U.S. Library of Medicine, National Center for Biotechnology Information
.
2019.
Pathogen browser to identify genetic markers
.
Available at: https://www.ncbi.nlm.nih.gov/pathogens/isolates/. Accessed 8 January 2020.
21.
Oliveira,
M.,
Vieir-Pinto
M.,
Martins da Costa
P.,
Vilela
C. L.,
Martins
C.,
and
Bernardo
F.
2012
.
Occurrence of Salmonella spp. in samples from pigs slaughtered for consumption: a comparison between ISO 6579:2002 and 23S rRNA fluorescent in situ hybridization method
.
Food Res. Int
.
45
:
984
988
.
22.
Patterson,
S. K.,
Kim
H. B.,
Borewicz
K.,
and
Isaacson
R. E.
2017
.
Towards an understanding of Salmonella enterica serovar Typhimurium persistence in swine
.
Anim. Health Res. Rev
.
17
:
159
168
.
23.
Petrovska,
L.,
Mather
A. E.,
AbuOun
M.,
Branchu
P.,
Harris
S. R.,
Connor
T.,
Hopkins
K. L.,
Underwood
A.,
Lettini
A. A.,
Page
A.,
Bagnall
M.,
Wain
J.,
Parkhill
J.,
Dougan
G.,
Davies
R.,
and
Kingsley
R. A.
2016
.
Microevolution of monophasic Salmonella Typhimurium during epidemic, United Kingdom, 2005–2010
.
Emerg. Infect. Dis
.
22
:
617
624
.
24.
Scallan,
E.,
Hoekstra
R. M.,
Angulo
F. J.,
Tauxe
R. V.,
Widdowson
M.,
Roy
S. L.,
Jones
J. L.,
and
Griffin
P. M.
2011
.
Foodborne illness acquired in the United States—major pathogens
.
Emerg. Infect. Dis
.
17
:
7
15
.
25.
Self,
J. L.,
Luna-Gierke
R. E.,
Fothergill
A.,
Holt
K. G.,
and
Vieira
A. R.
2017
.
Outbreaks attributed to pork in the United States, 1998–2015
.
Epidemiol. Infect
.
145
:
2980
2990
.
26.
Sexton,
T. Y.,
Geornaras
I.,
Belk
K.,
Bunning
M.,
and
Martin
J. N.
2018
.
Salmonella contamination in broiler synovial fluid: are we missing a potential reservoir?
J. Food Prot
.
81
:
1425
1431
.
27.
Soyer,
Y.,
Moreno Switt
A.,
Davis
M. A.,
Maurer
J.,
McDonough
P. L.,
Schoonmaker-Bopp
D. J.,
Dumas
N. B.,
Root
T.,
Warnick
L. D.,
Grohn
Y. T.,
and
Wiedmann
M.
2009
.
Salmonella enterica serotype 4,5,12:i:−, an emerging Salmonella serotype that represents multiple distinct clones
.
J. Clin. Microbiol
.
47
:
3546
3556
.
28.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
1996
.
Pathogen reduction, hazard analysis and critical control point (HACCP) systems. 9 CFR parts 304, 308, 310, 320, 327, 381, 416, & 417. Final rule
.
Fed. Regist
.
61
:
38805
38989
.
Available at: https://www.gpo.gov/fdsys/pkg/FR-1996-07-25/pdf/96-17837.pdf. Accessed 8 January 2020.
29.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2011
.
The nationwide microbiological baseline data collection program: Market hogs survey 2010–2011
.
30.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Microbiology laboratory guidebook
.
31.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Method for Salmonella detection
.
MLG 4.09. In Microbiology laboratory guidebook.
32.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Method for detection of non-O157 STEC from meat products and carcass and environmental sponges
.
MLG 5B.05. In Microbiology laboratory guidebook.
33.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Method for detection of E. coli O157 from meat products and carcass and environmental sponges
.
MLG 5.09. In Microbiology laboratory guidebook.
34.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Method for sanitary indicator estimates
.
MLG 3.02. In Microbiology laboratory guidebook.
35.
U.S. Department of Agriculture, Food Safety and Inspection Service
.
2015
.
Raw pork products exploratory sampling project phase I. Collection of data on presence of Salmonella and other pathogens and indicator organisms in various pork products and used to inform design of nationwide baseline study, phase II. May 2015 to November 2015.
36.
U.S. Department of Agriculture, Food Safety and Inspection Service. 2015. The nationwide microbiological baseline data collection program: beef-veal carcass survey. August
2014
–December 2015
.
37.
U.S. Department of Agriculture, Food Safety and Inspection Service. 2018. Food safety research priorities. Food Safety Research Studies
.
38.
Williams,
M. S.,
Ebel
E. D.,
and
Allender
H. D.
2015
.
Industry-level changes in microbial contamination on market hog and broiler chicken carcasses between two locations in the slaughter process
.
Food Control
51
:
361
370
.
39.
Zweifel,
C.,
Schumacher,
S.,
Beutin,
L.,
Blanco,
J.,
Stephan,
R.
2006
.
Virulence profiles of Shiga toxin 2e-producing Escherichia coli isolated from healthy pig at slaughter
.
Vet. Microbiol
.
117
:
328
332
.
This is an open access article

Supplementary data