Escherichia coli and Enterobacteriaceae counts on pig and ruminant carcasses along the

EFSA supporting publication 2014:EN-634

Any enquiries related to this output should be addressed to [email protected]

Suggested citation: Barco L, Belluco S, Roccato A and Ricci A, 2014. Escherichia coli and Enterobacteriaceae counts on pig

and ruminant carcasses along the slaughterline, factors influencing the counts and relationship between visual faecal

contamination of carcasses and counts: a review. EFSA supporting publication 2014:EN-634, 111 pp.

Available online: www.efsa.europa.eu/publications

© European Food Safety Authority, 2014

EXTERNAL SCIENTIFIC REPORT

Escherichia coli and Enterobacteriaceae counts on pig and ruminant

carcasses along the slaughterline, factors influencing the counts and

relationship between visual faecal contamination of carcasses and counts:

a review1

Lisa Barco, Simone Belluco, Anna Roccato and Antonia Ricci2

Istituto Zooprofilattico Sperimentale delle Venezie, Legnaro, Padova, Italy

ABSTRACT

A literature review was conducted covering the period 2000-2012 to gather information concerning the presence

and counts of E. coli and Enterobacteriaceae on carcasses of main livestock species during different stages of

the slaughterline (review question 1); risk factors that could explain the variability of the counts of the indicator

organisms (review question 2) and the relationship between the counts of indicator organisms and visual faecal

contamination on carcasses (review question 3). In total, 86 papers considering the main livestock species (cattle,

pigs, sheep and goats) with the exception of poultry, and providing pertinent data for the scopes of the search,

were retrieved. In relation to review question 1, the steps of the processing line where a decrease of indicator

bacteria was more evident were: sequential decontamination treatments such as pasteurization and hot water

washing applied before chilling for cattle; scalding and also according to some authors, pasteurization and

chilling for pigs, plus chilling and pasteurization for small ruminants. Concerning review question 2, most of the

retrieved studies investigated risk factors related to slaughtering process. Hot water washing and steam

pasteurization were clearly effective in reducing bacterial load on beef carcasses. Hot water treatments were

effective also for pig carcasses. The dressing technique and pasteurization treatment were described as factors

able to control bacterial contamination of small ruminant carcasses. In relation to review question 3, only studies

providing data about ruminants were available and the reported results confirmed that the presence of visible

faecal contamination led to higher bacterial loads on carcasses of dirty animals than those obtained from clean

animals and the application of additional hygienic measures can be effective in order to reduce bacterial load of

contaminated carcasses at the end of the processing line.

© Istituto Zooprofilattico Sperimentale delle Venezie, 2014

KEY WORDS

Escherichia coli, Enterobacteriaceae, counts, ruminant carcasses, pig carcasses, slaughterhouse, process hygiene

criteria

DISCLAIMER

The present document has been produced and adopted by the bodies identified above as author(s). This task has been carried

out exclusively by the author(s) in the context of a contract between the European Food Safety Authority and the author(s),

awarded following a tender procedure. The present document is published complying with the transparency principle to

which the Authority is subject. It may not be considered as an output adopted by the Authority. The European Food Safety

Authority reserves its rights, view and position as regards the issues addressed and the conclusions reached in the present

document, without prejudice to the rights of the authors.

1 Question No EFSA-Q-2014-00531. 2 Acknowledgement: The authors would like to thank EFSA staff: Winy Messens and Pablo Romero Barrios for the support

provided to this external scientific report submitted to EFSA.

E.

EFSA supporting publication 2014:EThe present document has been produced andby the author(s) in the context of a contractprocedure. The present document is publisheconsidered as an output adopted by the Authissues addressed and the conclusions reached

SUMMARY

A project entitled “Usefulness ofin poultry” was awarded by EFPadova, Italy) with the purposEnterobacteriaceae as Process Hliterature search and an experimeThe present document is the rEnterobacteriaceae counts on ppoultry carcasses and the experimexternal scientific reports (Barco

The extensive literature reviewindicator bacteria, E. coli and Ecarcasses during different stageexplain the variability of the counbetween the counts of indicatoquestion 3).

A worldwide literature search,databases (PubMed and Web of scholar was also carried out.

The principles of “systematic redefinition of the review questionthe studies for inclusion or exsynthesising data collected fromconclusions.

A total of 86 papers satisfied thprocess and were used to collecanimal species were concerned,about swine carcasses and 21 pap

A high level of variability amonof the slaughterlines, was evidenthe area of carcass sampled, the streatments applied along the slauto conflicting conclusions amoninvestigating the same risk factor

Among the indicator bacteria usthe hygiene of the entire meat interchangeable indicators useEnterobacteriaceae and E. coligenerally used to assess enteric which can be easily detected andenvironment or coming from the

In the context of the review, thlead to the same conclusions.

coli and Enterobacteriaceae counts on pig and rumin

EN-634 d adopted by the bodies identified above as author(s). This task hast between the European Food Safety Authority and the author(s),ed complying with the transparency principle to which the Autho

hority. The European Food Safety Authority reserves its rights, viewd in the present document, without prejudice to the rights of the auth

f Escherichia coli and Enterobacteriaceae as ProcFSA to Istituto Zooprofilattico Sperimentale dellse to collect available data on the indicator oHygiene Indicators (PHI) for the main livestockental study, in this case in broiler slaughterhousereport on the extensive literature review on E

pig and ruminant carcasses. The extensive literamental study in broiler slaughterhouses are publi et al., 2014; Cibin et al., 2014).

w was conducted to gather information concernEnterobacteriaceae and their counts on beef, smaes in the slaughterline (review question 1); risnts of the indicator organisms (review question 2

or organisms and visual faecal contamination o

covering the period 2000-2012, was conducScience) were consulted; in addition web-search

eview methodology” were applied and includedns and the eligibility criteria, searching for reseaxclusion in the review, collecting data from tm included studies, presenting data, interpreting

he eligible criteria considered at the different stact data for the three review questions. As far a, 41 papers provided pertinent data about beef pers about small ruminants.

ng the different studies, due to different aspects anced. Some variables, like the sampling and analspecific step of the slaughterline investigated andughterline, render the available data barely compang studies describing counts at the same stage ofr.

sed, aerobic plate counts are frequently used as production process, whereas Enterobacteriacea

ed to specifically address the level of fa, the two indicator bacteria investigated in thecontamination in foodstuffs. These are classified

d used as markers of pathogenic zoonotic agents panimals.

e studies considering both E. coli and EnterobaHowever, in case that different results were o

nant carcasses: a review

s been carried out exclusively , awarded following a tender rity is subject. It may not be w and position as regards the

hors.

2

cess Hygiene Criteria le Venezie (Legnaro, organisms E. coli or k species, based on a es, located in the EU. Escherichia coli and ature review covering ished as two separate

ning the presence of all ruminants and pig sk factors that could 2) and the relationship on carcasses (review

cted. Two electronic hing through Google-

d the following steps: arch studies, selecting the included studies,

results and drawing

ages of the screening as the different meat carcasses, 31 papers

and to the complexity lytical methods used,

d the decontamination arable and could lead f the slaughterline or

indicators to monitor ae or E. coli are two aecal contamination. e present review, are d as faecal indicators, present in processing

acteriaceae generally obtained for the two

E.


indicators, it should be pointed explaining the final outcome.

REVIEW QUESTION 1. Presetheir counts on carcasses durin

Cattle

Ten eligible papers provided perto identify the steps of the beef scounts on the carcasses was the dstage. Data have been generalldecontamination treatments havechange of bacterial loads was dapplied to the carcasses.

One study described the decreeviscerated, whereas another stuacid, no effect in terms of E. colbefore evisceration was effectivean increase of microbial load incorrelated to these slaughter phconsidered as an effective strategstudies demonstrated that washinThe application of different sequor washing with acids in all rebacteria. Finally, at the chilling chilling had higher E. coli and Edrop of the E. coli counts was rEnterobacteriaceae counts. Likethe counts of carcasses at the pre

Pigs

Fourteen papers dealing with coustages in the slaughterhouse wheslaughterline that lead to a decrchallenging because data have bdifferent operations have been usbacterial counts strongly relies on

Regarding both E. coli and Enteduring scalding (nine studies), wreduction in bacterial counts. Thlevel. The efficiency of highPasteurization was identified acommercial abattoirs are neededpoint in particular regarding Enstudies (described in five differejustifying its use as decontaminamong the selected studies focuEnterobacteriaceae during chilli




out that also additional variables could equally

ence of the indicator organisms E. coli and Entng different stages in the slaughterline

rtinent data concerning this issue on beef carcasslaughterline that lead to a decrease or an increasedifficulty of finding studies that provide data befly collected at distant sampling points, and ine been used. Hence, it was hard to identify if tdue to a specific phase of the slaughterline or

ease of E. coli counts after carcasses were wudy showed that when this was combined with li counts reduction was observed. Decontaminatioe in reducing bacterial load (1 study). Eviscerationn one study, whereas in five other studies, changhases were not observed. Washing treatment afgy to reduce the E. coli load of carcasses in one ng at this step had no effect on E. coli and Enterouential decontamination treatments, such as hot etrieved studies (5) led to reductions in numbestep, conflicting data were collected. In two stu

Enterobacteriaceae counts than in the previous phreported and for another study, this step did no

ely, the possibility of reducing contamination at cvious steps.

unts of E. coli and/or Enterobacteriaceae on pig ere retrieved. As with cattle, the identification orease or an increase of indicator bacteria countbeen generally collected at distant sampling poisually performed. Hence, the identification of stagn authors’ conclusions.

erobacteriaceae, a decrease in microbial contamiwhich is generally recognized as an important ophe planning and managing of this operation, henh temperature-based stages was demonstrateas effective in three retrieved studies; howevd to confirm this. Evisceration was confirmed as nterobacteriaceae. The washing process was inent papers) and never led to significant decreasation treatment. Regarding chilling, there was n

using on the effects of chilling on E. coli; in coning was observed in three out of four studies. Ho



hors.

3

y have had a role in

terobacteriaceae and

ses. A main challenge e of indicator bacteria fore and after a single n between, different

the effect in terms of a specific treatment

washed before being spraying with lactic

on treatments applied n and trimming led to ges of bacterial loads fter evisceration was study, whereas three

obacteriaceae counts. water, pasteurization

ers of both indicator udies, carcasses after hases, in one study, a ot have any effect on chilling step relies on

carcasses at different of the steps in the pig s on the carcasses is ints, and in between, ges that can influence

ination was observed peration to achieve a ce, is critical at plant ed against E. coli. ver more studies in a key contamination

nvestigated in seven se in bacterial counts no general agreement ntrast, a reduction of owever, the ability to

E.


assess chilling efficacy depends the possibility of reducing contaOther operations along the procein different studies with contrasti

Small ruminants

Six papers provided informationcarcasses at different stages in thstages of the slaughtering procesampled area varied consideraconclusions could only be drawn

An increasing level of contaminretrieved papers; in particular, thcounts of E. coli and Enterobacslaughtering process can have an

Along the slaughterline, the chilcan be reduced: two papers concreduce the counts; thus, chillinMoreover, two studies showed bacterial loads.

Finally, concerning the washingstudies, washing had no effect washing reduced the E. coli coun

REVIEW QUESTION 2. Riskindicator organisms

Cattle

According to the defined searchdealing with risk factors influenobtained. For papers providingcomparability of data was also cl

Season emerged as a risk factor wcounts on beef carcasses and in pseason (compared to wet season)

Indicator bacteria counts on caslaughtering process took placedesign of the plant, the throughpsome other cases differences areported without attributing thesindicator bacteria counts was wiin low-throughput plants, but slaughterhouses were significantthe bacterial counts of carcasse




on the study design and the location of samplinamination relies on the counts before the investessing line, such as polishing, scraping and singeiing results for both E. coli and Enterobacteriacea

n on counts of E. coli and/or Enterobacteriaceahe slaughterhouse. In five out of six papers, sam

ess distant to each other; this feature, coupled wably among studies make the comparison ran according to the authors’ observations.

nation along the slaughterline was recorded in he skinning and the evisceration steps contributecteriaceae. Furthermore, it seems clear that the n influence on counts.

lling step is the most effective point where miccluded that this phase seems to be the most effecng should be regarded as a control point alonthat carcass pasteurization was another import

g step, the results were not clear or unanimouin reducing the bacterial counts, while anothe

nts before the chilling step.

k factors that could explain the variability o

h process and the established eligibility criteria,ncing E. coli and Enterobacteriaceae counts on data for review question 2, a level of variablear.

which could have a direct impact on indicator baparticular the lowest levels of contamination were) and coldest months.

arcasses showed different values according to . In some studies, specific aspects related to thput, or the surveillance system in place, were emong plants in terms of bacterial loads of case findings to specific reasons. The effect of pidely investigated. Lower prevalences/counts we

in only one study the differences betweent. Hence, different slaughterhouses could induce

es, but the data collected from the retrieved stu



hors.

4

ng points. Moreover, tigated chilling stage. ing were investigated

ae.

ae on small ruminant mples were at several with the fact that the ather difficult. Thus,

three out of the six ed mostly to the final different steps of the

crobial contamination ctive point in order to ng the slaughterline. tant step in reducing

us; according to two er study showed that

of the counts of the

a total of 29 papers n beef carcasses were bility hampering the

acteria prevalence and e observed during dry

the plant where the his point, such as the evaluated, whereas in arcasses were simply plants’ throughput on ere generally reported n the two types of e different effects on udies did not clarify

E.


whether any particular aspect relcontext.

The effect of using physical and another aspect that was frequenteffective treatment to reduce E. tested was used both in high andindicator bacteria counts on carpasteurization was demonstratedcarcasses, but in this last case worsening of the organoleptic fea

The effect of washing with potabdifferent studies, and produced cthis treatment mainly depends on

The effect of chemical decontamperoxyacetic acid, nisin) was unceffectiveness in reducing bacterithe different chemicals tested, awhere the treatments were applie

The two studies addressing theextremely different treatments apossible to produce a definitivecarcasses.

Finally, only one study investigaof feed and water treatment or ancarcass swabs without evidencing

Pigs

Numerous risk factors were invetechnically feasible since few stuenvironments were very often noat farm level on carcass contampelleted five times a day followethan the other regimes examinedhandling nor batch size were idehad limited impact on carcass mi

Plant throughput and features obcontamination was not found to plants, Enterobacteriaceae wereweak points. However, some autcriteria, based on daily mean log

Several managerial factors thainvestigated and described in thefrequency of lairage disinfectionspraying live animals when exte




lated to different slaughterhouses could produce

chemical decontamination treatments on indicatotly investigated. Steam pasteurization was frequecoli and Enterobacteriaceae loads on beef carcad low-throughput plants, and all studies reportedrcasses due to such decontamination treatment. d as an effective way to improve microbiologthe improvement of microbiological quality w

atures.

ble water at environmental temperature was also tonflicting results. Some authors demonstrated tha

n the bacterial loads of carcasses to be treated.

mination treatments (e.g. washing-spraying with clear, since different studies described opposite rial loads on beef carcasses. These conflicting reas well as the procedures followed and the stepsed.

e effect of the chilling on bacterial load of cand therefore, their results cannot be comparede answer about the effect of the chilling on t

ated batch related risk factors, and in particular iny possible interactions on numbers of E. coli reg any correlation.

estigated in the eighteen selected papers, but comudies considered the same factors, and also in thisot comparable. Only one paper considered the infmination and observed that feeding/fasting regimed by a 24 h fast), resulted in lower E. coli countsd. Regarding animal management before slaughentified as risk factor. This suggested that the stricrobial quality.

bviously have an influence on bacterial counts, bube related to the throughput of the slaughterhous

e considered useful to provide indication of abattthors underline the ineffectiveness of EC-relatedE. coli value for carcasses.

at could influence microbial conditions of ce retrieved papers. The use of water during lairagn seemed to be protective against high E. coliernal temperature was considered by operators a



hors.

5

a main effect in this

or bacteria counts was ently described as an asses. The equipment d a clear reduction of

Similarly, hot water gical quality of beef

was associated with a

taken into account by at the effectiveness of

lactic acid, chlorine, results related to their sults could be due to s of the slaughterline

arcasses investigated d. Hence, it was not the bacterial load of

it evaluated the effect ecovered from hide or

mparisons were barely s case, the operational fluence of risk factors me (feeding the pigs s on the thoracic area

htering, neither rough ress condition applied

ut the level of carcass se. In low-throughput toir specific hygienic

d EU process hygiene

carcasses have been e cleaning and a high i counts. In contrast, as hot was correlated

E.


with an increase in carcass cproportionally with the length offactors in relation to E. coli immersion, the disinfection of thbefore chilling. As regards the eacidified sodium chlorite waterapproved according to EU legidecontamination intervention to r

Focusing on both E. coli and Entemperature used, in one study. implementation of good manufacwas recognised as effective. Bofound to be effective in decreasinselected ones. Pasteurization procounts, justifying the possibilityagreement with the results of retreatment for reducing Enterobdomestic steam cleaning systemprocessing plants.

Small ruminants

Among the retrieved papers, 1Enterobacteriaceae counts on smdata was hampered by the variab

Among the factors investigated, studies investigating this factor during the warm season comparaccount by three studies as a poplants recorded lower prevalenslaughterhouses.

The application of treatments alindicator bacteria were investigatpasteurisation of carcasses aftesignificant reduction of the prevAnother step that can have an experimental chilling treatments bacterial load; however, a loss of

Finally, the dressing technique hIn fact, inverted carcass dressingremoval, was considered by sevcontamination.




contamination. Moreover, microbial load on f processing time between killing and scalding. Icontamination were a scalding procedure usin

he splitting machine three times a day and changieffect of decontamination treatments, both hot wr solution (SANOVA) were effective, but sinislation, only the use of hot water could currereduce E. coli levels on slaughtered carcasses in t

terobacteriaceae, washing was considered effectHowever, this result was not confirmed in ano

cturing practices (GMP) during anal plugging at tth the trimming of contaminated sites and the cng E. coli counts on contaminated carcass sites asoduced significant decreases in both E. coli andy of using such a treatment in order to reduceview question 1. Other authors, after testing tbacteriaceae loads, suggested the possibility ms as a control measure in low and very lo

16 provided data on the risk factors that canmall ruminant carcasses. Also for this question, ility of the studies.

the season did not have a significant effect on thconcluded that there was no difference between

red to the cold season. The throughput of the abossible factor that could have an impact on cou

nce and counts of indicator bacteria compared

ong the slaughterline and the effect of the slaugted by eight papers. An effective treatment is repr

er dressing; several authors concluded that thisvalence and counts of E. coli and Enterobactereffect in reducing microbial load is the chillinwas effective in the rapid reduction of the carcas

f meat quality was frequently recorded.

had an effect on indicator bacteria counts on smalg, which minimizes the contact between hands anveral authors as the technique to be adopted in o



hors.

6

carcasses increased In contrast, protective ng steam instead of ing the carcass hooks

water and solution an nce the latter is not ently be an efficient the EU.

tive, regardless of the ther paper where the the evisceration stage cooling process were s well as on randomly d Enterobacteriaceae ce contamination, in the efficacy of steam of using household

ow throughput meat

n affect E. coli and the comparability of

he counts: in fact, two n counts on carcasses battoir was taken into unts: low throughput

d to high throughput

ghtering technique on resented by hot water s treatment led to a riaceae on carcasses. ng phase. The use of s temperature and the

ll ruminant carcasses. nd carcass during pelt order to limit carcass

E.


REVIEW QUESTION 3. The and visual faecal contamination

Cattle

Five papers provided pertinent carcasses and their E. coli and En

Clean cattle produced carcasses dirty animals. However, the ideneffective measures either on anican lead to a comparable contabacterial contamination level. Hslaughter visual evaluation of corrective measures for the initbacterial load at the end of the sla

Pigs

The literature research did not visual faecal contamination on ppossible to provide any informati

Small ruminants

Three papers investigated the reland/or Enterobacteriaceae count

1) the distinction between cleanimprove the hygiene of small rum

2) additional hygiene measures send of the day; reducing line spethe inverted dressing procedure;

3) modifications of the pelt remoof the slaughterman or the fleece




potential relationship between the counts of in on carcasses

information on the relationship between faecnterobacteriaceae counts.

with better microbiological quality than those ntification of the visibly contaminated animals aimals before entering the slaughterhouse and aloamination level to clean animals or in some caHence, the retrieved studies support the concthe level of animal contamination and the atially dirty carcasses can be an effective appraughterline.

provide any papers dealing with the possible ig carcasses and counts of indicator bacteria: theion about this topic in this species.

lationship between faecal contamination of carcats. The available data suggested that:

n and dirty carcasses could be an important startminant carcasses;

should be applied for high-risk (dirty) animals (i.eeed; thorough cleaning of operator hands, arms agreater spacing between carcasses);

oval methods reducing the contact between the ce can significantly improve gross visible contamin



hors.

7

indicator organisms

cal contamination of

derived from visibly and the application of ong the slaughterline

ases, even to a lower clusion that the pre-application of proper oach to reduce their

relationship between refore it has not been

asses and their E. coli

ting point in order to

e.: slaughtering at the and aprons; the use of

carcass and the hands nation.

E.


TABLE OF CONTENTS

Abstract ......................................Summary ....................................Table of contents ........................Background as provided by EFSASpecific objectives as provided bIntroduction and Objectives .......1. Materials and Methods ......

1.1. Defining the review q1.2. Searching for research

1.2.1. Search A: electroni1.2.2. Search B: Web-sea

1.3. Selecting studies for in1.3.1. Screening of the tit1.3.2. Examining full-text

1.4. Collecting data from t2. Results ...............................

2.1. Literature search and r2.2. Discussion of relevant2.3. Bovine .......................

2.3.1. General information2.3.2. General information2.3.3. Review question 1 2.3.4. Review question 2 2.3.5. Review question 3

2.4. Pigs ............................2.4.1. General information2.4.2. General information2.4.3. Review question 1 2.4.4. Review question 2

2.5. Small ruminants ........2.5.1. General information2.5.2. General information2.5.3. Review question 1 2.5.4. Review question 2 2.5.5. Review question 3

References ..................................Appendices .................................Appendix A. Relevance ScreAppendix B. Data CollectionAppendix C. General charactAppendix D. Counts at differAppendix E. Risk factors: deAppendix F. Faecal contaminAbbreviations .............................




..............................................................................

..............................................................................

..............................................................................A ...........................................................................

by EFSA ............................................................................................................................................................................................................................

questions and developing the eligibility criteria forh studies.................................................................ic databases ...........................................................arching ...................................................................nclusion or exclusion in the review ......................les and abstracts for the relevance to the study qut for the eligibility of studies ................................the included studies and creating evidence tables ...............................................................................relevance screening ..............................................t studies ...............................................................................................................................................n about the considered papers ..............................n about the slaughtering process ..................................................................................................................................................................................................................................................................................................................................................n about the considered papers ..............................n about the slaughtering process ....................................................................................................................................................................................................................................................................n about the considered papers ..............................n about the slaughtering process ................................................................................................................................................................................................................................................................................................................................................................................................................................

eening ....................................................................n .............................................................................teristics of the selected papers ..............................rent stages along the slaughterline ........................etailed results .........................................................nation ..................................................................................................................................................



hors.

8

................................ 1

................................ 2

................................ 8

................................ 9

................................ 9

.............................. 10

.............................. 13 r studies ................. 13 .............................. 14 .............................. 14 .............................. 15 .............................. 15

uestions .................. 16 .............................. 16 .............................. 17 .............................. 18 .............................. 18 .............................. 18 .............................. 20 .............................. 20 .............................. 20 .............................. 22 .............................. 27 .............................. 33 .............................. 35 .............................. 35 .............................. 36 .............................. 36 .............................. 42 .............................. 48 .............................. 48 .............................. 49 .............................. 49 .............................. 52 .............................. 57 .............................. 63 .............................. 69 .............................. 69 .............................. 72 .............................. 77 .............................. 82 .............................. 93 ............................ 109 ............................ 111

E.


BACKGROUND AS PROVIDED BY

Regulation (EC) No 854/2004 offor the organization of official coAmong others, inspection tasksinformation, ante-mortem inspec

EFSA received a mandate froinspection, requesting a series ofand rank the most relevant meatinspection system, to propose aloutline a generic framework forfor the prioritized hazards that ar

Several species were to be concovered by inspection of swine mwere published in 2011 and 20bovines/cattle (EFSA-Q-2011-00Q-2011-00365) and solipeds (EF

Current post-mortem visual inspas the main concerns for food smortem inspection can be an indi

The BIOHAZ Panel proposed rethe establishment of targets for tfood business operators own hygA potential approach for the latte

SPECIFIC OBJECTIVES AS PROVI

The purpose of the Service Coorganisms E. coli or Enterobactspecies. Based on this literature sEU should be designed and carriultimate aim is to support the puslaughterhouses.

According to the Technical Specthe tasks to be covered are as foll

• To carry out literature presence of the indicatcarcasses during differenvariability of the counts the counts of indicator or

• To perform an experimecollect relevant data on broiler carcasses after chvariability of these coucarcasses with and witho




Y EFSA

f the European Parliament and of the Council layontrols on products of animal origin intended for s within this Regulation include checks and antion and post-mortem inspection.

m the Commission in May 2010 on the mof scientific opinions. The main scope of these opit safety risks, to assess the strengths/weaknesseslternative approaches for addressing current mear inspection, prevention and control (including rre not (sufficiently) covered by the current system

nsidered. The scientific opinions on the public meat (EFSA-Q-2010-00886) and poultry meat (E012. Four more opinions concerning the inspe0365), farmed game (EFSA-Q-2011-00366), smaFSA-Q-2011-00367) were published in 2013.

ection is not able to detect any of the public heasafety. Visual detection of faecal contamination icator of slaughter hygiene.

ecommending that the current visual inspection pthe main biological hazards on the carcass and b

giene management through the use of Process Hyger is measuring E. coli or Enterobacteriaceae on c

IDED BY EFSA

ntract is to provide EFSA with the available dteriaceae as Process Hygiene Indicators (PHI) fosearch, an experimental study in broiler slaughteried out to collect relevant data on these two indiurpose of potential PHC for evaluating process c

cifications of the Service Contract CFT/EFSA/BIOlows:

searches for data related to the main livestocor organisms E. coli and Enterobacteriaceae nt stages in the slaughterline; (ii) information thof the indicator organisms and (iii) the potential rganisms and visual faecal contamination on carc

ental study in broiler slaughterhouses located in tthe variability of the counts of E. coli and En

hilling; (ii) collect information that could lead tounts and (iii) compare E. coli and Enterobacout visual faecal contamination.



hors.

9

ys down specific rules human consumption.

nalysis of food chain

odernization of meat inions was to identify s of the current meat at-safety risks, and to related methodology)

m.

health hazards to be EFSA-Q-2010-01469) ection of meat from all ruminants (EFSA-

alth hazards identified of carcasses at post-

process is replaced by by verification of the giene Criteria (PHC). carcasses.

data on the indicator or the main livestock rhouses located in the icator organisms. The control in EU broiler

OHAZ/2012/03-CT1,

ck species on (i) the and their counts on

hat could explain the relationship between

casses;

the EU in order to (i) nterobacteriaceae on o interpretation of the cteriaceae counts on

E.


The present document is the rEnterobacteriaceae on carcassespig and ruminant carcasses and tseparate external scientific report

This contract was awarded by EF

Contractor: Istituto Zooprofilattic

Contract title: Usefulness of Escpoultry

Contract number: CFT/EFSA/BI

INTRODUCTION AND OBJECTIV

Routine examination of meat salow level, inconsistent distributiomethods (Schaffner and Smith, 2on foods of animal origin, they 2010).

Ideally, an indicator bacterium sh

– rapidly detectable and ea

– strictly associated with tcomparable growth rates

However, controversy still remaican be indicative of the presencrecognition that indicator bacteridocumented (EFSA, 2012).

Several indicators can be useful colony count (ACC) is commonlEnterobacteriaceae and E. coli a2008). Psychrotrophic microorgindicators in products that are stoalteration of these products (Gon

Enterobacteriaceae are definedcatalase-positive and nitrate redufaeces, but also many non-faecalas Gram-negative bacilli fermenseveral bacteria, such as E. coli(Schaffner and Smith, 2004). Wito faecal contamination of foods(Smooth and Pierson, 1997). Ecounts, which are commonly usecontamination (Ghafir et al., 200




report on the literature search for available ds of pigs and ruminants. The extensive literature rthe experimental study in broiler slaughterhousests (Barco et al., 2014; Cibin et al., 2014).

FSA to:

co Sperimentale delle Venezie, Legnaro, Padova,

cherichia coli and Enterobacteriaceae as Process

OHAZ/2012/03

VES

amples for potential pathogens is impractical maon in meat samples and in some cases the need fo2004). Since indicator bacteria are found at mucare frequently used as indicators of pathogen pre

hould meet certain criteria (Jay et al., 2005); in pa

asily differentiable from other microorganisms pre

the pathogen whose presence it should indicate (s).

ins over the degree to which the presence/amountce of pathogens (Schaffner and Smith, 2004). ia are an effective tool in process hygiene assess

to evaluate hygiene levels during meat slaughterly used to evaluate the hygiene of the entire meaare more frequently used to assess enteric contamganisms, such as Pseudomonas, have great impored at low temperatures since they are responsib

nzalez-Miret et al., 2006).

d as Gram-negative, glucose fermenting, oxidaucing organisms. This family includes many bacl organisms (Schaffner and Smith, 2004). Faecal nting lactose within 48 h at 44.5 to 45.5 °C and, Enterobacter spp., Klebsiella pneumoniae and ithin this group, E. coli is the most relevant micro; thereby, it is the most widely used indicator of . coli counts are usually highly correlated withed in slaughterhouses as indicators of faecal as w

08). The proliferation of Enterobacteriaceae on liv



hors.

10

data on E. coli and review covering beef, are published as two

Italy

s Hygiene Criteria in

ainly because of their or laborious analytical ch higher prevalences esence (Matias et al.,

articular, it should be:

esent in the samples;

(e.g. correlate counts,

t of indicator bacteria On the contrary, the sment has been well-

ring process. Aerobic at production process.

mination (Ghafir et al., portance as spoilage ble for the superficial

ase negative, usually cteria associated with coliforms are defined d this group includes

d Citrobacter fruendii oorganism in relation faecal contamination

h Enterobacteriaceae well as environmental vestock carcasses has

E.


been routinely linked to inadeqconditions (Whyte et al., 2003).

In the European Union, Regulamicrobiological criteria for foodEnterobacteriaceae as process hy

These regulations introduced twoCriteria. An EU Food Safety Crmarket; if the criterion is not mProcess Hygiene Criterion (PHCCritical Control Points (HACCPit is applicable at the process levactions are required; if the critersummarizes the EC-related PHC

Table 1: Process Hygiene CriterNo 2073/2005 and Regulation (E

Food category Microor

Carcasses of cattle, sheep, goats and horses

Aerobic col

Enterobac

Salmo

Carcasses of pigs

Aerobic col

Enterobac

Salmo

3 Commission Regulation (EC) No 20

22.12.2005, p. 1-26. 4 Commission Regulation (EC) No

microbiological criteria for foodstuff




quate or unhygienic processing or inappropriate

ation (EC) No 2073/20053 and Regulation (ECd-stuffs has established the monitoring of Aerobygiene criteria for carcasses of cattle, sheep, goats

o different types of criteria: Food Safety Criteria riteria (FSC) defines the acceptability of food pr

met, the product/batch has to be withdrawn fromC) is an indicator of the acceptable functioning of

) system-based manufacturing handling and distrvel. It sets an indicative contamination value ab

rion is not met, the process has to be reviewed anto be applied in red meat slaughterhouses.

ria applied in red meat slaughterhouses as definedEC) No 1441/2007

rganism Limits

lony count

<3.5 log CFU/cm2: acceptable 3.5-5.0 log CFU/cm2: marginal

> 5.0 log CFU/cm2: unacceptable

cteriaceae

<1.5 log CFU/ cm2: acceptable 1.5-2.5 log CFU/ cm2: marginal

> 2.5 log CFU/ cm2 : unacceptable

onella up to 2 out of 50 samples can be positiv

lony count

< 4.0 log CFU/cm2: acceptable 4.0-5.0 log CFU/cm2: marginal

> 5.0 log CFU/ cm2: unacceptable

cteriaceae

< 2.0 log CFU/ cm2: acceptable 2.0-3.0 log CFU/ cm2: marginal

> 3.0 log CFU/ cm2: unacceptable

onella up to 5 out of 50 samples can be positiv

073/2005 of 15 November 2005 on microbiological criteria

1441/2007 of 5 December 2007 amending Regulation ffs. OJ L 322, 7.12.2007, p. 12–29.



hors.

11

e handling or storage

C) No 1441/20074 on bic colony count and s, horses and pigs.

and Process Hygiene roducts placed on the

m the market. An EU f Hazard Analysis and ribution processes, so ove which corrective nd improved. Table 1

d by Regulation (EC)

Stage where the criterion applies

Carcasses after dressing but

before chilling

ve

ve

a for foodstuffs. OJ L 338,

(EC) No 2073/2005 on

E.


In Europe, Regulation (EC) No animal origin, provides that foodwater to remove surface contamibeen approved in accordance wiSafety Authority (EFSA) adoptelactic acid for the removal oftrimmings. It was concluded thatand applied either by spraying ovariable, the microbial reductionunlikely that such treatments wou

This opinion is the scientific basbusiness operators are allowed tdomestic bovine carcasses or compliance with the conditions sinto good hygienic practices and

A literature search considering conducted to obtain data on:

• the presence of the indicarcasses during differen

• factors that could lead todesign of the slaughtetechniques, and any othe

• any potential relationshcontamination on carcass

The present report specifically c(cattle, pigs, small ruminants). Th

5 Commission Regulation (EC) No 8

origin. OJ L 139, 30.4.2004, p. 55–26 Commission Regulation (EC) No

microbiological surface contaminati




853/20045, which lays down specific rules on thd business operators should not use any substanination from products of animal origin, unless useith that Regulation. In this context, on July 201

ed a Scientific Opinion on the evaluation of the sf microbial surface contamination from beef t the use of lactic acid (from 2 to 5 %, at temperor misting) for decontamination is not a safety cns achieved by treatment of beef are generally siguld contribute to the development of microbial re

sis for the Regulation (EC) No. 101/20136, whichto use lactic acid to reduce microbiological surfahalf carcasses or quarters at the level of th

set out in the Annex of the Regulation and whenHACCP-based systems.

all the main livestock species (poultry, pigs

cator organisms E. coli and Enterobacteriaceaent stages in the slaughterline;

o the variability of the counts of these indicator oerhouses, the throughput of the slaughterho

er batch specific information;

hip between the counts of indicator organismses.

considers data related to the main livestock specihe review considered studies performed worldwid

853/2004 of 29 April 2004 laying down specific hygiene 205. o 101/2013 of 4 February 2013 concerning the use oon on bovine carcasses. OJ L 34, 5.2.2013, p. 1-3.



hors.

12

he hygiene of food of ce other than potable e of the substance has 1 the European Food safety and efficacy of

carcasses, cuts and ratures of up to 55 °C concern and although gnificant and that it is esistance.

h lays down that food ace contamination on he slaughterhouse in n the use is integrated

and ruminants) was

e and their counts on

rganisms, such as the ouse, the processing

ms and visual faecal

ies other than poultry de.

rules for food of animal

of lactic acid to reduce

E.


1. MATERIALS AND METHOD

The principles of “systematic literature search. This involved th

- defining the review ques

- searching for research stu

- selecting studies for inclu

- collecting data from the i

- assessing validity and qu

- synthesising data from in

- presenting data and resul

- interpreting results and d

1.1. Defining the review qu

The first step of the literature revorder to identify the key element

Review question 1 is relateEnterobacteriaceae, and their coelements of the question are:

• the population of interest

• the outcome, representeEnterobacteriaceae) on c

• the setting, represented b

Review question 2 is related to tterms of amount of the indicator


• the intervention strategrepresented by structurahandling of animals (ewithdrawal period before

• the scenarios against whbatches slaughtered incharacteristics, animals h





DS

review methodology” (EFSA, 2010) were aphe following steps:

tions and developing the eligibility criteria for stu

udies;

usion or exclusion in the review;

included studies and creating evidence tables;

uality of included studies;

ncluded studies;

lts;

drawing conclusions.

uestions and developing the eligibility criteria f

view process consisted of the analysis of the thres and to clarify their scopes.

ed to the presence of the indicator organounts on carcasses during different stages of the s

t, represented by the main livestock species (rumi

d by the presence and amounts of indicator orgcarcasses;

by the slaughterline.

the identification of the factors which could explbacteria on carcasses. The key elements of the qu


gies/scenarios that could influence the countal/managerial characteristics of the slaughterhoe.g. transport time and lairaging before slauge slaughter);

hich the interventions/strategies/scenarios have n different ways, in slaughterhouses with handled in different ways before being slaughtere

d by the presence and amounts of indicator orgcarcasses;



hors.

13

plied to the present

udies;

for studies

e review questions in

nisms, E. coli and laughterline. The key

inants and pigs);

ganisms (E. coli and

lain the differences in uestion are:

inants and pigs);

ts, which could be ouse or pre-slaughter ghter, diet and feed

been compared (e.g. different processing d);

ganisms (E. coli and

E.



Review question 3 is related to and visual faecal contamination o




The three key elements (populaunchanged also for review questaspects, such as variables influoutcome of interest and other fashared these main key elementsliterature search.

1.2. Searching for research

Electronic databases (search A) a

1.2.1. Search A: electronic d

For search A, the search termsquestions related to “outcome”, “

Terms related to the “populationas possible and then to select thethe screening process.

Regarding the “intervention strawere included in the search strselected in order to include as mconsidered the most important on

More specifically, the search strin

(E. coli) OR (Coliform*) OR ((Microbi*) AND (Slaughter) OR(Chill*) OR (Eviscerat*) OR (DeOR (Pre harvest) OR (Pre-harves

Before identifying the definitivetested, but they were not includein the retrieval of any additional

After having adjusted the search bibliographic databases: PubMed

The search covered the period Ja





the potential relationship between the counts ofof carcasses. The key elements of the question are


d by the presence and amounts of indicator orgcarcasses in relation to their visual faecal contami


ation, outcome and setting) identified for reviewtions 2 and 3, although for these last two reviewuencing the outcome of interest and the relat

actors, have to be considered respectively. Since s it was decided to combine the three review q

h studies

and Web-searching (search B) were used to retrie

databases

s used in relation to the specific key elements “setting” and “intervention strategy /scenarios”, a

n” were omitted from the search string in order toe relevant ones in terms of species of interest in t

ategy /scenarios” related to review question 2, noring. Conversely, for “outcome” and “setting” many synonymous terms as possible, since thesnes to retrieve relevant papers.

ng used was:

(Escheric*) OR (Enterobacter*) OR (Indicator)R (Slaughterhouse*) OR (Abattoir*) OR (Carcasefeathering) OR (De-feathering) OR (Post-harvest) OR (Holding pen*)

e search terms some other terms, such as “coli”d in the final search string since it was verified threlevant papers.

string for minor differences in syntax, it was appd and Web of Science.

anuary 2000 - December 2012 (01.01.2000 - 24.06



hors.

14

f indicator organisms e:

inants and pigs);

ganisms (E. coli and ination;

w question 1 remain w questions additional tionship between the the review questions

questions in a unique

eve pertinent studies.

of the three review are listed in Table 2.

o get as many papers the following steps of

ot all plausible terms the terms used were e two elements were

) OR (Hygien*) OR *) OR (HACCP) OR st) OR (Post harvest)

” and “process” were hat they did not result

plied to two electronic

6.2013).

E.


The bibliographic software RefWfrom the electronic databases.

Table 2: List of the terms includthree review questions

Outcome SettingE. coli SlaughEscheric* SlaughEnterobacter* AbattoColiform* CarcasIndicator Hygien* Microbi*

1.2.2. Search B: Web-search

Search B was conducted by using

Since Google Scholar looks foconducted by using a few veEnterobacteriaceae) and the “spertinent articles the terms O157

The search string used was:

((Escherichia coli) OR Enterobac

The titles of the first 500 returns

1.3. Selecting studies for in

For the purpose of the present rproviding data on the presence/cspecies of interest were considere

Moreover, the relevant studies w

– referring to slaughteringto the conclusion of the c

– providing data on the car

For both searches A and B, papeGeographical restrictions were no

The screening process was indep

In the following paragraphs the elevel are described. Moreover, in




fWorks was used to collect and manage the ref

ded in the search string in relation to the specific

g Intervention strategy /scenariohter HACCP hterhouse* Chill* oir* Eviscerat* s* Defeathering

De-feathering Post-harvest Post harvest Pre harvest Pre-harvest Holding pen*

hing

g the Internet search engine Google Scholar.

r the search terms in the entire document theery specific terms, related to the “outcome” setting” (slaughterhouse). Moreover, to limit t and resistance were excluded from results.

cteriaceae) AND (slaughterhouse)) NOT (O157) N

were assessed in order to identify the pertinent do

nclusion or exclusion in the review

eview, primary research studies performed at thcounts of E. coli and Enterobacteriaceae on carcaed.

were defined as the ones:

process from the point at which animals enter tchilling phase;

rcasses collected at the slaughterhouses.

ers/documents in English, French, Spanish or Italot imposed.

pendently carried out by three veterinarians.

eligibility criteria used to select relevant papers atn Appendix A the checklists developed to identi



hors.

15

ferences downloaded

c key elements of the

os

e search process was (Escherichia coli –

the retrieval of non-

NOT (resistance).

ocuments.

he slaughterhouse and asses of the livestock

the slaughterhouse up

lian were considered.

t abstract and full text ify relevant papers at

E.


the different steps of the screeniby the three reviewers involved i

1.3.1. Screening of the titles

For Search A, the first level asspapers. For Search B, the first lewas pertinent, also abstract (when

The first level assessment consconsidered in these steps were di

For the first step, it was definewritten in languages different fronot necessary to proceed with the

This first step consisted of selecti

1) were written in Engl

2) described data provid

3) provided data related

4) provided data on Enterobacteriaceae;

5) did not have the inve

The second step consisted of sele

1) provided data on minfluencing the load

2) provided data on csecondary process w

When the abstract screening dicriterion, the reviewers provideexclusion of the paper.

Each retrieved paper was individcase of disagreements between thto the next step of the screening p

1.3.2. Examining full-text fo

All retrieved articles (both from submitted to the second level ass

At this stage some eligibility crincluded since, in some cases, thand a final decision could not be

In particular, the following elassessment. Relevant papers:




ing process are reported. These checklists were n the screening process using a subset of 50 retrie

and abstracts for the relevance to the study qu

sessment was conducted considering the title anevel assessment was conducted considering title n available), or in the case of evaluating directly,

sisted of two steps. Papers that did not fulfil iscarded and considered ineligible.

ed that if the two initial criteria were not fulfilom English, French, Spanish or Italian, or were re screening and the article was considered ineligib

ing papers that:

ish, French, Spanish or Italian;

ded by primary research (review articles were exc

d to the main livestock species;

the presence and counts of generic Esch

estigation of antimicrobial resistance as their main

ecting papers that:

more than one stage of the slaughterline or ds of indicator bacteria on carcasses;

arcasses (papers considering parts of the carcawere excluded);

id not identify precise information concerning ed an inconclusive reply (e.g. unknown), that

dually evaluated by two independent reviewers (pahem, the paper was discussed to reach a consensprocess.

or the eligibility of studies

searches A and B) related to livestock species othessment, conducted examining the full-text of the

riteria already taken into account in the first lehe abstract analysis did not allow precise informmade.

ligibility criteria were used to select the pap



hors.

16

validated beforehand eved papers.

uestions

nd the abstract of the and, if the document the full text.

one or more criteria

led (the papers were review papers) it was ble.

cluded);

herichia coli and/or

n purpose.

data on risk factors

sses obtained after a

a specific eligibility did not lead to the

arallel review). In the sus before proceeding

her than poultry were e papers.

evel assessment were mation to be obtained,

pers at second level

E.


1) described data provid

2) considered main live

3) provided data obtain

4) provided data on the

5) did not provide data had been artificially

6) reported E. coli anslaughterline, or dEnterobacteriaceae, and E. coli and/or En

Moreover, at this stage, further study, at which stage of the slauwhich type of risk factor was invthese data was also useful to dpapers (Appendix B).

1.4. Collecting data from th

Two standardized forms were deB).

The first form was aimed at gastudies. It included informationproceeding), the aim of the stu(observational or experimentalspecifications on the sampling mthe carcass, significance and powcollecting pertinent analytical daspecies considered, the indicator counts of indicator bacteria at easteps of the slaughterline where tfactors on the counts of the indicand the counts.

Data related to multiple slaughtepresenting different scenarios we

Data were extracted from each pmethod). In cases where inconsithe forms were observed, the appropriate, they modified data in

In order to avoid double countinfor author names, geographic areduplicates were discarded.




ded by primary research (review articles were exc

estock species (cattle, sheep, goats, pigs);

ed from carcasses;

presence and counts of E. coli and/or Enterobact

about counts of E. coli and/or Enterobacteriaceacontaminated with these indicator bacteria;

nd/or Enterobacteriaceae counts at more thandescribed factors influencing the counts or considered the relationship between visual

nterobacteriaceae counts.

data (e.g. the country where the study was donughterline samples were collected, the procedure vestigated) were gathered from the screened pape

develop forms for the collection of pertinent da

he included studies and creating evidence table

esigned to extract pertinent data from the select

athering general information about the materialsn on the type of reference (e.g. article, technudy, where and when the study was conducted), and the number of plants involved in th

method and sample size (number of samples collecwer of the sample size) were collected. The seconata for the scope of the review. It included inform

bacteria investigated, the unit of enumeration, daach step of the slaughterline. Finally, it includedthe counts decreased or increased, or the effect ofcator bacteria or the relationship between visual

erhouses or data from different visits at the samere considered separately.

aper by one reviewer and then verified by a seconstencies between the data reported in the paper areviewers again verified the paper, but togethn the form accordingly.

ng of the studies published more than once, the pea where the study was conducted, sample size an



hors.

17

cluded);

teriaceae;

ae from carcasses that

n one stage of the of E. coli and/or faecal contamination

ne, the setting of the used to get samples,

ers. The collection of ata from the relevant

es

ted papers (Appendix

s and methods of the nical report, meeting d, the type of study he study. Moreover, cted, sampled area on nd form was aimed at mation concerning the ata on prevalence and d an evaluation of the f the investigated risk faecal contamination

me slaughterhouses or

nd person (sequential and those included in her this time, and if

apers were compared nd data reported. The

E.


The extraction forms were used collected data. The managementdone through the web-based softw

2. RESULTS

2.1. Literature search and

Figure 1 shows the flow chart PubMed and 5345 papers from Wthe screening since they were revFrench, Spanish or Italian. Goodocuments out of the first 500 reremained.

Of the 5500 retrieved papers, 46(thus 263 remaining) of the firstmore eligibility criteria considere

Of the 263 papers that passed livestock species different from p

After having examined the fulassessment because one or morepapers were selected since theyassessed in detail in order to colle

Two out of the 88 selected papesituation that was hardly comparnot possible to get the full-text pa

As a result, 86 papers moved fodata related to bovine slaughterruminant slaughterhouses.

2.2. Discussion of relevant

The studies described in the retroverall on the choice of effect simean with standard deviations. Mthe sum of counts recovered fromwere based on cm2 or 100 cm2 or

Authors’ conclusions about signibased on p-values. However, oftlacking. It's known that the p-valstatistically testing hypotheses (H

For these reasons studies werejudgement about significance oknowledge about all relevant info




to minimise the transcription errors and to obtat of references, the screening and the data extraware DistillerSR (Evidence Partners, Ottawa, Can

relevance screening

of the screening processes. Search A provideWeb of Science. A total of 578 papers were excview papers and/or were written in languages diogle Scholar search (search B) resulted in 73 eturns. After having eliminated the duplicates, a

646 were excluded at step 1 (thus 854 remainint level relevance screening assessment since theyed at that level.

both steps of the first level assessment 145 repoultry.

ll-text, 57 out of 145 papers were excluded e eligibility criteria were not fulfilled and 88 my provided appropriate data for the scope of tect pertinent data.

ers were eliminated since in one case the study rable with the European situation, whereas in thaper.

orward for data extraction; in particular 41 paperhouses, 31 papers to swine slaughterhouses and

studies

rieved papers differ notably in study design, nuizes. Effect sizes used were prevalence, mean wMoreover the unit of enumeration varied from lm different sampling sites or also from different r on the total area sampled.

ificance of results, in particular in the case of riskten, important data to interpret the quality of statlue alone is an unobjective and inadequate measu

Hubbard and Lindsay, 2008).

e mainly compared qualitatively (increase or df results on authors conclusions. It has been aormation, also if not reported, could support their



hors.

18

ain a record of all the action processes were nada).

ed 2148 papers from cluded before starting ifferent from English, potentially pertinent

a total of 5500 papers

ng) and 591 at step 2 y did not fulfil one or

eported data on main

at the second level main livestock-related the review and were

described a hygienic he second case it was

rs provided pertinent d 21 papers to small

umber of samples but ith standard errors or logCFU, CFU, log of carcasses and counts

k factors, were mainly tistical evaluation are ure of evidence when

decrease) basing the assumed that authors r conclusion.

E.


In any case, all the pertinent datand are available for further evalu

Figure 1: Flow-chart summarextraction process

The number of retrieved papers some papers provided data about

Search A - electronic database

N

N° N°

S

N° of p

Total number of papers obtain- PubMed: 2148

- Web of Science: 53(578 excluded: reviews and planguages different from EN, FR, S

N°

Bovine

41




ta extracted from the selected studies are availabuations.

rizing the results of the literature search – the sc

(86) was lower than the sum of the papers of tht more than one species.

es Search B – web se

Google Scho

N° of papers after having removed the duplicates: 5500

Screening of the first 500

Total number of docum

First level screening according to eligibility criteria

of papers that passed the 1° step: 854 of papers that passed the 2° step: 263

(145 main livestock species)

Second level screening according to eligibility criteria for papers

concerning main livestock species

Data extraction

papers that passed the second level: 88

ed by:

345 papers written in SP, IT)

of papers from which data were extracted: 86

Swine

31

Small Ruminants

21



hors.

19

ble in the appendices

creening and the data

he three species since

earching olar

0 returns (title)

ments selected: 73

s

E.


2.3. Bovine

2.3.1. General information a

Among the retrieved papers relathe three review questions.

Fifteen papers corresponded to outside Europe. All papers descresults obtained from an experim

Twenty papers provided data oconsidered both the indicator bac

Seven papers provided informaquestion 2, three papers dealt wreview question 3.

The most common sampling metand meat excision was used in thlog cfu/cm2. The sampling site oanalytical method used to culturplating system.

More details concerning the gen(Appendix C).


In beef slaughterhouses, the animwater, but in most cases are notdriven from the lairage to the splace. Moreover, along the slauggraph below (Figure 2), the treatmslaughterline where they are apdescribed.




about the considered papers

ated to bovine slaughterhouses, 41 papers provid

studies conducted in Europe, while the othercribed observational studies with the exception

mental design.

on E. coli, thirteen papers on Enterobacteriacecteria.

ation about review question 1, twenty-six papewith both review question 1 and 2 and five pape

thod was swabbing (36 papers), skin excision wahe last three papers. The unit of enumeration moston carcasses varied greatly among the retrieved sre the indicator organisms, which was most com

neral information about the retrieved papers are

about the slaughtering process

mals are received and kept in lairage facilities. Tt fed unless they are kept more than one day. Tslaughtering area, where the activities summarizghterline, different decontamination treatments cments described in the retrieved papers are listedpplied is also specified, in order to clarify th



hors.

20

ded pertinent data for

r 26 were conducted of seven presenting

eae and eight papers

ers addressed review ers provided data for

as used in three cases t commonly used was studies as well as the mmonly the Petrifilm

available in Table 6

The animals are given Then the animals are zed in Figure 2 take

can be applied. In the , while the step in the

he specific situations

E.


Figure 2: Flow-chart summardecontamination treatments desc

Since it has been demonstrated tto reduce the presence of microo(Ruby et al., 2007), in modern washes, organic acids washes pasteurization, can be applied at

holding

stunning

sticking and bleeding

hist

hide removal

opening carcass

evisceration

hova

splitting and cutting

chilling

wwwwwst




izing the activities carried out along the beeribed in the retrieved papers are also reported.

that the application of decontamination treatmenorganisms, especially when multiple approaches slaughterhouses different decontamination treatm(only lactic acid in EU slaughter houses), stedifferent steps of the slaughterline.

igh pressure water washing team vacuum

water washing lactic acid washing/spraying acetic acid washing steam vacuum

ot water washing acuum

washing with hot water washing with high pressure water washing with ambient temperature water washing with lactic acid washing with acetic acid steam pasteurization team vacuum



hors.

21

ef slaughterline. The

nts has great potential are sequentially used ments, such as water eam vacuuming and

E.


The retrieved studies described technologies. Hence, the data recannot be directly compared. Thone of the variables that shouldeffect of each step of the slaughte

Other variables, such as the samcarcass further hamper the compa

2.3.3. Review question 1

As regards the presence and amdifferent stages of the slaughterlon E. coli, three papers on Enterobtained through observational st

The main challenge in identifyinindicator bacteria counts on the cand after a single stage. Data havdifferent decontamination treatmthe single stages in a change in cnecessary to draw conclusions aamong studies.

In order to identify the steps obacterial loads, as reported in litetwo sampling points were consid

Gill et al. (2003) examined the trplant in Canada. These authors from less than 15 out of the 25 sbacterial load after pre-evisceratithe numbers recovered after hvacuum/hot water cleaning andreduced by the washing of the caspraying with lactic acid led to ahours E. coli counts increased removal and trimming had only tand steam pasteurization of evisFinally, after chilling, the numbapplying the decontamination trgrowth of injured bacterial cells r

Another study conducted in CanE. coli from carcasses after hicontamination was due to the de1.5 % sodium hydroxide at 55 carcasses were washed, a reductevisceration and trimming operaby spraying carcasses with lactiwere recovered from the sampled




the application of different decontamination treaeported in the different studies at the same stepshe effect of decontamination treatments applied d be taken into consideration before drawing cerline on the bacterial loads of carcasses.

mpling methods, the unit of enumeration and the arability of the data among the studies.

mount of E. coli and Enterobacteriaceae on bline, among the ten eligible papers, five papers probacteriaceae and two papers on both. All the tudies obtained in commercial slaughterhouses.

ng the steps of the slaughterline that led to a decrecarcasses was the paucity of studies that providedve generally been collected at distant sampling p

ments have usually been used. As a consequence,counts was not always evident. For this reason, iaccording to authors’ opinions instead of simply

of the slaughterline that could influence the verature (Gill et al., 2003), differences of less than

dered of no practical importance.

rend of E. coli counts along the slaughterline in a reported that after hide removal, indicator bac

sampled carcasses, at log total number of 3.16 loion washing and spraying with 2 % lactic acid soide removal, and the counts reported after ev

d trimming still remained very similar. The numarcass sides (about 1 log unit less), and the pasteua further reduction (about 2 log unit less). Finallyabout 2 log units. Hence, all operations on cartrivial effects on the microbiological condition of cerated carcasses appeared to be effective in redber of bacteria recovered was similar to the coreatments and authors hypothesised that this incrather than new contaminations of carcasses.

nada (Yang et al., 2012) recovered less than 2 ide removal. According to the authors, this

econtamination treatments applied on hide-on car°C and rinsing with chlorinated water). Then,

tion of bacterial load of more than 1 log unit wations the trend was reversed. Finally, the countsic acid and pasteurization and after these last trd carcasses.



hors.

22

atments and different s of the slaughterline along the process is

onclusions about the

area sampled on the

beef carcasses during provided information papers provided data

ease or an increase of d data obtained before oints and in between, , the effective role of it has been very often y comparing the data

ariation of carcasses n 0.5 log unit between

high-throughput beef cteria were recovered og cfu/2,500 cm2. The olution was similar to visceration, splitting, mber of E. coli was urisation followed by

y, after cooling for 24 rcasses between hide f meat. Then, washing ducing E. coli counts. unts recorded before

crease was due to re-

log cfu/2,500 cm2 of low level of initial

rcasses (washing with when uneviscerated

was reported, but after s were again reduced reatments, no E. coli

E.


Another study (McEvoy et alEnterobacteriaceae contaminatioand at different steps along the obtained from the hock, bung anfrom the cranial back. After evisthe previous step in all the sampwas reported. Trimming, splittin0.5 log unit for all sites apart froreported. Finally, chilling causedfrom the bung samples where asame study, Enterobacteriaceae and at the same points of the scranial back were more than 1 lounits was reported after splittinrecovered from all sites, and afobtained for all sites apart from Enterobacteriaceae counts werattributed the reductions to the intemperature.

The effectiveness of applying investigated in different retrieved

Bacon et al. (2000) conducteddecontamination technologies. Asampled, E. coli counts were in tthe application of multiple deconcarcass water and organic acid wacid washing) a reduction of the all plants. Finally, at the end of thlower than 1 log cfu/100 cm2

decontamination treatments along

Arthur et al. (2004) also evaluatein reducing the load of indicator The authors collected samples aafter the application of all antimithe following decontamination tr

- after hide removal the cu

- before evisceration the c

- after evisceration two otacid and a steam pasteuri

- at chilling, carcasses wer

At the beginning of the slaughcfu/100 cm2. After hide removacfu/100 cm2 and a slight increasewashed carcasses with hot wEnterobacteriaceae counts wereremained substantially unchange




l., 2004), conducted in Ireland, aimed to eson at different sites (hock, brisket, cranial back anprocessing line. After hide removal, comparablend brisket samples, whereas the indicator bacterceration, bacterial loads remained similar to the n

pled sites apart from the cranial back where an ig and washing with warm water led to a reductio

om the cranial back where a further increase (of ad a reduction of the E. coli counts of about 1 log ua bacterial load comparable to the previous step

counts were also registered at the same samplinslaughterline. After hide removal, Enterobacterog unit lower than at the other sampling sites. Anng at the cranial back site. After washing, sifter chilling, clear reductions of the Enterobactecranial back, where an opposite trend was repor

re reduced or remained unchanged after coolinjury of bacterial cells due to stresses from low w

sequential decontamination treatments along thd papers.

d a study involving eight different plants, whiAt the beginning of the process, after hide removthe range 2.6 – 5.3 log cfu/100 cm2. Along the slntamination treatments (including steam vacuum

washing, hot water washing, post-evisceration finE. coli counts ranging from 1.3 to 3.0 log cfu/100he process, after a 24-36 h chilling period, the E. . These authors demonstrated the effectivenesg the slaughterline.

ed the effectiveness of applying sequential antimbacteria on hides and carcasses at two commerc

a) after hide removal, b) after evisceration and icrobial treatments and d) at chilling. The carcas

reatments along the processing line:

ut lines were steam vacuumed,

arcasses were steam vacuumed again and washed

ther washing steps were carried out with hot waisation treatment was also used,

re sprayed for 29 hours before the collection of th

hterline Enterobacteriaceae counts on hides weal the bacterial load on the carcasses resulted ie (0.3 log cfu/100 cm2) was reported at post-evisc

water and peroxyacetic acid and after steame equal to 0.2 log cfu/100 cm2, then, after coolid.



hors.

23

stimate E. coli and nd bung) of carcasses e E. coli counts were ria were not detected numbers recovered in increase of 1 log unit on of E. coli of about about 1 log unit) was unit for all sites apart was reported. In the

ng sites of the carcass riaceae counts at the n increase of 1.45 log imilar numbers were eriaceae counts were rted. Both E. coli and ing and the authors

water activity (aw) and

he slaughterline was

ich applied different al, in the eight plants laughtering line, after

ming, pre-evisceration nal water and organic 0 cm2 was reported in coli load was always

ss of using different

microbial interventions cial processing plants.

carcass trimming, c) ses were treated with

d with lactic acid,

ater and peroxyacetic

he final sample.

ere equal to 6.2 log in a level of 1.4 log ceration. After having

m pasteurisation, the ing the bacterial load

E.


The same sampling points were loads on carcasses to verify the ehigh-throughput beef abattoirs. In

- before hide removal, carc

- after hide removal, carca

- after evisceration, carcawater, hot water and lact

Also in this study, as reported bwas quite high (5.28 log cfu/1001.04 cfu/cm2. A slight reductionevisceration was described and treatments (-0.38 log cfu/100Enterobacteriaceae counts (0.2 l

Bacon et al. (2000) reported thareducing or blocking the bacteriwhereas Ruby et al. (2007) desclast study, the bacterial load of three studies, authors demonstratreducing bacterial loads on carcaand before chilling. These studiechilled samples effectively reduc

Two other studies (Rigobelo etfrequency of recovery and the codetailed information about thedecontamination treatments usedhad an effect on the reported redthree plants. E. coli and Enterobaevisceration and after washing.slaughter process was describedinvestigated and samples were splitting and trimming (post-eviprocessing). The frequency of evisceration step (58 %) to the po

An opposite trend was describebrisket and renal sites of carcassIn this case also, detailed informtreatments used were not availabthen after washing a marginal deabout 1 log unit higher than thobacterial load remained essentialcarcass with cold water producremoving the contamination.

The same sampling points alongestimated the Enterobacteriaceaesplitting and after washing-befor




selected by Ruby et al. (2007), who estimated theffectiveness of the different decontamination tren this case the treatments used along the processin

casses were steam vacuumed,

asses were sprayed with lactic acid,

asses were steam vacuumed, then sprayed with tic acid.

by Arthur et al. (2004), the initial load of Entero cm2), and after hide removal bacterial load on cn down to an Enterobacteriaceae load of 0.8 loa further reduction was reported at the end of

0 cm2). Finally the chilling step led to aog cfu/100 cm2).

at the chilling step was an effective phase of thial load. These findings were confirmed also by cribed an increase of bacterial load during this st

carcasses before starting chilling were extremeted the effectiveness of using sequential decontamasses, especially when these treatments are appl

es showed that the overall intervention processes ced microbiological contamination of carcasses.

t al., 2008; Nero et al., 2012), conducted in Bounts of indicator bacteria at different points of e techniques applied along the processing ld were available, so it was not possible to infer duction. Nero et al. (2012) presented the results acteriaceae counts were estimated after bleeding A progressive decrease of the microbiologica

d. In the second study (Rigobelo et al., 2008),collected immediately after hide removal (pr

isceration) and after washing carcasses hangingE. coli positive carcasses progressively decre

ost-processing step (32 %).

ed by Barboza et al. (2002), who estimated E. es after hide removal, after trimming and splittin

mation about the stages of the processing line andble. A clear increase of the E. coli counts was reecrease of the bacterial load was registered, and se recovered after hide removal. Then, during thlly unchanged and the authors concluded that ates a redistribution of bacteria over the entire c

g the slaughterline were investigated by Maddee counts on brisket samples collected after hide rre chilling in different plants in Northern Ireland



hors.

24

e Enterobacteriaceae atments used in three ng line were:

ambient temperature

bacteriaceae on hide arcasses was equal to og cfu/100 cm2 after the decontamination

an increase of the

he processing line for Arthur et al. (2004),

tage; however, in this ely low. In these last mination treatments in ied after evisceration from hide-on to post-

Brazil, estimated the the slaughterline. No ine as well as the if specific treatments of a study involving

g, after skinning, after al counts during the only one plant was

re-evisceration), after g in the cooler (post-eased from the pre-

coli counts on neck, ng, and after washing. d the decontamination eported after splitting, the final counts were

he washing phase the t best, the washing of carcasses, rather than

en et al. (2004), who removal, after carcass . The counts reported

E.


after hide removal were comparaone of the sampled plants, the concluded that washing with pobacterial load on carcasses. Theythe posterior region (mainly faecthe anterior areas of the carcasses

In order to identify the steps obacterial loads, as reported in litsampling points were consideredunit were regarded as similar, wh

In the following graph (Figure 3of the slaughterline are summariz

It should be pointed out that in soor decrease of bacterial loads waapplied on carcasses since, as repdata before and after a specific pat distant points on the slaughterdata collected in the retrieved stu

One study described the decreeviscerated, whereas another stuacid, no effect in terms of E. colbefore evisceration was effectivean increase of microbial load incorrelated to these slaughter phconsidered as an effective strategstudies demonstrated that washinThe application of different sequor washing with acids in all caFinally, at the chilling step, conflhigher E. coli and EnterobacteriE. coli counts was reported Enterobacteriaceae counts.

The counts of E. coli on beef caslaughterline are reported in Tpresented in Table 10 (Appendixincrease or decrease of the countthese data were available.




able to the numbers obtained after carcass splittinwashing step led to an increase of the bacteri

otable water cannot be considered an effectivey remarked that during the washing process, the hcal contamination) may not only be removed, buts.

of the slaughterline that could influence the vterature (Gill et al., 2003), differences of < 0.5 ld of no practical importance. Therefore, values dhereas values differing by > 0.5 log unit were rega

3), trends of E. coli and Enterobacteriaceae counzed. Only studies providing quantitative data were

ome cases it was very hard to identify if the effecas due to a specific phase of the slaughterline orported above, in only very few cases did the retr

point of the processing line. Moreover, samples wrline. The trends summarised in Figure 3 were deudies as well as the considerations reported by the

ease of E. coli counts after carcasses were wudy showed that when this was combined with li counts reduction was observed. Decontaminatioe in reducing bacterial load (1 study). Eviscerationn one study, whereas in five other studies, changhases were not observed. Washing treatment afgy to reduce the E. coli load of carcasses in one ng at this step had no effect on E. coli and Enterouential decontamination treatments, such as hot ases (five) led to reductions in numbers of botflicting data were collected. In two studies, carcasiaceae counts than in the previous phases, in oneand for another study, this step did not h

arcasses described in the selected papers at the dable 9 (Appendix D), while the counts of En

x D). In these tables, the number of samples analys for each step, the mean and the standard deviati



hors.

25

ng, whereas in at least ial load. The authors e treatment to reduce heavy bacterial load at t also redistributed to

ariation of carcasses log unit between two differing by < 0.5 log arded as different.

nts at different stages e considered.

ct in terms of increase r a specific treatment

rieved studies provide were usually collected efined considering the eir authors.

washed before being spraying with lactic

on treatments applied n and trimming led to ges of bacterial loads fter evisceration was study, whereas three

obacteriaceae counts. water, pasteurization th indicator bacteria. sses after chilling had e study, a drop of the have any effect on

different stages of the nterobacteriaceae are ysed at each step, the ion are reported when

E.


Figure 3: Trends (increase and beef slaughterhouses (BE: before

Figure 4: Trends (increase andslaughterline in beef slaughterhou

0

1

2

3

4

Was

hing

BE

Dec

onta

min

atio

ntr

eatm

ents

BE

N°

of s

tudi

es

0

1

2

3

4

Evi

scer

atio

n -

split

ting

N°

of st

udie

s




decrease) of E. coli counts at different stages oe evisceration - AE: after evisceration).

d decrease) of Enterobacteriaceae counts at diuses (BE: before evisceration - AE: after eviscera

Evi

scer

atio

n -

split

ting

Was

hing

AE

Dec

onta

min

atio

ntr

eatm

ents

AE

Aft

er c

hilli

ng

Was

hing

AE

Dec

onta

min

atio

ntr

eatm

ents

AE

Chi

lling



hors.

26

f the slaughterline in

fferent stages of the ation).

No effectIncreaseDecrease


E.



According to the defined searchdealing with risk factors influenobtained. Fifteen papers descrEnterobacteriaceae counts and fi

As already mentioned for the rstudies was hampered by differearea sampled on the carcass andacross studies. Hence, differentconclusions drawn and not in tespecific risk factors investigated

The effect of the annual season odifferent studies (Rigobelo et alseason could have a direct impaparticular, it was demonstrated tduring the rainy season (RigobelJuly) could lead to higher EnterDecember and March) (Ruby et factors and also concluded that samples significantly influence th

E. coli and Enterobacteriaceae where the slaughtering took placdesign of the plant, the throughsome other cases differences amowithout attributing these findings

Zweifel et al. (2008) conducteanimals slaughtered annually). E55 % and some differences amostudy dealing with this point wasEnterobacteriaceae loads on carboth cases it was not possible tofor such differences.

The influence of plant featureBlagojevic et al. (2011), who inv(before dehiding) and at the end that the Enterobacteriaceae coun

Also the effect of the plant dEnterobacteriaceae loads of beewas a linear rail plant and had second abattoir had a serpentinelevel up to a higher level slaugclearly influenced the airbornEnterobacteriaceae load of the contamination is poor.




h process and the established eligibility criteria,ncing E. coli and Enterobacteriaceae counts onribed studies that reported E. coli counts, nfive papers considered both types of indicator bact

review question 1, the comparability of data pent aspects, such as the sampling method, the unid the decontamination treatments applied, whicht studies considering the same factor were corms of counts reported. However, many other acould influence the final results presented in the r

on the prevalence of E. coli on beef carcasses wal., 2006; Ruby et al., 2007; Rigobelo et al., 20act on indicator bacteria prevalence and counts that there is a higher probability of finding E. colo et al., 2006; Rigobelo et al., 2008), and that wrobacteriaceae counts on carcasses than colderal., 2007). These last authors investigated the inthe geographical position of the plant and the

he bacterial loads of beef carcasses (Ruby et al., 2

counts on carcasses varied, perhaps according tce. In some studies, specific aspects related to th

hput, or the surveillance system in place were eong plants in terms of bacterial loads of carcasses to some specific reasons.

d a survey in low-throughput abattoirs in SwEnterobacteriaceae prevalence on cattle carcassesong the slaughterhouses in terms of counts weres conducted in France and reported significant dircasses collected at four different plants (Collo identify which main features of the plants invol

es on microbial carcass contamination has alvestigated two different abattoirs collecting samof the slaughterline (before chilling). In this case

nts did not significantly differ between the two ab

design was investigated. Prendergast et al. (2ef carcasses obtained in two Irish slaughterhousslaughter, dressing and chilling systems on a sie rail on two floors and animals moved from thghter and dressing line. The authors concluded ne bacterial numbers, but it did not have

carcasses, and that the relationship between a



hors.

27

a total of 29 papers n beef carcasses were nine papers reported teria.

provided by different it of enumeration, the h were not consistent ompared in terms of spects apart from the retrieved papers.

as addressed by three 008), concluding that on beef carcasses. In oli positive carcasses

warmer months (June, r months (November, nfluence of other risk year of collection of

2007).

to the slaughterhouse his point, such as the evaluated, whereas in es are simply reported

witzerland (about 220 s ranged from 0 % to e reported. The other ifferences in terms of

obert et al., 2002). In lved were responsible

lso been studied by mples at the beginning e, the authors reported battoirs.

2004), compared the ses. The first abattoir ingle floor, while the he lairage on a lower that the plant design any effect on the

airborne and carcass

E.


The effect of plant throughput onAustralian studies addressed thiprevalence of E. coli on carcasssupplying the domestic market slaughterhouse differed signifiprocessing operations, the authorslaughterhouses. Opposite conclulower E. coli prevalences on cthroughput slaughterhouses (28.counts on carcasses from highthroughput plants (-0.54 log slaughterhouses were not signifibacteria (E. coli and Enterobacthigher (2.07 and 2.18 log cfu/ccfu/cm2). Also, in this last casestatistically significant, probably Finally, the last retrieved paperprevalence of E. coli positive car(41 %). Also in this last case statistically significant.

Authors generally reported lowerwere the differences between theplants applied more modern asatisfactory data were generallyslaughter rate that may result in b

Different slaughterhouses could data collected from the retrievedto different slaughterhouses coul

Also the effect of the surveillancinfluencing indicator bacteria coplants in Australia. E. coli was survey demonstrated that levelssystem were comparable to thosby government inspectors (Bass efaecal contamination caused by verify how it could influence capplication of this online monitoconsequently it was an effective m

The effect of using physical and another aspect that was commonconducted on developing pathoganimal carcasses. Among these recognised as two of the mostcarcasses.

Treatments with steam or hot wain a high-throughput beef slaughevaluated the decontamination e




n indicator bacteria counts was frequently investis point. Phillips et al. (2001) carried out a comses among slaughterhouses concentrating on exp

and low-throughput slaughterhouses. Althoughcantly for standards of construction, governrs concluded that the incidence of E. coli was cousions were drawn by Sumner et al. (2003), who rcarcasses at low-throughput slaughterhouses (4.4 %). Also Bohaychuk et al. (2011) reported h-throughput plants (-0.23 log cfu/cm2) than cfu/cm2), but the differences found betweencant. Then, Ozdemir et al. (2010), reported that t

teriaceae) on beef carcasses at high-throughput scm2) than at the low-throughput slaughterhousese, the difference between the two types of slaubecause the same slaughtering technique was use

r reported an opposite finding. Hansson et al. (2rcasses in high-throughput plants (34 %), than in the difference between the two types of slau

r prevalences/counts in lower throughput plants, e two types of slaughterhouses significant. Althoand specialized techniques than the low-throuy found at low-throughput plants, which could better hygiene (Sumner et al., 2003).

induce different effects on the bacterial counts d studies did not clarify which one among the difd have a main effect.

ce system in place in the slaughterhouses was invounts. For this scope a microbiological baseline

detected in 25 % of carcasses (mean log posits of indicator bacteria on carcasses processed ve established in abattoirs that operate the traditioet al., 2011). Moreover, the application of an onlidehairing and evisceration operations was also

carcass hygiene. Tergney and Bolton (2006) doring system led to a reduction of the faecal conmean of reducing the enteric counts.

chemical decontamination treatments on indicatonly investigated in the retrieved studies. Extensigen reduction technologies to improve the micrtechnologies, steam pasteurisation and steam v

t effective methods for decreasing bacterial po

ater have been tested by different authors. Retzlafhterhouse, a chamber steam pasteurisation unit ffect of seven different chamber temperatures (f



hors.

28

tigated. Two different mparison in terms of port, slaughterhouses h the three types of

nment oversight and omparable among the reported significantly 4.7 %) than at high-higher mean E. coli on those from low-

n the two types of the mean numbers of slaughterhouses were s (1.90 and 1.98 log ughterhouse was not ed in all plants. 2001) found a lower low-throughput ones

ughterhouse was not

but in only one study ough high-throughput ughput plants, more be due to the lower

of carcasses, but the fferent aspects related

vestigated as a factor was conducted in 16

tives -0.61/cm2). The via the co-regulatory onal system overseen ine system to monitor evaluated in order to

demonstrated that the ntamination rates and

or bacteria counts was ive research has been robiological safety of vacuuming have been opulations on animal

ff et al. (2005) tested, and they specifically

from 71.1 to 87.7 °C)

E.


at the pre-rigor step. When the trand Enterobacteriaceae were red

Also, Minihan et al. (2003) demoexposure time); the tested trEnterobacteriaceae at the more decontamination of the carcass.

The same conclusions were dradifferent decontaminant treatmewas always effective for substantPasteurisation equipment, consisside of carcass to be treated indunder steam pressure at an approbacterial loads were significantl(Trivedi et al., 2007), the effectmeat-processing plants (economilast case, the steam pasteurisatithree anatomical areas sampled o

The five studies addressing this pEnterobacteriaceae counts on bthroughput plants, and all studicarcasses due to this decontamina

Similarly, hot water pasteurisatioquality of the beef carcasses, bunegated by a worsening of the or

In a study conducted by Gill anfor different time frames (from reductions increased with longetreatment was applied for 11 or to be less desirable than that of th

Also Kiermeier et al. (2006), whprevalences reported in some Auwater decontamination systems registered in slaughterhouses tcontaminations of the livestock in

The use of hot water as a pre-evBosilevac et al. (2006), and its ehot water carcass wash cabinet counts by 2.7 log CFU/100 cmcabinet, applying 2 % lactic acidthis case the reduction was morwere used sequentially, a reductauthors concluded that hot watebeef carcasses before evisceration




reatment was conducted at temperatures higher thduced to undetectable levels (<0.4 CFU/cm2).

onstrated the positive effect of steam pasteurizatireatment significantly reduced the level of contaminated sites of the carcasses, but it did n

awn by Gill and Landers (2003a), who confirments they investigated, only pasteurisation with tially reducing the bacterial load of beef carcassessting of a fixed cabinet equipped with a carouseldividually was tested (Corantin et al., 2005). Caoximate operating temperature of 74.5 °C for 8 sly lower than those of untreated carcasses. Finativeness of a steam cleaner suitable for low-to ic and easy to operate and install equipment) wasion treatment significantly reduced Enterobacteon beef carcasses.

point showed that steam pasteurisation effectivelyeef carcasses. The equipment tested was used bies reported a clear reduction of both indicatoation treatment.

on was demonstrated as an effective way to imput in this case the improvement of microbiologrganoleptic features.

d Bryan (2000), beef carcasses were pasteurised8 to 12 seconds) and all treatments reduced E.

er treatment time and the highest reduction wa12 seconds. However, the appearance of treated he untreated ones.

ho conducted an investigation to clarify the reasoustralian slaughterhouses, concluded that the slaug

had lowest E. coli prevalences and this posithat had substantial incoming problems relatentroduced.

visceration wash in a commercial slaughterhouseffect was compared with that of lactic acid wash

tested, applying 74 °C water for 5.5 s, reducedm2 on the pre-evisceration carcasses. Similarly,d at approximately 42 °C, reduced Enterobacterre limited (1.0 log CFU/100 cm2). Moreover, wtion equal to 2.5 log CFU/100 cm2 was obtaineer washing was the more effective procedure forn.



hors.

29

han 85 °C both E. coli

ion (90 °C for 10 s of both E. coli and

not result in complete

med that among the steam and hot water s. that allows each full arcasses were treated seconds, and the final ally, in the last study very low-throughput

s verified. Also in this riaceae counts at all

y reduced E. coli and both in high and low or bacteria counts on

prove microbiological gical quality could be

d with water at 85 °C . coli counts, but the

as obtained when the carcasses was judged

ons for higher E. coli ghterhouses using hot itive result was also ed to high level of

se was tested also by hing. The commercial d Enterobacteriaceae , a lactic acid spray

riaceae counts, but in hen the two cabinets

ed. Consequently, the r decontamination of

E.


Hence, hot water washing as wreducing the bacterial load of bee

The effect of washing with potdifferent studies. Mies et al. (2ineffective in reducing bacterial potable water is a redistribution obe an efficacious or desirable wa

Gill and Landers (2003a) demonthe washing treatments applied. when the initial bacterial load onplausible explanation for this finassociated with particles, which aoperation. When numbers are lomay be refractory to physical rem

In the retrieved studies the usageresults obtained were rather contr

Mies et al., 2004 evaluated the double water wash, water wash reduce the bacterial contaminaapplication of each of the treatmGill and Landers (2003a), whotreatments applied at four slaughineffective treatment in reducingdrawn for peroxyacetic acid sprtreatments may be due to the dwashing or condensation from ste

Barboza et al., (2002) evaluated for beef carcasses. Lactic acid anpopulations on the carcasses, whThe conflicting results reportedslaughterline where the treatmenwashing procedure on hides of before carcasses moved into the c

Moreover, another study (Algdecontamination treatments in without water), spray-chilling, (commercial vinegar) and a mixAll the interventions considereEnterobacteriaceae counts on be

Also the application of other chebe equally effective as demonstrato clean cattle hides before capproximately 2 logs of Enteroconsisting in the removal of hairsubsequent neutralization with a




well as steam pasteurisation appears to be an efef carcasses at different stages of the slaughterline

table water at ambient temperature was also ta004) and Barboza et al. (2002) concluded thatloads on beef carcasses. The main effect observeof microbial contamination over the carcass surfa

ay to reduce bacterial loads.

nstrated that different effects could be accounted Moreover, the same authors indicated that was

n carcasses was low, but effective when the load wnding could be that when counts are high the bare washed from the carcass by the large volumesow, the bacteria are probably directly associatedmoval by washing (Gill and Landers, 2003a).

e of alternative decontaminant treatments was alsoroversial.

effectiveness of the following washing systemswith 0.5 % lactic acid and water wash with 50

ation on hides of cattle before entering the ent resulted in an increase in the bacterial loads oo investigated the antimicrobial effect of diff

hterhouses confirmed that spraying carcasses withg microbial contamination on carcasses. The samraying. The authors supposed that the apparent dilution of the applied solution by water preseneam rather than from bacterial resistance.

the effect of organic acids and bacteriocin washnd the mixture of lactic acid and nisin produced hereas when only nisin was used, no positive ed in the last cited papers could be also due

nts were applied. Mies et al. (2004) described datacattle before entering the slaughterhouse, Barbochilling area.

gino et al., 2007) investigated different phylow-throughput plants, such as dry aging (mlow pressure and high pressure hot water

xture of different acids (citric acid + ascorbic acied led to a significant decrease in mean leveef carcasses.

emicals as decontaminants at different steps of thated by Nou et al. (2003), who tested the chemicaarcass dressing. This treatment led to a signobacteriaceae counts immediately after hide rer and extraneous matter from hide with a sodium hydrogen peroxide solution, and a water washing



hors.

30

ffective treatment for e.

aken into account by t water washing was ed after washing with ace, so this would not

for by differences in shing was ineffective was relatively high. A bacteria are probably s of water in washing

d with tissues, and so

o investigated and the

s: single water wash, 0 ppm of chlorine, to slaughterhouse. The

of the carcasses. Also ferent decontaminant h 2 % lactic acid is an me conclusions were failure of these two

nt on carcasses from

hes as decontaminants a reduction of E. coli

effects were reported. e to the step of the a obtained applying a oza et al. (2002) just

ysical and chemical multiday refrigeration

, 2.5 % acetic acid id + erythorbic acid). vels of E. coli and

he slaughterline could al dehairing, designed nificant reduction of emoval. The process sulphite solution, the

g before dehiding.

E.


The effect of chemical decontaopposite results related to their ef

The effect of spray-chilling on thconsidered a crucial phase of theconsequently slowing proliferatiused for this purpose, and it has the carcass surface. However, thand result in a substantial reductithe process termed spray-chillingduring the first part of the cooling

One study evaluated the effect Landers, 2003b) in four plantslaughterline. This study demonscoli recovered from cooled carcadressed carcasses, for one plant of E. coli declined by about 1 lowere not due to the spray-cooldecontamination treatments applentering the cooling chamber.

Kinsella et al. (2006) investigatchamber to provide intermittent wCarcasses were sampled after drsuggested that this spray-chillingsurface counts of indicator bacter

The two studies addressing the extremely different treatments annot possible to produce a definiticarcasses.

Only one study considered batcarcasses. Feed regimen just beinfluencing bacterial load of caradministration of an experimentconfirmed that no main effects oon numbers of E. coli recovered

In the following tables (Tables 3ain terms of increase-decrease-nostudies which described differeinvestigating which risk factors c




amination treatments was unclear, since differeffectiveness in reducing bacterial loads on beef ca

he bacterial load of beef carcasses was quite cone slaughter process, reducing the internal temperaion of most mesophilic bacteria. Air cooling is the positive effect of limiting bacterial growth on

his methodology can provoke water evaporation fion in carcass weight (and therefore, carcass valug or spray-cooling, the carcasses are intermittentlg process in order to limit weight loss.

of a spray-cooling process on E. coli counts onts, which applied different decontamination tstrated that for two out of the four plants examinasses were at least 1 log unit higher than the numthe bacterial load remained unchanged and in th

og unit during carcass cooling. It is likely that theling step, but they could be partly explained blied along the line and the bacterial loads of the

ted the effect of a chilling system that consistedwater spraying of carcasses (spray cycle 2 min onressing and after 24 h in the spray-chilling unit, ag system can limit carcass shrinkage, without signria.

effect of spray-chilling on bacterial loads of cnd, therefore, the results presented cannot be comive indication about the effect of spray-chilling on

ch related risk factors on indicator bacteria coefore moving animals to the slaughterhouse isrcasses. Anderson et al. (2005) verified the efftal chlorate preparation on E. coli on beef carc

of feed and water treatment or any possible interafrom hide or carcasses.

a-3b-3c), the effect of the risk factors described in effect of indicator bacteria on beef carcasses arences between the variables investigated, but w

could influence that variability found, were not in



hors.

31

ent studies described arcasses.

ntroversial. Chilling is ature of carcasses and a process commonly n carcasses by drying from carcass surfaces

ue). For this reason, in ly sprayed with water

n carcasses (Gill and treatments along the ned, the number of E. mber recovered from

he last plant, numbers ese conflicting results by the effects of the carcasses just before

d of a humidification n, 1 min off) for 15 h. and the data obtained nificantly altering the

carcasses investigated mpared. Hence, it was n the bacterial load of

ontamination of beef s considered a factor fect of feed or water asses. Data collected

actions were observed

n the retrieved papers e summarized. Those

without specifying or ncluded in the list.

E.


Table 3a: Slaughterhouse and fadecrease/increase/or with no effe

Effect Risk factor/Treatm

Decrease dry season

low -throughput slaugh

on line monitoring fcontamination

decontamination – spasteurization temperatu

decontamination - stpasteurization

decontamination – hot

decontamination – lactic aacid + nisin

decontamination (low-thplants) – day aging – ace

mixture of different acidshigh pressure hot w

spray cooling (1 slaught

Decrease Total

Increase

decontamination - siwash/double wash/ water

0.5% lactic acid / water wappm of chlorine

spray cooling (2 slaughte

Increase Total

No effect high-throughput slaugh

co-regulatory inspection

washing with wat

decontamination: lactiperoxyacetic acid spr

diet - sodium chlorate feedchlorate water

spray cooling (1 slaught

No effect Total




arm related risk factors investigated in the retrievect on E. coli counts on beef carcasses

ment Comparison factor /control Refe

wet season Rigobelo et

Rigobelo et a

terhouse high-throughput Sumner et a

faecal control Tergney and Bo

steam re > 85°C temperature 71.1 - 82.2 °C Retzlaff et a

team control - unpasteurized

Minihan et a

Corantin et a

Gill and Lande

t water control - untreated Gill and Brya

Kiermeier et

acid – lactic control - untreated Barboza et a

hroughput etic acid – – low and

water

control - untreated Algino et al

terhouse) control - before cooling Gill and Lande

ingle wash with

ash with 50 e

control untreated Mies et al.,

erhouses) control - before cooling Gill and Lande

hterhouse low-throughput

Phillips et al

Bohaychuk et

Ozdemir et a

Hansson et a

n system governmental inspection system Bass et al.

ter control - other decontamination treatments

Mies et al.

Barboza et a

c acid - raying control Gill and Lande

d – sodium control Anderson et

terhouse) ²control - before cooling Gill and Lande



hors.

32

ved studies leading to

erence N° of studies

al. 2008 (BR) 2

al., 2006 (BR)

l., 2003 (AU) 1

olton, 2006 (TR) 1

al., 2005 (US) 1

al., 2003 (IE)

3 al., 2005 (CA)

ers, 2003a (CA)

ant, 2000 (CA) 2

al., 2006 (AU)

al., 2002(VE) 1

l., 2007 (US) 1

ers, 2003b (CA) 1

13

2004 (USA) 1

ers, 2003b (CA) 1

2

l., 2001 (AU)

4 t al., 2011 (CA)

al., 2010 (TR)

al., 2001 (SE)

2011 (AU) 1

, 2004 (US) 2

al., 2002 (VE)

ers, 2003a (CA) 1

al., 2005 (US) 1

ers, 2003b (CA) 1

10

E.


Table 3b: Slaughterhouse reladecrease/increase/or with no effe

Effect Risk factor/Treat

Decrease dry season

slaughter character

high-throughput slaugh

on line monitoring contamination

decontamination – steam ptemperature > 85

decontamination - steam p

decontamination - ho

decontamination (low-tplants) – day aging – ac

mixture of different acidhigh pressure hot w

decontamination - chemic

Decrease Total

No effect slaughterhouse design: lin

floor high-throughput slaugh

slaughter character

cooling–novel spray-

No effect Total


Five papers provided informatioand their E. coli and/or Enterobaon Enterobacteriaceae and one contaminated carcasses and were

It is generally accepted that the dthe amount of dirt transferred to tto meat can result in improvedinvestigated these points were ret

In the study conducted by Sarrcleanliness at the beginning of from 1 (very clean) to 5 (very selection of carcasses from each 4 and 5) animals had the highe(categories 1, 2 and 3). A direct cand the level of microbial conslaughter visual evaluation of measures (e.g. rejection of the anmicrobial contamination) couldslaughterline.




ated risk factors investigated in the retrievedect on Enterobacteriaceae counts on beef carcasse

tment Comparison factor /control Re

wet season Ruby et

ristics control Zweifel etCollobert e

hterhouse low-throughput Ozdemir efaecal

n control Tergney and

pasteurization 5°C temperature 71.1 - 82.2 °C Retzlaff et

pasteurization control - unpasteurized Minihan eTrivedi et

ot water lactic acid / lactic acid + hot water Bosilevac ethroughput cetic acid – ds – low and water

control - untreated Algino et

cal dehairing control - untreated Nou et a

near rail one serpentine rail and two floors Prendergast

hterhouse low-throughput Ozdemir e

ristics control Blagojevic

-cooling control Kinsella e

n on the relationship between faecal contaminatacteriaceae counts. Two papers provided data opaper on both the indicator bacteria. All papers

e conducted in commercial slaughterhouses.

dirtiness of animals at the time of slaughter is dirthe carcass, and that procedures avoiding the tran

d microbiological quality of the carcasses. Verytrieved.

raino et al. (2012), beef carcasses were classifthe slaughterline and were categorised accordindirty). Enterobacteriaceae and E. coli counts wcategory and the data collected confirmed that th

est bacterial loads on carcasses in comparison correlation was observed between visual cleanlin

ntamination of carcasses. Hence, the authors coanimals’ cleanliness and the consequent applinimals, washing or application of additional procd be effective to limit carcass contamination



hors.

33

d studies leading to es

eference N° of studies

al. 2007 (US) 1 t al., 2008 (CH)

2 et al., 2002 (FR) et al., 2010 (TR) 1

Bolton, 2006 (TR) 1

t al., 2005 (US) 1

et al., 2003 (IE) 2

t al., 2007 (US) et al., 2006 (US) 1

t al., 2007 (US) 1

al., 2003 (US) 1

11

t et al., 2004 (IE) 1

et al., 2010 (TR) 1

et al., 2011 (RS) 1

et al., 2006 (IE) 1

4

tion of beef carcasses on E. coli, two papers

considered naturally

rectly correlated with nsfer of dirt from hide y few studies which

fied in terms of hide ng to a scale ranging were estimated for a he dirtiest (categories to the cleanest ones

ness category of hides oncluded that a pre-ication of corrective cedures to reduce the

n at the end of the

E.


Similar conclusions were drawn that the simplification of the cattwo main categories, including tpractice to identify the carcaspotentially have the highest micthree different classes (1: clean, conducted in two beef abattoirs iafter hide removal, at the beginnwith visually clean hides prodHowever, at the next step of significantly higher levels of cacarcasses are treated more carefmicrobiological quality of the comparable E. coli counts regard The same findings were repocontamination of animals carcaconducted at four slaughterhouscollected on the detaining rail ancontaminated area and from anosame sampling scheme was repearrived at chilling after the slaughterhouses had been applieprocedures, bacterial loads of strimmed area or that were ranmicrobiological quality of the cabetter than, the other carcasses ro In the last study dealing with thimortem and post-mortem hide cllivestock policy was introduced i1-5, with animals in categoriesslaughter without taking further ante-mortem and taking additioconducted both in the lairage or Clipping procedures are aimedcontamination to the carcass durifrom 1 to 4, according to the levewas analysed for Enterobacterifurther precautions, whereas camortem on the slaughterline. Anbetween the Enterobacteriaceae mortem or post-mortem). Clippquality to those derived from clea

In conclusion, the studies retriebacterial load than those derivcontaminated cattle and applicaslaughterhouse and/or along thcontamination level of carcasses the studies support the conclusio




by Blagojevic et al. (2012), but in this case the ategorization scheme, and the classification of ththe very dirty animals and the less dirty ones, ses that should be managed with particular crobial loads. Similarly, a simplified classificati2: moderately dirty and 3: very dirty) was descrin Norway (Hauge et al., 2012). Samples were cning and at the end of the slaughterline. After h

duced carcasses with a lower amount of bactef the slaughterline, the dirtiest animals were arcass contamination. This could be due to the fully and such extra care might lead to a clear carcasses. At the end of the process all carc

dless of the initial level of contamination.

orted by Gill and Landers (2004b), who evasses detained for removal of visible contaminases in Canada. A selection of visibly contaminnd sampled within the contaminated area, from other non-contaminated area of the carcass, raneated on the carcasses after trimming, and again

routine washing and decontamination treatd. The data collected confirmed that after compl

sites that had been visibly contaminated, that wndomly selected were all similar. Moreover, arcasses that were initially detained was generallyoutinely processed at the same plant.

is point, authors (McCleery et al., 2008) evaluatlipping on the microbiological quality of carcassein 2007 based on the categorization of animals, as 1 and 2 being the cleanest and consequentlyprecautions, category 3 animals being rejected a

onal measures like clipping to facilitate cleanion the slaughterline after stunning and bleeding

d to remove visible dirt from the hide in oring slaughter process. In the study, a selection of el of visual contamination of hides before enteriniaceae counts. Category 1 and 2 animals were ategory 3 and 4 animals were either clipped aalysis of data obtained revealed no statistically spopulation recovered and animal category or cli

ped animals produced carcasses with a comparan animals.

eved confirmed that carcasses derived from cleved from visibly dirty animals. However, idenation of effective measures either on animals he slaughterline can lead to lower bacterial at the end of the process comparable to that of c

on that the pre-slaughter visual evaluation of the



hors.

34

authors demonstrated he carcasses into only could be effective in attention since they

ion system, based on ribed in another study collected immediately hide removal, animals eria than dirty ones. not associated with fact that the dirtiest improvement of the

casses had the same

aluated the level of ation. The study was nated carcasses were a site adjacent to the

ndomly selected. The n when the carcasses tments used in the letion of the dressing were adjacent to the it emerged that the

y comparable with, or

ted the effect of ante-es. In the UK, a clean ccording to a scale of

y considered safe for at first presentation at ng. Clipping can be

g but before dehiding. rder to reduce cross carcasses categorised

ng the slaughterhouse, slaughtered without

ante-mortem or post-ignificant association ipping location (ante-rable microbiological

ean cattle had lower ntification of visibly

before entering the contamination or a

clean animals. Hence, level of hide-carcass

E.


contamination and the applicatiothe initially contaminated animalend of the slaughterline.

Table 21 (Appendix F) summaralong the slaughterline in relation

2.4. Pigs


Among the retrieved papers, 31 pquestions. Seventeen papers werstudies identified described obseexperimental design. Eighteen pathree papers considered both indi

Fourteen papers provided inform2, two papers dealt with both revisual faecal contamination in oquestion 3) were retrieved.

The most common sampling metsampling method used in three stThe sampling site on carcasses method; in this case the most com

More information about the (Appendix C).




on of proper corrective measures along the slaughls can be an effective approach to reduce carcass

rizes the counts of E. coli and Enterobacteriacen to the level of visual faecal contamination.


papers were eligible to provide data on pigs regare conducted in Europe, and the other fourteen ouervational studies with the exception of five stuapers provided data on E. coli, ten papers on Enicator bacteria.

mation on review question 1, fifteen papers addreeview question 1 and 2, while no papers dealingrder to study the effect of this variable on indic

thod was swabbing (28 out of 31 papers), while studies. The unit of enumeration most commonly uvaried greatly among the retrieved studies as w

mmon was classified as ISO but also Petrifilm, an

characteristics of the selected papers are av



hors.

35

hterline especially for s contamination at the

eae on beef carcasses

rding the three review utside Europe. All the udies dealing with an nterobacteriaceae and

essed review question g with the scoring of cator bacteria (review

skin excision was the used was log cfu/cm2. well as the analytical nd MPN were used.

vailable in Table 7

E.



Figure 4: Flow-chart summarizin


This section is aimed at the evalcounts of indicator bacteria in swwith this issue. In particular, sevcounts and two counts of bothdescribed in the selected studies

The retrieved papers provided dathe steps of the slaughterline thacarcasses is the difficulty of finData have been generally collectsingle stage toward a change in b

E. coli

Different studies investigating thWarriner et al. (2002) evaluatedcarcass every 45-50 minutes. Tcarcasses according to the slauperformed after scalding, and byafter it. Moreover, carcasses unreduction of bacterial counts to s

Holding Stunni

EvisceraSplitting

Chilling

Head removal

PasteurizationTrimming

Lactic acid w





ng the activities carried out along the pig slaughte

luation of the effect of the different stages of thewine carcasses. Fourteen papers were retrieved frven papers reported Escherichia coli counts, fiveh indicator bacteria. In Figure 4, the main slaare presented.

ata that were not always comparable. A main chaat lead to a decrease or an increase of indicator bding studies providing data obtained before andted at distant sampling points; as a consequence tbacterial levels is not always evident.

he trend of E. coli counts along the slaughterlind the effect of slaughtering on E. coli counts b

This study evaluated also differences in terms ughter time in the day. The process was charay a double polishing technique performed dry befnderwent a washing treatment before chilling. Ascalding and singeing; however, since the sampl

ing Bleeding

Singeing

Sticking

Polishing ation

ScDry

S

Wa

n

Pasteurization Anal closure

wash



hors.

36

erline.

e slaughterline on the rom literature dealing e Enterobacteriaceae

aughtering operations

allenge in identifying bacteria counts on the d after a single stage. the effective role of a

ne are described here. by sampling a single of bacterial loads of

acterised by scraping fore singeing and wet

Authors attributed the ing of carcasses after

Scalding

Dehairing

craping polishing

Shaving

ashing

E.


these specific stages was not posreal sampling points (after scraobtained for Enterobacteriaceasignificant for E. coli.

Variations of E. coli counts wcharacterized by a dry and wet anal closure routinely performedcoli counts decreased during scaet al., 2000).

Nesbakken et al. (2008) evaluateE. coli, a slight reduction was reduction was stated as significan

A Danish study (Wu et al., 2009close to the anus, in a high thrreduced significantly (by 1-2 loattributed by authors to scaldingimmediately after the stages conafter splitting were compared to performed such as singeing, polis

Gill et al. (2000) investigated eigonly, polished carcasses were pSampling was performed by swagrid. Moreover, E. coli countsprocedure. From each plant, setscoli recovered indicated that subthe dressing processes at four ofreduced during the chilling witho

Tamplin et al. (2001) examined tthroughput (7,000 pigs/day) slaupoint-based inspection models pfollowing exsanguination and afcoli was found on all carcasses numbers recovered on different cfu/cm2).

In Canada, a study to evaluate thwas performed at Lacombe Resstudy was to compare the levinvestigated animals were raiseprovided. Samples were obtainesamples belonging to a commeexperimental design. In the LRsampled. However E. coli counCarcasses were sampled beforwashing. Random carcass samplimportance of pasteurisation wacase, carcasses had noticeable de




sible in the study, these results are presented in taaping and after polishing, respectively). In contae, the increase observed during evisceration

were also investigated in an Iberian study. Thscraping technique, as in the previously describ

d before evisceration in order to prevent gut conalding, but increased during evisceration despite

ed the effect of blast chilling on different bacterialobserved after the application of this chilling

nt.

9) investigated the behaviour of E. coli by samplroughput slaughterhouse (10,000 pigs per day). og units) at sequential stages of processing andg, singeing and chilling. However, sampling poisidered effective. As an example, bacterial loadsones collected after scalding. In between, severalshing, evisceration and splitting.

ght slaughterhouses processing 200 to 800 pigs ppasteurised before dressing commenced. Blast cabbing a random site on carcasses chosen from a ps were estimated by using the hydrophobic gs of 25 samples were collected at three differenbstantial numbers of those organisms were addedf the plants, and that bacterial levels on carcassout spraying process at two plants.

the prevalence and quantity of E. coli on swine caughterhouse operating under the hazard analysi

project (HIMP) program. Carcasses were sampledfter the carcasses were washed, eviscerated and sampled after bleeding and on 30.1 % of the chidays varied greatly, in particular after exsangui

he impact of slaughtering stages on E. coli countsearch Center (LRC) (Jones and Johns, 2012). vel of F-RNA coliphages with different inded and slaughtered in LRC pilot plant and Med from anal region and from random sites on ercial slaughterhouse were also analysed and u

RC pilot plant, E. coli numbers varied accordinnts in the anal region decreased at each samplinre pasteurisation, after pasteurisation, after evles showed a decrease in E. coli counts only aftes also confirmed by results obtained in the com

ecreases of E. coli counts after bleeding and after p



hors.

37

ables according to the trast with the results was not statistically

he slaughterline was bed paper, and by the ntent from leaking. E.

anal plugging (Rivas

l species. Concerning g technique and this

ling the left hind leg, E. coli counts were

d this reduction was ints were not located

s of samples collected l operations had been

per hour. At one plant chilling was applied. pre-designed 83 areas grid membrane filter nt sampling points. E. d to carcasses during es were substantially

arcasses at a medium-s and critical control d twice: immediately chilled overnight. E.

illed carcasses. Mean ination (53 to 11,000

ts on swine carcasses The objective of the

dicator bacteria. The MPN estimates were carcasses. Moreover, used to improve the

ng to the carcass site ng point investigated. visceration and after er pasteurisation. The

mmercial plant. In this pasteurisation.

E.


Another study was conducted in evaluate which processing resultcarcasses carried relatively largeof carcasses was applied in earmuscle surface; however therecontamination during head reremoval, instead of before. Howand washing) led to a new inccarcasses that were pasteurised aof E. coli.

Gill and Landers (2004a) analybefore and after cooling and dem

Namvar and Warriner (2006), inpig slaughterline, described thecarcasses. In particular, twelve visits after the following operatioE. coli numbers on carcasses folbut progressively decreased towsignificantly reduced by scaldingthe polishing process. E. coli wevisceration, splitting and the finlactic acid washed carcasses tevisceration or splitting. Accordcoli represented a stable endemabattoir.

Finally, in order to understand wpig carcasses, different papers inbelonging to the retrieved papers

Regarding scalding, all five studbacterial numbers (Wu et al., 202002). Namvar and Warriner (2water temperature of 65 °C. Warbut attributed this reduction to th

Singeing was investigated in onegeneral, the sampling of carcassconsidered study attributed to siscraping increased bacterial coun(Namvar and Warriner, 2002). reduce E. coli counts, but in additional operations such as pol

As regards pasteurisation, two preduce E. coli counts (Jones andescribed a reduction in E. coli cfrom a grid and the anal region. Ipolishing. In the second paper, tshould be noted that in this case,




the LRC and in this case, several operations wered in lower E. coli counts on carcasses (Bryant e

e numbers of E. coli and singeing did not reducerly stages of the dressing procedure to avoid th effectiveness of the treatment was impai

emoval. The pasteurisation process was then sever in this case also, the following operations (e

crease in bacterial counts. Finally, it was obserafter head removal and chilled without spraying h

ysed results obtained by swabbing routinely promonstrated the cooling process decreased E. coli co

n a study aimed at tracing the fate of E. coli withe changes in counts recovered from the briskanimals were sampled at 45 minutes intervals

ons: bleeding, scalding, scraping, evisceration, splowing bleeding, at the beginning of processing,

wards the end of the slaughterline. In particular,g but increased during scraping and were subsequwas sporadically recovered at low levels from

nal lactic acid washing. It is interesting to note thatested positive for E. coli compared to those

ding to genotype analysis, the authors suggested mic population, in particular belonging to the

which operation would be more effective in reducnvestigating the effects of slaughtering phases ws are summarized in Figure 5.

dies dealing with counts of Escherichia coli highl009; Namvar and Warriner, 2006; Rivas et al., 22006) in particular described a scalding process rriner et al. (2002) reported a decrease of E. coli che scalding that preceded this step.

e study that observed a decrease in counts (Warses after this stage was not performed in the retrngeing a reduction observed in carcasses after pnts according to two out of three studies, both deAccording to the same authors, the polishing this case carcasses underwent, after the previishing, washing and evisceration.

papers considering this step recognized it as annd Johns, 2012; Bryant et al., 2003). The first counts on both sampling sites investigated: a regiIn this case, carcasses were submitted to a pasteuthe pasteurisation process was applied after the the slaughtering process was performed in a pilo



hors.

38

re separately tested to t al., 2003). Dehaired

e them. Pasteurisation he degradation of cut ired by subsequent shifted to after head evisceration, splitting rved that chilled pig had acceptable levels

ocessed pig carcasses ounts.

hin a high throughput ket region of swine during two separate

plitting, and washing. were relatively high,

, E. coli counts were uently reduced during

m carcasses following at a greater number of e sampled following that the recovered E. holding area of the

cing E. coli counts on ere retrieved. Results

lighted a reduction in 2000; Warriner et al.,

of 5 minutes with a counts after scraping,

rriner et al., 2002). In rieved papers and the polishing. In contrast, escribed in one paper process was able to

ious sampling point,

n effective process to study, in particular,

on randomly selected urisation process after shaving operation. It t plant.

E.


The washing process was investled to significant results justifyinWarriner et al., 2002; Jones and underline that in the study perfoduring this step, but there was nonumber of lactic acid washed cevisceration and after splitting. Acoli over carcass surface. As prevevaluation of the slaughtering sequential processing stages alon

The final phase of the slaughterinbut within each study, the study reported bacterial counts beforedecrease E. coli counts (Nesbakchilling process was characteris2008). Two studies were not uobtained after chilling with coun(Tamplin et al., 2001) or after spin E. coli counts on carcasses a2000).

Figure 5: Trends (increase or deslaughterhouses.

0

1

2

3

4

5

6

7

8

Aft

er sc

aldi

ng

Aft

er d

ehai

ring

Aft

er si

ngei

ng

Aft

i

N° o

f stu

dies




igated in seven studies described in five differenng its use as decontamination treatment (NamvarJohns, 2012; Bryant et al., 2003; Gill et al., 2000

formed by Namvar and Warriner (2006), 1.5 % o significant effect on E. coli counts on carcassesarcasses tested positive for E. coli compared toAccording to authors this was probably due to viously stated, the described results were not obtastage, but were inferred from data belonging

ng the slaughterline.

ng operations is chilling. Five papers dealt with tdesigns were very different from each other. Twoe and after chilling: in both cases the chilling kken et al., 2008, Gill and Landers 2004). In onsed by a blast chilling phase (-21.9° for 70 minuseful to evaluate this stage because they compnts obtained at temporally/physically distant staglitting (Wu et al., 2009). The last retrieved paper

after chilling in only two out of eight plants inv

ecrease) of E. coli counts at different stages of th

Aft

er sc

rapi

ng

Aft

er p

olis

hing

Aft

er sh

avin

g

Aft

erpa

steu

riza

tion

Aft

erev

isce

ratio

n

Aft

er h

ead

rem

oval

Aft

er sp

littin

g



hors.

39

nt papers and it never r and Warriner, 2006; 0b). It is important to lactic acid was used

s. Moreover a greater o those sampled after a redistribution of E. ained from a punctual g from different, not

this slaughtering step, o studies in particular process was able to

ne of these cases, the n) (Nesbakken et al., pared E. coli counts

ges, i.e. after bleeding observed a reduction

vestigated (Gill et al.,

he slaughterline in pig

Aft

er w

ashi

ng

Aft

er c

hilli

ngNo effectIncreaseDecrease

E.


Enterobacteriaceae

One study (Spescha et al. 200analysed data belonging to diffecharacteristic of the first abattoirthere was a combined dehairingwater before chilling. In both abcarcass sites tested) and after chcounts decreased after singeing decreased after polishing.

Another study (Duggan et al. 201Salmonella presence. Results repof Enterobacteriaceae counts dur

Lenahan et al. (2006) also evalubelonging to four different plantwas not always demonstrated sconclusion was that chilling corelation to Decision 2001/471/EC

Hurd et al. (2008) evaluated thsamples and observed an increasevisceration. However, it shoulcomplex and had different aims, from the context. Moreover, samwere collected from bung and ple

The evisceration stage was evaluPDR processing sixty to eighty during evisceration (Inthavong ethygienic practices, if compared w

Warriner et al. (2002) evaluated tsingle pig carcass every 45-50 mtime of the day. The process wperformed dry before singeing chilling. The authors described aincrease in Enterobacteriaceae lecross-contamination operation.

The microbial contamination of paper (Rivas et al., 2000) dealingslaughtering process. The procestages before and after the siEnterobacteriaceae counts decreevisceration. According to authtechnique but to the singeing proof scraping. Interestingly, despobserved during evisceration.




06) providing data on Enterobacteriaceae aloerent stages in two abattoirs processing 250 and r was the use of blast chilling technique, while ig and singeing process. Both plants washed caattoirs, Enterobacteriaceae counts decreased aftehilling and increased after dehairing. Moreover,and slightly increased after polishing, while in t

10) evaluated the relationship between Enterobacported in graphics also showed a non statistically ring chilling.

uated the effect of chilling on Enterobacteriaceats. They observed an average reduction even if, since some carcasses showed a slight increase ould be an important step to improve the carcC.

he Enterobacteriaceae presence in terms of prese in prevalence of Enterobacteriaceae on carcald be noticed that the experimental design of so these results in terms of counts at different st

mpling sites on carcasses are quite different froeura during and after evisceration.

uated in a study performed in a low-throughput spigs per day. In this case the authors observed at al., 2006), but in general the situation was quite with other slaughterhouses.

the effect of slaughtering on Enterobacteriaceae minutes in order to also determine differences re

was characterised by scraping after scalding and and wet after it. Moreover carcasses underwe

a reduction in counts after scalding and singeing. Oevels was observed after evisceration, which was

pig carcasses in an Iberian slaughterhouse was g with variation of Enterobacteriaceae counts onss encompassed a dry and wet scraping techniqngeing phase, respectively, and anal closure

eased during scalding and scraping but increased hors, the unexpected decrease during scraping ocess that, as described above, was performed betpite anal plugging, an increase in Enterobacte



hors.

40

ong the slaughterline 160 pigs per hour. A in the second abattoir arcasses with potable er scalding (in all the , in the first abattoir, the second plant they

cteriaceae counts and y significant reduction

ae counts in carcasses at carcass level, this during chilling. The ass categorisation in

evalence in pools of asses during and after this study was quite tages are extrapolated om usual since swabs

slaughterhouse in Lao a significant increase different, in terms of

counts by sampling a lated to the slaughter by double polishing

ent a washing before On the other hand, an s identified as the key

the study object of a n carcasses during the que performed in two

before evisceration. during dehairing and was not due to this tween the two phases eriaceae counts was

E.


The effect of different operationalso shown in Figure 6.

As already observed for E. coli, studies (Spescha et al., 2006; Rivcritical to reduce initial bactermanaging of activity at plant leve

The same studies mentioned befof contamination in two out of thet al. (2006) determined the jointan increase in Enterobacteriaceathree different studies (Spescha eEvisceration is a key contaminat2002; Inthavong et al., 2006; Hubacterial contamination only in te

The last operation that, accordingis chilling. A decrease in EnterSpescha et al., 2006) out of four al., 2010). The fourth study (Dreported results of Enterobacteribacteria variations between differ

Figure 6: Trends (increase or slaughterline in pig slaughterhou

The counts of E. coli on small stages of the slaughterline ar

0

1

2

3

4

5

Aft

er sc

aldi

ng

Aft

er d

ehai

ring

Aft

er si

ngei

ng

N°

of st

udie

s




ns on Enterobacteriaceae counts are summarize

Enterobacteriaceae levels also decreased duringvas et al., 2000; Warriner et al., 2002). This step

rial counts and must be carefully considered el.

fore highlighted that the dehairing process was rehree studies. In particular, one study described in t effects of the singeing and dehairing phases toae counts. In contrast singeing reduced Enterobaet al., 2006; Rivas et al., 2000; Warriner et al., 200ion point according to four studies (Rivas et al., 2

urd et al., 2008); however Hurd et al. (2008) evalerms of prevalence in the tested pig carcasses.

g to the literature, results in notable variation of crobacteriaceae counts was described in three (Lselected studies (Lenahan et al., 2009; Spescha et

Duggan et al., 2010) describing variation in coiaceae counts only in graphics and did not analyrent stages.

decrease) of Enterobacteriaceae counts at difuses

ruminant carcasses described in the selected pare reported in Table 11 (Appendix D), w

Aft

er p

olis

hing

Dur

ing

evis

cera

tion

Aft

erev

isce

ratio

n

Afte

r tr

imm

ing

Aft

er w

ashi

ng

Aft

er h

ead

rem

oval



hors.

41

ed below; results are

g scalding in the four p of pig processing is in the planning and

ecognised as a source the paper by Spescha

ogether, and observed acteriaceae counts in 02). 2000; Warriner et al., luated the increase in

carcass contamination Lenahan et al., 2009; t al., 2006; Duggan et ounts during chilling yse in depth indicator

fferent stages of the

apers at the different while the counts of

rem

oval

Aft

er c

hilli

ng


E.


Enterobacteriaceae are presenteanalysed at each step, the increadeviation are reported when these


A total of 18 papers were collectswine carcasses. In particular, 1Enterobacteriaceae and two cons

As already mentioned for reviewwas hampered by different aspecwere not consistent across studiein terms of conclusions drawn an

Since some papers provided datfactor influencing bacterial couninstead of the papers. In Tables indicator bacteria.

E. coli

A total of 30 risk factors or trecarcasses at slaughterhouse in the

Only one paper considered theconsidered 12 treatment groupsversus 5), the feed type (pelleted to a 2×2×3 factorial design. Ttreatment, according to the statisthe thoracic area was feeding thcontrast, the highest counts werefast. A statistical difference wasanalysis. No correlation was fobetween E. coli levels on carcass2007).

Attention has been paid also to(2007) evaluated the influence oIn particular, pigs were dividedhandling. The rough treatment gentle handling was characteriseNeither handling nor group dimeof differences between groups scarcass microbial quality.

The effect of the plant throughcompared four low throughputpresumptive E. coli (74 %) repopigs/year) was higher comparedyear). These differences were staevaluation of other bacteria, an




d in Table 12 (Appendix D). In these tables, thase or decrease of the counts for each step, the me data were available.

ted dealing with risk factors potentially influencin11 papers described studies reporting data on E.sidered both bacteria.

w question 1, the comparability of data providedcts, such as the sampling method and the unit ofs. Hence, different studies considering the same f

nd not in terms of counts reported.

ta on both types of indicator bacteria, or considnts, it was more convenient to consider the indiv

4a and 4b, the risk factors investigated are sum

eatments were investigated in relation to Eschere collected papers.

e influence of risk factors at farm level. This and different conditions regarding the numbervs mash) and the fasting time before slaughter (4

The slaughter process took place in an experimstical model performed, which resulted in the lowhe pigs pelleted feed five times a day followee recovered from pigs fed mash, five times a das observed between these two treatments with

ound between stomach weight and E. coli counses and the weight of faeces collected on the truck

the management of the animals before slaughtof handling and group dimension on microbial cad in groups of 10 and 30 individuals subjected consisted in a quickly handling with electric p

ed by slow handling managed with the use of a pension significantly influenced E. coli counts onsuggested that the stress condition applied had

hput on bacterial counts was investigated by Ht with four high throughput slaughterhouses. orted on pig carcasses in high throughput plantsd to the percentage (58 %) in low throughput patistically significant. However, since the study w

nd these did not significantly differ among slau



hors.

42

e number of samples mean and the standard

ng bacterial counts on . coli, five papers on

d by different studies f enumeration, which factor were compared

dered more than one vidual studies (trials),

mmarised for both the

richia coli counts on

s study in particular r of daily feeding (2 4, 14, 24 h) according mental abattoir. The wer E. coli counts on d by a 24 h fast. In ay followed by a 4 h an a posteriori Tobit nts on carcasses and k floor (Saucier et al.,

tering. Rabaste et al. arcass contamination. to rough and gentle

pod; on the contrary, plastic board or whip. n carcasses. This lack

a limited impact on

Hansson (2000), who The percentage of s (250,000-1,300,000 plants (450-800 pigs was aimed also at the ughterhouses, authors

E.


concluded that level of carcasslaughterhouse.

In contrast, another study dealiCanada found no relation betwebacterial counts (Bohaychuk et samples was reported. Most of cfu/cm2.

The effect of the surveillance sysinfluencing the counts of indicaconducted in 7 plants located incarcasses (mean log of positive scoli on carcasses processed viaabattoirs that operate the tradition

In another paper, Lindblad et al. Swedish slaughterhouses. Both Ebaseline value, but only E. coli which slaughtering was performcarcasses were obtained in the pperiod, so thus the seasonal varia

Delhalle et al. (2008) studied secarcasses in the 10 largest Belgvariables through a questionnaire(univariate mixed logistic regresmodels with microbiological datparticular that the use of water resulted in lower E. coli counts.was considered hot by operators practice of refreshing animals aftwere spraying 75 % of the timtemperature. According to the acarefully. Moreover, the E. colprocessing time between killingcontamination, were a scalding splitting machine three times a da

Hamilton et al. (2009) evaluatecarcasses. The study focused in two different high throughput abwith hot water (15 sec with wawater (SANOVA solution) wersampled using a meat excision teabattoir was significantly and compared to the control. This recontamination. Further, the twoalthough SANOVA treatment iregarding the use of hot water, reduce E. coli levels on slaughter




ss contamination could not be related to the

ing with plant throughput and investigating 50een abattoir throughput (8000 pigs/year was thal., 2011). An overall prevalence of 33.7 % Ethe carcasses (92.7 %) had levels of E. coli l

stem in place in slaughterhouses was investigatedator bacteria. For this scope, a microbiologicaln New South Wales, Australia. E. coli was detesamples was -0.23 cfu/cm2). This survey demonsta the co-regulatory system were comparable to nal system overseen by government inspectors (B

(2006) described the results of a microbiologicalE. coli and Enterobacteriaceae were taken into data have been analysed in order to study the efmed. Authors observed that the highest E. coleriod June-July; however, the study was conduct

ation has not been confirmed for data belonging to

everal managerial factors that could influence mgian slaughterhouses. In particular, they investie, data from which were then incorporated into tssion and multivariate mixed linear regression) aa obtained from a total of 584 carcass samples. Rduring lairage cleaning and a high frequency o

. In contrast, spraying of live animals when the was correlated with an increase in carcass E. col

fter transport is common. Other possibilities conteme during lairage and spraying automatically inauthors, however, this conclusion about sprayingli counts on carcasses increased proportionally

g and scalding. In contrast, protective factors, wprocedure using steam instead of immersion, thay and the change of carcass hooks before chillin

d the effect of decontamination treatments on Eparticular on the effect of two decontamination tbattoirs over three days. In each trial, two treatmater 81.9-83.5 °C) and one performed with acidire compared with the standard hygiene procedechnique at belly level. The prevalence of E. colisubstantially reduced for both hot water and

esult was also achieved if one abattoir had higheo treatments did not differ significantly. It shois not approved according to EU legislation, and thus, this could be an effective efficient altred pig carcasses.



hors.

43

e throughput of the

0 slaughterhouses in he cut-off value) and E. coli-positive swine lower or equal to 10

d as a potential factor l baseline study was ected in 63 % of pig trated that levels of E.

those established in Bass et al., 2011).

l survey performed at account to assess the ffect of the season in li counts on the pig ted during a one year o other years.

microbial status of pig igated different plant two statistical models and they then fed the Results highlighted in f lairage disinfection external temperature

li contamination. The emplated in the study

n relation to external g must be considered y with the length of

which reduced E. coli he disinfection of the ng.

E. coli counts in pig techniques applied in

ments, one performed ified sodium chlorite

dure. Carcasses were i on carcasses in each SANOVA treatment

er levels of microbial ould be noticed that no restriction exists

ternative operation to

E.


Different treatment conditions ostudy reported data aimed at evE. coli counts on pig carcasses. throughput slaughterhouses in Wthere was a reduction in E. cowashing treatment, it was foundwater temperature applied. In franging from <12.8 °C to >32.2 days) effectively reduced E. coli

In contrast, another study investeffective lowering of E. coli levnormally processed (anal plugManufacturing Practices (GMP)carcasses (which were processedevisceration to avoid intestinal rwater at the end of the slaughteduring evisceration in order to reuse of carcass washing (Rivas et

Since tissues exposed during theslaughtering, it is required to triminvestigated the effect of this prabefore and after their applicationreduce the numbers of E. coliineffective for that purpose proba

The technique of trimming ccontamination was investigatedcontaminated site on the microbsites and a random site on the cwas investigated through the pharboured high level of bacteriaprocessing. According to the aufound to be effective in decreasinselected ones.

Conter et al. (2006) investigatedwhich a prototype steam decoslaughter process. This treatment




occurring during slaughter were investigated byvaluating the effect of different risk factors towSamples were collected before washing and afte

Wisconsin (Algino et al., 2009). Between the selecli levels on both skinned and unskinned carca

d to be always effective in reducing E. coli counfact, in this study, washing was performed with

°C in different trials. Also both the tested chillelevels on the carcasses.

tigating the effect of pig carcass washing failedvels. More specifically, the authors divided carcagging was routinely applied before eviscerat) applied at this step to avoid intestinal ruptu

d with particular attention to the Good Manufacturuptures) and washed GMP carcasses (25 sec oerline). Results highlighted the effectiveness ofeduce of E. coli counts. In contrast, no effect waal., 2000).

e sticking of pigs are likely to be contaminated m the tissue around wound before evisceration. Giactice as well as the effect of pasteurisation by ann. The authors concluded that while the pasteurisai on the cut tissue of sticking wounds, the trably because of contamination during the routine

ontaminated sites on pig carcasses detained d. One paper in particular described the effebial level of carcasses by sampling on both the scarcasses (Gill and Landers, 2004a). Moreover, previously mentioned sampling scheme. Visiblya and were considered potential sources of bacteuthors, the trimming technique, as well as the cng E. coli counts on contaminated carcass sites as

d the effect of pasteurisation treatment in an Italintamination unit was installed before eviscerat provided a significant reduction of E. coli count



hors.

44

y several papers. One ward the variations in er chilling in 10 low-cted sampling points,

asses. As regards the nts, regardless of the h water temperatures er hold times (1 or 2

d to demonstrate any asses in three groups: tion), but no Good

ures; unwashed GMP uring Practices during of high pressure cold f implementing GMP as associated with the

with bacteria during ill and Badoni (2001) nalysing these tissues ation of carcasses can rimming is probably operation.

because of visible ect of trimming the specific contaminated the effect of chilling y contaminated sites

erial spreading during cooling process, were s well as on randomly

ian slaughterhouse in ation during the pig ts on pig carcasses.

E.


Table 4a: Slaughterhouse and fadecrease/increase/or with no effe

Effect Risk factor/Treatment

Decrease Acidified sodium chlorite Chilling Chilling contaminated site

Cleaning and disinfection times a day

Frequency of lairage disin

Hot water carcass washing(83.5°C)

Lairage cleaning with wat

Low throughput slaughter

Meal frequency 5, Feed ty(pelleted), fasting time ( 2

New hooks for carcasses bchilling

Pasteurized sticking woun

Scalding with steam

Steam treatment (before pasteurization)

Trimming of the contamin Unwashed+GMP Washing Decrease Total Increase June-July Skinned

Spraying when ext temp isconsidered hot

Time between killing and Increase Total

No effect Co-regulatory system

Gentle handling

Low throughput slaughter

Small group (10 animals) Sticking wounds trimming Washed+GMP Washing T No effect Total Total

* Risk factor referring to farm managem




arm related risk factors investigated in the retrievect on E. coli counts on swine carcasses

Compared factor/ Control Reference

washing Standard hygiene procedure Hamilton etNo chilling Gill and LanPrewashed carcass Algino et al

e No chilling Gill and Lanthree Others regimen Delhalle et

nfection Other frequencies Delhalle et g Standard hygiene procedure Hamilton et

ter Other cleaning solutions Delhalle et

rhouses High throughput slaughterhouses Hansson 20

ype 4 h)

Meal frequency (2; 5), Feed type (pelleted; mash), fasting time (4; 14; 24 h)

Saucier et a

before Same hooks Delhalle et

nds Not pasteurized sticking wounds Gill and Ba

Immersion scalding Delhalle et

No steam treatment Conter et al

nated site No trimming Gill and LanUnwashed Rivas et al. Prewashed carcass Algino et al Other months Lindblad et Unskinned Algino et al

s Spraying with others rules Delhalle et

scalding Continuous variable Delhalle et Governmental inspection system Bass et al. 2

Rough handling Rabaste et a

rhouses High throughput slaughterhouses Bohaychuk

Large group (30 animals) Rabaste et ag Not trimmed sticking wounds Gill and Ba

Unwashed+GMP Rivas et al. Prewashed carcass Algino et al

ment; ** Risk factor referring to pre slaughter operations



hors.

45

ved studies leading to

N° of studies

t al. 2009 (AU) 1 nders 2004 (CA) 1 l. 2009 (US) 1 nders 2004 (CA) 1

al., 2008 (BE) 1

al., 2008 (BE) 1

t al. 2009 (AU) 1

al., 2008 (BE) 1

000 (SE) 1

al. 2007 (CA)* 1

al., 2008 (BE) 1

doni 2001 (CA) 1

al., 2008 (BE) 1

l., 2006 (IT) 1

nders 2004 (CA) 1 2000 (ES) 1

l. 2009 (US) 1 17

al. 2007 (SE) 1 l. 2009 (US) 1

al., 2008 (BE) 1

al., 2008 (BE) 1 4

2011 (AU) 1

al. 2006 (CA)** 1

et al. 2011 (CA) 1

al. 2006 (CA)** 1 doni 2001 (CA) 1 2000 (ES) 1

l. 2009 (US) 1 7

28

E.


Enterobacteriaceae

A total of 19 papers were rEnterobacteriaceae counts on pig

Zweifel et al. (2007) evaluated throughput Swiss abattoirs withwere collected from 750 carcassmall subset of microorganisms.comparison (log of the total nimportant difference between carof positive sites on carcasses ranuseful indicator of abattoir-specif

The influence of plant features oal. (2011), who investigated twofrom randomly selected pig carcskin microflora was calculated itransfer of incoming microbEnterobacteriaceae load was sigthat, despite the EU process hygdiscriminating between the prochygienic according to the compaevaluated in the study was sugge

Both papers described the impacother studies tried to demonstramicrobial contamination.

The use of anal plugging prior throughput abattoir (Purnell et abacon market and were consEnterobacteriaceae counts arounwas observed, while slight non sto the authors, the results, togethleakage can occur during these reduce pig carcass contamination

Different risk factors were invesskinning, washing temperature an(Algino et al., 2009). In each slother halves of the correspondindecreased in both skinned and leaving the carcasses unskinnEnterobacteriaceae. Washing wawater temperature ranged from <supposed that the highest water and underlying tissue thus enhaholding times (1 or 2 days) wasswine carcasses.




retrieved dealing with risk factors or treatmg carcasses at slaughter.

the microbiological contamination of pig carcash a median annual slaughter of 600 animals. In sses. Authors noticed that Enterobacteriaceae a Because log normality was not ensured, log N number recovered per square centimetre). Rercass contamination within the selected abattoirsnging from 9 % to 22 %. Thus Enterobacteriacefic hygienic weak points in low throughput plants

on microbial carcass contamination was also studo different abattoirs (A and B), through the collecasses. In particular, the ratio between carcass in order to assess the process with respect to its bial loads onto dressed carcasses. In bognificantly reduced during processing. However, giene criteria (PHC, based on daily mean log valcesses in the two abattoirs, the process in plantarison of individual carcasses. Consequently a PHsted.

ct of slaughterhouse characteristics on the proceate which specific stages and factors exerted d

to scalding and dehairing was investigated in al., 2010). These carcasses were produced for thesequently not singed after dehairing. A signd the anal areas of unplugged carcasses after scsignificant decreases were observed on plugged

her with visual observation during the work, clearprocessing stages and that methods avoiding i

n.

stigated in another study in which the authors connd chilling duration in 10 very low-throughput alaughterhouse, carcass halves were sampled befng carcasses were sampled after chilling. Entero

unskinned carcasses between pre-washing anned during chilling was associated with lowas effective in reducing Enterobacteriaceae leve<12.8° to 32.2 °C, but not when temperature watemperature resulted in the least chilling and co

ancing bacterial attachment. Also, chilling for bs effective in significantly reducing Enterobacter



hors.

46

ments in relation to

sses in different low-total, 3,000 samples

accounted only for a values were used for

esults highlighted an s, with the percentage eae was considered a s.

died by Blagojevic et ection of 100 samples (before chilling) and ability to reduce the

oth slaughterhouses, the authors observed

lue for carcasses) not t B was clearly more HC based on the ratio

essing hygiene, while deleterious effects on

a EU licensed low-e pork rather than the nificant increase in

calding and dehairing carcasses. According

rly showed that faecal it could be useful to

nsidered the effect of abattoirs in Wisconsin fore washing and the obacteriaceae counts

nd chilling; however, wer mean levels of els on carcasses when as >32.2 °C. Authors ontraction of the skin oth the tested chiller riaceae levels on the

E.


Trivedi et al. (2006) evaluated tfor reducing naturally occurring or very low throughput abattoirsbefore any organic acid treatmeeffect on Enterobacteriaceae levcold storage. The steam treatmlocations with no differences in th after the application of steamcleaning systems as a critical con

The effect of GMP and washinauthors divided carcasses in thrbefore evisceration, but GMP w(which were processed with pevisceration) and washed GMPslaughterline). Interestingly, reevisceration thus demonstrating tlevels when GMP are implemEnterobacteriaceae levels after t2000).

Finally, Tomovic et al. (2011) epopulations. Two regimens of rathen 2-4 °C until 8 h post-mortethen carcasses were left at 2-4 °CEnterobacteriaceae counts in a decrease them. Authors concludea rapid chilling treatment is impl

Tables 17 and 18 (Appendix E) rwith the effect of the slaughterhslaughtering line) on the indicato




the effectiveness of a commercially-available dobacterial populations on freshly slaughtered pig

s. A 60 second steam treatment was applied afteent on three anatomical sites on the right half ovel was evaluated immediately after the treatmenent significantly reduced microbial contaminatiterm of mean Enterobacteriaceae population imm

m. The authors thus suggested the use of housentrol measure in low and very low throughput mea

ng during processing was investigated by Rivas ee groups: normally processed (anal plugging w

was not implemented during evisceration); unwaparticular attention to the Good Manufacturi

P carcasses (25 sec of high pressure cold wateesults showed the effectiveness of implemethat anal plugging could be more efficient agains

mented. In contrast no significant effect was othe addition of the washing treatment during pro

evaluated the effect of rapid chilling of swine capid chilling were applied. Treatment 1 consistedem, while the first phase of the second treatmenC for 24 h. Interestingly, results highlighted that t

highly significant way; in contrast treatment 1 ed that rapid chilling can improve the microbiologemented together with a 24 h duration of the entir

report E. coli and Enterobacteriaceae counts fromhouse and farm related features (general and at dor bacteria on pig carcasses.



hors.

47

omestic steam cleaner carcasses in four low er the final wash and of the carcasses. The nt and 24-h later after ion at all anatomical

mediately after and 24 hold domestic steam at processing plants.

et al. (2000). These was routinely applied ashed GMP carcasses ing Practices during er at the end of the enting GMP during st Enterobacteriaceae observed on carcass

ocessing (Rivas et al.,

carcasses on bacterial d of 3h at -31 °C and

nt was equivalent, but treatment 2 decreased did not significantly

gical safety of pork if re chilling operation.

m the studies dealing different stages of the

E.



Effect Treatment/R

Decrease Anal plugging

GMP appliedevisceration truptures

Chilling time

Chilling time+ 24 h 2-4 °C

Steam decont Washing T° (Decrease Total

Effect Plant features

Plant featuresEffect Total

No effect Chilling time31 °C+ 8 h 2-

Washing No effect Total Total

2.5. Small ruminants


Among the retrieved papers, 21regarding the three review questi

Half of the papers (11) were con(2), Canada (1), Brazil (1), Morothe exception of one paper; ten and three papers considered both

Four papers provided informatioone paper dealt with review ques

Twenty out of twenty-one papepapers, samples were collected fr

The most common sampling methe unit of enumeration mostly u

Alternative plate count was the aby Petrifilm (seven papers).

Concerning the sampled area, it surface of the sampled area range

Table 8 (Appendix C) reports gequestions (1, 2, 3) for which each




ated risk factors investigated in the retrievedect on Enterobacteriaceae counts on swine carcas

Risk Factor Compared factor/Control Ref

g No anal plugging Purnell et ald during to reduce No such GMP applied Rivas et al.

Pre-chilling Algino et al (3h -31°C

C) Pre chilling Tomovic et

tamination No steam treatment Trivedi et al(< 32.2 °C) Pre-washing Algino et al

s Another plant Blagojevic e(SRB)

s Different plants Zweifel et a

(3 h --4 °C) Pre-chilling Tomovic et

Unwashing Rivas et al.


1 papers were eligible in order to provide dataions.

nducted in Europe, while the remainder were fromocco (1) and India (1). All papers described obserpapers provided data on E. coli, eight papers on

h indicator bacteria.

on on review question 1, thirteen papers addressestion 3, while three papers answered all three revi

ers described studies that analysed pooled samrom two to five regions of the carcass.

ethod was swabbing (16 out of 21 papers), follosed was log cfu/cm2 (15 out of 21 papers).

analytical method most frequently used (10 out of

should be pointed out that a wide variability amoed from 5 to 100 cm2.

eneral information concerning the selected 21 ph paper provided pertinent data.



hors.

48

d studies leading to sses

ference N° of studies

l. 2010 (UK) 1

2000 (ES) 1

l. 2009 (US) 2

al., 2011 (RS) 1

l. 2006 (US) 1 l. 2009 (US) 4

11 et al. 2011 1

al. 2008 (CH) 1 2

al., 2011 (RS) 1

2000 (ES) 1 2 15

a on small ruminants

m Australia (4), USA rvational studies with n Enterobacteriaceae

ed review question 2, ew questions.

mples; in most of the

owed by excision and

f 21 papers) followed

ong studies exists; the

papers and the review

E.



The six papers providing informacarcasses along the slaughterindifferent steps of the slaughterlicattle (see Figure 2, paragraph 2and non-penetrative) with a captfollowed by bleeding in a hangin

An important step takes place afttypes of dressing can be used acccarcasses are shackled by first onskinned. Then the carcasses areforequarters and evisceration. In for pelt removal; head and feet arflaying completed. An alternaticarcasses being skinned while adopted only after skinning is improvement compared to the pskinning and reduces the contam

After dressing, evisceration takmaterial is separated and disposchilling, standing in a tunnel foBefore the chilling step, carcasse

In some slaughterhouses, washinwashing is to remove visible soilSoiled carcasses should be spraysoiling material dries, thus minim

Along the slaughterline, the uncwhile the clean operations are ev


Among the six papers, three Enterobacteriaceae. Concerning which samples were collected apapers reported data on carcasse(i.e. chilling). The retrieved papers provided slaughter processes described arused vary considerably. For thesthe different studies, as it wouldtrend of the indicator bacteria cothe studies, according to the coneffect in terms of decrease, increand Enterobacteriaceae counts o





ation on the presence and counts of E. coli and Eng process seldom reported information about ne. The slaughter procedure for small ruminants2.2.2), and it commonly starts with mechanical tive bolt pistol. Electrical stunning is increasing

ng position.

fter bleeding: the dressing (or skinning or flayingcording to the features of the abattoir: in conventne then the other hind leg while the unshackled he suspended from a gambrel by both rear legslow-throughput slaughterhouses, the carcass is l

re removed and dressing initiated. Then the carcaive type of process, referred to as inverted drsuspended by the forelegs, with suspension bycompleted. This second type of dressing is u

previous one, due to the fact that it reduces themination of the carcasses.

kes place: offals are separated, inspected and csed of in a sanitary manner. Finally, the carcasor 2 h at 2-4 °C and then stored at 4±1 °C (coes can be split and cut.

ng is an additional step before chilling: the primling and bloodstains and to improve carcass appeyed/washed with potable water immediately aftemizing the time for bacterial growth.

clean operations, therefore, include stunning, blevisceration, carcass splitting and carcass dressing.

papers provided data on E. coli counts anboth the indicator bacteria, five out of six papers

at more than one step along the slaughterline, wes sampled immediately before and after one step

data that were not always comparable due to re rather heterogeneous, the sampled area and thse reasons, it is not appropriate to provide a direcd not be meaningful. Rather, it seems more informontamination at the different stages of the slaughtnclusions drawn by the authors. Figures 7 and 8 sease or no effect of the different stages of the slaon small ruminant carcasses, as reported in the elig



hors.

49

Enterobacteriaceae on the features of the

s is similar to that of stunning (penetrative

gly an option. This is

) operation. Different tional dressing, sheep hind leg and rump are s for skinning of the lowered onto a cradle ass is raised again and ressing, involves the y the rear legs being sually considered an e labour required for

cleaned. Condemned sses are submitted to onventional chilling).

mary object of carcass earance after chilling. er dressing before the

eeding and dehiding,

nd three papers on s performed studies in while one out of six p of the slaughterline

several reasons: the he analytical methods ct comparison among mative to analyse the terline within each of show respectively the aughterline on E. coli gible studies.

E.


Figure 7: Trends (increase and small ruminant slaughterhouses

Figure 8: Trends (increase andslaughterline in small ruminant s

As far as E. coli counts are concsamples from the brisket area (10

0

1

2

3

Afte

rsk

inni

ng

Afte

r

N°

of st

udie

s

0

1

2

3

Afte

rsk

inni

ng

N°

of st

udie

s




decrease) of E. coli counts at different stages o

d decrease) of Enterobactreiaceae counts at dilaughterhouses

cerned, one study performed by Hauge et al. (2000 cm2) of 35 lambs at two different points of the

Afte

rev

iscer

atio

n

Afte

rw

ashi

ng

Bef

ore

chill

ing

NID

Afte

rev

iscer

atio

n

Afte

rw

ashi

ng

Afte

rch

illin

g



hors.

50

of the slaughterline in

ifferent stages of the

011b), collected swab e slaughterline: at the

No effect IncreaseDecrease

No effect IncreaseDecrease

E.


beginning (after fleece removal) in order to investigate the microbcounts expressed as log cfu/100 the end of the slaughter process,of contamination of carcasses aslaughterline, just before chilling

The microbial load of sheep/goa(Bhandare et al., 2007), in whslaughterline: after flaying, after sites of the carcass (neck, shouldand mean counts were reported aflaying; 21 % and 3.9 log cfu/cm

Finally, differences in the levelsslaughterline but also at differenswab samples from an area of 2after skinning, evisceration and wforequarter (0.31 log cfu/cm2) agradual reduction (not significan0.31 after skinning to 0.07 log cf

Concerning Enterobacteriaceae represented an important point iYalcin et al. (2003) collected samareas of the same carcass at fourwashing and after chilling. AltoEnterobacteriaceae counts, lead(after dressing); 0.75 (after evisstated that even if not significanreduction of Enterobacteriaceae the most important steps in impro

The same conclusion about the which swab samples were takeabattoirs in Ireland. In this case lateral thorax, breast, brisket) an51 % of carcasses tested, counts26 %. This may be due to re-grow

A third paper by Milios et al. (2the process. Swab samples fromAfter pelt removal of hind and fevisceration steps (C- After echilling). The study showed subsequent step after the first saThe recorded counts (log cfu/cmsampling point C-2.68; samplingremoval was the determinant stathe lamb slaughterline.




and at the end (before chilling). The two samplinbial transfer from fleece to the carcass surface ducm2, had the following values: 1.78 and 2.71 at

, respectively. The authors concluded that there ialong the slaughtering process and that the coung were significantly (p<0.05) higher than at skinni

at carcasses along the slaughterline was investigahich swab samples were taken at three differewashing and after evisceration. Samples were co

der, flank, rib, brisket, rump) and analysed in pooat the three different sampling points: 11 % and

m2 after evisceration; 8.3 % and 3.1 log cfu/cm2 aft

s of indicator bacteria can be found not only alnt sampling regions of the carcass. Martineli et20 cm2 of 30 carcasses from the forequarter andwashing processes. The highest value of E. coli and the lowest on hindquarter (0.03 log cfu/cmntly different) of these counts was observed in tfu/cm2 after washing).

counts, two out of three papers concluded thin reducing the load of these indicator bacteria omples (by the excision technique - 10 cm2 of sur different slaughterline stages: after dressing, aftogether, 176 samples were analysed in pools inding to the following mean values (expressed asceration); 0.58 (after washing) and 0.11 after c

nt (p>0.05), evisceration brought an increase, whcounts. However, the study concluded that chill

oving the hygienic quality of carcasses.

chilling step was drawn by another study (Lenen from lamb carcasses before and after chillalso, samples were collected from several areas

nd analysed in pools. Chilling reduced Enterobas remained unchanged on 23 % of carcasses, whwth or contamination just before or during the chi

011), investigated the hygienic status of a lamb m different sites of the carcass were collected dur

forelegs/ before pulling; B- After pulling/before evisceration/before pluck removal; D- After p

that Enterobacteriaceae counts progressivelyampling point and evisceration contributed most

m2) were the following: sampling point A-0.76; sag point D-2.90. In particular, the stage after evisage for the prediction of the carcasses contaminat



hors.

51

ng points were chosen uring skinning. E. coli

the beginning and at is an increasing level nts at the end of the ing.

ated by another study ent points along the llected from different

ols. E. coli prevalence 3.5 log cfu/cm2 after

fter washing.

long the steps of the t al. (2011) collected d hindquarter surface counts was found on

m2) after skinning. A the forequarter (from

hat the chilling step on carcasses. In fact, rface area) from four

fter evisceration, after n order to determine as log cfu/cm2): 0.38 chilling. The authors

hile chilling brought a ling represents one of

nahan et al., 2010) in ing in five different of the carcass (flank,

acteriaceae counts on hile they increased on illing process.

slaughterhouse along ring pelt removal (A- evisceration) and the

pluck removal/before y increased in each tly to the final count. ampling point B-2.27; ceration/before pluck tion rate at the end of

E.


According to the retrieved paperecorded by three out of six papcaused increases of E. coli and Eof the slaughterline can have anbacteria can change according tohad a higher level of E. coli comp

Along the slaughterline, the mosthe chilling step: two papers conreduce the counts, thus being a co

Finally, concerning the washing washing had no effect in reducinthe chilling step.

The counts of E. coli on small stages of the slaughterline arEnterobacteriaceae are presenteanalysed at each step, the increadeviation are reported when these


The aim of this section is to management and slaughter charaloads on small ruminant carcasse

The great variability among the different studies and to the coindicator organism considered, thwere carried out and the specifbarely comparable.

According to the defined searchdealing with risk factors were obacteria, or considered more thaconsider the individual studies (tr

Altogether, 19 studies related to small ruminant carcasses were de

Studies were divided accordingresulting in a total of 11 studies dwhile ten studies examined slaugcounts and risk factors, eight stud

As already mentioned for reviewwas hampered by several aspectwere not consistent across studiein terms of conclusions drawn an




ers, an increasing level of contamination along tpers; in particular, the skinning and the eviscera

Enterobacteriaceae. Furthermore, it seems clear thn influence on counts, but on the other hand, tho the sampled region. In fact, one paper reportedpared to the hindquarters.

st effective point in order to reduce the microbiancluded that this phase seems to be the most effecontrol point along the slaughterline.

step, the results are not clear or unanimous; accong the counts, while in another study it reduced

ruminant carcasses described in the selected pare reported in Table 13 (Appendix D), wd in Table 14 (Appendix D). In these tables, th

ase or decrease of the counts for each step, the me data were available.

describe results of the papers investigating acteristics, as well as other factors, on E. coli andes.

selected papers is due to the multitude of aspeomplexity of the slaughterline. Moreover, othehe sampling and analytical methods used, the settfic steps of the slaughterline investigated, rend

h process and the established eligibility criteria,btained. Since some papers provided data on bo

an one factor influencing bacterial counts, it wasrials), instead of the papers.

factors which could potentially influence indicatescribed in the retrieved papers.

g to the indicator bacteria considered and thedealing with E. coli. Of these, one study focused oghtering techniques or annual season. Concerningdies investigated the slaughtering techniques.

w question 1, the comparability of data providedts, such as the sampling method and the unit ofs. Hence, different studies considering the same f

nd not in terms of counts reported.



hors.

52

the slaughterline was ation steps frequently hat the different steps he levels of indicator d that the forequarters

al contamination was ctive point in order to

ording to two studies, E. coli counts before

apers at the different while the counts of

e number of samples mean and the standard

the effects of farm d Enterobacteriaceae

ects considered in the er variables like the ting where the studies er the available data

a total of 16 papers oth types of indicator s more convenient to

tor bacteria counts on

e factor investigated, on batch information, g Enterobacteriaceae

d by different studies f enumeration, which factor were compared

E.


Among the retrieved papers, onHauge et al. (2011b) investigatregimes as: unshorn lambs; lambfarm three days before slaughtersamples collected from the briskand before chilling) led to the fol

- at the skinning point, Eslaughter compared with

- lower levels of E. colskinning; at this samplin

- increasing contaminatreduction of E. coli on shthat E. coli numbers on send of the slaughterline.

Consequently, the application ofattention to slaughter hygiene dimplemented and monitored.

Among factors related to the slalymph nodes for caseous lymphpapers.

As far as the season was concernE. coli counts on carcasses in thstudy swab samples were collectthree of the six plants applied oret al., 2006), raw meat sampledifferent sampling periods: a hot The authors concluded that seaso

Concerning the microbial contamadult sheep carcasses for caseoufrom two regions (rump and scashowed that the contribution ofEnterobacteriaceae) was statisticfor CLA had a detrimental impacimprovement in live animal manslaughter could improve the hygi

Three studies investigated the effcarcasses. In an Australian studywere collected from differingslaughterhouses and Low-Througmicroorganisms including genercarcasses from LTSs (21.4 %) coA similar conclusion was drawcarcasses after chilling from mefrequently on carcasses from LT




nly one study dealt with factors related to the ted E. coli counts on carcasses associated witbs shorn in the abattoir immediately before slaugr; lambs shorn on farm seven days before slaug

ket area at two points of the slaughterline (after rllowing conclusions:

E. coli counts were higher on the carcasses shornh those on farm;

li were detected on shorn lambs compared to ug point, shearing was effective in reducing E. col

tion of carcasses was recorded along the slauhorn lambs at the skinning point was negated furtshorn lambs were not significantly different from

f a shearing regime can be unsuccessful if other cduring the process (especially during evisceratio

aughterhouse, three factors, the effect of the seashadenitis and the plant features were investiga

ned, Duffy et al. (2001) did not find any differenche spring versus the fall or winter seasons. In tted from six different lamb slaughterhouses at 24rganic acid rinse to carcasses before chilling. In aes were collected from several slaughterhouses

season (April to September) and a cold season (Non had no effect on the recorded E. coli counts.

mination due to the handling of lymph nodes durus lymphadenitis (CLA), Jordan et al. (2012) colapular region) of carcasses before and after inspf inspection to higher counts of both indicator cally significant (p<0.001). In fact, the routine inct on carcass hygiene; curtailed inspection procednagement and the use of data regarding the occienic status of carcasses.

ffect of slaughterhouse throughput on E. coli couny (Phillips et al., 2001), swab samples of sheep cag throughput slaughterhouses: domestic slaughput Slaughterhouses (LTSs). The samples wereric E. coli counts, which were found to be lessompared with domestic (32.7 %) and export (35.2

wn by Sumner et al. (2003), who analysed samedium and low-throughput slaughterhouses. E. cTSs (18.5 %) compared with medium throughpu



hors.

53

batch characteristics. th different shearing

ghter; lambs shorn on ghter. Analysis of the removal of the fleece

n immediately before

unshorn lambs at the li loads on carcasses;

ughterline; thus, any ther along the line, so unshorn lambs at the

control measures, like on), are not properly

son, the inspection of ated by the retrieved

ces (p>0.05) between the above mentioned 4 hours after chilling; another study (Cohen

in Morocco in two November to March).

ring the inspection of llected swab samples

pection. Data analysis bacteria (E. coli and

nspection of carcasses dures coupled with an currence of lesions at

nts on small ruminant arcasses after chilling ughterhouses, export e analysed for several s prevalent on sheep 2 %) slaughterhouses. mples collected from coli was detected less ut (61.5 %) abattoirs.

E.


However, within LTSs there wcarcasses. Finally, another Austrorder to examine carcasses fromthus differing in terms of compleof prevalences (from 8.7 % to 8taken alone, the "snapshot" of thintegration of each plant's micrprocess elements should be comb

Among the retrieved studies, oslaughterhouse throughput (low lamb carcasses after washing/bcarcasses from the low-throughpE. coli counts compared to thosehigh-throughput slaughterhouse between the two slaughterhousesIn fact significant variations wethroughput slaughterhouse, with versus 3.18 log cfu/cm2 in winterslaughterhouse, lower counts weautumn).

Eight papers studied the influslaughterline on indicator bacteri

The results reported in differenprocessing operations. In particubacteria will be discussed: theprocedure; skin-on versus convetreatment.

Concerning the application of a peffect of this treatment on gocontamination on skin/hides of athe treated group (1 min spray wdid not significantly decrease dur

The type of dressing was investicarcasses in two studies. Gill et adressing might reduce of 1.5 log2002) comparing different types method) found that the Hybrid significantly lower EnterobacterThus, the greatest reduction in mit minimized hand contact during

The microbiological status of conventionally dressed carcasseproduction of ruminant carcassestatutory requirements. Howeverorganoleptic qualities) by severawell occur in other European




was a considerable in-plant disparity between thralian study (Bass et al., 2011) conducted a micr

m slaughterhouses that processed a broad range oexity of construction and processing. The results 9.5 %) of E. coli among the different slaughterhhe slaughterhouse may be misleading. Those au

robiological data with information involving livbined in order to be effective.

one study (Feizullah and Daskalov, 2010) comor high throughput) with the season. Samples

efore chilling. Concerning E. coli counts, the put slaughterhouse during the spring had significae collected from the low-throughput slaughterhouin the spring. With regards to Enterobacteriaceas were significant (P<0.001) for the autumn, wintere reported between the spring and the winter lower counts found on carcasses in spring (1.30

r); between the winter and the autumn seasons forere found in winter (1.27 cm2 in winter versus

ence of specific practices applied at one or ia counts on carcasses.

nt studies are presented following the normal orular, the effect of the following practices on the application of a pre-slaughter washing treaentionally dressed carcasses; pasteurisation treatm

pre-slaughter washing treatment, Kannan et al. (2oat carcass bacterial counts, as a strategy to

animals. The comparison between the control grouwashing with potable water) led to the conclusioring spray washing.

igated as a potential risk factor influencing indicaal. (2000a) concluded that the substitution of inveg units E. coli counts on sheep carcasses. Anothe

of pelt removal methods (Cradle dressing; Hybriand the Frame systems, considered as inverted

iaceae counts (p<0.01) compared to the conventimicrobiological contamination was achieved usingg pelt removal.

skin-on sheep carcasses was studied and disces by Fisher et al. (2007). Current EU legiss with the skin left on and flaying during the drr, demand for these skin-on, singed products (aal ethnic groups resident in the United Kingdomcountries. Skin-on carcasses had lower Entero



hors.

54

he hygienic status of robiological survey in of slaughter volumes, showed a wide range

houses: consequently, uthors concluded that vestock handling and

mbined together the were collected from authors found those

antly (P<0.001) lower use in the winter and

ae counts, differences ter and spring season. seasons for the low-log cfu/cm2 in spring r the high-throughput 6.05 log cfu/ cm2 in

more stages of the

rder of the slaughter he levels of indicator atment; the dressing ment and the chilling

2007) investigated the o reduce the faecal up (no washing) with on that E. coli counts

ator bacteria loads on erted for conventional er study (White et al., id method and Frame d dressing, produced ional Cradle dressing. g inverted dressing as

cussed in relation to slation prohibits the essing procedure is a associated to specific m is evident and may obacteriaceae counts

E.


(p<0.001) before chilling thaneliminated by the application odressed carcasses. Moreover, theon skin-on carcasses after evichilling, thus toasting is a recomm

Hauge et al. (2011a) evaluatedcarcasses. The application of dressing/before chilling led to a 9of 1.85 log CFU per carcass (P<02.37 log CFU per carcass after application of steam was also repasses of steam spraying pistoreduced Enterobacteriaceae coorganoleptic characteristics of lam

Finally, the chilling step represecontrol the growth of bacteria combinations can be used to coinvestigated by Brown et al. (20and water circulated through thesystem involves attaching a speca chilled isotonic solution of sugapproximately 3 min thereby remThe study concluded that VPC comparison with air chilling; hcooling period. Concerning Enteand the treatment group (VPCEnterobacteriaceae on the poolebefore chilling and at 24 h webetween three different chilling ttreatments differed in the folcarcasses at 2 °C for 24h; Ultra-rtreatment - carcasses at 12 °C folowest Enterobacteriaceae countto cold shortening and consequen

According to the retrieved paperrisk factors on indicator bacteria two papers investigating this faccarcasses during the warm seasoslaughterhouse throughput coupindicator bacteria counts, withslaughterhouses and in winter for

A process that can lead to an innodes during the inspection oprocedures coupled with the usebacteria loads on carcasses.




n conventionally dressed carcasses. However, of a chilling step that reduced the counts mainle application of a toasting step (an additional exisceration significantly reduced Enterobacteriamended step for these kind of products.

d the microbiological effects of hot water paa pasteurising treatment (82 °C± for 8 s)

99.5 % reduction of E. coli counts, corresponding0.001). Also Enterobacteriaceae counts were sigpasteurisation. The reduction of Enterobacteriaeported by Milios et al. (2011). The use of steal on each side of the carcass) after pluck remounts (almost 1 log CFU/cm2) without advmb carcasses.

ented an important point along the slaughteringand to ensure food safety. Several methods a

ool the carcass. The vascular perfusion chilling009), in which very fine ice particles in a solutioe vascular system, offering significant reductions cially designed catheter to the carotid artery in thgars and salts is then pumped through the arterimoving as much residual blood as possible fromwas capable of a rapid initial reduction of ca

owever uptake of perfusate into the carcass ocrobacteriaceae counts on the carcasses belonging

C), no conclusions can be drawn due to the fed surface samples and in the deep tissue samplere below the detection limits for all but two streatments on lamb carcasses was studied by Rublowing time-temperature combinations: Converapid treatment - carcasses at 20 °C for 3.5h then or 7h then 2 °C until 24 h. Carcasses of ultra-rapts, however, the carcasses subjected to this treatmntly to a loss of meat quality.

rs several conclusions can be drawn on the effectcounts. Annual season had no significant effect otor concluded that there was no difference betwe

on compared to the cold season. However, one papled with the season, concluded that season h

h the lowest counts being detected in spring r high-throughput slaughterhouses.

ncreasing of the counts of indicator bacteria is thof carcasses for caseous lymphadenitis; thus, of data regarding the management of animals co



hors.

55

this difference was ly on conventionally

xposure to gas flame) aceae counts before

asteurisation of lamb on carcasses after

g to a mean reduction gnificantly reduced by ceae counts after the am application (8-10

moval/before chilling erse effects on the

g process in order to and time-temperature g (VPC) is a method on of sodium chloride

in cooling time. The he neck of the animal, ial/venous system for m within the carcass. arcass temperature in ccurred, reducing the g to the control group fact that numbers of es taken from the loin samples. Comparison bio et al. (2012). The entional treatment - 2 °C until 24 h; Slow pid treatment had the ment were susceptible

t of different types of on the counts: in fact, een bacterial loads on aper investigating the has an influence on

for low-throughput

he handling of lymph reduced inspection ould reduce indicator

E.


Another factor that can have an iconcluded that low-throughput sof indicator bacteria compared to

The application of treatments alindicator bacteria levels on carslaughter does not significantly rof carcasses after dressing is esignificant reduction of the prevAnother step that can have an efof experimental chilling treatmenbacterial load; however a loss of

Finally, the dressing technique hincrease or reduction. In fact, induring pelt removal, is considerecarcass contamination.

Tables 5a and 5b summarize therelated to the slaughterhouse on E

Finally, Tables 19 and 20 (Appendealing with the effect of theslaughterline) on the counts of in

Table 5a: Slaughterhouse reladecrease/increase/or with no effe


Decrease low throughput-sp

Low ThroughpuSlaughterhouses (L

LTSs

low throughput–sampoint

Inverted dressin Hot water pasteurizDecrease Total Increase Pre-InspectionIncrease Total No effect Spring

Hot season Slaughterhouse throu Pre-slaughter wash

No effect Total




mpact on counts is the throughput of the slaughteslaughterhouses produce carcasses with lower pro high-throughput slaughterhouses.

ong the slaughterline and the effect of the slaugrcasses were investigated by eight papers. Wasreduce the microbial load. On the other hand, hoteffective; several authors concluded that this tvalence and counts of E. coli and Enterobacterffect in reducing carcass microbial load is the chnts can be effective in order to quickly reduce carmeat quality is recorded.

has an effect on indicator bacteria counts on carverted dressing, which minimizes contact betweed by several authors as the technique to be ado

e effect (in terms of decrease, increase and no eE. coli and Enterobacteriaceae counts.

ndix E) report E. coli and Enterobacteriaceae coe slaughterhouse features (general and at diffndicator bacteria.

ated risk factors investigated in the retrievedect on E. coli counts on small ruminant carcasses

ment Compared factor/Control Ref

pring low throughput-winter; high throughput-spring

Feizullah (201

ut LTSs)

Domestic-export slaughterhouses

Phillips et a

Medium throughput slaughterhouses

Sumner (

mpling high throughput-sampling point Feizullah (201

ng Conventional dressing Gill et al.zation Untreated Hauge et al

n Post-inspection Jordan et a

Fall or winter Duffy et aCold season Cohen et a

ughput broad range of volume Bass et alhing no washing Kannan et



hors.

56

erhouse. Three papers evalences and counts

ghtering technique on shing animals before t water pasteurisation treatment leads to a riaceae on carcasses. hilling phase. The use rcass temperature and

rcasses, leading to an en hands and carcass pted in order to limit

effect) of risk factors

ounts from the studies fferent stages of the

d studies leading to

ference N° of studies

and Daskalov 10) (BG)

1

al. (2001) (AU) 1

et al. (2003) (AU)

1

and Daskalov 10) (BG)

1

. (2000) (CA) 1 l. (2011a) (NO) 1

6 al. (2012) (AU) 1

1 al. (2001) (US) 1 al. (2006) (MA) 1 . (2011) (AU) 1 al. (2007) (US) 1

4

E.




Decrease Inverted dressing Skin on carcasse Ultra rapid chillin Hot water pasteuriza Steam applicatio low throughput-spr

Decrease Total Increase low throughput-win

Pre-Inspection Increase total No effect Vascular perfusion chNo effect Total


Three papers investigated the reland/or Enterobacteriaceae countal., 2011b), visual faecal contam2006), visual faecal contaminatiolast one (Whyte et al., 2002), thvisible contamination on carcasse

Concerning the sampling pointsslaughterline, mainly after pelt rtook samples at several sites oanalytical method of enumeratinplate count. Counts were express

Hauge et al. (2011b) investigatetotal, 140 lambs of 5 months of apasture were slaughtered in a com

The animals were divided into fabattoir immediately before slaudays); 35 lambs shorn on-farm slaughter (unshorn).

After stunning and bleeding (befcleanliness of the fleece (0– 3 scfleece (minor faecal material or mthe belly, legs, and tail; a scorepresented a very dirty fleece (fa

Immediately after the removal ocm2) for analysis of E. coli (pou




ated risk factors investigated in the retrievedect on Enterobacteriaceae counts on small rumina

ment Compared factor/Control Referen

g Conventional dressing White et aes conventionally dressed Fisher et ang Conventional and slow chilling Rubio et aation Untreated Hauge et aln Untreated Milios et aring high throughput-winter Feizullah

(201

nter high throughput-autumn Feizullah (201

Post-inspection Jordan et a

hilling conventional chilling Brown et a

lationship between faecal contamination of carcats: all the studies were conducted in Europe. In

mination related to E. coli counts was studied, in aon in relation to Enterobacteriaceae counts was he focus was more on the development of a mees according to different types of pelt removal tec

s, two out of three papers collected samples at removal, while regarding the sampled areas, twoof the carcass. The sampling method used wang the bacteria (E. coli and/or Enterobacteriaceaesed as log cfu/cm2 or log cfu/total sampled area.

d E. coli carcass contamination associated with age, grazed in the hills for 3-4 months and finishmmercial slaughterhouse.

four groups according to the shearing regimes: 3ughter (day 0); 35 lambs shorn on-farm three day

m seven days before slaughter (7 days); 35 lam

fore fleece removal), all the lambs were assessed cale) by a skilled observer. The score ‘0’ represemud in the fleece); a score of ‘1’ represented smare of ‘2’ represented a generally dirty fleece;

faecal material or mud under the belly, legs, and ta

of the fleece, swab samples were collected from ur plate according to NMKL method No. 125 - N



hors.

57

d studies leading to ant carcasses

nce (study) N° of studies

al. (2002) (GB) 1 al. (2007) (GB) 1 al. (2012) (ES) 1 l. (2011a) (NO) 1 al. (2011) (GR) 1 and Daskalov

10) (BG) 1

6 and Daskalov

10) (BG) 1

al. (2012) (AU) 1 2

al. (2009) (GB) 1 1

asses and their E. coli one paper (Hauge et

another (Byrne et al., studied, while in the

ethod to assess gross chniques.

one point along the o out of three papers as swabbing and the e) was the alternative

fleece cleanliness. In hed on grass on home

35 lambs shorn in the ys before slaughter (3 mbs not shorn before

d and scored on visual ented a visually clean all spots of dirt under and a score of ‘3’ ail).

the brisket area (100 Nordic Committee on

E.


Food Analysis, 2005). Mean clea(3 days), 0.63 (7 day), and 2.49 (value than shorn lambs (p<0.05)log cfu/100 cm2 for carcasses wnot significantly different.

This study demonstrated that, oncontaminated carcasses than dirskinning are undeniably importanmade by controlling cleanliness o

Byrne et al. (2006) studied the rithe fleece to the ovine carcass, “clean sheep policy”. In this studat the slaughterhouse and gradedclean and dry; (B) clean and wet;

Microbiological evaluation of Enterobacteriaceae counts were shoulder, flank and rump) immed

Enterobacteriaceae were detectewere: 2.7, 2.9, 4.4, 3.9 and 4.4 sampled sites (brisket, shoulderecovered from dirty sheep were or wetness. Moreover, the paramcount on the carcass when the flecount when the fleece is dirty, rethat dirt was a contributing risk fpolicy should, therefore, differenmeasures for the latter.

Whyte et al. (2002) investigatedabattoirs (no more than 20 livestmodified cradle design and sevMoreover, a method of scoring gorder to measure gross visible co

Gross, visible contamination (stregions of the carcass using a mfilm (90 by 130 mm – 3M Pat-ifixed to a light blue cartridge contamination was based on thebetween 0 (no visible contaminat

Concerning the modified pelt (mainly hair) at the shoulder andsignificantly better than the convinverted dressing procedures (Crproduced the most visibly contawhich a similar contamination sc




anliness scores for the four shearing groups were(unshorn); thus the unshorn lambs were dirtier an). Mean E. coli values at skinning were of 1.65,

with cleanliness score ‘0’, ‘1’, ‘2’ and ‘3’; howev

n average, visually clean animals tended to prodrty animals. While efforts to improve hygiene nt, significant improvements in the hygiene of ovof the live animals.

sk factors associated with the transfer of bacteriawith the aim to provide a scientific basis for t

dy, sheep in lairage were visually inspected by thed (based on the visual inspection of the fleece) int; (C) dirty and dry; (D) dirty and wet and (E) with

the carcasses was conducted using the swabobtained from 40 animals per category at four s

diately after pelt removal.

ed in 37.6 % of samples tested; the mean values log cfu/ 4.000 cm2. Contamination levels were

er, flank and rump). In this study, the Enterohigher than the counts found on the clean sheep

meter “fleece dryness/wetness” did not affect theeece was clean, while this parameter did influencesulting in overall higher counts. Enterobacteriacfactor regardless of wetness or dryness of the anintiate between clean and dirty sheep and requir

d methods of reducing lamb carcass contaminatiotock units/week) where cradle dressing was emplveral improved pelt removal methods were degross visible contamination on the depelted carc

ontamination when comparing different methods o

traw, wool, dirt and faecal material) was quantmethod based on the use of sheets of colourless, c

it). The adhesive sheets were applied to the carcpaper background. Quantification of the deg

e quantity of material adhering to the adhesive fition) and 10 (maximum visible contamination).

removal procedure, measurements of gross vd abdomen regions showed that a modified pelt remventional method (p<0.01 and p<0.05 respectiveradle, Hybrid and Frame methods), the Cradle meaminated carcasses for all carcass sites apart frcore to that of the Hybrid method was recorded.



hors.

58

e: 0.66 (0 days), 0.60 nd had a higher score 1.88, 2.16, and 2.49

ver these values were

duce less microbially-during slaughter and vine carcasses can be

al contamination from the development of a e veterinary inspector to five categories: (A) h visible faecal dags.

b sampling method. eparate sites (brisket,

in the five categories similar over the four

obacteriaceae counts regardless of dryness e Enterobacteriaceae e Enterobacteriaceae

ceae counts suggested imal. The clean sheep re additional hygiene

on in low-throughput loyed. In the study, a eveloped and tested. ass was developed in of pelt removal.

tified from the same clear adhesive plastic cass surface and then gree of gross visible film. This was graded

visible contamination moval procedure was

ely). In relation to the ethod of pelt removal om the shoulder, for

E.


Correlation between the microcontamination was not possibEnterobacteriaceae (less than 1contamination scores.

In conclusion, the degree of grosthe contact that was likely to ofleece. Significant improvement dressing, as this minimizes hand

Despite the fact that the numberlimited, the available data sugges

1) the distinction between cleanimprove the hygiene of ovine car

2) additional hygiene measures send of the day; reduced line speinverted dressing; greater spacing

3) modifications to pelt removalthe slaughterman or the fleece ca

Table 22 (Appendix F) summaricarcasses along the slaughterline

CONCLUSIONS AND REMAR

• A total of 87 papers werprovided pertinent data aabout small ruminants.

• A high level of variabilcomplexity of the slauganalytical methods usedspecific step of the slaualong the slaughterline,conflicting conclusions aor investigating the same

• Further studies are needEnterobacteriaceae counalso those with contrastinand after specific steps in

• Studies dealing with riskto the choice of control g

• In general a better agreemregarding the presentatireaching this objective.among studies and resear




obiological contamination of the carcass and le because minor changes in TVC (Total V log cfu/cm2) were accompanied by large chan

ss visible contamination at the carcass sampling sccur between the carcass and the hands of the in gross visible contamination was achieved by contact with the carcass during pelt removal.

r of retrieved studies providing data for review sted that:

n and dirty carcasses could be an important startrcasses;

should be applied for high-risk (dirty) animals (i.eed; thorough cleaning of operator hands, arms ang between carcasses);

methods which reduce contact between the carcan significantly improve gross visible contaminati

izes the counts of E. coli and Enterobacteriaceain relation to the level of visual faecal contamina

RKS

e used to collect data for the three review questioabout beef carcasses, 31 papers about swine car

lity among the different studies, due to differenghterlines, was evidenced. Some variables, likd, the area of carcass sampled, the unit of enughterline investigated and the decontaminatio, render the available data barely comparableamong studies describing counts at the same stage risk factor.

ded to state precisely the effect of slaughterlinents on carcasses regarding in particular poorly ng results. Studies should be performed by sampn order to avoid confounding.

k factor should be addressed in particular by posingroups in order to better appreciate results.

ment should be obtained within the scientific comion of results. A standardized unit of enumer Confidence interval should be preferred to frch synthesis.



hors.

59

the level of gross Viable Counts) and nges in gross visible

sites closely reflected slaughterman or the adoption of inverted

question 3 was quite

ting point in order to

e.: slaughtering at the nd aprons; the use of

cass and the hands of ion.

ae on small ruminant ation.

ons. Forty-two papers rcasses and 21 papers

nt aspects and to the ke the sampling and numeration used, the n treatments applied

e and could lead to ge of the slaughterline

e step on E. coli and investigated steps as

pling carcasses before

ng particular attention

mmunity in particular ration could help in facilitate comparison

E.


REVIEW QUESTION 1. Presetheir counts on carcasses durin

• The main challenge in iincrease of indicator bacthat provided data obtainat distant sampling poinconsequence, the effectievident.

Beef carcasses

• Evisceration and trimmistudy, whereas in othecorrelated to these slaugh

• Washing after evisceratiostudy (reduction up to abwith potable water had nolog compared to the previ

• The application of diffpasteurisation and washinlog) of both E. coli and En

• Chilling does not effectivconflicting data in terms chilling treatment of beef

Pigs

• Scalding effectively reduboth indicator bacteria obwere described in all considered in the implem

• Pasteurisation effectivelyunder experimental slauslaughterhouses.

• Carcass washing does no

• Chilling is effective in rconsequently, it should Achieved reductions vari




ence of the indicator organisms E. coli and Entng different stages in the slaughterline

identifying the steps of the slaughterline that ledcteria counts on pig and ruminant carcasses was tned before and after a single stage. Data have gennts and in between, different treatments have usuive role of the single stages in a change in co

ng led to an increase of E. coli counts up to 1 er studies (5), changes of E. coli and Enterohter phases were trivial (within 0.5 log).

on was effective to reduce the E. coli load of bbout 1 log), whereas some other studies demono effect on E. coli and Enterobacteriaceae countsious step of the slaughterline).

ferent sequential decontamination treatments, ng with acids after evisceration always led to anterobacteriaceae counts on beef carcasses.

vely reduce bacterial load on beef carcasses. Diffeof reduction/increase of E. coli and Enterobacte

f carcasses at the end of the slaughterline.

uces both E. coli and Enterobacteriaceae counts bserved reductions were in some cases higher thasampling regions tested. Therefore, scalding

mentation of GMP and HACCP within pig slaught

y reduces E. coli counts on pig carcasses up to 1.ughter conditions. However, more studies are ne

ot effectively reduce microbial contamination on p

reducing E. coli and Enterobacteriaceae counts obe carefully implemented in the context of

ied greatly according to sampling site on carcasse



hors.

60

terobacteriaceae and

d to a decrease or an the paucity of studies nerally been collected ually been used. As a unts was not always

log (cfu/cm2) in one obacteriaceae counts

beef carcasses in one nstrated that washing s (variation within 0.5

such as hot water, a drop (higher than 1

ferent studies reported eriaceae loads due to

on pig carcasses. For an 3 log (cfu/cm2) and

should be carefully terhouses.

.86 log (cfu/100 cm2) eeded in commercial

pig carcasses.

on pig carcasses, and slaughter operations. es.

E.


Small ruminants

• Skinning and evisceratio(from 0.5-1.0 to 2.0 logreported by the four avai

• Carcass washing does ncarcasses.

• Chilling is effective in rcarcasses, as it reduced m

REVIEW QUESTION 2. InfoEnterobacteriaceae counts alon

Beef carcasses

• Pre-slaughter diet on farmbeef carcasses. However,

• Annual season has a dicarcasses; the lowest levedry season.

• Slaughterhouse characterdifficult to assess which throughput was not foundon beef carcasses. Althouobtained in low andprevalences/counts were g

• Steam pasteurisation and quality of beef carcasseprocedures tested were ureported a clear reductionundetectable levels.

• The effect of washing cmentioned for review quewashing treatments appliewashing seems to be ineffective when the load is

• The effect of chemical dchlorine, peroxyacetic acrelated to their effectivenresults could be due to ththe steps of the slaughterl




on are the two main steps along the slaughterg cfu/cm2) of E. coli and Enterobacteriaceae onlable papers.

not effectively reduce microbial contamination

reducing Enterobacteriaceae counts (two papermicrobial counts by more than 0.5 log cfu/cm2.

ormation that could explain the variability g the slaughterline

m (batch-related risk factor) does not affect microbthis was addressed in only one retrieved study.

irect impact on indicator bacteria prevalence els of carcass contamination were obtained durin

ristics have an influence on carcass contaminatifactors had the greatest impact on the counts.

d to be clearly correlated to the counts of E. coli ough in only one study differences between indid high throughput slaughterhouses were generally reported in lower throughput plants.

hot water washing are effective ways to improves, as reported for review question 1. The sed both in high and low throughput slaughterho

n for both E. coli or Enterobacteriaceae counts on

carcasses with potable water at ambient tempeestion 1. Different effects could be accounted fored as well as differences in the initial bacterial loneffective when the initial bacterial load on cs relatively high.

decontamination treatments (e.g. washing-sprayid, nisin) is unclear, since different studies descrness in reducing bacterial loads on beef carcasshe different chemicals tested, as well as the procline where the treatments were applied.



hors.

61

rline where increases n carcasses occur, as

n on small ruminant

rs) of small ruminant

of the E. coli and

bial contamination of

and counts on beef g coldest months and

ion levels but it was . The slaughterhouse

or Enterobacteriaceae icator bacteria counts

significant, lower

e the microbiological different equipment-ouses, and all studies n beef carcasses up to

erature is unclear, as r by differences in the ad of carcasses, since

carcasses is low, but

ying with lactic acid, ribed opposite results es. These conflicting cedures followed and

E.


Pigs

• A high number of diffcomparisons are not possstudy design and operatio

• Hot water (>80 °C) baseE. coli counts (three studand on specific sites. The

• Anal plugging was foundone study. However otheplugging to maximize the

Small ruminants

• Carcasses of shorn lambbeginning of the slaughteof this batch-related risprocess if other control m

• Annual season has no However, the associationto the conclusion that inslaughterhouses and duri

• The effect of the slaugcarcasses was unclear.slaughterhouses had lowanother study observed tplants.

• Hot water pasteurisation prevalence and level (frcarcasses.

• Modified dressing proceas they reduce the indconventional dressing.

REVIEW QUESTION 3. ThEnterobacteriaceae and visuslaughterline

Beef carcasses

• Clean cattle produce cacarcasses.




ferent risk factors were investigated in the rible since few studies considered the same factor

onal environments were uncomparable.

ed treatments, such as pasteurization, are effectdies) and Enterobacteriaceae counts (one studyese results agree with papers described in review

d effective in reducing carcass contamination der authors suggested the importance of GMP due prevention of contamination.

bs had significantly lower (1.0 log cfu/cm2) couerline compared to those of unshorn lambs (in thesk factor). However, this advantage is lost durmeasures are not applied.

effect on microbial contamination of small n of the annual season with slaughterhouse througndicator bacteria counts are lower during springing winter for high-throughput slaughterhouses.

ghterhouse throughput on the prevalence of i. Some studies concluded that carcasses fr

wer prevalence of E. coli than those from high-thrthat the prevalence of E. coli differed greatly am

of carcasses after skinning is effective, as it signrom 1.0 to 2.0 log cfu/cm2) of indicator bacteri

edures, which reduce contact between hands anddicator bacterial load by 1.5 log unit on car

he potential relationship between the counal faecal contamination on ruminant ca

arcasses with lower bacterial loads (about 0.5-



hors.

62

retrieved papers but r and also in this case

tive in reducing both ) on whole carcasses question 1.

during evisceration in uring evisceration and

unts of E. coli at the e only retrieved study ring the slaughtering

ruminant carcasses. ghput (one study) led g for low-throughput

ndicator bacteria on from low-throughput roughput plants while mong the investigated

nificantly reduces the ia on small ruminant

d pelts, are advisable, rcasses compared to

nts of E. coli and arcasses along the

1 log) than do dirty

E.


• Pre-slaughter visual classcorrective measures wademonstrated by the retri

• Application of effective and/or on dirty carcassescomparable to or lower th

Pig carcasses

• The literature research between visual faecal con

Small ruminants

• Visually clean small rucontaminated than dirty dirty animals and apply a

• Modified dressing methocan improve gross visiblthat methods to assess gpossible to compare data

REFERENCES

Algino R, Ingham S, Zhu J. 2treatments in very small State

Algino R, Badtram G, Ingham Bpork carcasses in very small a

Anderson R, Carr M, Miller R, KMcReynolds J, et al. 2005.supplements on Escherichia cMicrobiol 22(5):439-47.

Arthur TM, Bosilevac JM, Nou XDM, Koohmaraie M. 2004. EEnterobacteriaceae, and Escplants. J Food Prot 67(4):658-

Bacon RT, Belk KE, Sofos JN, animal hides and beef carcasequential interventions for de

Barboza de Martinez Y, Ferrer Kin reducing levels of micr65(11):1780-3.

Barco L, Belluco S, Roccato A apoultry carcasses along the




sification of the level of animal dirtiness and the as effective in reducing microbial contaminatiieved studies addressing this issue.

measures either on dirty animals before enterins along the slaughterline can lead to contaminatihat of clean animals at the end of the slaughterlin

did not provide any papers dealing with the ntamination on pig carcasses and counts of indica

uminants inspected at the lairage or before fleeanimals. Thus, an effort should be made in orderadditional hygienic measures for the latter group.

ods which reduce contact between hands and flee contamination of small ruminant carcasses. How

gross visible contamination on carcasses differ ama provided by the available studies.

2007. Survey of antimicrobial effects of beef e Inspected slaughter plants. J Food Sci 72(5):M17

B, Ingham S. 2009. Factors associated with salmabattoirs in Wisconsin. Journal of Food Protection

King D, Carstens G, Genovese K, Callaway T, E Effects of experimental chlorate preparationscoli colonization and contamination of beef cattle

X, Shackelford SD, Wheeler TL, Kent MP, JaronEscherichia coli O157 prevalence and enumeratiocherichia coli O157 at various steps in commer-65.

Clayton RP, Reagan JO, Smith GC. 2000. Micrasses at different stages of slaughter in plants econtamination. J Food Prot 63(8):1080-6.

K, Salas EM. 2002. Combined effects of lactic acrobiological contamination in red meat carc

and Ricci A, 2014. Escherichia coli and Enterobae slaughter processing line, factors influenci



hors.

63

application of proper ion of carcasses, as

ng the slaughterhouse ion level of carcasses ne.

possible relationship ator bacteria.

ece removal, are less r to classify clean and

eece with the carcass, wever, due to the fact mong studies, it's not

carcass intervention 73-9.

monella prevalence on n® 72(4):714-21.

Edrington T, Jung Y, s as feed and water e and carcasses. Food

ni D, Pauling B, Allen on of aerobic bacteria, rcial beef processing

robial populations on employing multiple-

cid and nisin solution casses. J Food Prot

acteriaceae counts on ing the counts and

E.


relationship between visual supporting publication 2014:E

Bass C, Crick P, Cusack D, Lockthe co-regulation of abattoirs

Bhandare SG, Sherikar AT, Patucontamination on sheep/goat cControl 18(7):854-8.

Blagojevic B, Antic D, Ducic Mmicrobial loads on the hides a

Blagojevic B, Antic D, Ducic Mabattoir process hygiene indic

Bohaychuk VM, Gensler GE, Bcarcasses from provincially in

Bosilevac JM, Nou X, Barkocy-hot water instead of lactic acidthe prevalence of Escherich69(8):1808-13.

Bryant J, Brereton DA, Gill CO.of microbiological contamina

Brown T, Ian Richardson R, Wcarcasses - A feasibility study

Byrne B, Dunne G, Lyng J, Boltwith the new hygiene regulati

Cibin V, Mancin M, Pedersen KLau Baggesen D, Ricci A, 20Hygiene Criteria in poultry121 pp.

Cohen N, Ennaji H, Hassa M, Ka(morocco). Molecular Nutritio

Collobert J, Dorey F, Dieulevecarcasses. Sci Aliments 22(3)

Conter M, Zanardi E, Ghidini S,and equipment in a pig slaugJournal of Food Science 18(4)

Corantin H, Quessy S, Gaucher pasteurization in controlling mCan J Vet Res 69(3):200-7.

Delhalle L, De Sadeleer L, BollDaube G. 2008. Risk factorsslaughterhouses. J Food Prot 7

Duffy EA, Belk KE, Sofos JN, LMicrobial contamination occu64(4):503-8.




faecal contamination of carcasses and countEN-636, 107 pp.

ke G, Sumner J. 2011. The use of microbiologicain new south wales, Australia. Food Control 22(6

urkar AM, Waskar VS, Zende RJ. 2007. A comcarcasses in a modern Indian abattoir and traditio

M, Buncic S. 2012. Visual cleanliness scores of cand the carcases. Vet Rec 170(22):563.

M, Buncic S. 2011. Ratio between carcass-and scator. Food Control 22(2):186-90.

Barrios PR. 2011. Microbiological baseline stunspected abattoirs in Alberta, Canada. Can Vet J 5

-Gallagher GA, Arthur TM, Koohmaraie M. 200d reduce levels of aerobic bacteria and Enterobachia coli O157:H7 on preevisceration beef car

. 2003. Implementation of a validated HACCP sation of pig carcasses at a small abattoir. Can Vet

Wilkin CA, Evans JA. 2009. Vascular perfusion y. Meat Sci 83(4):666-71.

on DJ. 2007. The development of a 'clean sheep pion (EC) 853/2004 (hygiene 2). Food Microbiol 2

K, Barrucci F, Belluco S, Roccato A, Cocola F, F14. Usefulness of Escherichia coli and Enteroba: experimental study. EFSA supporting public

arib H. 2006. The bacterial quality of red meat anon & Food Research 50(6):557-62.

eux V, Quillien N. 2002. Microbiological cont:327-34.

Guidi E, Campanini G. 2006. Microbiological cghterhouse and evaluation of a steam decontamin):387-96.

ML, Lessard L, Leblanc D, Houde A. 2005. Efmicrobiological hazards of cull cow carcasses in

laerts K, Farnir F, Saegerman C, Korsak N, De for salmonella and hygiene indicators in the 1071(7):1320-9.

LeValley SB, Kain ML, Tatum JD, Smith GC, Kurring on lamb carcasses processed in the unite



hors.

64

ts: a review. EFSA

al surveys to evaluate 6):959-63.

mparison of microbial onal meat shops. Food

cattle at slaughter and

skin-microflora as an

dy of beef and pork 52(10):1095-100.

06. Treatments using cteriaceae and reduce rcasses. J Food Prot

ystem for the control J 44.

chilling of red meat

policy' in compliance 24(3):301-4.

Ferrarini S, Sandri A, cteriaceae as Process cation 2014:EN-635,

nd offal in Casablanca

tamination of bovine

condition of carcasses nation system. Italian

ffectiveness of steam n a commercial plant.

ewulf J, De Zutter L, 0 largest Belgian pig

Kimberling CV. 2001. ed states. J Food Prot

E.


Duggan SJ, Mannion C, PrenderF, Duffy G. 2010. Trackingslaughter process in the Repub

Feizullah F and Daskalov H. 20low and high production capa

Fisher A, Wilkin CA, Purnell Gcarcasses. Meat Sci 77(4):467

Gill CO and Landers C. 2004aroutinely processed and detain8.

Gill CO and Landers C. 2004b. Mremoval of visible contaminat

Gill CO, Landers C. 2003a Micrpacking plants. Meat Sci 65(4

Gill CO and Landers C. 2003b. of decontaminated beef carcas

Gill C, Bryant J, Landers Cmicrobiological contaminatioplant. Food Microbiol 20(6):6

Gill CO and Badoni M. 2001. Thpig carcasses. Can Vet J 42(4)

Gill CO, Bryant J, Brereton conventional or inverted dress

Gill CO, Dussault F, Holley REvaluation of the hygienic pcarcasses at eight packing pla

Gill C and Bryant J. 2000. Throutinely in a commercial bee

Hamilton D, Holds G, Lorimer decontamination of pork carcAustralian abattoirs. Zoonose

Hansson IB. 2001. MicrobiologicJ Food Prot 64(6):820-5.

Hauge SJ, Nafstad O, Rotterud Oby hide cleanliness in the abat

Hauge SJ, Wahlgren M, Rotterucarcasses: Microbial effects an

Hauge SJ, Nafstad O, Skjerve Ecleanliness on microbiologic3):178-83.

Hurd HS, Brudvig J, Dickson J, impact on carcass contaminat




rgast DM, Leonard N, Fanning S, Gonzales-Barrg the salmonella status of pigs and pork fromblic of Ireland. J Food Prot 73(12):2148-60.

010. Investigation on lamb meat production hygiacity. Bulgarian Journal of Veterinary Medicine 1

G. 2007. The production and microbiological sta7-73.

a. Proximate sources of bacteria on boneless ned carcasses at a pork packing plant. Int J Food

Microbiological conditions of detained beef carcation. Meat Sci 66(2):335-42.

robiological effects of carcass decontaminating tr4):1005-1011.

Effects of spray-cooling processes on the microsses. J Food Prot 66(7):1247-52.

C. 2003. Identification of critical control poon in processes leading to the production of grou641-50.

he microbiological effects of trimming sticking w):297-9.

DA. 2000a. Microbiological conditions of shsing processes. J Food Prot 63(9):1291-4.

RA, Houde A, Jones T, Rheault N, Rosales Aerformances of the processes for cleaning, dres

ants. Int J Food Microbiol 58(1-2):65-72.

he effects on product of a hot water pasteurizinef carcass dressing process. Food Microbiol 17(5)

M, Kiermeier A, Kidd C, Slade J, Pointon A. ases with hot water or acidified sodium chlorite -s Public Health 57 Suppl 1:16-22.

cal meat quality in high- and low-capacity slaugh

O, Nesbakken T. 2012. The hygienic impact of cattoir. Food Control 27(1):100-7.

ud OJ, Nesbakken T. 2011a. Hot water surface pnd cost-benefit considerations. Int J Food Microb

E, Rotterud OJ, Nesbakken T. 2011b. Effects ofcal contamination of lamb carcasses. Int J Foo

Mirceta J, Polovinski M, Matthews N, Griffith Rion and human foodborne risk. Public Health Rep



hors.

65

ron U, Egan J, Butler m lairage through the

iene in facilities with 3(4):252-8.

atus of skin-on sheep

loins prepared from Microbiol 97(2):171-

asses before and after

reatments at four beef

obiological conditions

oints for control of und beef at a packing

wounds in pasteurized

heep carcasses from

A, Quessy S. 2000b. ssing and cooling pig

ng treatment applied ):495-504.

2010. Slaughterfloor - a comparison in two

hterhouses in Sweden.

ategorisation of cattle

pasteurisation of lamb biol 146(1):69-75.

f shearing and fleece od Microbiol 150(2-

R. 2008. Swine health p 123(3):343-51.

E.


Inthavong P, Srikitjakarn L, KyMicrobial contamination of Southeast Asian J Trop Med P

Jones TH and Johns MW. 2012virus contamination of hog ca

Jordan D, Sentance CB, Spoonccaseous lymphadenitis and i92(4):837-40.

Kannan G, Jenkins AK, Eega Keffects on physiological stresRuminant Research 67(1):14-

Kiermeier A, Bobbitt J, Vandercarcase Escherichia coli monincoming stock and processin

Kinsella KJ, Sheridan JJ, RoweDA. 2006. Impact of a novel loss during beef carcass chilli

Lenahan M, O'Brien SB, Kinsehygiene before and after chilli

Lenahan M, Crowley H, O'Brienchilling to control the growthcoli O157:H7 in pigs. J Appl

Lindblad M, Lindmark H, Lambcarcasses at Swedish slaughte

Madden RH, Murray KA, Gilcontamination during beef ca6.

Martineli T, Rossi Junior O, Cevaluation of the muscle slaughterhouse/Avaliação mioperações de abate em matado

McCleery DR, Stirling JM, Mcclipping on the microbiologicEC-approved abattoir. J Appl

McEvoy JM, Sheridan JJ, Blair Ito hygiene assessment based 92(2):217-25.

Mies PD, Covington BR, Harricattle hides prior to slaughte67(3):579-82.

Milios K, Mataragas M, Pantouover the microbiological conwith or without pasteurizing t




yule M, Zessin KH, Baumann M, Douangngeupig carcasses at a slaughterhouse in Vientian

Public Health 37(6):1237-41.

. Assessment of F-RNA coliphage as a potentiaarcasses. J Food Prot 75(8):1492-500.

cer WF, Balan JA, Morris SM. 2012. Inspectionits effect on the density of microbes on sheep

KR, Kouakou B, McCommon GW. 2007. Preslauss responses and skin and carcass microbial co-9.

rlinde P, Higgs G, Pointon A, Sumner J. 2006.nitoring data to investigate the relationship betweng efficacy. Int J Food Microbiol 111(3):263-9.

e TA, Butler F, Delgado A, Quispe-Ramirez A, spray-chilling system on surface microflora, wateng. Food Microbiol 23(5):483-90.

ella K, Sweeney T, Sheridan JJ. 2010. Assessming at five Irish abattoirs. Food Control 21(3):313

n SB, Byrne C, Sweeney T, Sheridan JJ. 2009. h of Enterobacteriaceae on porcine carcasses anMicrobiol 106(5):1512-20.

bertz ST, Lindqvist R. 2007. Microbiological baserhouses. J Food Prot 70(8):1790-7.

mour A. 2004. Determination of the principaarcass dressing processes in northern Ireland. J F

Cereser N, Cardozo MV, Pinto F, Perri S. 20surface of lamb carcasses during slaught

crobiológica da superfície muscular de carcaçaouro. Ars Vet 27(3):145-51.

cIvor K, Patterson MF. 2008. Effect of ante- ancal quality and safety and ultimate pH value of Microbiol 104(5):1471-9.

IS, McDowell DA. 2004. Microbial contaminatioon criteria used in EU decision 2001/471/EC. I

s KB, Lucia LM, Acuff GR, Savell JW. 2004.er using washes with and without antimicrobial

uvakis A, Drosinos EH, Zoiopoulos PE. 2011. Entamination of carcasses in a lamb carcass dressreatment. Int J Food Microbiol 146(2):170-5.



hors.

66

un B, Fries R. 2006. e capital, Lao PDR.

al indicator of enteric

n of lymph nodes for carcasses. Meat Sci

ughter spray-washing ounts in goats. Small

Use of routine beef een hygiene status of

Blair IS, McDowell er activity and weight

ment of lamb carcass 3-8.

The potential use of nd the incidence of E.

seline study of swine

al points of product ood Prot 67(7):1494-

011. Microbiological ter process in the as ovinas durante as

nd post-mortem hide f beef carcasses in an

on on beef in relation Int J Food Microbiol

Decontamination of agents. J Food Prot

Evaluation of control sing process operated

E.


Minihan D, Whyte P, O'Mahonyon the levels of EnterobacteriJ Vet Med B Infect Dis Vet Pu

Namvar A and Warriner K. 20polymerase chain reaction to slaughter line. Int J Food Mic

Nero L. A. 2012. Monitoring oslaughterhouses located in Mi

Nesbakken T, Eckner K, Rotterpathogenic Yersinia enterocoindicators on pig carcasses. In

Nou X, Rivera-Betancourt M, BoKoohmaraie M. 2003. Effect and the levels of aerobic baprocessing plant. J Food Prot

Ozdemir H, Sireli UT, Can HYcarcasses. Archiv Fur Lebensm

Phillips D, Sumner J, Alexandermeat. J Food Prot 64(5):697-7

Prendergast D, Daly D, Sheridancontamination levels and the Irish beef abattoirs. Food Mic

Purnell G, James C, Wilkin CA,through the use of anal plug73(6):1108-10.

Rabaste C, Faucitano L, Saucier of handling and group size ocarcass microbial contaminati

Retzlaff D, Phebus R, Kastner Cfor a high-volume static chamsupport HACCP programs. Fo

Rigobelo EC, Santo E, Marin Escherichia coli strains in andrugs. Foodborne Pathog Dis

Rigobelo EC, Stella AE, Avila Fisolated from carcasses of beMicrobiol 110(2):194-8.

Rivas T, Vizcaino JA, Herrera FJIberian pig slaughterhouse. J F

Rubio B, Vieira C, Martinez B, Fcarcass quality of suckling lam

Ruby JR, Zhu J, Ingham SC. 20large-scale beef abattoirs oveplan carcass testing for salmon




y M, Collins JD. 2003. The effect of commercialiaceae and Escherichia coli on naturally contamiublic Health 50(7):352-6.

006. Application of enterobacterial repetitive intrace the fate of generic Escherichia coli within robiol 108(2):155-63.

of hygiene indicator microorganisms in bovine inas Gerais State, Brazil. 2012 annual meeting Iaf

rud OJ. 2008. The effect of blast chilling on oolitica compared to Campylobacter spp. and nnt J Food Microbiol 123(1-2):130-3.

osilevac JM, Wheeler TL, Shackelford SD, Gwarof chemical dehairing on the prevalence of Escheacteria and Enterobacteriaceae on carcasses in66(11):2005-9.

Y. 2010. Determination of microbial surface comittelhygiene 61(1):27-30.

r JF, Dutton KM. 2001. Microbiological quality700.

n J, McDowell D, Blair I. 2004. The effect of abarelationship between aerial and carcass contam

crobiol 21(5):589-96.

, James SJ. 2010. An evaluation of improvemengging of pig carcasses prior to scalding and de

L, Mormede P, Correa JA, Giguere A, Bergeronon welfare of pigs in lairage and their influenceion and meat quality. Canadian Journal of Anima

C, Marsden J. 2005. Establishment of minimum opmber steam pasteurization system (SPS 400-SC) oodborne Pathog Dis 2(2):146-51.

JM. 2008. Beef carcass contamination by shn abattoir in brazil: Characterization and resista5(6):811-7.

FA, Macedo C, Marin JM. 2006. Characterizatioeef cattle during their processing at an abattoir

J. 2000. Microbial contamination of carcasses anFood Prot 63(12):1670-5.

Fernandez AM. 2013. Effect of different post-momb. Food Sci Technol Int.

007. Using indicator bacteria and Salmonella teer an 18-month period to evaluate intervention nella. J Food Prot 70(12):2732-40.



hors.

67

l steam pasteurization inated beef carcasses.

ntergenic consensus-a high capacity pork

carcasses from three fp.

occurrence of human numbers of hygienic

rtney BL, Reagan JO, erichia coli O157:H7 n a commercial beef

ontamination on beef

y of Australian sheep

attoir design on aerial mination levels in two

nts in carcass hygiene ehairing. J Food Prot

n R. 2007. The effects e on stomach weight, l Science 87(1):3-12.

perational parameters for beef carcasses to

higa toxin-producing ance to antimicrobial

n of Escherichia coli in brazil. Int J Food

nd equipment from an

rtem temperatures on

est results from three system efficacy and

E.


Saucier L, Bernier D, Bergeron meal frequency and pre-slauquality in pigs. Canadian Jour

Serraino A, Bardasi L, Riu R, PVisual evaluation of cattle clslaughtering. Meat Sci 90(2):5

Spescha C, Stephan R, Zweifel stages of slaughter in two Eur

Sumner J, Petrenas E, Dean contamination on beef and she

Tamplin ML, Feder I, PalumboEscherichia coli biotype I ocontrol point-based inspection

Tergney A and Bolton D. 2006. Vand microbial contamination o

Tomovic VM, Petrovic LS, JokRapid chilling effect on the Letters 16(6):6766-75.

Trivedi S, Reynolds AE, Chen decontaminate beef and hog cin Georgia. J Food Prot 70(3)

Warriner K, Aldsworth TG, Kauduring pork processing. J App

Whyte RT, Holder JS, Tinker development of procedures aremoval in low-throughput ab

Wu S, Dalsgaard A, Vieira AR, and genotypic analysis of pprocessed at a pig slaughterho

Yalcin S, Nizamlioglu M, Gurbuslaughtering process. J Food S

Yang X, Badoni M, Youssef MKof product at a large beef pack

Zweifel C, Fischer R, Stephan Rdifferent small-scale Swiss ab




R, Giguere A, Methot S, Faucitano L. 2007. Eughter fasting on behaviour, stomach content anrnal of Animal Science 87(4):479-87.

Pizzamiglio V, Liuzzo G, Galletti G, Giacometti leanliness and correlation to carcass microbial c502-6.

C. 2006. Microbiological contamination of pig ropean union-approved abattoirs. J Food Prot 69(

P, Dowsett P, West G, Wiering R, Raven eep carcases in south Australia. Int J Food Microb

o SA, Oser A, Yoder L, Luchansky JB. 2001. on swine carcasses processed under the hazard n models project. J Food Prot 64(9):1305-8.

Validation studies on an online monitoring systemon beef carcasses. Food Control 17(5):378-82.

kanovic MR, Tomovic MS, Tasic TA, Ikonic PMbacterial populations of pork carcasses. Roman

J. 2007. Use of a commercial household steamcarcasses processed by four small or very small m:635-40.

ur S, Dodd CE. 2002. Cross-contamination of carcpl Microbiol 93(1):169-77.

DB, Allen VM, White RP, Hinton MH. 20and apparatus to reduce contamination of lamb battoirs. J Food Prot 65(1):41-9.

Emborg HD, Jensen LB. 2009. Prevalence of tepopulations of Escherichia coli from animals,ouse. Int J Food Microbiol 135(3):254-9.

uz U. 2004. Microbiological conditions of sheepSaf 24(2):87-93.

K, Gill CO. 2012. Enhanced control of microbiolking plant. J Food Prot 75(1):144-9.

R. 2008. Microbiological contamination of pig anbattoirs. Meat Sci 78(3):225-31.



hors.

68

Effect of feed texture, nd carcass microbial

F, Merialdi G. 2012. contamination during

carcasses at different 11):2568-75.

G. 2003. Microbial biol 81(3):255-60.

Salmonella spp. and analysis and critical

m for reducing faecal

M, Sumic ZM. 2011. nian Biotechnological

m cleaning system to meat processing plants

casses and equipment

002. Assessment and carcasses during pelt

etracycline resistance , carcasses and cuts

p carcasses during the

logical contamination

nd cattle carcasses in

E.


APPENDICES

Appendix A. Relevance Scree

The checklists used for the asseNeutral means that the answer w

First level assessment – step 1

Question

Language of the paper

Is the paper a primary research pape

Does the paper consider main livest

Does the paper provides data on prepathogenic) Escherichia coli or Ente

Is the main aim of the paper the inveresistance and does the paper providscope of the review?

First level assessment – step 2

Questions

Which animal species are considere




ening

essment of the relevance of the retrieved papersas informative only and did not lead to paper exc

ns Answers Incl

English

French

Italian

Spanish

German

Other

er?

Yes

No

Unknown

tock species?

Yes

No

Unknown

esence/counts of (non erobacteriaceae?

Yes

No

Unknown

estigation of antimicrobial de no pertinent data for the

Yes

No

Unknown

Answers Inclusion

ed in the paper ?

Ruminants Ne

Swine Ne

Poultry Ne

Horses Ne

Others Ne



hors.

69

s are reported below. lusion.

lusion/Exclusion

Inclusion

Inclusion

Inclusion

Inclusion

Exclusion

Exclusion

Inclusion

Exclusion

Inclusion

Inclusion

Exclusion

Inclusion

Inclusion

Exclusion

Inclusion

Exclusion

Inclusion

Inclusion

n/Exclusion

eutral

eutral

eutral

eutral

eutral

E.


Does the paper provide data on morslaughterline or data on risk factors of indicator bacteria on carcasses?

Does the paper provide data on carc

Does the paper consider a real settin

Second level assessment

Questions

Where was the paper conducted?

Does the paper provide original data

Which species is/are considered?

Does the study provide data on carc

Which indicator organism/s is/are copaper?




re than one stage of the influencing the loads

Yes Inc

No Exc

Unknown Inc

casses?

Yes Inc

No Exc

Unknown Inc

ng?

Yes Ne

No Ne

Unknown Ne

Answers Inclusio

Europe Neutral

Neutra

Central and south America

Neutral

Africa Neutral

Asia Neutral

Oceania Neutral

a? Yes Inclusio

No Exclusi

Cattle Inclusio

Sheep Inclusi

Goat Inclusio

Swine Inclusio

Horse Inclusio

Other Exclusi

asses? Yes Inclusio

No Exclusi

onsidered in the Escherichia coli Inclusio

Enterobacteriaceae Inclusio

None of them Exclusi

Real setting Inclusio



hors.

70

lusion

clusion

lusion

lusion

clusion

lusion

eutral

eutral

eutral

on/Exclusion

l

al

l

l

l

l

on

ion

on

ion

on

on

on

ion

on

ion

on

on

ion

on

E.


Which kind of scenario is considere

Which kind of sample is collected?

Which is the study aim?

If a factor influencing the counts wafactor related to:




ed?

Pilot slaughterhouse

Inclusio

Artificial contamination with E. coli and/or Enterobacteriaceae strains

Exclusi

Swab Neutral

Skin Excision Neutral

Meat excision Neutral

Other Neutral

Counts at different stages

Inclusio

Factors influencing the counts

Inclusio

Relation between visual faecal contamination and counts of indicator bacteria

Inclusio

None Exclusi

as considered, is that

Batch characteristics

Neutral

Slaughterhouse features

Neutral

Sampling method Neutral



hors.

71

on

ion

l

l

l

l

on

on

on

ion

l

l

l

E.


Appendix B. Data Collection

The forms used for the collection

General information

Questions

Reference ID

Type of reference

Start year of the study

End year of the study

Country where the study was conducted

Type of study

Study aim

Species considered in the study

Age (months)

Number of plants involved

Plant 1 - Number of animal slaughtered/hour

Plant 1 - Number of animals slaughtered/day

Plant 1 - Number of animal slaughtered/year





n

n of data from the selected papers are reported bel

Answers

Paper

Other

(YYYY)

(YYYY)

Observational

Experimental



Relationship between visual faecal contamination and counts

Cattle

Swine

Sheep

Goat

Horses

1

2

3

More



hors.

72

low.

E.







Does the paper describe the use of any decontaminant treatments?

Sampling method

If a swab method was used, specify:

Sampling area on the carcasses:

Type of sample

Sampled area on the carcass

Dimension of the sampled area (cm2

Sample size (number)




Yes

No

NR

Swab

Skin excision

Meat excision

:

Sponge

Dry and wet

Gauze

Delimited

Undelimited

Not specified

Single

Pool

Neck

Brisket

Flank

Rump

Jowl

Belly

Back

Ham

Other 2) Single sample area (cm2)

E. coli



hors.

73

E.


Indicator bacteria considered

Analytical method

Significance level (alfa) e.g. 0.05

Power (1-beta) e.g. 0.80

Effect size (delta)

Data collection

Questions

Reference ID

Study aim

Species considered

Risk factor considered

Classification of faecal contamination (how carcasses were classified)

When was faecal contamination evaluated?

Indicator bacteria considered

Unit of enumeration




Enterobacteriaceae

Both

ISO

Alternative plate count

Petrifilm

MPN

Other

Answers



Relationship between visual faecal contamination and counts

Bovine

Swine

Goat

Sheep

Ante-mortem

Post-mortem

Enterobacteriaceae

E. coli

CFU/g

CFU/cm2

logCFU/g



hors.

74

E.


Does the paper describe how negative results have been considered?

Experimental group n°

Slaughtering stage

Slaughtering stage features

Sampling region

Counts

Sampling region

Counts

Sampling region

Counts




logCFU/cm2

Other

Yes

No

No negative results

N

Mean

SD

Min

Max

5 percentile

95 percentile

Prevalence (%)

other

N

Mean

SD

Min

Max

5 percentile

95 percentile

Prevalence (%)

other

N

Mean

SD

Min

Max



hors.

75

E.


Sampling region

Counts

Stage/s where counts decrease

Stage/s where counts increase

Effect of the RF on carcass counts

Effect of faecal contamination on carcass counts

Final considerations (comments)




5 percentile

95 percentile

Prevalence (%)

other

N

Mean

SD

Min

Max

5 percentile

95 percentile

Prevalence (%)

other

Increase

Decrease

None

Increase

Decrease

None



hors.

76

EFSA supporting publication 2014:EN-634 The present document has been produced and adopted by the bodies idenSafety Authority and the author(s), awarded following a tender procedurean output adopted by the Authority. The European Food Safety Authorityto the rights of the authors.

Appendix C. General characteristics of the selec

Table 6: General characteristics of the 41 paperEB: Enterobacteriaceae)

Reference (country) Type of study

Sam

Type of sample

Blagojevic et al., 2012 (RS) O Swab (sponge)

Yang et al., 2012 (CA) O Swab (sponge)

Nero 2012 (BR) O Swab

Serraino et al., 2012 (IT) O Swab (sponge)

Hauge et al., 2012 (NO) O Swab (sponge)

Bass et al., 2011 (AU) O Swab (sponge)


Bohaychuk et al., 2011 (CA) O Swab (sponge)

Ozdemir et al., 2010 (TR) O Skin excision

McCleery et al., 2008 (GB) O Meat excision

Rigobelo et al., 2008 (BR) O Swab (sponge)

Zweifel et al., 2008 (CH) O Skin excision

Algino et al., 2007 (US) E Swab (sponge)

Ruby et al., 2007 (US) O Swab (sponge)

Trivedi et al., 2007 (US) E Swab (sponge)

Tergney and Bolton, 2006 (TR) O Swab (sponge)

Bosilevac et al., 2006 (US) E Swab (sponge)

Kiermeier et al., 2006 (AU) O Swab

Kinsella et al., 2006 (IE) O Meat excision

Rigobelo et al., 2006 (BR) O Swab (sponge)

Anderson et al., 2005 (US) E Swab (sponge)

E. coli and Enterobacteriaceae counts on pi

ntified above as author(s). This task has been carried out exclusively by the author(s) in the cone. The present document is published complying with the transparency principle to which the Ay reserves its rights, view and position as regards the issues addressed and the conclusions reach

cted papers

rs providing data on beef carcasses. (E: experimental; O: obser

mple characteristics Indicator bacteria

Method characterist

Region sampled Analytical method Unit o

brisket, flank, rump, neck EB Petrifilm lo

randomly selected EC Membrane filter log

not specified EC - EB Petrifilm lo

brisket, groin, hock EC - EB ISO lo

brisket, belly EC NMKL log brisket, flank, rump, jowl,

midloin EC Petrifilm lo

brisket, flank, rump, perianal region, neck EB Petrifilm

not specified EC Alternative plate count

brisket, flank, rump EC - EB ISO lo

brisket, flank, rump, neck EB ISO lo

rump EC Qualitative analysis pres

brisket, flank, rump, neck EB Alternative plate count lo

brisket, flank, rump EC - EB Alternative plate count loinside - outside area + navel-

plate-brisket-shank area EB Petrifilm log

rump, midline, neck EB Alternative plate count lo

brisket, flank, rump, anus, hock EC - EB ISO - Alternative plate count

brisket, foreshank, top round surface, anus hock EB Petrifilm - Bactometer log

not specified EC Qualitative analysis pres

brisket, flank, rump, neck EB ISO lo

near the anus EC Qualitative analysis pres

between the bung and the hock EC Alternative plate count log

ig and ruminant carcasses: a review

ntext of a contract between the European Food Authority is subject. It may not be considered as

hed in the present document, without prejudice

77

rvational; EC: Escherichia coli;

ics Review question

of enumeration

ogCFU/cm2 3

CFU/100 cm2 1

ogCFU/ cm2 1

ogCFU/ cm2 3

CFU/100 cm2 3

ogCFU/ cm2 2

CFU/ cm2 2

CFU/ cm2 2

ogCFU/ cm2 2

ogCFU/ cm2 3

sence/absence 1-2

ogCFU/ cm2 2

ogCFU/ cm2 2

CFU/100 cm2 1-2

ogCFU/ cm2 2

NR 2

CFU/100 cm2 2

sence/absence 2

ogCFU/ cm2 2

sence/absence 2

g CFU/swab 2

EFSA supporting publication 2014:EN-634 The present document has been produced and adopted by the bodies ideFood Safety Authority and the author(s), awarded following a tender prconsidered as an output adopted by the Authority. The European Food Swithout prejudice to the rights of the authors.

Table 6 (Continued): General characteristics of the 4EB: Enterobacteriaceae; NR: not reported)

Reference (country) Type of study Sa

Type of sampleCorantin et al., 2005 (CA) O Swab (gauze)

Retzlaff et al., 2005 (US) E Meat excision

Arthur et al., 2004 (US) O Swab (sponge)

Gill and Landers, 2004b (CA) O Swab (sponge)

Madden et al., 2004 (IE) O Swab (sponge)

McEvoy et al., 2004 (GB) O Swab (dry andwet)

Mies et al., 2004 (US) E Swab (sponge)

Prendergast et al., 2004 (IE) O Swab (dry andwet)

Gill et al., 2003 (CA) O Swab (gauze)

Gill and Landers, 2003a (CA) O Swab (gauze)

Gill and Landers, 2003b (CA) O Swab (gauze)

Minihan et al., 2003 (IE) O Swab (sponge)

Nou et al., 2003 (US) O Swab (sponge)

Sumner et al., 2003 (AU) O Swab (sponge)

Barboza et al., 2002 (VE) O-E Swab (sponge)

Collobert et al., 2002 (FR) O Skin excision

Hansson, 2001 (SE) O Swab (sponge)

Phillips et al., 2001 (AU) O Swab (sponge)

Gill and Bryant, 2000 (CA) E Swab (gauze)

Bacon et al., 2000 (US) O Swab (sponge)


entified above as author(s). This task has been carried out exclusively by the author(s) in the rocedure. The present document is published complying with the transparency principle to wh

Safety Authority reserves its rights, view and position as regards the issues addressed and the co

41 papers providing data on beef carcasses. (E: experimental; O: obse

ample characteristics Method characteris

e Region sampled Indicator bacteria Analytical method Unit not specified EC Petrifilm l

ventral midline area EC - EB Petrifilm l

) inside - outside area + plate-brisket-shank area EB Petrifilm - Bactometer log

) randomly selected +

contaminated area + adjacent to the contaminated area

EC Membrane filter C

) brisket EB ISO ld brisket, hock, cranial back,

bung EC - EB Alternative plate count l

) brisket, belly, round EC Petrifilm ld brisket EB ISO l

randomly selected EC Membrane filter



) rump, midline, neck EC - EB ISO log

) anal-hock area EB Petrifilm log

) Brisket, flunk, rump EC Petrifilm L

) brisket, renal site, neck EC Petrifilm l

not specified EB Alternative plate count l

) loin, sternum EC Alternative plate count

) brisket, flank, rump EC Petrifilm pre

randomly selected EC Membrane filter log ofrecov

) brisket, flank, rump EC Petrifilm log


context of a contract between the European hich the Authority is subject. It may not be onclusions reached in the present document,

78

ervational; EC: Escherichia coli;

stics Review question

of enumeration logCFU/cm2 2

logCFU/cm2 2

gCFU/100 cm2 1

CFU/100 cm2 3

logCFU/cm2 1

logCFU/cm2 1

logCFU/cm2 2

logCFU/cm2 2

cfu/100cm2 1

g CFU/100cm2 2

g CFU/100cm2 2

CFU/1000 cm2 2

g CFU/100 cm2 2

Log CFU/cm2

logCFU/cm2 1-2

logCFU/cm2 2

CFU/cm2 2

esence/absence 2 f the total number vered from 2500

cm2 2

g CFU/100 cm2 1


Table 7: General characteristics of the 31 papersEB: Enterobacteriaceae; HGMFT: Hydrophobic Gri

Reference (country) Type of study Type of sample

Jones and Johns, 2012 (CA) O -E Swab (gauze)


Bohaychuk et al., 2011 (CA) O Swab (sponge)


Tomovic et al., 2011 (RS) E Swab (sponge)

Duggan et al., 2010 (IE) O Swab (sponge)

Hamilton et al., 2010 (AU) E Meat excision

Purnell et al., 2010 (UK) E Swab (sponge)

Wu et al., 2009 (DK) O Swab (sponge)

Algino et al., 2009 (US) O-E Swab (sponge)

Lenahan et al., 2009 (IE) O Swab (sponge)

Hurd et al., 2008 (US) O Swab (sponge)

Delhalle et al., 2008 (BE) O Swab

Zweifel et al., 2008 (CH) O Skin excision

Nesbakken et al., 2008 (NO) O Swab

Lindblad et al., 2007 (S) O Swab (sponge)

Rabaste et al., 2007 (CA) O Swab (sponge)

Saucier et al., 2007 (CA) E Swab (gauze)

Trivedi et al., 2007 (US) E Swab (sponge)

Conter et al., 2006 (I) E Skin excision

Inthavong et al., 2006 (LA) O Swab




s providing data on swine carcasses. (E: experimental; O: observd Membrane Filtration technique)

Sample characteristics Indicator bacteria

Method cha

e Region sampled Analytical method

random, anal EC HGMFT

) brisket, rump, ventral jowl (if possible) EC Petrifilm

) adjacent areas same carcass side EC ISO

) jowl, belly, ham, perianal EB Petrifilm

) jowl, belly, back, ham EB ISO

) jowl, belly, back, ham EB ISO

Belly strip EC Petrifilm

) anus EB ISO

) left leg close to the anus EC Petrifilm

) jowl, belly, ham EC - EB Petrifilm

) whole side EB ISO

) bung, skin, pleura EB ISO

ham, pelvis, forelimb, sternum EC ISO

neck, belly, back, ham EB ISO

ham EC ISO

) neck, belly, back, ham EC ISO

) brisket, internal rib cage, front feet top EC HGMFT

flank, thoracic inside EC HGMFT

) jowl, belly, ham EB ISO

jowl, belly, back, ham EC ISO

jowl, belly, ham EB ISO



79

vational; EC: Escherichia coli;

racteristics Review question

Unit of enumeration logCFU/100 cm2 (MPN) 1

logCFU/cm2 2

CFU/cm2 2

CFU/cm2 2

logCFU/cm2 2

logCFU/g 1

logCFU/g 2

logCFU/ cm2 2 logCFU/100

cm2 1

logCFU/ cm2 2

logCFU/ cm2 1

Prevalence 1

logCFU/ cm2 2

logCFU/ cm2 2

logCFU/ cm2 1

logCFU/ cm2 2 logCFU/983 cm2 (MPN) 2

logMPN 2

logCFU/ cm2 2

logCFU/ cm2 2

logCFU/cm2 1


Table 7 (Continued): General characteristics of thecoli; EB: Enterobacteriaceae; HGMFT: Hydrophobi

Reference (country) Type of study Type of sam

Spescha et al., 2006 (CH) O Swab (dry and

Namvar and Warriner et al., 2006 (CA) O Swab (spong

Gill and Landers, 2004a (CA) O Swab (gauz

Warriner et al., 2002 (UK) O Swab (spong

Tamplin et al., 2001 (US) O Swab (spong

Hansson, 2001 (S) O Swab (spong

Gill and Badoni, 2001 (CA) O Swab (spong

Gill et al., 2000b (CA) O Swab (spong

Rivas et al., 2000 (ES) O-E Swab

Bryant et al., 2003 (CA) E Swab




e 31 papers providing data on swine carcasses. (E: experimental; O: c Grid Membrane Filtration technique)


Method cha

mple Region sampled Analytical method

wet) neck, belly, back, ham EB ISO

ge) brisket EC Petrifilm

ze) contaminated + randomly selected EC HGMFT

ge) brisket EC-EB ISO

ge) neck, belly, ham EC Petrifilm

ge) loin, sternum EC ISO

ge) sticking wound EC HGMFT

ge) random EC HGMFT

neck, abdomen EC - EB ISO

back, jowl, ham, belly EC HGMFT



80

observational; EC: Escherichia

aracteristics Review question Unit of

enumeration

logCFU/ cm2 1

logCFU/100cm2 1

logTotal (MPN) 1 -2

CFU/100 cm2 1

CFU/ cm2 1

CFU/ cm2 2

MPN 2

MPN 1

logCFU/ cm2 1-2

MPN/100 cm2 1


Table 8: General characteristics of the 21 papers proEB: Enterobacteriaceae)

Reference (country) Type of study Type of sample

Rubio et al., 2013 (ES) O Skin excision

Jordan et al., 2012 (AU) O Swab (sponge)


Hauge et al., 2011a (NO) O Swab (sponge)

Hauge et al., 2011b (NO) O Swab (sponge)

Martineli et al., 2011 (BR) O Swab

Milios et al., 2011 (GR) O Swab (sponge)

Feizullah and Daskalov, 2010 (BG) O Swab (sponge)

Lenahan et al., 2010 (IE) O Swab (dry and we

Brown et al., 2009 (GB) E Meat excisionBhandare et al., 2007 (IN) O Swab Byrne et al., 2007 (IE) O Swab (sponge)

Fisher et al., 2007 (GB) O Skin excision Kannan et al., 2007 (US) O Swab (sponge)

Cohen et al., 2006 (MA) O Meat excision

Yalcin et al., 2004 (TR) O Skin excision Sumner et al., 2003 (AU) O Swab (sponge)

Whyte et al., 2002 (GB) O Swab (dry and we

Phillips et al., 2001 (AU) O Swab (sponge)Duffy et al., 2001 (US) O Swab (sponge)

Gill et al., 2000a (CA) O Swab (sponge)




oviding data on small ruminant carcasses. (E: experimental; O: obse


Method chara

Region sampled Analytical method

brisket, rump EB Petrifilm

rump and scapular region EC-EB Petrifilm

brisket, flank, midloin EC Petrifilm outside (mid-line of the abdomen, under the forelegs, around rectum and hind legs) and

inside the carcass EC - EB Alternative plate count

Brisket EC Alternative plate count

forequarter - hindquarter leg EC MPN

brisket, flank, rump, shoulder EB Alternative plate count

leg, chest-outer and inner surface, neck EC - EB ISO

et) brisket, flank, lateral thorax, breast EB ISO

lateral thorax, flank, brisket, breast EB Alternative plate count neck, brisket, flank, rump, shoulder EC Alternative plate count

brisket, flank, rump, shoulder EB Alternative plate count

rump, belly, flank, brisket, shoulder, neck EB Alternative plate count brisket, flank, leg EC Petrifilm

not specified EC Alternative plate count

neck, brisket, leg, shoulder EB Alternative plate count brisket, flank, rump EC Petrifilm

et) shoulder, abdomen and lateral surface of the rear leg EB Alternative plate count

brisket, flank, rump EC Petrifilm flank, breast, leg EC Petrifilm

two sites for each carcass selected from a grid of 43 areas on one side of the carcass

surface EC Membrane filter



81

ervational; EC: Escherichia coli;

acteristics Review question Unit of

enumeration log CFU/cm2 2

CFU/cm2 2

log CFU/cm2 2

log CFU/carcass 2

log CFU/100 cm2 1-2-3

log CFU/cm2 1

log CFU/cm2 1-2

log CFU/cm2 2

log CFU/cm2 1

log CFU/cm2 2 log CFU/cm2 1

log CFU/4000cm2 3


log CFU/g 2


log CFU/cm2 2-3


CFU/100 cm2 2


Appendix D. Counts at different stages along the

Table 9: Beef - Counts of E. coli reported at different

Reference Stages of the slaughterline

Additional operations - decontamination treatments

Gill

et a

l., 2

003

(CA

)

Killing Hide removal

washing carcasses and spray with lactic acid (2%)

Evisceration vacuum/hot water clean

Trimming washing

pasteurization and spray with 2% lactic acid

Cooling

Yan

g et

al.,

201

2 (C

A)


washing spraying with lactic acid (5%)

Evisceration Trimming

washing spraying with lactic acid (5%) pasteurization (steam at > 90°C)

Cooling

* hide-on carcasses were washed with 1.5% of sodium hy ND: none detected




e slaughterline

t stages of the slaughterline. (N: number of samples; D: decrease; I: in

Sampling points Sampled region N % Counts

randomly selected 425

1) after hide removal 52 3.16

2) after spraying with lactic acid 76 3.41

3) before washing 68 3.43

4) after washing 64 2.57

5) after pasteurization and spraying with lactic acid 4 0.3

6) after chilling 32 2.18

randomly selected 200

1) after hide removal* 12 1.62

2) after washing 8 0.303) after spraying with lactic acid 0 ND

4) after evisceration 0 ND 5) after trimming 20 1.436) after washing 8 1.00

7) after spraying with lactic acid 0 ND 8) after pasteurization 0 ND

ydroxide at 55°C and then washed with chlorine water



82

ncrease; ND: not detected)

s Unit of enumeration D I

log

of th

e to

tal n

umbe

r rec

over

ed fr

om 2

5 sa

mpl

es (c

fu/1

00 c

m2 )

X

X

X

log

of th

e to

tal n

umbe

r rec

over

ed fr

om

25 sa

mpl

es (c

fu/1

00 c

m2 )

X X

X X


Table 9 (Continued): Beef - Counts of E. coli report


Additional operations - decontamination treatments

McE

voy

et a

l., 2

004

(GB

)

Killing

Hide removal

Evisceration

Trimming splitting

washing with warm water (35-40°C)

Cooling 24 h Killing

Hide removal

Evisceration

Trimming splitting


Cooling 24 h

Killing

Hide removal Evisceration Trimming splitting


Cooling 24 h Killing

Hide removal

Evisceration

Trimming splitting


Cooling 24 h




ed at different stages of the slaughterline. (N: number of samples; D:

Sampling points Sampled region N % Counts Un

enum

hock

1728

2

1) after hide removal 41.7 2.91

2) after evisceration 25 2.74

3) after splitting

4) after washing 16.7 2.02

5) after chilling 5.9 1.3

brisket


2) after evisceration 30.6 2.79

3) after splitting



cranial back

1) after hide removal 0 0 2) after evisceration 5.6 1

3) after splitting



bung


2) after evisceration 33.4 2.67

3) after splitting

4) after washing 26.6 2




83

decrease; I: increase)

nit of meration D I

log

N: l

og o

f the

tota

l num

ber o

f rec

over

ed/c

m2

X

X

X

X

X

X

X

X


Table 9 (Continued): Beef - Counts of E. coli reporreported) * min - max 8 plants sampled; ** min - max


Additional operationsdecontamination treatm

Bac

on e

t al.,

200

0 (U

S)


steam vacuum of spot contam(plants 1,2,3,4,5,6,7,8

washing (water 29 - 38°(plants 1,2,3,4,5)

lactic (1,6 - 2,6%) or acetirinsing (plants 1,2,3,4


thermal pasteurization(plants 1,2,3,4,5,6,7,8

washing (16 to 32°C)(plants 1,2,3,4,5,6,7,8

lactic (1,6 - 2,6%) or acetirinsing (plants 1,2,3,4,6,

Cooling 24 h (plant 5,6,7,8) 36(plants 1,2,3,4)

Bar

boza

et a

l., 2

002

(VE)

Killing Hide removal Evisceration

Trimming

Cooling

Ner

o et

al.,

201

2

(BR

)

Killing Hide removal Evisceration Trimming

Cooling

Rig

obel

o et

al

., 20

08 (B

R) Killing

Hide removal Evisceration Trimming Cooling




rted at different stages of the slaughterline. (N: number of samples; Dx considering the three sampling area s -

ments Sampling points Sampled region N % Counts enu

flank, brisket, rump 1280

1) after hide removal 2.6 - 5.3*

mination 8)

°C)

c acid 4)

n 8)

) 8) 2) after final washing 1.0 - 3.0*

c acid 7,8)

h 3) after chilling 0.9*

neck, brisket, renal site 192

1) after hide removal 0.3 - 0.5**

2) after splitting 1.3 - 1.7**

3) after washing 1.2 - 1.4**

1) after bleeding

NR 65

2) after hide removal 3) after evisceration NR

4) after last washing

rump near the anus 216

1) after hide removal 58

NR 2) after splitting and trimming 38

3) after last washing in the cooler 32



84

D: decrease; I: increase; NR: not

Unit of umeration D I

leas

t squ

ares

mea

ns fo

r the

log1

0 va

lues

(lo

g C

FU/1

00cm

2)

X

X

X

log

CFU

/cm

2 X

X

NR

NR X X


Table 10: Beef - Counts of Enterobacteriaceae repor


Additional operations - decontamination treatment

Arth

ur e

t al.,

200

4 (U

S)

Killing

high-pressure water rinses ansteam vacuum

Hide removal steam vacuum

wash cabinet - cold water + lacacid (2-3%)


wash cabinet (water 90°C) wash cabinet (peroxyacetic aci steam pasteurization cabinet

Cooling

Rub

y et

al.,

200

7 (U

S)

Killing

Hide removal trim and steam vacuum spraying with lactic acid (5%


steam vacuum spray with ambient temp. Wat hot water lactic acid (5%)

Cooling




rted at different stages of the slaughterline. (N: number of samples; D:

ts Sampling points Sampled region N % Count

inside and outside area + navel-plate-brisket-foreshank

area

288

nd 1) before hide removal 6.2

(on thhide)

2) after hide removal 1.4

ctic

3) after evisceration 1.7

id) t

4) after decontamination treatments 0.2

5) after spray chilling (29 h) 0.4

inside and outside + navel-plate-

brisket-shank area 18989

1) before hide removal 4.9

2) after hide removal 1.04%)

3) after evisceration 0.8ter

4) after decontamination treatments - 0.38

5) after chilling (36-48h) 0.2



85

: decrease; I: increase)

ts Unit of enumeration D I

he )

log

cfu/

100

cm2

X

X

log

CFU

/100

cm

2

X

8 X

X


Table 10 (Continued): Beef - Counts of EnterobaI: increase)


Additional operations - decontamination treatmen

McE

voy

et a

l., 2

004

(GB

)

Killing Hide removal Evisceration Trimming splitting

washing with warm water (35-4Cooling 24 h Killing






washing with warm water (35-4Cooling 24 h




acteriaceae reported at different stages of the slaughterline. (N: nu

ts Sampling points Sampled region N % Count

hock

1728

1) after hide removal 69.4 3.442) after evisceration 58.3 2.75

3) after splitting 40°C) 4) after washing 44.5 3.03


brisket

1) after hide removal 75 2.842) after evisceration 63.9 2.82

3) after splitting 63.9 2.7140°C) 4) after washing 63.9 2.45


cranial back




bung






86

umber of samples; D: decrease;

ts Unit of enumeration D I

log

N: l

og o

f the

tota

l num

ber o

f rec

over

ed/c

m2

X

X

X X

X

X


Table 10 (Continued): Beef - Counts of EnterobaI: increase; NR: not reported)


Additional operations - decontamination treatmen

Ner

o et

al.,

201

2 (B

R)

Killing Hide removal Evisceration Trimming

Cooling

Mad

den

et a

l., 2

004

(IE)



high-pressure washing with hot

Cooling




acteriaceae reported at different stages of the slaughterline. (N: nu

ts Sampling points Sampled region N % Coun

1) after bleeding

not reported 65

NR2) after hide removal NR3) after evisceration NR

NR4) after last washing NR

NR

brisket 100

1) after hide removal 0.7

2) after splitting 0.6water

3) after washing 1.02



87

umber of samples; D: decrease;

nts Unit of enumeration D I

R R X R NR R R R

log

cfu/

cm2

7

3

2


Table 11: Swine - Counts of E. coli reported at D: decrease; I: increase; NR: not reported) * reduction as

Reference Operations

Jones and Johns 2012 (CA)

10 sec 86°

10 sec 86°

Wu et al. 2009 (DK)

(singeing, polishing, evisceration)

Nesbakken et al. 2008 (NO) Air Blast. -21.9° 1h

Namvar and Warriner 2006 (CA)

CO2 (tunnel) 65° for 5 min

Dry polishing and wash

Lactic acid 1,5%

CO2 (tunnel) 65° for 5 min

Dry polishing and wash

Lactic acid 1,5%

Gill and Landers 2004a (CA)

Warriner et al. 2002 (UK)

Scalding

Dry pol./singeing/wet pol.

Tamplin et al. 2001 (US)

Gill et al. 2000b (CA)

After evisceration




different stages of the slaughterline. (N: number of samples; M: ssessed according to the t-paired test

Sampling points Sampled region N % Counts (M and SD)

Before pasteurization Anal 25 100 3.80 (0.93) After pasteurization Anal 25 100 2.42 (0.94) After evisceration Anal 25 84 1.32 (1.40)

After washing Anal 25 64 -0.09 (1.27) Before pasteurization Random 25 92 1.79 (1.13) After pasteurization Random 25 64 -0.07 (1.11) After evisceration Random 25 56 -0.33 (1.17)

After washing Random 25 44 -0.67 (0.97) After stunning Left leg

105

5.07 (4.95-5.18After scalding Left leg 4.14 (3.92-4.36After splitting Left leg 2.03 (1.89-2.18After cooling Left leg NR

Before chilling Ham 60 NR After chilling Ham 60 NR After bleeding Brisket 12 4.84 +/- 0.85After scalding Brisket 12 <1,17 After scraping Brisket 12 4.01+/-1.23

After evisceration Brisket 12 <1,17 After splitting Brisket 12 <1,17 After washing Brisket 12 <1,17 After bleeding Brisket 12 4.48+/- 0.65After scalding Brisket 12 <1.17 After scraping Brisket 12 3.51 +/- 0.88

After evisceration Brisket 12 <1.17 After splitting Brisket 12 <1.17 After washing Brisket 12 <1.17 Before chilling Random 25 64 1.64

After chilling Random 25 64 1.11 After bleeding Brisket 10 90 3.86×10^4 After scraping Brisket 10 100 6.39×10^3 After polishing Brisket 10 100 5.18×10^2 After washing Brisket 10 100 5.44×10^2 After bleeding Neck, belly, ham 100 100 1700 After chilling Neck, belly, ham 122 30.1 1.1

After polishing Random 200 NR After washing Random 200 NR After chilling Random 200 NR



88

mean; SD: standard deviation;

Unit of enumeration D I

Log CFU/100 cm2

X X X

Log CFU/100 cm2

X

) Geometric mean log CFU/100 cm2 (CI)

- ) X ) X

X

Log CFU/cm2 X*

Log CFU/100 cm2

X X

X

X X

X

log cfu of the total number for 2500

cm2

X

CFU/100 cm2

X X

CFU/ cm2

MPN CFU 100 cm2


Table 11 (Continued): Swine - Counts of E. colideviation; D: decrease; I: increase; NR: not reported)

Reference Operations

Rivas et al. 2000 (ES)

Bryant et al. 2003 (CA)




reported at different stages of the slaughterline. (N: number of sa


After bleeding Neck, abdomen

216

3.36 (0.45) After scalding Neck, abdomen 0.10 (0.16) After dehairing Neck, abdomen 0.45 (0.42) After scraping Neck, abdomen 0.05 (0.12)

After evisceration Neck, abdomen 1.06 (0.98) End Neck, abdomen 1.16 (0.97)

After dehairing Random 25 92 1.25(1.02) After shaving Random 25 100 1.29 (0.81)

Before head removal Random 25 4 NR After head removal Random 25 56 NR after pasteurization Random 25 12 NR

After washing Random 25 16 NR



89

amples; M: mean; SD: standard


logCFU/ cm2

X X

MPN logCFU/ cm2 (SD)

X

X


Table 12: Swine - Counts of Enterobacteriaceae repoD: decrease; I: increase; NR: not reported)

Reference Sampling points Sa

Duggan et al. 2010 (IE)

Before chilling HamAfter chilling Ham

Lenahan et al., 2009 (IE)

Before chilling After chilling

Hurd et al. 2008 (US) Before scalding

During evisceration After evisceration

Inthavong et al. 2006 (LA)

Before evisceration HamAfter evisceration Ham

Spescha et al. 2006 (CH)

After bleeding NeckAfter scalding NeckAfter dehairing NeckAfter singeing NeckAfter polishing NeckAfter trimming NeckAfter washing Neck

After head removal NeckAfter chilling NeckAfter bleeding NeckAfter scalding Neck

After dehairing and singeing NeckAfter polishing NeckAfter trimming NeckAfter washing NeckAfter chilling Neck

Warriner et al. 2002 (UK)

After bleeding After scraping After polishing

After evisceration and washing


After bleeding NAfter scalding NAfter dehairing N

After singeing and scraping NAfter evisceration N

End N



orted at different stages of the slaughterline. (N: number of samples;

ampled region N % Counts (M and SD)

m, back, belly, jowl 30 NR m, back, belly, jowl NR

Whole side 480 Not aggregated Whole side 480 Not aggregated

Skin 7 82.1 Bung 7 pool of 5 96.4 Pleura 7 pool of 5 92.9

m, back, belly, jowl 62 2.81 (23-3.1) m, back, belly, jowl 2.98 (2.1-3.3) k, belly, back, ham 100 100 4.57; 4.51; 4.57; 4.65 k, belly, back, ham 100 0-6 1.38; 1.94; 1.,20; NR k, belly, back, ham 100 82-98 3.95; 3.84; 4.47; 4.56 k, belly, back, ham 100 12-42 2.51; 1.86; 3.05; 2.62 k, belly, back, ham 100 7-27 2.60; 2.02; 1.51; 2.20 k, belly, back, ham 100 13-49 2.96; 2.28; 2.51, 1.,94 k, belly, back, ham 100 18-36 2.68; 2.47; 2.18; 2.08 k, belly, back, ham 100 19-36 2.74; 2.16; 3.29; 2.23 k, belly, back, ham 100 0-14 1.98; 0.60; 0.09; NR k, belly, back, ham 100 99-100 6.10; 6.05; 6.08; 6.11 k, belly, back, ham 100 2-22 3.26; 1.68; 1.64; 1.64 k, belly, back, ham 100 56-86 6.09; 5.20; 5.55; 5.78 k, belly, back, ham 100 76-87 4.09; 4.00; 3.68; 3.56 k, belly, back, ham 100 83-92 4.65; 4.59; 4.27; 4.40 k, belly, back, ham 100 69-85 4.35; 3.,86; 3.89; 3.87 k, belly, back, ham 100 17-43 3.62; 2.68; 3.00; 2.,84

Brisket 10 90 1.05 x 106

Brisket 10 100 1.32 x 104 Brisket 10 100 1.,84 x 103 Brisket 10 100 4.29 x 103

Neck, abdomen - 3.54 (0.20) Neck, abdomen - 0.12 (0.29) Neck, abdomen - 0.84 (0.40) Neck, abdomen - 0.23 (0.19) Neck, abdomen - 1.18 (0.84) Neck, abdomen - 1.39 (0.98)




90

M: mean; SD: standard deviation;


Log CFU/cm2

Log CFU/ cm2 X

X

Log CFU/ cm2 X

Log CFU/ cm2 logN(Neck,

Belly,Back, Ham)

X X

X X

X

X X

X X

X

CFU/100 cm2

X X X

Log CFU/ cm2

X X

X X


Table 13: Small ruminants - Counts of E. coli reporNR: not reported)

Reference (country) Stages of the slaughterline Operations

Bhandare et al., 2007 (IN) NR NR

Hauge et al., 2011b (NO)

Skinning

Fleece removal -manual and

mechanical (invertedressing)

Cooling After trimming,

grading and steamvacuum treatment

Martineli et al., 2011 (BR)

Pelt removal

NR Evisceration

Washing



rted at different stages of the slaughterline. (N: number of samples; M


After flying Pool (neck, shoulder, flank, brisket, rump)

32 11.1 3.55 (0.08)

After washing 32 8.3 3.11 (0.05)

After evisceration Pool (brisket, rib and flank) 32 20.8 3.95 (0.06)

ed After removal of fleece

(start of slaughter) Brisket

35

NR

2.79

m t

Before chilling (End of slaughter) 35 2.99

After skinning Forequarter legs 30

NR

0.31 (0.84) Hindquarter legs 30 0.03 (0.16)

After evisceration Forequarter legs 30 0.08 (0.21) Hindquarter legs 30 0.10 (0.25)

After washing Forequarter legs 30 0.07 (0.00) Hindquarter legs 30 0.00 (0.00)




91

M: mean; SD: standard deviation;

Unit of enumeration Increase Decrease

log CFU/cm2

- -

- -

X -

log CFU/100 cm2

- -

X -

log CFU/cm2

X - - - - - X - - X - X


Table 14: Small ruminants - Counts of Enterobacterdeviation; NR: not reported)

Reference (country) Stages of the slaughterline Operations

Yalcin et al., 2004 (TR)

Pelt removal

NR Evisceration Washing Cooling

Lenahan et al., 2010 (IE) 5 Abattoirs NR

Milios et al., 2011 (GR)

Pelt removal

NR

Evisceration



riaceae reported at different stages of the slaughterline. (N: number of

Sampling points Sampled region N % Counts (M

and SD) After dressing Pool (Leg;

shoulder, brisket, neck)

44

NR

0.38 (0.16) After evisceration 44 0.75 (0.21)

After washing 44 0.58 (0.21) After chilling (24 h) 44 0.11 (0.08)

Before chilling Pool (flank, lateral thorax, breast, brisket)

400

NR

-0.35-1.16

After chilling 400 -0.32-0.43

A- After pelt removal of hind and forelegs/ before

final pulling Pool (rump,

flank, brisket,

shoulder)

60

NR

0.76 (0.80)

B-After pulling/before evisceration 60 2.27 (0.53)

C-After evisceration/before pluck removal 60 2.68 (0.62)

D-After pluck removal/before chilling 60 2.90 (0.55)




92

f samples; M: mean; SD: standard

Unit of enumeration Increase Decrease

log CFU/cm2

- - X - - - - X

log CFU/cm2

- -

- X

log CFU/cm2

- -

X -

X -

X -


Appendix E. Risk factors: detailed results

Table 15: Beef - Counts at different stages of the slaNR: not reported; SD: standard deviation)

Reference (country) Treatment/ Risk factor Specifications E

Rigobelo et al., 2006 (BR)

sampling season Rigobelo et al., 2008 (BR)

Phillips et al., 2001 (AU)

slaughterhouse

plant throughput

slaughslaught

loSumner et al., 2003 (AU) lo

Bohaychuk et al., 2011 (CA)

lohi

Hansson, 2001 (SE) hilo

Ozdemir et al., 2010 (TR)

Bass et al., 2011 (AU)

type of surveillance

applied in the plant

co

Tergney and Bolton, 2006 (TR)

application of an online

monitoring system to

monitor faecal contamination

applic



aughterline reported in papers describing risk factors influencing E. co

Experimental group Sampling points Sampled region N % Co

rain season pre-evisceration - post evisceration - post processing

rump, near the anus 80

30 - 70 - 27.5

dry season 22.5 - 55 - 17.5

rain season pre-evisceration - post evisceration - post processing

rump, near the anus 216

44

dry season 20

hterhoused for the export after chilling

(12 h)

rump, flank, brisket

1275

11.3 terhouses for the domestic

market 8.8

ow throughput plants 7.9 slaughterhouse after chilling

(8-48 h) flank, brisket 159

28.4 ow throughput plants 4.7 ow throughput plants

during chilling NR 1036 -0.54

gh throughput plants -0.23 gh throughput plants at the end of the

slaughterline loin,

sternum 200 34

ow throughput plants 41 high throughput

NR rump, flank, brisket

120 - 2.0

low throughput - 1.9

o-regulatory system after chilling (at

least 4 hours) flank, brisket NR

- rep

aggrtraditional system -

cation of the monitoring - yes

final inspection stand (before

trimming)

anus, rump,

brisket, flank, hock

180

-

no d-




93

oli counts (N: number of samples;

unts (SD) Unit of enumeration

Treatment/ RF effect

- - yes

- -

- - yes

- -

- -

no - -

- - - -

yes - -

(-0.68/-0.40) log CFU/cm2 no

(-0.43/-0.02) - -

no - -

07 (0.13) log CFU/cm2 no

90 (0.08)

ported as regated data

- no

-

etailed data log CFU/cm2 yes


Table 15 (Continued): Beef - Counts at different stagof samples; SD: standard deviation)* log of the total nReference (country) Treatment/

Risk factor Specifications Exp

Retzlaff et al., 2005 (US)

decontamination

steam pasteurization /temperature

71.71.173.73.976.76.779.479.482.282.285.85.087.87.8

Minihan et al., 2003 (IE)steam

pasteurization

pu

pa

Corantin et al., 2005 (CA)

b

Gill and Bryant, 2000 (CA)

hot water pasteurization / temperature

8 sec8 s

9 sec9 s

10 se10 s

11 se11 s

12 se12 s



ges of the slaughterline reported in papers describing risk factors influnumber recovered from 2500 cm2 perimental group Sampling

points Sampled region N % Co

1 °C pre-treatment

pre rigor - before/after treatment

ventral midline area 280

5 Min: 01 °C post-treatment 20 Min: 09 °C pre-treatment 15 Min: 09 °C post-treatment 5 Min: 07 °C pre-treatment 5 Min: 07 °C post-treatment 0 4 °C pre-treatment 20 Min: 04 °C post-treatment 10 Min: 02 °C pre-treatment 10 Min: 02 °C post-treatment 0 0 °C pre-treatment 30 Min: 00 °C post-treatment 0 8 °C pre-treatment 5 Min: 08 °C post-treatment 0 untreated neck

pre chilling neck, midline, rump 30

100 0.pasteurized neck 97 0.7untreated midline 93 0.7asteurized midline 77 0.6untreated rump 90 0.9treated rump 67 0.4

before treatment Not reported randomly

selected sites 1003 14.2 0.0

after treatment 1.8 0.0c - before treatment

pre chilling randomly selected sites 250

96 ec- after treatment 64 c - before treatment 84 ec- after treatment 52 ec - before treatment 80 sec- after treatment 12 ec - before treatment 76 sec- after treatment 32 ec - before treatment 76 sec- after treatment 12




94

uencing E. coli counts (N: number



0.4 Max: 0.8

log CFU/cm2 no

0.4 Max: 5.0 0.4 Max: 1.7

no 0.4 Max: 0.8 0.4 Max: 0.8

no <0.4

0.4 Max: 6.6

no 0.4 Max: 4.1 0.4 Max: 1.7 marginally

effective <0.4 0.4 Max: 2.5

yes <0.4

0.4 Max: 0.8 yes

<0.4 84 (0.45)

log CFU/1000 cm2 yes

79 (0.37) 76 (0.39) 60 (0.44) 98 (0.77) 47 (0.34) 06 (0.19)

log CFU/cm2 yes 01 (0.05)

2.95

CFU/100 cm2* yes

1.81 2.91 1.66 3.08 0.95 3.88 1.08 3.58 0.85


Table 15 (Continued): Beef - Counts at different stagof samples; SD: standard deviation)

Reference (country) Treatment/ Risk factor Specifications

Mies et al., 2004 (US)

decontamination

different treatments

Barboza et al., 2002 (VE) different treatments

Algino et al., 2007 (US) different treatments

miml

hh



ges of the slaughterline reported in papers describing risk factors influ

Experimental group Sampling points Sampled region N % Co

(single water wash

pre slaughter - before/after treatment

brisket - belly -

inside round 120 -

double water wash water wash with 0,5% lactic acid water wash with 50 ppm chlorine

lactic acid (before treatment)

after washing - before/after treatment

neck 192 -

lactic acid (after treatment) nisin (before treatment) nisin (after treatment)

lactic acid + nisin (before treatment) lactic acid + nisin (after treatment) <dry-aging 4 days before treatment

prechilling before/after treatment

flank, brisket, rump

265

9 0dry-aging 4 days after treatment 0 -

dry-aging 6 days before treatment 61 -0dry-aging 6 days after treatment 3 -

dry-aging 7 days before treatment 21 0dry-aging 4 days after treatment 11 -0

acetic acid spray before treatment 47 -0acetic acid spray after treatment 13 -

ixture of different acids before treatment 24 -0mixture of different acids after treatment 7 -low pressure hot water before treatment 24 0low pressure hot water after treatment 3 -

high pressure hot water before treatment 18 -0high pressure hot water after treatment 3 -




95

uencing E. coli counts (N: number

ounts SD)

Unit of enumeration


- log CFU/cm2 yes

1.1

log CFU/cm2

yes 0.5 1.0

no 1.0 1.0

yes <0.2 0.64

log CFU/cm2 yes

1.38 0.76 1.35 0.10 0.66 0.34 1.05 0.33 1.23 0.13 1.24 0.13 1.29


Table 15 (Continued): Beef - Counts at different(N: number of samples; SD: standard deviation)


Gill and Landers, 2003a (CA)

decontamination treatments

comparison of 4 plants

plant

plant

plant

plan

plant

plan

* log10 of the total number of E. coli recovered from 25 samples



t stages of the slaughterline reported in papers describing risk fa

Experimental group Sampling points

Sampled region N % Cou

plant A after skinning

- randomly selected

sites 500

60 A after washing and spraying with lactic acid 56

plant A before washing 52 plant A after washing 52

t A after spraying with peroxyacetic acid then steam pasteurization 0 None

plant B after skinning 20 B after washing and spraying with lactic acid 52

plant B before washing 92 plant B after washing 92

nt B after steam pasteurization and spraying with lactic acid 0 None

plant C after skinning 48 C after washing and spraying with lactic acid 60

plant C before washing 60 plant C after washing 32

nt C after steam pasteurization and spraying with lactic acid 0 None

plant D after skinning 64 plant D before washing 68 plant D after washing 60

plant C after spraying with lactic acid 32




96

ctors influencing E. coli counts.


Treatment/RF effect

2.97

CFU/100 cm2* Yes

2.76 2.63 2.51

e detected

2.20 2.04 3.56 3.19

e detected

3.99 2.74 2.78 2.45

e detected

4.01 4.16 3.16 2.93


Table 15 (Continued): Beef - Counts at different (N: number of samples; NR: not reported; SD: standa

Reference (country) Treatment/Risk factor Specifications

Gill and Landers, 2003b (CA) spray chilling comparison of 4 plants

Anderson et al., 2005 (US) diet different feed-water treatments




t stages of the slaughterline reported in papers describing risk facard deviation) * log10 of the total number of E. coli recovered from

Experimental group Sampling points Sampled region N %

plant A before cooling

cooling - before/after treatment

randomly selected

sites 200

24plant A after cooling 4

plant B before cooling 0plant B after cooling 4

plant C before cooling 12plant C after cooling 16

plant D before cooling 44plant D after cooling 68

administration of sodium chlorate - water NR bung and

hock 64 -administration of sodium chlorate - water no treatment



97

ctors influencing E. coli counts m 25 samples

Counts (SD) Unit of enumeration


1.08

CFU/100 cm2*

controversial effect

0.0 none detected

0.0 0.48 1.40 1.59 2.98

reported as aggregated

data s

- no -

-


Table 16: Beef - Counts at different stages of th(N: number of samples; NR: not reported; SD: standsamples; * prevalence obtained on the entire carcassReference (country) Treatment/Risk factor Specifi

Ruby et al., 2007 (US) sampling sea

samplin/ye

Ruby et al., 2007 (US)

slaughterhouse

loca

Prendergast et al., 2004 (IE) plant d

Ozdemir et al., 2010 (TR) plathrou

Collobert et al., 2002 (FR) pla

Blagojevich et al., 2011 (RS) pla

Zweifel et al., 2008 (CH) pla




he slaughterline reported in papers describing risk factors influencdard deviation) § Data reported for different stages of the slaughterlin

ications Experimental group Sampling points

Sampled region N % Counts (S

ason colder months

different stages of the slaughterline

inside outside round +

navel-plate-brisker-

shank area

18989

- Aggregatdatawarmers months

ng time ear

2005 - 0.972006 - 1.04

ation East Aggregat

data§Midwest - Southwest

design

A: linear rail - one floor end of the

slaughterline, after washing

brisket NR - -0,61

B: serpentine rail - two floors - -0,57

ant ughput

high throughput NR

rump, flank, brisket

120 2.18 (0.1

low throughput 1.98 (0.1

ant

A end of the

slaughterline

233

- 0.70B NR - 1.30-1.4C - 0.60-0.6D - 1.18-1.6

ant

A hide before dehiding

rump, flank,

brisket, neck

100

1.97 x 1B hide 2.92 x 1

A dressed carcass before chilling

1.06 x 1B dressed carcass 0.59 x 1

ant

A

NR neck,

brisket, flank, rump

50 30* not detecB 50 20* Max: 2.2C 50 55* Max: 2.3D 50 33.3* Max: 1.6E 50 23.3* Max: 1.9F 50 33.3* Max: 1.6G 50 8.3* Max: 1.6H 50 6.7* not detecI 50 16.7* not detecJ 50 3.3* not detecK 50 16.7* Max: 1.9L 50 0.0* not detecM 50 11.7* Max: 1.6P 50 8.3* not detecQ 50 36.7* Max: 4.2



98

cing Enterobacteriaceae counts ne; ^ counts obtained on brisket

SD) Unit of enumeration


ted

logCFU/100cm2

yes

yes

ted yes

logCFU/cm2 no

11) no 12)

log/CFU/cm2 yes 43 66 65 02

mean CFU/cm2 no 02 01 01 ted

log CFU/cm2

some differences

were detected among plants

20^ 30^ 60^ 90^ 60^ 6^ ted ted ted 9^ ted 6^ ted

24^


Table 16 (Continued): Beef - Counts at different stcounts (N: number of samples; SD: standard deviatio

Reference (country) Treatment/ Risk factor Specifications Expe

Retzlaff et al., 2005 (US)

decontamination

steam pasteurization /

temperature

71.1 °71.1 °73.9 °73.9 °76.7 °76.7 °79.4 °79.4 °82.2 °82.2 °85.0 °85.0 °87.8 °87.8 °

Minihan et al., 2003 (IE)

steam pasteurization

unpasunt

pastunt

Trivedi et al., 2007 (US)

steam pasteurization

befimmedia

24 h

Nou et al., 2003 (US) chemical dehairing

tre

untr




tages of the slaughterline reported in papers describing risk factors n)

erimental group Sampling points

Sampled region N % Counts (SD

°C pre-treatment

pre rigor - before/after treatment

ventral midline

area 280

20 Min: 0.4 Max: °C post-treatment 25 Min: 0.4 Max: 2°C pre-treatment 25 Min: 0.4 Max: °C post-treatment 5 Min: 0.4 Max: °C pre-treatment 25 Min: 0.4 Max: °C post-treatment 15 Min: 0.4 Max: °C pre-treatment 30 Min: 0.4 Max: °C post-treatment 10 Min: 0.4 Max: °C pre-treatment 5 Min: 0.4 Max: °C post-treatment 5 Min: 0.4 Max: °C pre-treatment 45 Min: 0.4 Max: °C post-treatment 0 <0.4 °C pre-treatment 15 Min: 0.4 Max:°C post-treatment 0 <0.4 ntreated neck

prechilling neck,

midline, rump

30

90 1.71 (1.06)steurized neck 57 0.85 (0.95)treated midline 97 1.92 (0.73)eurized midline 70 0.96 (0.84)ntreated rump 97 2.25 (1.20)treated rump 87 1.46 (0.90)fore treatment

before/after treatment

neck, midline,

rump 72

- 1.36 ately after treatment - 0.52 h after treatment - 0.50 ated carcasses pre

evisceration before/after treatment

anal-hock area 480

- 1.4 (0.7)

reated carcasses - 3.2 (1.0)



99

influencing Enterobacteriaceae

D) Unit of

enumeration


9.1

log CFU/cm2

no 24.8 1.7

no 1.7 3.3

no 17.3 9.9

no 4.1

17.3 marginally effective 1.7

9.9 yes

5.0 yes

log CFU/1000

cm2 yes

log/CFU/ cm2 yes

log CFU/ 100 cm2 yes


Table 16 (Continued): Beef - Counts at different stcounts (N: number of samples; SD: standard deviatio


Bosilevac et al., 2006 (US)

decontamination

Comparison of different

treatments lactilacti

Algino et al., 2007 (US) different treatments

drd

drd

drd

acam

mixtulow plow

high high

Kinsella et al., 2006 (IE) chilling spray chilling concon




tages of the slaughterline reported in papers describing risk factors n) #logN: log of the total number recovered/cm2 - mean of the four si

Experimental group Sampling points Sampled region N %

lactic acid before treatment

pre evisceration before/after treatment

brisket, foreshank,

anus-hock, top round

768

- lactic acid after treatment - hot water before treatment - hot water after treatment -

ic acid + hot water before treatment - ic acid + hot water before treatment - ry-aging 4 days before treatment

prechilling before/after treatment

flank, brisket, rump 265

9 dry-aging 4 days after treatment 3 ry-aging 6 days before treatment 84 dry-aging 6 days after treatment 33 ry-aging 7 days before treatment 26 dry-aging 4 days after treatment 16 cetic acid spray before treatment 58 acetic acid spray after treatment 30 mixture of different acids before

treatment 28

ure of different acids after treatment 22 pressure hot water before treatment 27

w pressure hot water after treatment 12 pressure hot water before treatment 19

h pressure hot water after treatment 15 spray chilling before treatment

before/after chilling neck, flank, brisket, rump 30

- spray chilling after treatment -

nventional chilling before treatment - nventional chilling after treatment -



100

influencing Enterobacteriaceae tes sampled

Counts (SD)

Unit of enumeration

Treatment/RF effect

4.0

log CFU/100 cm2 yes

3.0 4.4 1.7 4.7 2.2

0.96

log CFU/cm2 yes

-1.21 -0.08 -0.87 0.78 0.37 0.15 -0.84

0.04

-0.57 0.57 -0.86 0.22 -0.66 2.30#

log CFU/cm2 no 2.25# 2.55# 2.39#


Table 17: Swine - Counts at different stages of thsamples; NR: not reported) Reference Treatment/Risk

Factor Experimental g

Bohaychuk et al. 2011 (CA)

Plant throughput (cut-off 8000)

Low throughpHigh throughp

Hamilton et al. 2010 (AU)

Hot water-acidified sodium chlorite Abattoir

A and B

A-ControlB-Hot water(83C-Acidic treatm

A-ControlB-Hot water(81C-Acidic treatm

Algino et al. 2009 (US)

Skinning Unskinned

Skinned

Washing T°

Washing T <12

Washing T 12.8°

Washing T 21.1

Washing T >3

Chilling time 1 day chillin

2 days chillin

Delhalle et al., 2008 (BE)

Mixed models considering different RF

Spraying if ext temTime between killing a

Scalding with sWashing and disinfecti

machine three timeLairage cleaning wi

Frequency of lairage dNew hooks for carcasses

Lindblad et al. 2007 (S) Season

Gill and Landers 2004a (CA)

Detained carcass, contamination Visibly contaminated



he slaughterline reported in papers describing risk factors influencin

group Sampling points Sampled region N %

put During chilling NR 1069 33.7% 0.0put During chilling NR 0.2

Belly 42 92.9% 03.5°) Belly 41 9.8% -ment Belly 40 12.5% -

Belly 150 69.3% 01.9°) Belly 150 22% -ment Belly 100 30% 0

d Before washing Bell, ham, jowl 121 After chilling Bell, ham, jowl

Before washing Bell, ham, jowl 60 After chilling Bell, ham, jowl

2.8° Before washing Bell, ham, jowl 26 After chilling Bell, ham, jowl

°-21.1° Before washing Bell, ham, jowl 97 After chilling Bell, ham, jowl

-32.2 Before washing Bell, ham, jowl 42 After chilling Bell, ham, jowl

32.2 Before washing Bell, ham, jowl 16 After chilling Bell, ham, jowl

ng Before washing Bell, ham, jowl 112 After chilling Bell, ham, jowl

ng Before washing Bell, ham, jowl 69 After chilling Bell, ham, jowl

mp hot During chilling Ham, back, forelimb, sternum

584

0and scalding During chilling Ham, back, forelimb, sternum team During chilling Ham, back, forelimb, sternum -ion splitting es a day During chilling Ham, back, forelimb, sternum -

ith water During chilling Ham, back, forelimb, sternum -disinfection During chilling Ham, back, forelimb, sternum -before chilling During chilling Ham, back, forelimb, sternum -

Before chilling Ham,back,Belly,Neck 541 57% 0

d carcasses

Before trimming Contaminated site 25 60% After trimming Contaminated site 25 80% After chilling Contaminated site 25 32%

Before trimming Random 25 72% After trimming Random 25 76% After chilling Random 25 32%




101

ng E. coli counts (N: number of

Counts Unit of enumeration RF effect

1(-0.15-0.16) LogCFU/cm2 23(0.06-0.40) 0.89 (0.11)

LogCFU/g

0.83 (0.21) X 0.75 (0.19) X 0.45 (0.08) 0.65 (0.11) X

0.60 (0.,13) X -0.3

logCFU/ cm2

X -1. 2 0.5

-0.61 1.01

X

-0.71 -0.43 -1.24 0.09 -0.91 0.34 -0.31 -0.29

X -1 -0.04 -1.01

0.59 (0.22)

model parameter

logCFU/ cm2

0.23 (0.1) 0.65 (0.29) X

0.89 (0.37) X

0.56 (0.08) X 0.76 (0.15) X 0.69 (0.08) X 0.05 (0.58) logCFU/ cm2 X

4.78

log of the total number for 2500 cm2

2.92 X 1.7 X

2.37 2.24 1.69 X


Table 17 (Continued): Swine - Counts at differen(N: number of samples; NR: not reported)

Reference Treatment/Risk

Factor Experimental group

Hansson 2001 (S) Plant throughput Low throughput (450-80

pigs/year) High throughput >250,0

Gill and Badoni 2001 (CA)

Pasteurization, sticking wound

trimming


Washing and GMP (anal plugging

before evisceration)

Unwashed-non GMPUnwashed, GMP Washed and GMP

Bass et al. 2011 (AU) Plants features

B D L M N O P

Saucier et al. 2007 (CA)

Meal frequency (2; 5), Feed type

(pelleted; mash), fasting time (4; 14;

24 h)

P-2-4h P-2-4h

P-2-14h P-2-24h P-5-4h

P-5- 14h P-5- 24h M-2-4h M-2-14h M-2-24h M-5-4h M-5-14h M-5-24h

Rabaste et al. 2006 (CA)

Handling Rough Gentle

group size 10 30

Conter et al. 2006 (I)

Steam treatment (after evisceration)

Steam treatment Control



nt stages of the slaughterline reported in papers describing risk fa

Sampling points Sampled region N %

00 Before chilling Loin, sternum 100 58%

000 Before chilling Loin, sternum 100 74% Before pasteurization Sticking wound 25 80% After pasteurization Sticking wound 25 28%

After trimming Sticking wound 25 40% Before chilling Neck, abdomen 36 1Before chilling Neck, abdomen 54 0Before chilling Neck, abdomen 35 1During chilling Rump, brisket, jowl 19 42% During chilling Rump, brisket, jowl 5 40% During chilling Rump, brisket, jowl 17 29% During chilling Rump, brisket, jowl 8 12.5% During chilling Rump, brisket, jowl 5 60% During chilling Rump, brisket, jowl 20 95% During chilling Rump, brisket, jowl 3 10%

After evisceration Thoracic area, flanks 8 100% 1After evisceration Thoracic area, flanks 8 37.5% 1After evisceration Thoracic area, flanks 8 25% 1After evisceration Thoracic area, flanks 8 25% 1After evisceration Thoracic area, flanks 8 37.5% 1After evisceration Thoracic area, flanks 8 25% After evisceration Thoracic area, flanks 8 87.5% 2After evisceration Thoracic area, flanks 8 37.5% 1After evisceration Thoracic area, flanks 8 50% 2After evisceration Thoracic area, flanks 8 25% 1After evisceration Thoracic area, flanks 8 12.5% After evisceration Thoracic area, flanks 8 62.5% 1After evisceration Thoracic area, flanks 8 25%

NR Int Rib cage, brisket, feet 10 52% 1NR Int Rib cage, brisket, feet 10 68% 1NR Int Rib cage, brisket, feet 10 65% 1NR Int Rib cage, brisket, feet 10 55% 1

After evisceration Jowl, belly, back, ham 54 After evisceration Jowl, belly, back, ham 32




102

actors influencing E. coli counts

Counts Unit of enumeration

RF effect

logCFU/ cm2 X

2.09 logCFU/100 cm2

0.85 X 1.56

1.20 (0.72) logCFU/ cm2

0.24 (0.43) X 1.13 (0.34)

-0.6

logCFU/ cm2

-0.6 0.06 -0.45 -0.27 -0.17 -0.55

1.31 (0.43)

CFU/926 cm2 from MPN

estimate X

1.77 (0.14) 1.89 (0.54) 1.99 (0.55) 1.79 (0.25) 1.8 (0.30) 2.07 (0.49) 1.81 (0.22) 2.18 (0.60) 1.83 (0.25) 1.8 (0.28) 1.92 (0.34) 2.1 (1.12) 1.45 (0.69)

CFU/983 cm2

from MPN estimate

1.49 (0.57) 1.48 (0.58) 1.46 (0.67)

NR logCFU/ cm2 NR


Table 18: Swine - Counts at different stages of th(N: number of samples; D: decrease; I: increase)

Reference Treatment/ Risk Factor Experimental group

Purnell et al. 2010 (UK) Anal plugging

Plugged Plugged Aft

Unplugged Unplugged Aft

Algino et al. 2009 (US)

Skinning Unskinned

Skinned

Washing T°

Washing T <12.8°

Washing T 12.8°-21.1°

Washing T 21.1-32.2

Washing T >32.2

Chilling time 1 day chilling

2 days chilling

Zweifel et al. 2008 (CH) Plant features

A B C D E F G H I J K L M N O




he slaughterline reported in papers describing risk factors influenc

Sampling points Sampled area on the carcass N % Coun

Before scalding Anal region 34 2.18 (1ter scalding and dehairing Anal region 34 1.91 (0

Before scalding Anal region 34 1.86 (1ter scalding and dehairing Anal region 34 3.01 (1

Before washing Bell, ham, jowl 121 0.57After chilling Bell, ham, jowl -0.86

Before washing Bell, ham, jowl 60 1.28After chilling Bell, ham, jowl 0.41


Before washing Bell, ham, jowl 97 0.31After chilling Bell, ham, jowl -1





Before/during chilling Neck, belly, back, ham 50 22% 2.30;2.92;2Before/during chilling Neck, belly, back, ham 50 22% 2.60;2.45;2Before/during chilling Neck, belly, back, ham 50 32% 2.20;NR;3.Before/during chilling Neck, belly, back, ham 50 44% 3.84;2.56;3Before/during chilling Neck, belly, back, ham 50 8% 3.17;2.38;1Before/during chilling Neck, belly, back, ham 50 14% 2.08;2.45;2Before/during chilling Neck, belly, back, ham 50 56% 3.15;2.51;3Before/during chilling Neck, belly, back, ham 50 14% 2.51;ND;NBefore/during chilling Neck, belly, back, ham 50 14% 1.90;1.60;1Before/during chilling Neck, belly, back, ham 50 26% 1.90;2.20;3Before/during chilling Neck, belly, back, ham 50 10% 2.38;1.60;2Before/during chilling Neck, belly, back, ham 50 22% 2.72;2.45;3Before/during chilling Neck, belly, back, ham 50 16% 2.30;2.38;2Before/during chilling Neck, belly, back, ham 50 56% 2.72;2.98;3Before/during chilling Neck, belly, back, ham 50 2% 1.60; ND; N



103

cing Enterobacteriaceae counts

nts Unit of enumeration

RF Effect

.28)

logCFU/cm2

0.95)

.01)

.08) X 7

logCFU/cm2

6 X 8 1 X 5 2 X 1

X 6 4 X 9 1 4 3 X 8 5 X

2.38;1.90

logCFU/cm2 (log Total Number

recovered)

2.80;2.45 3.1;3.56 .44;3.17 .90;2.08

2.20;1.60 .08;3.30

ND;1.60 .90;2.64 .09;2.51

2.45;1.60 .29;2.20

2.60;2.08 .45;3.20 ND; ND


Table 18 (Continued): Swine - Counts at different counts (N: number of samples; D: decrease; I: increas

Reference Treatment/ Risk Factor Experimental group S

Trivedi et al. 2006 (US)

Steam decontamination 60 sec 82-85° on

3 anatomical areas

25 % of carcasses acid treated Befo

25 % of carcasses acid treated Aft

75% of carcasses acid treated 24h a


Washing and GMP (anal

plugging before evisceration)

Unwashed-non GMP Unwashed, GMP

Washed and GMP

Tomovic et al. 2011 (RS)

Chilling length (3 h at -31.1°

followed by 2-4° chill room (8 or

24h)

8 h

24 h

Blagojevic et al. 2011 (RS) Plant features

A

B




stages of the slaughterline reported in papers describing risk factors se)

Sampling points Sampled area on the carcass N % Counts

fore steam treatment

Ham 72 134 Belly 72 219 Jowl 72 212 All 72 61,6% 188

ter steam treatment


after steam treatment


Before chilling Neck, abdomen 36 121 (074) Before chilling Neck, abdomen 54 025 (045)

Before chilling Neck, abdomen 35 014 (032)

Before chilling Ham, back, belly, jowl 24 70% 053 After chilling Ham, back, belly, jowl 54% 039

Before chilling Ham, back, belly, jowl 24

58% 04

After chilling Ham, back, belly, jowl 25% 009

After stunning Ham, belly, jowl 50 778x103(162x105-537x10)

Before chilling Ham, belly, jowl 50 894x10(141x103-063x10)

After stunning Ham, belly, jowl 50 419x103(204x105-468x10)

Before chilling Ham, belly, jowl 50 097x10(955x102-009x10)



104

influencing Enterobacteriaceae

Unit of enumeration

RF effect

logCFU/cm2

X

logCFU/cm2

X

logCFU/cm2

X

-

CFU/cm2 geometric

mean

X

-

X


Table 19: Small ruminants - Counts at different stagE. coli counts (N: number of samples; NR: not report

Reference (country)

Treatment/Risk Factor Experimental group

Gill et al., 2000a (CA) Dressing procedure

Conventional set 1 (carcasuspended by the rear le

Conventional set 2

Inverted dressing 1 (carcasuspended by the forele

Inverted 2

Duffy et al., 2001 (US) Season Spring

Fall or winter

Phillips et al., 2001 (AU) Plant throughput

All the slaughterhouseExport slaughterhouse

Domestic slaughterhousLow throughput plant

Sumner et al., 2003 (AU) Plant throughput

Medium throughput plan(4)

Low throughput plants (

Cohen et al., 2006 (MA) Seasonal effect

Hot season (April to September)

Cold season (NovemberMarch)

Kannan et al., 2007 (US)

Pre-slaughter spray washing

Control (no wash) Washed (1 min spray was

- potable water)

Feizullah and Daskalov, 2010 (BG)

Capacity and Season

Low throughput (15-2animals per day)/SprinLow throughput/Winte

High throughput (100-2animals per day)/Sprin

Sampling point High throughput Low throughput




ges of the slaughterline reported in papers describing factors related toted; M: mean; SD: standard deviation)

p Sampling point/s Sampled region/s N % Counts (M and SD)

asses egs)

After the final wash in the

dressing procedure

Pool (Two sites for each carcass selected

from a grid that specifies 43 areas on

one side of the carcass surface)

25 15/25 total number o

bacteria recover4.40



asses eg) 25 11/25

total number obacteria recover

2.58



post 24h carcass chilling Pool (flank, breast, leg) 1259 NR 0.63

1261 0.76 es

After chilling Pool (rump, flank and brisket)

921 29.2 0.17 (0.60)es 270 35.2 NR ses 306 32.7 NR ts 345 21.4 NR nts After chilling (8-

48 h) Pool (rump, flank and

brisket) 148 61.5 0.39

(11) 216 18.5 -0.01

NR Raw meat 26 44.2 (on the

total samples)

1.2 (0.6)

r to 26 1.3 (0.6)

After skinning/before

evisceration Pool (flank, brisket, leg)

10 NR

2.1 shed 10 2.1

0 ng

After washing/Before

chilling

Pool (leg, chest -outer and inner surface, neck)

15

NR

1.00

er 15 1.65 200 ng 15 2.00

Leg (lateral) 12 2.00 12 0.7



105

o the slaughterhouse influencing

Unit of enumeration

Treatment /RF effect

of red:

CFU/100 cm2

-

of red: -

of red:

Yes of red:

log CFU/cm2 No

log CFU/cm2

- - -

Yes

log cfu/cm2 No

log CFU/g No

log CFU/cm2 No

log CFU/cm2

Marginally effective

Yes


Table 19 (Continued): Small ruminants - Counts at dand to the batch (one study) influencing E. coli countReference (country)

Treatment/Risk Factor Experimental group

Bass et al., 2011 (AU) Plant throughput

Plant A (goat) Plant C (sheep)

Plant D Plant E Plant F Plant G Plant H Plant I Plant J Plant K Plant L Plant M Plant N Plant O Plant P

Hauge et al., 2011a (NO)

Pasteurisation (82°C+-1 for

8 s): after dressing and

grading/ before chilling

Control A

Pasteurisation

A

Hauge et al., 2011b (NO) Shearing regime

Unshorn B

Shorn 0 days abattoir B

Shorn 3 days on farm B

Shorn 7 days on farm B

Jordan et al., 2012 (AU)

Inspection of lymph nodes

Pre-inspection

Post-inspection




different stages of the slaughterline reported in papers describing factts (N: number of samples; NR: not reported; M: mean; SD: standard d

Sampling point/s Sampled region/s N % Count

(M and

After chilling midloin, flank, brisket

29 29 -0.38

323

68.9 -0.2889.5 0.4126.8 -0.1641.4 0.0887.5 -0.2240 -0.20

57.9 -0.3360 -0.10

24.1 -0.3854.5 -0.0542.1 -0.3570 0.128.7 0.40

86.2 0.02After evisceration- dressing/before

chilling Outside (mid-

line of the abdomen, under

the forelegs, around rectum and hind legs) and inside the

carcass

90 66 (59/90) 2.39

24 h after chilling 90 43 (39/90) 0.98After evisceration-dressing/before

chilling 90 26 (23/90) 0.54

24 h after chilling 90 21 (19/90) 0.37

After removal of fleece (start of slaughter)

Brisket

35 80 2.79

Before chilling (End of slaughter) 35 2.99After removal of fleece (start of

slaughter) 35

62

1.78


slaughter) 35 1.49


slaughter) 35 1.73

Before chilling (End of slaughter) 35 2.69

NR Scapula 148 35 7 Rump 148 84 42

NR Scapula 148 67 22Rump 148 93 52



106

tors related to the slaughterhouse deviation). ts SD)

Unit of enumeration

Treatment /RF effect

8

log CFU/cm2

some differences

were detected among plants -

8

6

2 0 3 0 8 5 5

log CFU/carcass (4500 cm2)

-

Yes

Yes

Yes

log CFU/100 cm2

-

-

No

CFU/cm2 Yes


Table 20: Small ruminants - Counts at different stagEnterobacteriaceae counts (N: number of samples; N

Reference (country) Treatment/Risk Factor Experimen

Whyte et al., 2002 (GB) Dressing procedure

Cradle dressing (the carrear legs before the p

manuaHybrid method (use of th

lamb while the pelt waforelegs and brisket; then

a vertical positioFrame method (carcass c

an optimum working popelt to hang down and aduring the pelt removal

Fisher et al., 2007 (GB) Skin on vs

conventionally dressed

No toast (singe, wasToa

Toast beforeToast after

Toast before Toast after i

Skin Conventional (s

Brown et al., 2009 (GB)

Vascular perfusion chilling

(very fine ice particles in a solution of

sodium chloride and water)

Cont

Treatment (4 mi

Feizullah and Daskalov, 2010 (BG)

Slaughterhouse throughput and

Season

Low throughpLow throughpHigh throughpHigh throughp




ges of the slaughterline reported in papers describing factors related toNR: not reported; M: mean; SD: standard deviation)

tal group Sampling point/s Sampled region/s N %

rcass was hung by the pelt was pulled off ally)

After skin removal

Pool (Shoulder,

abdomen and lateral surface of the rear leg)

48

NR

he cradle to support the as released from the n the carcass is lifted in on (inverted))

48

arcass manipulation at osition and encourage away from the carcass l - inverted dressing)

48

sh and eviscerate) Before evisceration

Pool (rump, belly, flank,

brisket, shoulder, neck)

60 24/60 ast After evisceration 60 3/60 e splitting Before splitting 60 10/60 splitting After splitting 60 7/60 inspection Toasted and then inspected 60 6/60

inspection Inspected and then toasted 60 4/60 on After singeing/pre chilling 60 2/60

skin removed) After dressing/pre-chilling 60 26/60

trol After dressing/Before chilling Pooled surface

samples (lateral thorax, flank, brisket,

breast)

10 1/10

ins perfusion) At 24 h after held in the chill-room 10 1/10

put/ Spring

After washing/Before chilling

Pool (leg, chest -outer and

inner surface, neck)

15

NR put/ Winter 15 put/ Winter 15 put/ Autumn 15



107

o the slaughterhouse influencing

Counts (M and SD)

Unit of enumeration

Treatment/ RF Effect

0.47

log CFU/cm2 Yes 0.14

0.03

0.964

log CFU/cm2

Yes 0.433 0.569

No 0.509 0.424

No 0.429 0.420

Yes 0.955

1.4

log CFU/cm2 No 2.0

1.30

log CFU/cm2

Yes 3.18 1.27 Yes 6.05


Table 20 (Continued): Small ruminants - Counts at dinfluencing Enterobacteriaceae counts (N: number of

Reference (country) Treatment/Risk Factor Experimen

Hauge et al., 2011a (NO)

Pasteurisation (82°C+-1 for 8 s): after dressing and

grading/ before chilling

Cont

Pasteuri

Milios et al., 2011 (GR)

Steam application (8-10 passes of steam spraying

pistol at each side of the carcass)

Before Steam

After Steam A

Jordan et al., 2012 (AU)

Inspection of lymph nodes

Pre-insp

Post-insp

Rubio et al., 2013 (ES) Chilling

CT: Conventional trea

UT: Ultra rapid treatmen2°C until

ST: Slow treatment (12 °24h




different stages of the slaughterline reported in papers describing factf samples; NR: not reported; M: mean; SD: standard deviation)

tal group Sampling point/s Sampled region/s N %

trol After evisceration-

dressing/before chilling Outside (mid-

line of the abdomen, under the forelegs,

around rectum and hind legs) and inside the

carcass

90 100

24 h after chilling 90 NR

sation

After evisceration- dressing/before chilling 90 66

24 h after chilling 90 NR

Application After pluck removal/Before

chilling

Pool (rump, flank, brisket,

shoulder)

60

NR Application 60

pection NR Scapula 148 46 Rump 148 87

pection NR Scapula 148 77 Rump 148 96

atment (2°C for 24h) Before post -treatment (after

dressing)

Rump and brisket

20

NR

Post-mortem treatment 20

nt (-20°C for 3.5h then l 24 h)

Before post-mortem treatment (after dressing) 20


C for 7h then 2°C until h)

Before post-mortem treatment (after dressing) 20




108

tors related to the slaughterhouse

Counts (M and SD)

Unit of enumeration

Treatment/ RF

Effectiveness

3.78

log CFU/carcass (4500 cm2)

Yes

1.94

1.41

0.49

3.74 (0.51)

log CFU/cm2 Yes 2.67 (0.60)

11

CFU/cm2 Yes 66 26 57

3.24 (0.16)

log CFU/cm2

-

3.31 (0.16) -

3.04 (0.17) Yes

2.43 (0.17)

3.44 (0.17) -

3.22 (0.17) -


Appendix F. Faecal contamination

Table 21: Beef - Relationship between visual faecal of samples; NR: not reported; O: observational; SD: C1: category 1 - C2: category 2 - C3: category 3 - C4

Indicator bacteria

Reference (country) Study Plants Decontaminants S

EB McCleery et al., 2008 (GB) O 1 NR

EB Blagojevic et al., 2012 (RS) O 2 No

EB

Serraino et al., 2012 (IT) O 1 No b

EC

EC Gill and Landers, 2004b (CA) O 4

plant A: washing with 200 ppm

peroxyacetic acid; plant B: steam + 2% lactic acid; plant C:

hot water + 2% lactic acid; plant D:

2% lactic acid

c

c

EC Hauge et al., 2012 (NO) O 2 No




contamination and E. coli (EC), Enterobacteriaceae (EB) counts alonstandard deviation) * min - max values obtained for the 4 sampled sl

4: category 4 - C5: category 5

Sampled area on the carcass Classification of faecal contamination N Unit of

enumeration

neck, brisket, flank, rump

1-2 clean animals - 3-4 animals classed as dirty and requiring special provisions to be

taken before being slaughtered (clipped ante-mortem or online) - 5 animal rejected

362 Log CFU/cm2

neck, brisket, flank, rump

1: clean and dry; 2: slightly dirty; 3: dirty; 4: very dirty 100 Log CFU/cm2

brisket, groin, flank, hock 1 (clean-dry) to 5 (filthy-wet)

75

Log CFU/cm2

75

contaminated area + adjacent to the

contaminated area + randomly selected

areas

NR 75 log/ 2500 cm2

A

brisket, belly 0:clean; 1: moderately dirty; 2: very dirty 324 Log CFU/100 cm2



109

ng the slaughterline. (N: number laughterhouses; C0: category 0 -

Sampling point Experimental group

Mean (SD)

NR No punctual data were available

before chilling

C1 0.81 (0.74) C2 0.78 (0.63) C3 0.83 (0.68) C4 1.49 (0.60)

before chilling

C1 0.3 (0.01) C2 0.3 (0.01) C3 1.8 (2.2) C4 1.0 (1) C5 1.4 (0.9) C1 0.3 (0.01) C2 0.3 (0.01) C3 1.1 (1.1) C4 0.9 (1.1) C5 1.1 (0.8)

CA - before trimming

plant A, B, C, D

2.67* - 6.32* CA - after trimming 1.54 * - 5* CA - after dressing none detected* - 2.8*

ACA -before trimming 1.4* - 3-94* ACA after trimming 1.23* - 3.63* ACA after dressing none detected* - 3.09*

RS - before trimming 1.65* - 4.46* RS - after trimming 1* - 2.54* RS - after dressing none detected* - 1.91*

after dehiding C0 0.97 end of slaughterline C0 0.28




Table 22: Small ruminants - Relationship between v(N: number of samples; M: mean; O: observational)

Indicator bacteria Reference (country) Study Plant D

EC Hauge et al., 2011b (NO) O 1

EB Byrne et al., 2007 (IE) O 1

EB Whyte et al., 2002 (GB) O 1

* mean gross contamination score




visual faecal contamination and E. coli (EC), Enterobacteriaceae (EB

Decontaminants Sampled region Classification of faecal contamination N Unit of

enumeration

No brisket

0: visually clean fleece; 1: small spots of dirt; 2: generally dirty fleece;

3: very dirty fleece

35

log CFU/100 cm2 35 35 35

No brisket, flank, rump, shoulder

(A) clean and dry; (B) clean and wet; (C) dirty and dry; (D) dirty and wet; (E) with visible

faecal dags

40

log CFU/4000cm2

40 40 40 40

No

Shoulder

Score between 0 (no visible contamination)

and 10 (maximum visible contamination)

9

log CFU/cm2

Flank 9 Hindquarters 9 Lower foreleg 9

Shoulder 9 Flank 9

Hindquarters 9 Lower foreleg 9

Shoulder 9 Flank 9

Hindquarters 9 Lower foreleg 9



110

B) counts along the slaughterline

Sampling point

Experimental group M

At skinning

C0 1.65

C1 1.88 C2 2.16 C3 2.49

After skinning

A 2.7

B 2.9 C 4.4 D 3.9 E 4.4

After skinning

Cradle dressing 0.47

4.3* 2.7* 5.6* 9.0*

Hybrid dressing 0.14

3.4* 1.4* 0.8* 6.3*

Frame dressing 0.03

1.7* 1.6* 0.7* 4.4*

E.


ABBREVIATIONS

ACC: Aerobic colony count

AT: Austria

AU: Australia

BR: Brazil

BG: Bulgaria

CA: Canada

CZ: Czech Republic

ES: Spain

GB: United Kingdom

IE: Ireland

IN: India

LA: Lao People's Democratic Re

IT: Italy

MA: Morocco

NO: Norway

PHC: Process Hygiene Criteria

PHI: Process Hygiene Indicators

RS: Serbia

SE: Sweden

TR: Turkey

TW: Taiwan

US: United States

VE: Venezuela

ZA: South Africa




epublic



hors.

111

Escherichia coli and Enterobacteriaceae counts on pig and ruminant carcasses along the

Documents

Transcript of Escherichia coli and Enterobacteriaceae counts on pig and ruminant carcasses along the