295

Journalof Food Protection, Vol. 67, No. 2, 2004, Pages 295– 302

Prevalence of Escherichia coli O157:H7, Listeria monocytogenes, and Salmonella inTwo Geographically Distant Commercial Beef Processing Plantsin the UnitedStates †

MILDRED RIVERA-BETANCOURT, 1* STEVEN D.SHACKELFORD, 1 TERRANCE M.ARTHUR, 1 KURT E.WESTMORELAND, 2 GINA BELLINGER, 3 MICHELLE ROSSMAN, 4 JAMES O.REAGAN, 4 AND MOHAMMAD KOOHMARAIE 1

1RomanL. HruskaU.S. Meat Animal Research Center,Agricultural Research Service, U.S.Department of Agriculture, Clay Center, Nebraska 68933-0166; 2SillikerLaboratories, 2100 North Highway 360, Suite 2006, Grand Prairie, Texas; 3FoodSafety Net Services, LTD,221 West Rhapsody,San Antonio, Texas 78216;and 4NationalCattlemen’ s Beef Association,9110 East Nichols Avenue, Centennial, Colorado 80112, USA

MS03-293:Received 30June 2003/ Accepted 18September 2003

ABSTRACT

Fortwo large beef processing plants, one located in thesouthern United States (plant A) andone located in the northern UnitedStates (plant B), prevalenceof Escherichiacoli O157:H7, Listeria spp., Listeriamonocytogenes, and Salmonella was determinedfor hide, carcass, and facility environmental samples over the course of 5 months.The prevalence of E. coli O157: H7(68.1 versus 55.9%) and Salmonella (91.8versus 50.3%) was higher ( P , 0.05),and the prevalence of Listeria spp. (37.7 versus75.5%) and L.monocytogenes (0.8versus 18.7%) was lower ( P , 0.05)for the hides of cattle slaughtered at plant A versusplant B. Similarly,the prevalence of Salmonella (52.0versus 25.3%) was higher ( P , 0.05)and the prevalence of Listeria spp.(12.0 versus 40.0%) and L.monocytogenes (1.3versus 14.7%) was lower ( P , 0.05)for the fence panels of the holdingpens of plant A versusplant B. Theprevalence of E. coli O157:H7(3.1 versus 10.9%), Listeria spp.(4.5 versus 14.6%), and L.monocytogenes (0.0versus 1.1%) was lower ( P , 0.05)for preevisceration carcasses sampled at plantA versus plant B. Salmonella (bothplants), Listeria spp.(plant B), and L.monocytogenes (plantB) weredetected on fabrication  oor conveyorbelts (product contact surfaces) late during the production day. For plant B, 21of 148 (14.2%) late-operational fabrication oorconveyor belt samples were L.monocytogenes positive.For plant B, E. coli O157:H7 and L.monocytogenes weredetected in preoperational fabrication  oorconveyor belt samples. Overall results suggest that there are regional differ- encesin the prevalence of pathogens on the hides of cattlepresented for harvest at commercial beef processing plants. While hidedata may re ect the regional prevalence, the carcass data is indicative of differencesin harvest practices and procedures inthese plants.

Inthe United States, foodborne pathogens have been carcasses (23). Thesepathogens have been associated with estimatedto cause 7 millionillnesses and up to 9,000 thehide, the intestinal tract of healthy animals, and the deathsannually, with a resultingeconomic loss of six bil- environment (7,11, 15, 31, 32, 36). Recentstudies have liondollars (14). Also,bacterial pathogens account for 60% shownhigh prevalence of E. coli O157:H7 and Salmonella ofhospitalizations attributable to foodborne transmission on hides (3,4, 5,11,22), whichserve as apotentialsource and67% of estimatedfood-related deaths. Mead et al. (28) ofindirectcarcass contamination during hide removal. Gen- havereported that 90% of theestimated food-related deaths erally,the muscle surfaces of the carcasses are sterile and involvethe pathogens Salmonella (31%), Listeriamono- contaminationoccurs as a resultof microbial transfer dur- cytogenes (28%), Toxoplasma (21%),Norwalk-like viruses inghide removal and dressing defects during the slaugh- (7%), Campylobacter (5%), and Escherichiacoli O157:H7 teringprocess. (3%). Sourcesof in-plant carcass microbial contamination Twomajor foodborne bacterial pathogens, E. coli duringslaughter include those associated with the process- O157:H7 and Salmonella, havean animal reservoir and ingpractices, slaughter plant facilities, and plant personnel havebeen associated with the contamination of meat and (7, 15, 36). Theslaughter plants in the United States have meatproducts (7, 9, 38). Recently, L.monocytogenes has developedHazard Analysis Critical Control Points plans to alsobeen identiŽ ed as a seriousfoodborne pathogen (38) decreasethe risk of foodborne illnesses by intervening at andhas been demonstrated to be a contaminantof beef stagesof processing that pose a plausiblerisk of carcass contamination (40). Theseplans require adequate microbi- *Authorfor correspondence. Tel: 402-762-4226; Fax: 402-762-4149; ologicaldata to beableto assessthe effectiveness of control E-mail: [email protected]. programsfor foodbornepathogens. †Brandnames are necessary toreport factually on available data; how- Carcasscontamination is a functionof incoming bac- ever,the U.S. Department ofAgriculture neither guarantees nor warrants thestandards of anyproduct mentioned, and the use ofthename bythe terialload on the cattle hides and in-plant harvest practices U.S.Department ofAgriculture implies no approval of the product to andprocedures. Therefore, the present study was conducted theexclusion of others that may alsobe suitable. todetermine differences in the prevalence of E. coli O157: 296 RIVERA-BETANCOURT ETAL. J.FoodProt., Vol. 67, No. 2

TABLE 1. Environmentalsampling sites for each processing plant (plant A andplant B) Numberof samples Area sampled (cm 2) Site perplant per month (plantA/ plantB)

Beforeoperation a Slaughter oor Floordrains 10 413–1,006/ 162–366 b PCSc 10 206.5/318–488 Brisketsaws 4 205.5/83 Split saws 4 205.5/194 Fabrication oor PCS-conveyorbelts 20 1,463/660–828.5 Anytimeduring operation d Locker room Floordrains 6 642/32–180 Knobson external doors 6 826/195–320 Knobson toilet stalls 6 58/125–137 Knobson lockers 6 148/32 Knobson soap dispensers 6 77/41 Holdingpen fences Panels 15 774/549 Latein operation d Slaughter oor Trolleys 15 535.5/19 Floordrains 15 413–1,006/ 162–366 PCS 15 200/244–488 Brisketsaws 6 206.5/83 Split saws 6 206.5/194 Fabrication oor PCS-conveyorbelts 30 1,463/600–828.5 a Sampleswere taken on the Ž rstand third week of the month. b Areasampled was a functionof the size and shape of the item to be sampled. Thus, the area for a givensample type varied within andbetween plants. c Productcontact surface. d Sampleswere taken on the Ž rst,second, and third week of the month.

H7, L.monocytogenes, and Salmonella betweenincoming eration),35 samples each of hides, preevisceration carcasses (be- bacterialload on the cattle’ s hides(index of regional dif- foreany application of antimicrobial treatments), and postinter- ferences)and carcasses (index of the plant’ s practicesand vention(after full complement of antimicrobial intervention) car- procedures)in two geographically distant commercial beef casseswere taken at each plant. Thus, a totalof 105 animals per processingplants in the United States. monthwere sampled and the sampling of hidesand carcasses was performedas previously described (5, 11), withthe modiŽ cation MATERIALS AND METHODS thatall hides and carcass samples were collected using a HydraSponge(cat. no. HS-10BPW/ 2G,International BioProducts, Samplingprocedure. Sampleswere collected for a period Bothell,W ash.)premoistened with 10 mlof bufferedpeptone wa- of5months(April, May, July, August, and October) in three trips ter. permonth to each of the two large steer/ heifercommercial beef Environmentalsamples were taken by swabbingthe site area processingplants. One plant was located in thesouthern (plant A) (Table1) using a HydraSponge(cat. no. HS-10NB/ 2G,Interna- UnitedStates while the other was located in the northern (plant tionalBioProducts) premoistened with 10 mlofneutralizingbuff- B)UnitedStates. For both of these plants, slaughter chain speed er.Environmental,hide, and carcass samples were transported on rangedfrom 250 to 300 animals per hour .Thecarcass decontam- ice packs (1) tothe corresponding laboratories and processed inationtreatments used by both plants is a combinationof avail- within20 to 24 h. ableinterventions including but not limited to (i) steam vacu- uming,(ii) knife trimming, (iii) preevisceration carcass washing, Preenrichmentof the samples. Inthe laboratory, once the and(iv) Ž nalpostevisceration carcass rinses. sampleswere received, 90 ml of Trypticasesoy broth (TSB; Dif- Dependingon which environmental site was to be sampled, co,Becton Dickinson, Sparks, Md.) was added to each sponge sampleswere taken either before operation, anytime during op- sample,stomached for 1 min,and incubated at 25 8C for 2 to 3 h. eration,or late in operation (T able1). At thesame time that the Theresulting preenrichment culture was divided into three por- environmentalsampling was being conducted (anytime during op- tionsfor three enrichment procedures, which were handled as fol- J.FoodProt., Vol. 67, No. 2 PREVALENCEOF PATHOGENSIN PROCESSINGPLANTS 297 lows.For the enrichment of E. coli O157:H7,10 ml ofthepreen- sampleswere conŽ rmed using culture, biochemical, and serolog- richedsample was transferred into a sterile15-ml conical tube, icalmethods as previously described (10). Isolatesgiving typical incubatedfor 6 hat42 8C,andheld at 4 8C overnight (6). These Salmonella reactionswere tested serologically for the presence of sampleswere then shipped on ice packs to the Roman L. Hruska Poly(O) andPoly (H) antigensusing the Salmonella O Antiserum U.S. MeatAnimal Research Center via overnight mail (1) for the PolyA-I and Vi (Difco2264-47) and H AntiserumPoly a-z (Difco isolationand characterization of E. coli O157:H7.For the enrich- 2406-47)following the manufacturer’ s instructions. ment of Listeria spp.,4 mlof the preenriched sample was added Statisticalanalysis. Foreach type of sample (hide, carcass, to36 ml of Fraser broth (Difco) and incubated at 30 8C for 18 to andenvironmental), prevalence of each pathogen was estimated 24h. Finally, the remaining 76 ml of the preenriched sponge forboth processing plants by dividingthe number of positive sam- samplewas incubated at 35 8Cfor22 to 24 h forthe screening plesby the total number of samples tested. The exact binomial andrecovery of Salmonella. 95%conŽ dence intervals (CI) were calculated for each prevalence pointestimate using PEPI ( (13);‹http://www.sagebrushpress.com/ Recoveryand conŽrmation of E. coli O157:H7. Upon ar- pepibook.html›). Inorder to test for pathogen- and sample type- rivalof the samples at the Roman L. HruskaU.S. MeatAnimal speciŽc prevalencedifferences between plants, DIFFER procedure ResearchCenter, E. coli O157:H7was isolated from the enriched ofPEPI wasused to calculate a continuity-adjustedchi-square for samplesusing immuno-magnetic separation as previously de- thedifference between plants. scribed (6). Characteristiccolonies on either ctSMAC ornt- Forhide data, the interaction of processingplant with month Rainbowplates were tested (up to 6 coloniesper sample) using ofsampling was also tested by multiple continuity-adjuste dchi- the Dryspot E. coli O157latex test (Oxoid, Inc., Ogdensburg, squaretests. T oavoidType I errorrates due to multiple compar- N.Y.).Latex-positive colonies were streaked onto ctSMAC plate isons,the pair-wise P valueswere adjusted using Hommel’ s mod- forpurity and single colonies were stored in nutrient agar stabs iŽcation of the Bonferroni procedure (18) usingPEPI software. (33) forfurther testing. Subsequently, broth cultures of each iso- latewere stored at 2708Casglycerol stocks. Suspect colonies RESULTS AND DISCUSSION wereconŽ rmed using an indirect ELISA (O157and H7 antigens) (11) andmultiplex polymerase chain reaction to detect the pres- Thedata collected and reported in this study is unique inthat a numberof key pathogens were simultaneously ence of stx1, stx2, eaeA,  iCH7, and rfbEO157 genefragments (19). Isolateswere further veriŽ ed as E. coli speciesusing the Sensititre isolatedfrom thesame sample. This is in contrastwith tak- Gram-negativeAutoIdentiŽ cation (AP80) system (Accumed In- ingindividual samples for eachpathogen. The nonrandom ternational,Westlake, Ohio). Samples were considered positive if distributionof pathogenson sites examined in theseexper- atleast one isolate recovered in the sample met the following imentsand in general requires simultaneous isolation (30). requirements:either expressed the O157 antigen or carried the Apotentialdrawback to this approach could be that the rfbEO157 gene,expressed the H7antigen or carriedthe  iCH7 gene, prevalenceof pathogens at sampling sites with low cell carriedat leastone stx gene,and was determined to be E. coli by densitymay be underestimated. However ,theadvantages theAP80 system. (eliminationof nonrandomdistribution as a sourceof error) Enrichment,recovery, and conŽrmation of Listeria spp. inour assessment far outweighthe potential disadvantage. Afterincubation of thesamples in Fraser broth, a 1-mlaliquot of Prevalenceof pathogenic bacteria on hides, car- eachsample was transferred into 9 mlof buffered Listeria en- richmentbroth (BLEB, Difco)containing 8.5 g of3-(N-morphol- casses,and holding pen environment. Thesample period ino)propanesulfonic acid (MOPS), and13.7 g ofMOPS sodium thatwas chosento carry out this study was basedon pre- saltin 1 literpuriŽ ed water.These secondary enrichment (MOPS- viousdata on fecal shedding and prevalence on hide of E. BLEB) sampleswere incubated at 30 6 28C for 20 6 2 h. There- coli O157:H7,which has been shown to peak during sum- after,1 mlwas removed from each sample and the detection of mer andearly fall (5, 11, 24). Overallprevalence of the Listeria spp.was performed according to the AOAC OfŽcial threepathogens tested on hides was statisticallydifferent Method996.14 (2) usingthe Assurance EIA Listeria enzyme im- (P , 0.001)between plants (T able2). The overall preva- munoassaytest kit (BioControl System, Inc., Belleview, W ash.). lenceof Shiga toxin-producing E. coli O157:H7 and Sal- Theremaining enrichment broth was refrigerated and used for monella onhides was signiŽcantly higher for plantA, conŽrmation of presumptive positives, as described elsewhere whereasthe prevalence of Listeria spp. and L.monocyto- (10). Typicalhemolytic colonies of Listeria werestreaked to a genes onhides was higherfor plantB. Thedata suggest purityplate (T rypticasesoy agar– yeast extract or bilayer horse bloodplates) to obtain a pureisolate. If no hemolysis was ob- thathide is a principalsource of E. coli O157:H7, Listeria served,the typical nonhemolytic Listeria colonieswere also spp., L.monocytogenes, and Salmonella. Tothebest of our streakedto a purityplate. The pure culture was then conŽ rmed knowledge,the prevalence of Listeria spp. and L. mono- for Listeria spp. and/or L. monocytogenes usingthe biochemical cytogenes oncattle hides has not been determined in the identiŽcation method (AOAC ofŽcial method 992.19) (2), the UnitedStates. A limiteddocumentation on the prevalence semisolidmotility agar for motility (41), or the API-Listeria ofthese pathogens in cattle feces has been done in Europe (bioMe´rieux-Vitek,Hazelwood, Miss.) rapid test strip. (12, 34) and Brazil (17). Theprevalence of Listeria spp. and L.monocytogenes Enrichment,recovery, and conŽrmation of Salmonella. onfencepanels in the holding pens was signiŽcant between Afterthe 22 to 24 h incubationat 35 8Cofthe samples in TSB, selectiveenrichment and detection of motile and nonmotile Sal- plants,with plant B showinghigher prevalence (40.0 and monella wasperformed following the AOAC OfŽcial Method 14.7%,respectively), while the prevalence of Salmonella 992.11protocol (2). TheM-broth cultures as wellas the selective onfence panels was signiŽcantly higher ( P , 0.01) for mediacultures were retained for conŽ rmation of presumptivepos- plantA (Table2). There were nosigniŽ cant differences itiveson the enzyme immunoassay test kit. Presumptive-positiv e betweenplants in the prevalence of E. coli O157:H7 for 298 RIVERA-BETANCOURT ETAL. J.FoodProt., Vol. 67, No. 2

TABLE 2. Prevalence(%) ofpathogens on the holding pen fences and on animals’ hides and carcasses in plant A andplant B Salmonella Listeria spp. L.monocytogenes E. coli O157:H7

Site/plant n % 95% CI n % 95% CI n % 95% CI n % 95% CI

Hides Plant A 510 91.8a 89.0–94.0 512 37.7a 33.5–42.1 510 0.8a 0.2–7.0 511 68.1a 63.9–72.1 Plant B 523 50.3 45.9–54.7 523 75.5 71.6–79.2 523 18.7 15.5–22.4 522 55.9 51.6–60.2 Preeviscerationcarcasses Plant A 511 23.3 19.7–27.2 511 4.5a 2.9–6.7 511 0.0b 0.0–0.7 510 3.1a 1.8–5.0 Plant B 522 26.8 23.1–30.9 522 14.6 11.7–17.9 522 1.1 0.4–2.5 523 10.9 8.4–13.9 Postinterventioncarcasses Plant A 499 0.0 0.0–0.7 499 0.0 0.0–0.7 499 0.0 0.0–0.7 497 0.0c 0.0–0.7 Plant B 520 0.8 0.2–2.0 520 0.2 0.0–0.1 520 0.0 0.0–0.7 520 1.0 0.3–2.2 Fencepanels in the holding pens Plant A 75 52.0d 40.2–63.7 75 12.0a 5.6–21.6 75 1.3d 0.0–7.2 75 13.3 6.6–23.2 Plant B 75 25.3 16.0–36.7 75 40.0 28.9–52.0 75 14.7 7.6–24.7 75 5.3 1.5–13.1 a Withinan organism and sample type, prevalence differed between plants ( P , 0.001). b Withinan organism and sample type, prevalence differed between plants ( P , 0.05). c Withinan organism and sample type, prevalence differed between plants ( P , 0.10). d Withinan organism and sample type, prevalence differed between plants ( P , 0.01). thissite. The incidence of these pathogens on fence panels April,July, August, and October .Theseasonal onset of was lowerthan on hides with the exception of L. mono- high E. coli O157:H7prevalence occurred sooner for plant cytogenes inplantA, whichis in agreement with a previous A.Theprevalence of Salmonella was higherfor plantA study (35). Furthermore,the lower incidence on the fence duringall sampling months. However ,themagnitude of the panelsin our study may be a functionof the difference in differencebetween plants was lowestduring August. The samplingarea between hides ( ;1,700 cm2)andthe fence prevalenceof L.monocytogenes was #2.0%during each panels(774 or 549 cm 2;Table1). The data in this study samplingmonth for plantA, butfor plantB, theprevalence showsthat all three pathogens tested were recoveredfrom of L.monocytogenes rangedfrom 3.8to 48.6% during the thefence panels in the holding pens (T able2). Others have 5samplingmonths. The cattle presented for slaughterusu- reportedsimilar observations in the lairage environment at allycome from approximately150 miles radius of aplant. abattoirsin Europe (3, 12, 35). Becauseplant A andplant B arelocated far from each Therewas aninteraction( P , 0.05)between plant and other,theprevalence of pathogens on theanimal’ s hiderep- samplingmonths in the presence of all three pathogens on resentan index of regionaldifferences in the prevalence of hides(T able3). Whereas prevalence of E. coli O157:H7 on thesemicroorganisms. hideswas higherfor plantA duringMay, July, and August, Despitethe higher incidence of E. coli O157:H7 and theplants did not differ ( P , 0.05)during the months of Salmonella onhidesof cattleharvested at plant A, theprev-

TABLE 3. Interactionof plant (plant A versusplant B) andsampling month on the prevalence (%) ofpathogens on hides

Prevalence bymonth

Organism/plant April May July August October

E. coli O157:H7a Plant A 46.5 C 79.0 AB 72.5 AB 81.8 A 60.6 BC Plant B 40.4 C 44.8 C 71.4 AB 82.7 A 40.4 C Listeria spp.a Plant A 39.6 CD 42.9 CD 32.7 D 13.9 E 58.7 C Plant B 82.9 B 93.3 AB 46.7 CD 57.7C 97.1 A L.monocytogenes a Plant A 2.0 C 1.0 C 0.0 C 0.0 C 1.0 C Plant B 16.2 B 48.6 A 9.5 BC 3.8 BC 15.4 B Salmonellaa Plant A 98.0 AB 87.6 BC 95.0 AB 99.0 A 79.8 C Plant B 26.7 E 49.5 D 45.7 DE 77.9 C 51.9 D a Withinan organism, means with different letters are signiŽ cantly ( P , 0.05)different. J.FoodProt., Vol. 67, No. 2 PREVALENCEOF PATHOGENSIN PROCESSINGPLANTS 299

TABLE 4. Differencesbetween plant A andplant B forprevalence (%) ofpathogens in slaughter  oorenvironmental samples Microorganism

Site/planta Salmonella Listeria spp. L.monocytogenes E. coli O157:H7

Floordrains, before operation Plant A 2/50 (4.0) 0/50 (0.0) 0/50 (0.0) 0/50 (0.0) Plant B 2/50 (4.0) 0/50 (0.0) 0/50 (0.0) 0/50 (0.0) Floordrains, late during operation Plant A 17/74(23.0) 2/74 (2.7) 0/74 (0.0) 0/75 (0.0) Plant B 10/75(13.3) 4/75 (5.3) 1/75 (1.3) 0/75 (0.0) Productcontact surfaces, before operation Plant A 0/49 (0.0) 0/49 (0.0) 0/49 (0.0) 0/49 (0.0) Plant B 0/50 (0.0) 0/50 (0.0) 0/50 (0.0) 0/50 (0.0) Productcontact surfaces, late during operation Plant A 0/74 (0.0) 0/74 (0.0) 0/74 (0.0) 1/74 (1.4) Plant B 1/75 (1.3) 1/75 (1.3) 0/75 (0.0) 0/75 (0.0) Brisketsaw, before operation Plant A 0/20 (0.0) 0/20 (0.0) 0/20 (0.0) 1/20 (5.0) Plant B 0/11 (0.0) 1/11 (9.1) 1/11 (9.1) 0/11 (0.0) Brisketsaw, during break/ lunch Plant A 0/30 (0.0) 0/30 (0.0) 0/30 (0.0) 0/30 (0.0) Plant B 0/17 (0.0) 0/17 (0.0) 0/17 (0.0) 0/16 (0.0) Splittingsaw, before operation Plant A 0/20 (0.0) 0/20 (0.0) 0/20 (0.0) 0/20 (0.0) Plant B 0/20 (0.0) 0/20 (0.0) 0/20 (0.0) 0/20 (0.0) Splittingsaw, during break/ lunch Plant A 0/27 (0.0) 0/27 (0.0) 0/27 (0.0) 0/27 (0.0) Plant B 0/29 (0.0) 0/29 (0.0) 0/29 (0.0) 0/29 (0.0) Trolleys,late during operation Plant A 0/73 (0.0) 1/73 (1.4) 0/73 (0.0) 0/73 (0.0) Plant B 0/73 (0.0) 2/73 (2.7) 0/73 (0.0) 0/73 (0.0) a Plantsdid not differ for any comparison ( P . 0.10). alenceof thesetwo pathogens on preeviscerationand post- whichpathogens are transferred onto carcasses (15, 36). interventioncarcasses was signiŽcantly lower (T able2). Escherichiacoli O157:H7, Salmonella,Listeria spp., and Similartrends were observedfor plantB. Moreimportant, L.monocytogenes havebeen found to be contaminants of for bothplants, the antimicrobial interventions used during carcassesafter dehiding and other slaughtering/ dressing carcassprocessing substantially reduced the prevalence of processes (4,5, 11, 12, 20, 21, 23, 26, 27, 37, 39, 42). E. coli O157:H7, Salmonella,Listeria spp., and L. mono- Anoverall (both plants, see T able6) evaluationof the cytogenes onpostinterventio ncarcasses(T able2). Hide re- bacterialproŽ le ofthesamples collected from hides,preev- movaland evisceration are considered to be procedures in iscerationcarcasses, and from panelsin the holding pens

TABLE 5. Differencesbetween plant A andplant B forprevalence (%) ofpathogens in fabrication  oorenvironmental samples Microorganism

Site/plant Salmonella Listeria spp. L.monocytogenes E. coli O157:H7

Productcontact surfaces, before operation Plant A 0/100 (0.0) 0/100 (0.0)a 0/100 (0.0)a 0/100 (0.0) Plant B 0/99 (0.0) 6/99 (6.1) 6/99 (6.1) 2/98 (2.0) Productcontact surfaces, late during operation Plant A 4/150 (2.7) 0/150 (0.0)b 0/150 (0.0)b 0/150 (0.0) Plant B 3/148 (2.0) 25/148(16.9) 21/148(14.2) 0/148 (0.0) a P , 0.05. b P , 0.001. 300 RIVERA-BETANCOURT ETAL. J.FoodProt., Vol. 67, No. 2

TABLE 6. Numberof hide, carcass, and plant environmental samples containing one or more bacterial species a Numberof pathogens

Site 0b 1 2 3

Beforeoperation Slaughter oor Drains 96 4 0 0 Productcontact surfaces 99 0 0 0 Brisketsaws 28 2 0 0 Splittingsaws 40 0 0 0 Fabrication oor Productcontact surfaces 190 8 0 0 Anytimeduring operation Slaughter oor Hides 38 274 479 241 Carcasses(pre-evisceration) 657 324 46 5 Carcasscooler Carcass(postevisceration) 1,007 10 0 0 Locker room Drains 47 9 1 0 Knobson external doors 59 1 0 0 Knobson lockers 58 0 0 0 Knobson soap dispensers 58 0 0 0 Pens Panels 68 57 21 4 Latein operation Slaughter oor Trolleys 143 3 0 0 Drains 117 31 1 0 Productcontact surfaces 146 3 0 0 Brisketsaws 46 0 0 0 Split saws 56 0 0 0 Fabrication oor Productcontact surfaces 268 28 2 0 a E. coli O157:H7, Salmonella,Listeria spp., and L.monocytogenes. b Negativeby method. revealedthat a proportionof the samples contained more thebrisket saw (beforeoperation). There were nosigniŽ - thanone bacterial species. cantdifferences ( P . 0.10)between plants for allthe path- ogenstested. Overall, the data indicates that sources of Prevalenceof pathogenicbacteria on environmental pathogencontamination were identiŽed on the processing samples. Atotalof 812 and 797 samples were obtained from the11 environmental sites (not including pens) sam- equipmentand surfaces (brisket saw andPCS) inaddition pledat plant A andplant B, respectively,representing the tothe  oordrains at different time points. Moreover ,the slaughter oor,fabrication oor,andlocker rooms at dif- pathogencontamination observed on the slaughter  oorof ferenttimes during operation (T ables4 and5). A totalof eachplant emulates the prevalence of these pathogens on 417and 400 samples (plant A andplant B, respectively) hidesand preevisceration carcasses. from theslaughter  oorwere testedfor allthree pathogens Onthe fabrication  oor,prevalenceof pathogenic bac- (Table4). Only two positive samples of E. coli O157:H7 teriawas onlytested on conveyor belts (T able5). E. coli were obtainedfrom plantA andwere identiŽed on the O157:H7 and L.monocytogenes were recoveredon thissite productcontact surfaces (PCS), lateduring operation, and for plantB beforeoperation. Interestingly, late during op- onthe brisket saw (beforeoperation). As for Salmonella, eration,the incidence of L.monocytogenes on conveyor the oordrains were positivefor bothplants (before and beltswas foundin 14.2% of the samples tested. Also, Sal- lateduring operation); plant B alsotested positive for the monella was foundon the conveyor belts late during op- PCS(lateduring operation). L.monocytogenes was also erationin both plants, which suggests that there was a foundin plant B oordrains (late during operation) and on sourceof contaminationduring processing. Once more, this J.FoodProt., Vol. 67, No. 2 PREVALENCEOF PATHOGENSIN PROCESSINGPLANTS 301 differencein the prevalence of thesepathogens on the con- KoleEwoldt, and Lauren Sammel fromthe University of Wisconsin, Mad- veyorbelts in each plant is consistent with the preeviscer- ison,for their assistance insampling. ationcarcass data. REFERENCES Anevaluation of atotalof 145and 150 samples, from plantA andplant B, respectively,for thepresence of path- 1.Anonymous. 1998. FSIS Directive 10,230.5,Section 9. Sample ship- ogenson differentsites within the locker room environment ment,p. 9-1 to 9-3. U.S. Department ofAgriculture, Food Safety revealsthat Salmonella (10%), L.monocytogenes (3.3%), andInspection Service, Washington, D.C. 2.Anonymous. 2000. Compendium of microbiological methods for the and E. coli O157:H7(3.3%) were presenton the  oor analysisof food and agricultural products. AOAC International,Gai- drainsfrom plantB, whereasin plant A, only Salmonella thersburg,Md. (22.2%)was presenton this site. Furthermore, the locker 3.Avery, S. M.,A.Small,C.-A. Reid, and S. Buncic.2002. Pulsed- rooms’door knobs in plant A testedpositive for E. coli Želdgel electrophoresis characterization ofShiga toxin-producing O157:H7and the toilet stalls’ door knobs in plant B tested Escherichiacoli O157from hides of cattle at slaughter. J.FoodProt. 65:1172–1176. positivefor thispathogen. There were nosigniŽ cant dif- 4.Bacon, R. T.,J.N.Sofos,K. EBelk,D. R.Hyatt,and G. C.Smith. ferences (P . 0.10)between plants for allthe pathogens 2002.Prevalence andantibiotic susceptibility of Salmonella isolated tested.An assessment of the pathogenic microbial proŽ le frombeef animal hidesand carcasses. J.FoodProt. 65:284–290. inthe overall environmental samples from bothplants 5.Barkocy-Gallagher, G. A.,T .M.Arthur,M.Rivera-Betancourt,X. showedthat pathogens recovered from 5.5%of thesamples Nou,S. D.Shackelford,T .L.Wheeler,and M. Koohmaraie.2003. Seasonalprevalence ofShiga toxin-producing Escherichiacoli, in- containedonly one bacterial species whereas four samples cludingO157:H7 and non-O157 serotypes, and Salmonella in com- containedtwo species (T able6). The data suggests that con- mercial beef processingplants. J. FoodProt. 66:1978–1986. taminationcan occur in theenvironment of theplant as well 6.Barkocy-Gallagher, G. A.,E. D.Berry,M. Rivera-Betancourt,T .M. asby the personnel anytime during processing. Arthur,X.Nou,and M. Koohmaraie.2002. Development of methods Few studieshave addressed the prevalence of multiple forthe recovery of Escherichiacoli O157:H7 and Salmonella from beef carcass spongesamples andbovine fecal andhide samples. J. pathogenicbacteria species in commercial beef processing Food Prot. 65:1527–1534. plants;therefore, reported data regarding the prevalence of 7.Bell, C. 2002.Approach to the control of entero-haemorrhagic Esch- E. coli O157:H7, Salmonella, and Listeria specieswithin erichiacoli (EHEC). Int.J. FoodMicrobiol. 78:197–216. theenvironment of beef slaughtering plants in the United 8.Byrne, C. M.,D.J.Bolton,J. J.Sheridan,D. A.McDowell,and I. Statesat different time points of beef processing is scarce S.Blair.2000. The effects ofpreslaughterwashing on the reduction ornonexistent. Studies conducted in beef abattoirs in Eu- of Escherichiacoli O157:H7transfer fromcattle hidesto carcasses duringslaughter . Lett.Appl. Microbiol. 30:142–145. ropehave revealed that the points of pathogen contamina- 9.Department ofHealth andHuman Services, Centers forDisease tioninclude the processing equipment (8,12, 16, 20, 25, Controland Prevention. 2001. Salmonella:Annualsummary, 2001. 29), plantoperation personnel (8, 16, 20), andthe plant’ s Availableat: http://www.cdc.gov/ncidod/dbmd/phlisdata/salmonella. environment (16,20, 25). TheseŽ ndings,as well as those htm.Accessed 27December 2002. ofthepresent study, suggest that bacterial contaminants can 10.Dey, B. P.(ed.).1998. Microbiology laboratory guidebook, 3rd ed. U.S.Department ofAgriculture, Food Safety, and Inspection Ser- occuron mechanical equipment and PCS andthat human vice,Government Printing OfŽ ce, Pittsburgh,Pa. factorscan contribute to contamination of beef carcasses. 11.Elder ,R.O.,J.E.Keen, G.R.Siragusa,G. A.Barkocy-Gallagher, Inconclusion, this research suggests that hides and M.Koohmaraie,and W .W.Laegreid.2000. Correlation of entero- fencepanels in the holding pens are sources of contami- hemorrhagic Escherichiacoli O157prevalence infeces, hides,and nationin the preslaughter environment and can be used to carcasses ofbeef cattle duringprocessing. Proc.Natl. Acad. Sci. determinethe contamination pattern of the carcasses. For USA 97:2999–3003. 12.Fenlon, D. R.,J. Wilson,and W .Donachie.1996. The incidence and instance,in plant A, Salmonella and E. coli O157:H7were level of Listeriamonocytogenes contaminationof food sources at predominantin this environment whereas Listeria spp. and primaryproduction and initial processing. J.Appl.Bacteriol. 81: L.monocytogenes were predominantin plant B, andsimilar 641–650. patternsof the incidence of the speciŽ c pathogenswere 13.Fleiss, J. L.1981.Statistical methods for rates andproportions, 2nd observedon carcasses (T able2) and in-plant environment ed.John Wiley & Sons,Inc., New York. 14.Food and Drug Administration, U.S. Department ofAgriculture,U.S. (Tables4 and5). Furthermore, seasonality is a factorthat EnvironmentalProtection Agency, and Centers forDisease Control canbe used as predictors of these pathogens because the andPrevention. 1997. Food safety fromfarm totable: a national prevalenceof E. coli O157:H7 and Salmonella on hides foodsafety initiativereport to the President. Available at: http:// was sustainedfor alongerperiod of timein plant A (May, www.cfsan.fda.gov/ ;dms/fsreport.html.Accessed 27December July,August, and October; T able3). 2002. 15.Galland, J. C. 1997.Risks and prevention of contaminationof beef ACKNOWLEDGMENTS carcasses duringthe slaughter process inthe United States ofAmer- ica. Rev. Sci.Tech. Off. Int. Epiz. 16:395–404. Thisproject was fundedin part by beef andveal producersand 16.Guyon, R., F.Dorey,J. P.Malas, andA. Leclercq. 2001.Hazard importersthrough their $1-per-head checkoff and was producedfor the analysis of Escherichiacoli O157:H7contamination during beef Cattlemen’s Beef Boardand state councilsby the National Cattlemen’ s slaughteringin Calvados, France. J.FoodProt. 64:1341–1345. Beef Association.We thankPatty Beska, JulieDyer ,TroyGramke, Sue 17.Hofer ,N.1983.Bacteriologic and epidemiologic studies on the oc- Hauver,Bruce Jasch, Dee Kucera, Kim Kucera, FrankReno, Audrey currence of Listeriamonocytogenes inhealthycattle. Zbl.Bakt. Hyg. Smidt,and Pat Tammen fortheir technical assistance andMarilyn Bierman A 256:175–183. andCarol Grummert forsecretarial assistance. Also,we thankMindy 18.Hommel, G.1988.A stagewise rejective multipletest procedure Brashears, JasonMann, and Mark Miller from Texas Tech University, basedon a modiŽed Bonferroni test. Biometrika 75:383–386. Lubbock;and Dennis Buege, Steve Ingham, Jill Losinski, Cara Cross, 19.Hu, Y .,Q.Zhang,and J. C.Meitzler.1999. Rapid and sensitive 302 RIVERA-BETANCOURT ETAL. J.FoodProt., Vol. 67, No. 2

detectionof Escherichiacoli O157:H7in bovine faeces bymultiplex 32.Sheridan, J. J.1998.Source of contamination during slaughter and PCR. J.Appl.Microbiol. 87:867–876. measures forcontrol. J. Food Saf. 18:321–339. 20.Johnson, J. L.,M. P.Doyle,and R. G.Cassens. 1990. Listeriamon- 33.Silhavy, T .J.,M. L.Berman, andL. W.Enquist(ed.). 1984. Exper- ocytogenes and other Listeria spp.in meat andmeat products:a iments withgene fusions. Cold Spring Harbor Laboratory, Cold review. J. FoodProt. 53:81–91. SpringHarbor ,N.Y. 21.Karr ,K.J.,E. A.E.Boyle,C. L.Kastner, J.L.Marsden,R. K. 34.Skovgaard, N., and C.-A. Morgen. 1988. Detection of Listeria spp. Phebus,R. K.Prasal, W.P.Pruett,Jr .,andC. M.Garcia-Zepeda. infaeces fromanimals, infeeds, andin raw foodsof animal origin. 1996.Standardized microbiological sampling and testing procedures Int.J. FoodMicrobiol. 6:229–242. forthe beef industry. J.FoodProt. 59:778–780. 35.Small, A., C.-A. Reid, S. M.Avery,N. Karabasil, C.Crowley,and 22.Keen, J.E.,andR. O.Elder.2002. Isolation of Shiga-toxigenic Esch- S.Buncic.2002. Potential spread of Escherichiacoli, Salmonella, erichiacoli O157from hide surfaces andthe oral cavity of Ž nished and Campylobacter inlairage environmentat abattoirs. J.FoodProt. beef feedlotcattle. J.Am.Vet. Med. Assoc. 220:756–763. 65:931–936. 23.Korsak, N., G.Daube,Y .GhaŽr, A. Chahed,S. Jolly,and H. Vin- 36.Sofos, J. N.1994.Microbial growth and its control in meat, poultry, devogel.1998. An efŽ cient sampling technique used to detect four andŽ sh,p. 359 –403. In A.M.Pearsonand T .R.Dutson(ed.), foodborne pathogens on pork and beef carcasses innine Belgian Advancesin meat research, vol.9. Qualityattributes and their mea- abattoirs. J.FoodProt. 61:535–541. surement inmeat, poultry,and Ž shproducts. Chapman and Hall, 24.Laegreid, W .W.,R.O.Elder,andJ. E.Keen. 1999.Prevalence of Glasgow,United Kingdom. Escherichiacoli O157:H7in range beef calves at weaning. Epide- 37.Sofos, J. N.,S. L.Kochevar,G.R.Bellinger,D. R.Buege,D. D. miol.Infect. 123:291–298. Hancock,S. C.Ingham,J. B.Morgan,J. O.Reagan,and C. Smith. 25.Lou, Y .,andA. E.Yousef.1999. Characteristics of Listeriamono- 1999.Sources and extent of microbiologicalcontamination in seven cytogenes importantto food processors, p. 131– 224. In E. T. Ryser UnitedStates slaughteringplants. J.FoodProt. 62:140–145. andE. H.Marth(ed.), Listeria, listeriosisand food safety. Marcel 38.Tauxe, R. V.1997.Emerging food borne diseases: anevolving public Dekker,Inc.,New York. healthchallenge. Emerg.Inf. Dis. 3:425–434. 26.Madden, R. H.,W .E.Espie,L. Moran,J. McBride,and P .Scates. 39.U.S. Department ofAgriculture, Food Safety, and Inspection Ser- 2001.Occurrence of Escherichiacoli O157:H7, Listeriamonocyto- vice.1994. Nationwide beef microbiologicalbaseline data collection genes,Salmonella and Campylobacter spp.on beef carcasses in program:steers andheifers, October 1992– September 1993. Wash- NorthernIreland. Meat Sci. 58:343–346. ington,D.C. Available: FSIS, OfŽ ce ofPublic Health andScience, 27.McNamara, A.M.1995.Establishment of baseline data onthe mi- http://www.fsis.usda.gov/OPHS/baseline/contents.htm. crobiotaof meats. J. Food Saf. 15:113–119. 40.U.S. Department ofAgriculture,Food Safety and Inspection Service. 28.Mead, P .S.,L.Slutsker,V.Dietz, L.F.McCaig,J. S.Bresee, C. 1996.Pathogen reduction; hazard analysisand critical controlpoint Shapiro,P .M.GrifŽn, and R. V.Tauxe.1999. Food-related illness (HACCP) systems; Žnalrule. Fed.Regist. 61(144):38806–38989. anddeath in the United States. Emerg.Inf. Dis. 5:607–625. 41.U.S. Food and Drug Administration. 2001. Center forFood Safety 29.Nesbakken, T .,G.Kapperrud,and D. A.Caugant.1996. Pathways andApplied Nutrition: bacteriological analytical manual. Online, 8th of Listeriamonocytogenes contaminationin the meat industry. Int. ed.Available at: http://www.cfsan.fda.gov/ ;ebam/bam-toc.htm.Ac- J.FoodMicrobiol. 31:161–171. cessed 30November 2001. 30.Nou, X. 2003.Personal communication. 42.V anderlinde,P .B.,S.Shay,and J. Murray.1998. Microbiological 31.Rebhun, W .C.1987.Listeriosis. Vet.Clin. North Am. Food Anim. qualityof Australianbeef carcass meat andfrozen bulk packed beef. Pract. 3:75–83. J.FoodProt. 61:437–443.