The Pennsylvania State University

The Graduate School

College of Agricultural Sciences

IDENTIFICATION AND VALIDATION OF ANTIMICROBIAL

INTERVENTIONS FOR RED MEAT CARCASSES PROCESSED IN

VERY SMALL MEAT ESTABLISHMENTS

A Thesis in

Food Science

Sally Lucile Flowers

Submitted in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

August 2006 The thesis of Sally Lucile Flowers was reviewed and approved* by the following:

Catherine N. Cutter Associate Professor of Food Science Thesis Advisor Chair of Committee

Stephanie Doores Associate Professor of Food Science

William R. Henning Professor of Animal Science

Edward W. Mills Associate Professor of Dairy and Animal Science

Nancy Ostiguy Associate Professor of Entomology

John D. Floros Professor of Food Science Head of the Department of Food Science

*Signatures are on file in the Graduate School.

ABSTRACT

Carcass decontamination strategies for very small meat plants (VSMP) were developed by systematic experimentation. First, a survey of VSMP in Pennsylvania,

Washington and Idaho determined that knife trimming and cold water washing were the most common antimicrobial interventions in use. Secondly, red meat carcasses (n = 866) processed in VSMP in Pennsylvania, Washington, Idaho, and Texas were swabbed to determine prevalences of Salmonella spp., Campylobacter spp., and E. coli O157:H7 and enumerate hygiene indicators (mesophilic aerobic plate count, coliforms, and generic E. coli ). Thirdly, manual water washes (various temperatures, pressures, drip times, applications, and distances) and chemical rinses at various concentrations (acetic acid, citric acid, lactic acid, peroxyacetic acid, aqueous ozone, sodium hypochlorite, chlorine dioxide, and acidified sodium chlorite) were applied in inoculation challenge studies to reduce bacteria on inoculated beef surfaces. While hot water (77°C) yielded log reductions of 2.7 to 5.1

CFU/cm 2, there were quality issues with this intervention. Relative antimicrobial effectiveness was determined: organic acids > peroxyacetic acid > chlorinated compounds > ozone. Using this information, individual and combined effectiveness of water washing and/or rinsing with an antimicrobial compound was measured in a laboratory setting. The effectiveness of a portable, pressurized stainless steel spray tank to apply a 2% lactic acid rinse was compared with a garden sprayer, retrofitted garden sprayer, and motorized backpack sprayer. The findings from these studies indicated that a manual warm water wash

(54°C), followed by a 5 minute drip and 2% lactic acid rinse applied with a portable, pressurized steel tank, was the most effective multi-step antimicrobial carcass intervention

iii

(MSACI) with reductions of pathogens between 3.5 and 5.0 log CFU/cm 2 and 3.8 to 8.9 log

CFU/cm 2 for hygiene indicators. Instructional materials were developed to train VSMP employees how to implement MSACI in their plants. Lastly, red meat carcasses (n = 747) were swabbed to generate a second microbiological baseline after MSACI implementation.

The first and second baselines were statistically compared to validate the reduction in carcass hygiene indicators and prevalence of harmful pathogens. Overall, MSACI was deemed an effective and feasible method of improving the microbiological safety of carcasses processed in VSMP.

TABLE OF CONTENTS

List of figures ……………………………………………………………………………….xiv

List of tables ………………………………………………………………………………..xvi

Acknowledgments …………………………………………………………………….…xxviii

Chapter One: Preliminary Survey ………………………………………………………… 1

Statement of the problem……………………………………………..…………….....2

Research objective……………………………………………………..………….…..3

Experimental plan in brief………………………………………………..…………...4

Chapter Two: Literature Review ……………………………………..………………...… 6

Introduction………………………………………………………..……………...... 7

Large, small, and very small plants………………………..………………... 7

HACCP implementation……………………………………..……………… 8

Primary processing…………………………………………….……………. 9

Types of interventions………………………………………….…………... 11

Chemical and physical hazards………………………………………….…………. 14

Chemical hazards……………………………………………….………….. 14

Physical hazards………………………………………………….………… 16

Biological hazards……………………………………………………….………..... 18

Escherichia coli O157:H7……………………………………….………..... 18

Salmonella spp…………………………………………………….….……...22

Campylobacter spp………………………………………………….….……25

Bacterial attachment to meat surfaces…………………………………….…28

Antimicrobial interventions……………………………………………………..…...31

Water washing………………………………………………………….…....31

Rinsing with antimicrobial compounds…………………………………...... 34

Lactic acid…………………………….………………………….…..34

Acetic acid…………………………………….…………………..…37

Citric acid…………………………………………….……………...38

Peroxyacetic acid………………………………………………...…..40

Chlorinated compounds……………………………………………...41

Sodium hypochlorite………………………………………...….……41

Acidified sodium chlorite………………..………………….……….42

Chlorine dioxide…………………………..…………………...…….43

Aqueous ozone……………………………..……….…………..…...44

Steam vacuuming……………………………………..…………..……….....47

Steam pasteurization…………………………………..…………….…….....48

Chemical dehairing……………………………………...... ………………49

Knife trimming…………………………………………..…………………...50

Antimicrobial interventions used in combination………..…………………..51

Physical attributes of interventions………………………..…………………52

Future/potential interventions………………………………..…………...….53

Conclusions…………………………………………………………..………………56

References………………………………………………………………….………...57

Chapter Three: Preliminary Survey ………………………………………………………80

Introduction……………………………………………………………...…………..81

Methods…………………………………………………………………..………….83

Results…………………………………………………………………….………….85

Discussion………………………………………………………………………….. 93

References………………………………………………………………………….. 97

Chapter Four: Investigation of water washes on the elimination of meat borne pathogens from inoculated beef surfaces …………………………………………..…….127

Introduction………………………………………………………………..….…….128

Types of interventions………………………………………………..……..129

Selection of antimicrobial treatments………………………………..……..129

Water washing…………………………………………….………….….....130

Physical attributes of interventions…………………………………………131

Research Objectives………………………………………………………………...132

Methods………………………………………………………………………….…133

Preparation of fecal slurry…………………………………………………..133

Inoculation of beef plates with fecal slurry…………………..……………..134

Treatment of inoculated plates with cold water washes…….……………...135

Treatment of inoculated plates with cold, warm or hot water washes……..136

Enumeration/isolation of pathogens and hygiene indicators.………………………137

Detection of pathogens at low levels…………………….…………………138

E. coli O157:H7………………………………….…………………138

Salmonella spp. ………………………………….…………………139

Campylobacter spp. ……………………………….………………..139

Enumeration of hygiene indicators…………………………………………140

vii

Statistical analysis of cold water washes…………………….……………..140

Statistical analysis of cold, warm, and hot water washes…..………………141

Results……………………………………………………………….….………….142

Cold water washes…………………………………………..….…………..142

Cold, warm, and hot water washes………………………….….…………..144

Discussion………………………………………………………….….……………156

References…………………………………………………………….…………….152

Chapter Five: Investigation of chemical rinses on the elimination of meat borne pathogens from inoculated beef surfaces ………………………………….……………..172

Introduction…………………………………………………………………………173

Rinsing with antimicrobial compounds………………….…………....……173

Lactic, citric, and acetic acids………………….………………..….174

Peroxyacetic acid……………………………….………………..…175

Sodium hypochlorite……………………………….………….…....176

Acidified sodium chlorite…………………………………….…….177

Chlorine dioxide…………………………………….………….…...178

Aqueous ozone….……………………………………………….….179

Types of interventions……………………………………………..……..…180

Research objectives…………………………………………………………………182

Methods……………………………………………………………………………..183

Preparation of fecal slurry…………………………………………………..183

Inoculation of beef plates with fecal slurry…………………..……………..184

Treatment of inoculated plates with chemical rinses………………..……...185

viii

Monitoring of chemical residues……………………………………...……189

Enumeration of pathogens…………………………………………………190

Detection of pathogens at low levels………………………………………191

E. coli O157:H7……………………………………………………191

Salmonella spp. ……………………………………………………192

Campylobacter spp. ………………………………………………..192

Enumeration of hygiene indicators…………………………………………193

Statistical analysis of chemical rinses………………………………..……..193

Results………………………………………………………………………………195

Discussion…………………………………………………………………….…….201

References………………………………………………………………….….……209

Chapter Six: Investigation of water washes combined with chemical rinses on the elimination of meat borne pathogens from inoculated beef surfaces ………………..…232

Introduction…………………………………………………………………………233

Research objectives…………………………………………………………………235

Methods………………………………………………………………………...... 237

Preparation of fecal slurry…………………………………………………..237

Inoculation of beef plates with fecal slurry…………………..……………..238

Combined treatments of inoculated plates ………..……………………...... 239

pH of combination washes using different spraying systems………….…...242

Enumeration of pathogens……………………………………………….…242

Detection of pathogens at low levels…………………………………….…243

E. coli O157:H7………………………………………………….…243 ix

Salmonella spp. ………………………………………………….…243

Campylobacter spp. ………………………………………………..244

Enumeration of hygiene indicators…………………………………………245

Statistical analysis of combination treatments……………………………...245

Results…………………………………………………………………………...….247

Discussion…………………………………………………………………………..249

References…………………………………………………………………………..251

Chapter Seven: Establishment of microbiological baselines of red meat carcasses to validate a multi-step antimicrobial carcass intervention in very small meat plants ………………………………………………………………………………..…….. 266

Introduction…………………………………………………………………….…. .267

Biological hazards…………………………………………………….….... 269

Escherichia coli O157:H7……………………………………….….269

Salmonella spp. ……………………………………………….…... 273

Campylobacter spp. …………………………………………….… 276

Research Objective………………………………………………………………... 279

Methods…………………………………………………………………………… 280

Preparation of positive controls…………………………………………… 282

Collection of carcass swabs……………………………………………….. 283

Preparation of homogenate………………………………………...……… 284

Enumeration of hygiene indicators…………………………………………285

Detection of E. coli O157:H7………………………………………………286

Detection of Salmonella spp. ………………………………………………287

Detection of Campylobacter spp. ………………………………………….288

Statistical analysis…………………………………………………………. 289

Results……………………………………………………………………………... 290

Pathogen prevalence………………………………………………………. 291

Hygiene indicators………………………………………………………… 293

Discussion…………………………………………………………………………. 295

References…………………………………………………………………………. 302

Chapter Eight: Future Research …………………………………………………………332

References…………………………………………………………………………..337

Appendix A: Survey Questionnaire …………………………………………..………….338

Appendix B: Establishment of growth curves for Escherichia coli O157:H7, Salmonella

Typhimurium, Campylobacter coli , and Campylobacter jejuni …….………………..….344

Introduction……………………………………………………………..….……... 345

Methods……………………………………………………………………...…… 346

Results and discussion…………………………………………………………….. 350

References………………………………………………………………………… 352

Appendix C: Effect of diluent selection on pH of homogenized beef brisket treated with a 2% lactic acid rinse and on populations of Escherichia coli O157:H7, Salmonella

Typhimurium, and Campylobacter spp. …………………………………………………361

Introduction……………………………………………………………………….. 362

Methods……………………………………………………………….………...….363

Results and discussion…………………………………………………….………..367

References……………………………………………………………………….….370

Appendix D: Determination of carcass surface areas and application time for water washing and rinsing with 2% lactic acid ……………………………………………….. 375

Introduction………………………………………………………………….….... 376

Methods………………………………………………………………………..…. 378

Results and discussion………………………………………………………..….....379

References………………………………………………………………….……... 382

Appendix E: Video brochure ………………………………………………….……...... 386

Background and purpose………………………………………………….………..389

Step 1. Water wash………………………………………………………………...390

Step 2. Five-minute drip…………………………………………………………...394

Step 3. Antimicrobial rinse………………………………………………………...396

Suggestions for establishing a critical limit………….………………………….…404

Suggestions for monitoring a critical limit…………………………...…………....405

Suggestions for corrective actions……………………………………………....…405

Spray equipment selection…………………………………………………………406

Summary…………………………………………………………………………...415

Disclaimer………………………………………………………………………….416

Contact information………………………………………………………………..416

Appendix F: Cost comparison of chemical treatments used to decontaminate carcass

surfaces in very small meat establishments ……………………………………………..418

Introduction………………………………………………………………………..419

Spraying equipment………………………………………………………………..421

Antimicrobial compounds………………………………………………………….423

xii

Aqueous ozone……………………………………………………………...423

Chlorine dioxide……………………………………………………………425

Acidified sodium chlorite…………………………………………………..428

Sodium hypochlorite………………………………………………………..431

Peroxyacetic acid……………………….…………………………………..433

Citric, acetic, and lactic acids……………….……………………………...434

Acknowledgments…………………………………….……………………………439

References………………………………………………..…………………………440

Appendix G: Questions and comments from processors related to the implementation of a multi-step carcass decontamination treatment in very small meat plants ………..441

Introduction…………………………………………………………………….…..442

Question 1………………………………………………………………………….442

Question 2………………………………………………………………….………443

Question 3…………………………………………………………………….……444

Question 4………………………………………………………………….………444

Comment 1…………………………………………………………………….…...445

Comment 2………………………………………………………………………....445

Comment 3…………………………………………………………………….…...446

Comment 4…………………………………………………………………….…...446

References………………………………………………………………….………447

xiii

LIST OF FIGURES

Chapter Seven

Figure 1. Sampling locations for testing of cattle carcasses…………………..………..…. 314

Figure 2. Sampling locations for testing of swine carcasses………………….………..…. 315

Appendix B

Figure 1. Escherichia coli O157:H7 (ATCC 43889) growth curve (n = 2) and absorbance readings (n = 2)……………………………………………………….……………….…....353

Figure 2. Escherichia coli O157:H7 (PSUGDC 93-0133) growth curve (n = 2) and absorbance readings (n = 2).………………………………………………………….…….354

Figure 3. Salmonella Typhimurium (ATCC 14028) growth curve (n = 2) and absorbance readings (n = 2)………………………………………………………………………….….355

Figure 4. Salmonella Typhimurium (ATCC 13311) growth curve (n = 2) and absorbance readings (n = 2)……………………………………………………………………….….....356

Figure 5. Campylobacter coli (ATCC 33559) growth curve (n = 2) and absorbance readings

(n = 2)………………………………………………………..……………………….…...... 357

Figure 6. Campylobacter coli (ATCC 33560) growth curve (n = 2) and absorbance readings

(n = 2)………………………………………………………...……………………….….…358

xiv

Appendix E

Figure 1. Bacteria and water film on a carcass surface…………………………..………..394

Figure 2. 2% lactic acid rinse being applied to a carcass surface that has not been allowed to drip adequately…………………………………………….………………………………395

Figure 3. 2% lactic acid is diluted as it mixes with the water film on a carcass surface…..395

Figure 4. Heavy-duty stainless steel tank………………………………………….………406

Figure 5. Close-up view of top of heavy-duty stainless steel tank…….…………………..408

Figure 6. Garden sprayer…………………………………………………………………..410

Figure 7. Retrofitted garden sprayer…………………………………………….…...……411

Figure 8. Close-up view of retrofitted garden sprayer………………………….………....412

Figure 9. Backpack sprayer………………………………………………………………..414

Figure 10. Spray tank on wheels…………………………………………………………..414

LIST OF TABLES

Chapter Three

Table 1. Slaughter interventions used to reduce microbial contamination on carcass surfaces

in very small meat establishments in Pennsylvania…………………………….………...... 99

Table 2. Approximate application time of slaughter intervention to carcass surfaces in very

small meat establishments in Pennsylvania ……….………………………………….……100

Table 3. Types of equipment used to apply slaughter interventions in very small meat establishments in Pennsylvania …………....……………………………………………….101

Table 4. Timing of the application of slaughter interventions to carcass surfaces in very small meat establishments in Pennsylvania ………………………………………………………102

Table 5. Type of nozzle used in washing cabinets to apply hot water or organic acids in very small meat establishments in Pennsylvania …………………………………………….….103

Table 6. Application pressure of hand-held garden sprayer or spray cabinet used to apply hot water or organic acids to carcass surfaces in very small meat establishments in

Pennsylvania……………………………………..…………………………………………104

Table 7. Distance between hand-held hose or hand-held sprayer and carcass surfaces during application of hot water or organic acids in very small meat establishments in

Pennsylvania …………..……………………………………………………...……………105

Table 8. Proportion of very small meat establishments in Pennsylvania that apply more than one antimicrobial compound to carcass surfaces as a slaughter intervention……...... 106

Table 9. Method for monitoring the concentration of antimicrobial compounds used for slaughter interventions in very small meat establishments in Pennsylvania ………………107

xvi

Table 10. Interest in participating in a microbial sample study for evaluation of slaughter

interventions in very small meat establishments in Pennsylvania ……………………...….108

Table 11. Interest in training for new slaughter interventions in very small meat

establishments in Pennsylvania …………………………………………………….………109

Table 12. Interest in various types of instructional methods of processors in

Pennsylvania………………………………………………………………………………..110

Table 13. Availability of Internet access to survey participants in Pennsylvania for educational purposes………………………………………………………………………..111

Table 14. Availability of CD-ROM access to survey participants in Pennsylvania for educational purposes…………………………………………………………………..……112

Table 15. Most suitable time of year for employees of very small meat plant establishments in Pennsylvania to attend a training program………………………..…………………..…113

Table 16. Best times and days of the week for employees of very small meat establishments in Pennsylvania to attend training programs………….………………………………….…114

Table 17. Highest level of education completed by processors in Pennsylvania…...……...115

Table 18. Job title of survey participants in Pennsylvania………………………………….116

Table 19. Number of years worked in meat and/or poultry processing industry by survey participants in Pennsylvania………………………………………....…117

Table 20. Number of employees who work in surveyed establishments in

Pennsylvania………………………………………………………………………………..118

Table 21. Slaughter interventions used to reduce microbial contaminants on carcass surfaces in very small meat establishments in Washington and Idaho………………………………119

xvii

Table 22. Approximate application time of slaughter intervention to carcass surfaces in very small meat establishments in Washington and Idaho…………………………...... …120

Table 23. Types of equipment used to apply slaughter interventions in very small meat establishments in Washington and Idaho……………………………………………...……121

Table 24. Timing of the application of slaughter interventions to carcass surfaces in very small meat establishments in Washington and Idaho…………………………………...….122

Table 25. Type of nozzle used in washing cabinets to apply hot water or organic acids in very small meat establishments in Washington and Idaho…………………………….………...123

Table 26. Application pressure of hand-held garden sprayer or spray cabinet used to apply hot water or organic acids to carcass surfaces in very small meat establishments.….……..124

Table 27. Distance between hand-held hose or hand-held sprayer and carcass surfaces during application of hot water or organic acids in very small meat establishments in Washington and Idaho………………………………………………………………………………....…125

Table 28. Method for monitoring the concentration of antimicrobial compounds used for slaughter interventions in very small meat establishments in Washington and Idaho……..126

Chapter Four

2 Table 1. Mean populations and reductions of Salmonella Typhimurium (log 10 CFU/cm ±

SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures,

spray distances, application times, and drip times …………………………………………155

2 Table 2. Mean populations and reductions of Campylobacter spp. (log 10 CFU/cm ± SE) on

inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray

distances, application times, and drip times………………………………………………...156

xviii

2 Table 3. Mean populations and reductions of Escherichia coli O157:H7 (log 10 CFU/cm ±

SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times…………………..……………………...157

2 Table 4. Mean mesophilic aerobic plate counts and reductions (log 10 CFU/cm ± SE) on

inoculated beef plates following treatment with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times………………………………158

2 Table 5. Mean populations and reductions of generic E. coli (log 10 CFU/cm ± SE) on

inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray

distances, application times, and drip times…………………………………………...... 159

2 Table 6. Mean populations and reductions of coliforms (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances,

application times, and drip times…………………………………………………………...160

2 Table 7. Mean populations and reductions (log 10 CFU/cm ± SE) of Salmonella

Typhimurium on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C)

water wash (30 psi, 20 s application time) with varying drip times and spray distances..…161

2 Table 8. Mean populations and reductions (log 10 CFU/cm ± SE) of Campylobacter spp. on

inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances………..…………162

2 Table 9. Mean populations and reductions (log 10 CFU/cm ± SE) of Escherichia coli

O157:H7 on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C)

water wash (30 psi, 20 s application time) with varying drip times and spray distances….163

xix

2 Table 10. Mean mesophilic aerobic plate counts and reductions (log 10 CFU/cm ± SE) on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances……………….…...164

2 Table 11. Mean populations and reductions (log 10 CFU/cm ± SE) of generic E. coli on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances……………….…...165

2 Table 12. Mean populations and reductions (log 10 CFU/cm ± SE) of coliforms on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances..……………………………166

Table 13. Mean temperature of lean tissue surfaces of beef plates manually washed with warm (54°C) or hot (77°C) water in a benchtop washing cabinet ……………………...….167

Table 14. Mean log reductions of pathogens and hygiene indicators on inoculated beef plates

following cold water (15°C) washing (first, flawed analysis of cold water washing data)...168

2 Table 15. Mean populations (log 10 CFU/cm ) of bacterial populations on inoculated beef plates following cold water (15°C) washing under a variety of pressures, spray distances,

application times, and drip times (second analysis of data)……………………………...…170

Chapter Five 2 Table 1. Mean populations and reductions of Salmonella Typhimurium (log 10 CFU/cm ±

SE) on inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5 cm, 15 s application, 5 min drip)…………..……………………………………..215

2 Table 2. Mean populations and reductions of Campylobacter spp. (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5 cm, 15 s application, 5 min drip)…………………………………………………………...217

2 Table 3. Mean populations and reductions of E. coli O157:H7. (log 10 CFU/cm ± SE) on

inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5

cm, 15 s application, 5 min drip)…………………………………………………………...219

2 Table 4. Mean mesophilic aerobic plate count reductions (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5 cm, 15 s application, 5 min drip) …………………………………………………………………….221

2 Table 5. Mean populations and reductions of generic E. coli (log 10 CFU/cm ± SE) on

inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5

cm, 15 s application, 5 min drip)……………………………………………..…………….223

2 Table 6. Mean populations and reductions of coliforms (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with water or an antimicrobial rinse (40 psi, 30.5 cm, 15 s

application, 5 min drip)……………………………………………………………………..225

2 Table 7. Mean reductions of bacterial populations (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with chlorinated rinses…...………………………………….…227

2 Table 8. Mean reductions of bacterial populations (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with 2% organic acid rinses…...…………………………...…..228

Table 9. pH, concentration, and pressure observations of antimicrobial compounds before and after decontamination of inoculated beef plates………………….……………………229

Table 10. Aqueous ozone concentration before and after decontamination of inoculated beef plates………………………………………………………………….……………………231

xxi

Chapter Six

2 Table 1. Mean bacterial populations and reductions (log 10 CFU/cm ± SE) on inoculated beef plates before (control) and after treatment with water washing, water rinsing, or 2% lactic

acid rinsing alone or in combination using a portable, stainless steel tank ..………..……..252

2 Table 2. Mean bacterial populations and reductions (log 10 CFU/cm ± SE) on inoculated beef plates before (control) and after treatment with water washing, water rinsing, or 2% lactic

acid rinsing alone or in combination using a garden sprayer……………………………….255

2 Table 3. Mean bacterial populations and reductions (log 10 CFU/cm ± SE) on inoculated beef plates before (control) and after treatment with water washing, water rinsing, or 2% lactic

acid rinsing alone or in combination using a motorized backpack sprayer...………………258

2 Table 4. Mean bacterial populations and reductions (log 10 CFU/cm ± SE) on inoculated beef plates before (control) and after treatment with water washing, water rinsing, or 2% lactic

acid rinsing alone or in combination using a retrofitted garden sprayer …………….…….261

2 Table 5. Mean reductions of bacterial populations (log 10 CFU/cm ± SE) on inoculated beef plates after water washing combined with 2% lactic acid rinsing using a garden sprayer (GS),

retrofitted garden sprayer (RF), portable stainless steel tank (SS), or motorized backpack

sprayer (BP)………………………………………………………………………………...264

Table 6. pH of inoculated beef plates homogenized after treatment with a warm water wash

(54°C) and/or water or 2% lactic acid rinse….....………………………………………..…265

xxii

Chapter Seven

Table 1. Prevalence of pathogens on red meat carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention……….….316

Table 2. Prevalence of pathogens on red meat carcasses processed in very small meat

establishments in Pennsylvania before (first baseline) and after (second baseline) the

implementation of a multi-step antimicrobial carcass intervention………………………...317

Table 3. Prevalence of pathogens on red meat carcasses processed in very small meat

establishments in Washington and Idaho before (first baseline) and after (second baseline)

the implementation of a multi-step antimicrobial carcass intervention………………….....318

Table 4. Prevalence of pathogens on red meat carcasses processed in very small meat

establishments in Texas before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention………………………………………….319

Table 5. Prevalence of pathogens on beef carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention…………..320

Table 6. Prevalence of pathogens on pork carcasses processed in very small meat

establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention…………..321

Table 7. Prevalence of pathogens on lamb carcasses processed in very small meat

establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention………..…322

xxiii

Table 8. Prevalence of pathogens on bob veal carcasses processed in very small meat

establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention……….….323

Table 9. Median populations (log CFU/cm 2) of hygiene indicators on red meat carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass

intervention………………………………………………………………………………....324

Table 10. Median populations (log CFU/cm 2) of hygiene indicators on red meat carcasses processed in very small meat establishments in Pennsylvania before (first baseline) and after

(second baseline) the implementation of a multi-step antimicrobial carcass intervention....325

Table 11. Median populations (log CFU/cm 2) of hygiene indicators on red meat carcasses processed in very small meat establishments in Washington and Idaho before (first baseline)

and after (second baseline) the implementation of a multi-step antimicrobial carcass

intervention………………………………………………………………………………....326

Table 12. Median populations (log CFU/cm 2) of hygiene indicators on red meat carcasses processed in very small meat establishments in Texas before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention……..…....327

Table 13. Median populations (log CFU/cm 2) of hygiene indicators on beef carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass

intervention…………………………………………………………………………………328

xxiv

Table 14. Median populations (log CFU/cm 2) of hygiene indicators on pork carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass

intervention………………………………………………………………………………....329

Table 15. Median populations (log CFU/cm 2) of hygiene indicators on lamb carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step antimicrobial carcass intervention………………………………………………………..………………………..330

Table 16. Mean (± SE) or median populations (log CFU/cm 2) of hygiene indicators on bob

veal carcasses processed in very small meat establishments in three geographical regions before (first baseline) and after (second baseline) the implementation of a multi-step

antimicrobial carcass intervention……………………………………...…………………..331

Appendix B

Table 1. Pearson correlation (r) of bacterial populations to absorbance readings…………359

Table 2. Regression coefficients (R 2) of bacterial populations versus

absorbance readings………………………………………………………………………..360

Appendix C

Table 1. Mean pH of diluents before and after use……………………………………..…372

Table 2. Populations (log CFU/ml) of pathogens before and after recovery from acid-treated beef brisket homogenized in Butterfields phosphate diluent (BPD), buffered peptone water

(BPW) or 0.1% peptone water (peptone)……………………………………………..…..373

xxv

Table 3. Average reductions in bacterial populations (log CFU/cm 2) recovered from acid- treated beef briskets homogenized in Butterfields phosphate diluent (BPD), buffered peptone water (BPW), or 0.1% peptone water (peptone)……………………………………….…...374

Appendix D

Table 1. Average surface areas of red meat carcasses from previous studies……………...383

Table 2. Surface areas of a small number of lamb pelts and pork skins determined by plastic sheeting method in a very small meat plant………………………………………………..384

Table 3. Average time necessary to apply a water wash or 2% lactic acid rinse to small, red meat carcasses in a very small meat establishment………………..…………………….…385

Appendix E

Table 1. Cost comparison assumptions………………………………………….…..….....392

Table 2. Cost comparison of cold, warm and hot water on per gallon and per carcass bases……………………………………………………………………..………………...393

Table 3. Antimicrobial effectiveness of several food-safe compounds used to eliminate meatborne pathogens from experimentally inoculated beef surfaces……………….……398

Table 4. Antimicrobial effectiveness of several food-safe compounds used to eliminate hygiene indicators from experimentally inoculated beef surfaces.……..……………..…400

Table 5. Average log reductions of common meat pathogens and hygiene indicators on beef surfaces…………………………………………………………..……………………….404

xxvi

Appendix F

Table 1. Estimated flow rate and cost of spraying equipment that can be used to apply antimicrobial rinses to carcass surfaces in very small meat establishments…………….….422

Table 2. Estimated per usage costs of aqueous ozone to decontaminate carcass surfaces in a very small meat establishment………………………………….………………………..…424

Table 3. Estimated costs of obtaining two chlorine dioxide generators………...…..…..….426

Table 4. Estimated per usage costs of chlorine dioxide to decontaminate carcass surfaces in a very small meat establishment………………………………………………………..…….427

Table 5. Cost of reagents needed to prepare acidified sodium chlorite……………..……...429

Table 6. Estimated per usage costs of acidified sodium chlorite to decontaminate carcass surfaces in a very small meat establishment………………………………………..………430

Table 7. Estimated per usage costs of sodium hypochlorite (NaOCl) to decontaminate carcass surfaces in a very small meat establishment………………………………………..………432

Table 8. Cost of concentrated organic acids used to prepare antimicrobial carcass rinses…………………………………………………………………………………..……435

Table 9. Estimated per usage costs of acetic acid to decontaminate carcass surfaces in a very small meat establishment……………………………………………………………...……436

Table 10. Estimated per usage costs of citric acid to decontaminate carcass surfaces in a very small meat establishment…………………………………………………………………..437

Table 11. Estimated per usage costs of lactic acid to decontaminate carcass surfaces in a very small meat establishment…………………………………………...... 438

xxvii

ACKNOWLEDGMENTS

A dissertation of this size simply does not happen without help from a number of people. Firstly, I want to thank the people who helped me gather or analyze information to

fulfill numerous research objectives. Dr. Naana Nti from the Penn State Office of Outreach

and Marketing calculated and tabulated the results of survey questionnaire that is described

in Chapter One. Chapter Two was my literature review, which was adapted from a book

chapter (from Handbook of Beef Safety and Quality ; The Haworth Press, Inc.; Binghamton,

NY) co-authored by myself, my major advisor, Dr. Catherine Cutter, Dr. William R.

Henning, who is also on my doctoral committee, and Dr. Margaret Hardin, who currently works for Boars Head Provisions. Their technical prowess has helped me deepen my understanding of meat safety during slaughter. As I round out my doctoral committee, I consider myself extremely fortunate to have been guided by Dr. Edward Mills (my major advisor for my Master of Science degree), Dr. Stephanie Doores, and Dr. Nancy Ostiguy. I attribute the quality of the documents that follow directly to their thoughtful attention to detail. Additionally, I wish to acknowledge the National Integrated Food Safety Initiative, which is administered by the Cooperative State Research, Education, and Extension Service under the United States Department of Agriculture, for funding this project.

There are numerous individuals who were absolutely essential to my success. I wish to thank each of them for their willingness to work with me and for providing me with a wide range of resources from technical knowledge to workspace. Of the Penn State Meats Lab, I am grateful to Glenn Myers, Barrie Moser, Jason Monn and Katie Logan. Furthermore, I appreciate Jon Cofer and Lucas Parker for filming and producing a 17-minute training video and to Bill Houser for narrating it. Another large component of my project was the collection

xxviii

of hundreds of carcass samples from very small meat establishments in Pennsylvania, Idaho,

Washington, and Texas with the assistance of Washington State University and Texas Tech

University. I received a mountain of data from Peter Gray at Washington State University

under the direction of Dr. Dong-Hyun Kang, a co-investigator of this project. I genuinely

appreciate everything that Peter has done for me. He has approached each aspect of this project with an admirable sense of integrity and professionalism. Alejandro Echeverry, Jason

Mann, and other students from Texas Tech University also provided me with the data that we

needed to answer our research questions. Dr. Mindy Brashears, also a co-investigator, provided much guidance to the Texas Tech crew and was quick to answer all of my

questions.

In order to develop recommendations to very small meat processors, I spent countless

hours in the laboratory studying the antimicrobial efficacies of eight chemicals. As these

experiments were costly to perform, Dr. Cutter and I were fortunate to benefit from the

generosity of the following individuals and their companies for donating and loaning supplies

and/or equipment: Rick Hess of Hess Machines, Ephrata, Pennsylvania; Joy Herdt and Keith

Johnson of ECOLAB, St. Paul, Minnesota; Rich O’Reskie of Halox Technologies,

Bridgeport, Connecticut; and Dr. Neeraj Khana of Bio-Cide International, Norman,

Oklahoma. These professionals have donated their time and industrial expertise and some of

them traveled to University Park to set up equipment and provide one-on-one training.

Additionally, I called upon several experts to help me answer tough questions. I am thankful

for lactic acid experts at Birko Corporation in Henderson, Colorado and PURAC America in

Lincolnshire, Illinois. Also, I want to thank several individuals from the Food Safety and

Inspection Service of the United States Department of Agriculture: Ms. Patricia White, a

xxix

meat product labeling expert; Jan Behney, District Manager of District 60; and Dr. Ronald

Jones, District Manager of District 15.

Although I am unable to reveal the identities of the very small meat establishments that allowed us to gather our carcass samples, I do wish it to be known that their participation in this project is tremendously appreciated. Personally, I am very grateful to all of them for the privileges that they extended to me so that I could earn this degree. Several of these companies have had long-term partnerships with the meat research programs at Penn State,

Washington State, and Texas Tech and I certainly hope that these special relationships continue to thrive.

Next, I thank my wonderful labmates and research assistants who have given me advice, helped me prepare media and reagents, washed my dirty labware, and performed various unpleasant laboratory tasks so that I could do my job better. Since January 2002, I have had the pleasure of working with Donna Miller, Naveen Chikthimmah, Kerry Fabrizio,

Peyman Fatemi, Mary Nguyen, Niraja Ramesh, Renee Britton, Amie Geiger, Tim

Strittmatter, Ben Joyce, Ginger Fenton, Matt Fenton, Nathan Angstadt, Donna Abdullah,

Kristin Wernosky, Rich Sweich, Alison Bell-St. John, Steve Adoff, Kate Lamm, Aubri

Carmichael, Drew Aurand and Kim Norton in the food microbiology facilities of Borland

Laboratory. Many of these individuals were also volunteers for Borland Day Care, a wonderful group of students, faculty, and staff who watched Hazel, my daughter, for a few minutes or for several hours while I carried out my research tasks under hazardous conditions.

I would be remiss not to recognize the administrative support that I have received from the Penn State Department of Food Science. Dr. John Floros, the head of the

xxx

department, has been especially supportive of my academic success. Furthermore, there is a

long list of wonderful ladies, who have always been willing to point me in the right direction

regarding matters of the Graduate School and use of budget funds. Juanita Wolfe, Amy

Larimer, Donna Merrill, Martha Neiheisel, Melissa Strouse, Ellie Chapman, and Carole

Donald, thank you for all of your help. I wish everyone knew how much your efforts impact

the success of each graduate student.

On a personal note, I want to recognize my family for their unconditional support of me throughout this degree program. I began this program of study in January 2002 and I graduated four and half years later. My parents, Rose and Dale, always supported my decision to move forward with my research through some of the most difficult phases of my life. My sister, Susan, has also been steadfast in her support of my academic objectives. She and her husband, Nate, willingly cared for Hazel for six months of her infancy so that I could complete the most labor-intensive segments of my research. It is amazing how much love this precious, little gem has brought into my life. Finally, I want to acknowledge Tim Yoder, the love of my life. Since we have been together, Tim has shown me more patience, thoughtfulness, and genuine kindness than I feel I deserve, especially during the final stretches of research and writing. Without my family, my failsafe support network, I would still be in the laboratory conducting washing experiments or collecting samples from meat establishments.

I devotedly recognize the unconditional support that Dr. Cutter has given to me during my time in the Food Science Department. As her first doctoral student, I provided her with an abundance of mentoring challenges. I can think of several people who would have shown me to the door. Instead, Dr. Cutter advised me to make the right choices and for that I

xxxi

am forever grateful to her. It is my hope to pattern my professional and personal life after hers – one distinguished by integrity, goodwill, and resolve. Thank you, Cathy.

xxxii

CHAPTER ONE

INTRODUCTION

STATEMENT OF THE PROBLEM

While meat and poultry establishments in the United States were in the midst of developing and implementing HACCP plans as required by the Pathogen Reduction Act of

1996, the Food Safety and Inspection Service openly admitted that more research efforts would be needed to provide documentation for the validation of critical control points in very small plants (USDA-FSIS, 1996b). As the decisions were made to establish acceptable levels for generic E. coli and Salmonella spp. on raw meat and poultry, FSIS depended mostly on data that had been collected from large or small establishments. While these performance standards may certainly apply to processing in the large or small plant, it is not clear that these same standards are suitable objectives for very small meat and poultry plants. Though the data do not exist, the levels of pathogenic microorganisms on carcasses processed in very small meat and poultry plants may be as low as those detected in large plants. Even in the absence of adequate data, very small establishments are required to cite reliable sources, typically from refereed journals, in their HACCP plans to demonstrate that antimicrobial interventions to reduce microbiological hazards are, indeed, effective.

RESEARCH OBJECTIVE

Given the labor constraints to implement and monitor elaborate decontamination methods and limited financial resources for capital investment, it is necessary to identify and validate alternative methods of reducing bacterial populations on carcass surfaces to meet the unique needs of very small plants in the United States. Very small meat plants do not have the available resources to dedicate to a third-party for the generation of sufficient and valid data. When valid techniques for carcass decontamination and supporting documentation become available, then there will be a logical basis to assess the prevalence of pathogenic microorganisms on carcass surfaces in very small meat plants. With such a basis for process control, appropriate decontamination techniques can be implemented in very small meat plants and the meat industry should come one step closer to preventing food borne illness in

humans.

EXPERIMENTAL PLAN IN BRIEF

Described below is a series of five research objectives that achieve the overall goal of providing scientific documentation for the validation of HACCP plans of very small meat plants. While the study of a subset of very small plants may not precisely represent all very small meat and poultry establishments (approximately 3,400), the selection of three geographical regions in the United States represented by Pennsylvania, Texas, Washington, and Idaho should illustrate the general preferences for and characteristics of carcass decontamination techniques practiced in the United States. To represent all very small meat plants in the United States, three state universities (The Pennsylvania State University, Texas

Tech University, and Washington State University) collaborated to survey the decontamination practices and sample red meat carcasses in very small establishments in the states listed above.

In the first phase of this project, very small meat plants in three states were surveyed to ascertain which antimicrobial interventions were being used to eliminate bacteria from carcass surfaces. The next objective involved the sampling of red meat carcasses processed in very small meat plants to determine a baseline for Salmonella spp., Campylobacter spp., E.

coli O157:H7, aerobic plate counts, and counts of generic E. coli and coliforms. For the third research objective, a variety of antimicrobial compounds to reduce bacteria on meat surfaces was selected for laboratory study based on survey results, peer-reviewed literature, and personal communication. Next, the individual and combined effectiveness of water washing

and/or rinsing with an antimicrobial compound was measured in a laboratory setting. The

results of this fourth research objective were used to produce a training video and

instructional booklet to transfer technical knowledge to employees of very small plants. The fifth objective served as a real-world validation of the decontamination strategies that were investigated under laboratory conditions. After implementation of a multi-step antimicrobial intervention in selected very small plants, carcasses were sampled again to generate a second microbiological baseline to determine whether the implementation of this intervention reduced the prevalence of target bacterial populations. The entire project will be summarized into a report and distributed to all very small meat plants. In the near future, the very small plants that were sampled during the secondary baseline are to be surveyed again to assess the adoption rate of the new technology and gather employee opinions of the feasibility and success of the multi-step antimicrobial intervention.

CHAPTER TWO

LITERATURE REVIEW

INTRODUCTION

Meat processors have an ethical responsibility to provide wholesome products to consumers. From the moment the live animal sets foot on the slaughter floor until the beef carcass is fabricated into primals and throughout processing, there are hazards that must be controlled. To fulfill this responsibility, plant employees must be aware that the procedures performed on the slaughter floor and in the fabrication and processing rooms can have a tremendous impact on the safety of the meat products that reach the end user. Currently, every meat plant under federal inspection conducts daily operations under Hazard Analysis and Critical Control Point (HACCP) systems. Each plant develops its own HACCP plan focused on the prevention of problems to assure the safety of meat products through control of chemical, physical, and biological hazards.

Large, small, and very small plants

The United States meat and poultry industry comprises approximately 6,000

establishments dedicated to harvest, fabrication, processing, and/or purveying. The USDA

classifies plants into one of three categories based on the number of employees and annual

revenue. About 5% or 300 of these plants are classified as large plants as they employ more

This literature review was adapted from: S. Flowers Yoder, M. D. Hardin, W. R. Henning, C. N. Cutter, IN

PRESS. Beef safety during slaughter, fabrication, and further processing. Ch. 3 in Handbook of Beef Safety and

Quality . D. (Roeber) Van Overbeke (Ed.), The Haworth Press, Inc., Binghamton, NY.

than 500 workers, whereas small plants (38%, 2,300 establishments) employ between 10 and

500 people. Plants considered very small employ 10 or fewer employees and generate average annual revenue of $2.5 million or less (USDA-FSIS, 1996). Given these definitions, it is obvious that very small plants have limited labor and finances for the purchase, operation and maintenance of elaborate decontamination methods such as carcass cabinet washers and steam vacuum sanitizers.

HACCP implementation

In the interest of public health, it has become unacceptable for contamination with pathogens to persist in muscle foods. The Pathogen Reduction: Hazard Analysis and Critical

Control Point Systems final rule of 1996 mandated the implementation of HACCP plans for all meat and poultry establishments, established performance standards for the incidence of

Salmonella in raw and ground meat products, required periodic testing by the company for generic E. coli , and required the development and implementation of Sanitation Standard

Operating Procedures (SSOPs) in each plant (USDA-FSIS, 1996c). Employees also should

follow the good manufacturing practices (GMPs) that are prescribed within individual meat plants (Katsuyama and Humm, 1995). By doing these practices and adhering closely to the plant’s HACCP plan, the chemical, physical, and biological hazards associated with meat

should be under control at all times. Generally, biological hazards require the most

consideration because of the potential for widespread outbreaks of illness in humans.

Primary processing

Several steps must take place in the slaughter facility to convert a live animal to a

meat carcass. These steps are described in a general manner, although there is some variation

among plants due to existing workspace, available equipment, religious practices and other

factors. During slaughter, exsanguination takes place after stunning and it is the first step in

the conversion of muscle to meat. Exsanguination also is the first opportunity during

slaughter for microbes to contaminate the carcass, unless captive bolt stunning or some other

invasive procedure is used to render the animal unconscious (Aberle et al., 2001). Moreover,

microbes can gain access to the bloodstream through the opening made by severing the major blood vessels with a contaminated knife.

Next, the integument is removed. In some plants, chemical dehairing or hide washing

may take place prior to exsanguination; then the hide is removed. Hide or pelt removal can be achieved by manual skinning or an automated puller. Pork carcasses may be scalded,

dehaired, and singed in place of skinning. When performed manually, the employee should

minimize cross-contamination and be careful to keep the skinning knife clean and avoid

letting the skin touch the carcass surface as it is being cut away. Automated removal of hides

can release dust and debris into the ambient air as the skin is pulled away from the carcass by

a hydraulic arm. Some of these particles could be desiccated manure that have the potential

to carry viable cells of Escherichia coli O157:H7, which could be deposited on an otherwise

clean carcass surface or processing equipment (Delazari et al., 1998; Elder et al., 2000). In

fact, the bovine hide is known to be a major source for carcass contamination with E. coli

O157:H7 (Elder et al., 2000 and Reid et al., 2002).

Typically, the head is removed after the carcass has been skinned. At some point, the

tail, if present, also is removed. Removal of viscera (stomach, spleen, intestines, and anus)

and pluck (tongue, esophagus, heart, diaphragm, lungs, and liver) from the body cavity, or

evisceration, follows. Fecal matter can be released onto carcass surfaces if the intestinal tract

is punctured or the bung is not properly tied. In some plants, carcasses are steam vacuumed

for removal of fecal matter and other debris. The beef carcass is then sawed into halves while

other red meat species may be left intact until fabrication. As a preventive measure for

controlling bovine spongiform encephalopathy (BSE), specified risk materials (SRM), such

as the spinal cord, brain and skull, must be removed from cattle that are 30 months of age or

older (USDA-FSIS, 2004).

At the end of primary processing, the carcass is visually inspected for contamination by employees, undergoes knife trimming as necessary and is then washed thoroughly with

water. In most plants, an intervention step follows the water wash. Interventions may include

a sanitizing step with an antimicrobial substance or surface pasteurization. After the final

wash, carcasses are placed in a chill cooler or hot box to control the lowering of carcass

temperature.

In a very small meat plant, the final washing step prior to chilling provides the best

opportunity to decontaminate carcass surfaces. While an intervention (washing, knife

trimming) may take place during slaughter and provide an effective means of removing bacteria, carcass surfaces could become recontaminated before the final wash. The debris that may have accumulated on meat surfaces during the slaughter sequence is most effectively removed by a worker who is profoundly familiar with the effort that is necessary for each carcass to pass federal inspection. Given the relatively small labor force in these plants, one

or two workers are usually dedicated to the quality assurance tasks of visual inspection, knife

trimming, washing, and application of interventions to each carcass before it enters the chill

cooler.

Types of interventions

Currently, there are several types of antimicrobial treatments or interventions shown to substantially reduce bacteria levels on red meat carcasses and are approved for use to meet

HACCP requirements. The most common carcass decontamination strategies include water washing, knife trimming, chemical treatments, and moist heat under pressure or vacuum.

When considering which interventions to implement, plants of all sizes should consider the expenses of purchase, maintenance, and daily operation (fixed costs); available space and manpower; and documented antimicrobial effectiveness. Some interventions, which are well- suited for large plants, are difficult or impossible to implement in small or very small plants.

As part of HACCP, FSIS requires the removal of all visible fecal contamination, ingesta, or milk from carcasses, by knife trimming or vacuuming with hot water or steam

(USDA-FSIS, 1996b). Carcasses are then subject to re-inspection and must meet finished product standards applied by plant employees and verified by FSIS inspectors before the carcasses can enter the chiller. Despite the use of sanitary dressing procedures to reduce or eliminate contamination during slaughter and processing, pathogenic bacteria still may be present on red meat carcasses (USDA-FSIS, 1996a). Therefore, antimicrobial treatments or interventions are recommended by USDA to reduce the prevalence of pathogens in raw products.

Some plants decide to integrate antimicrobial interventions into their HACCP plans

as control points (CPs) or critical control points (CCPs). Both CPs and CCPs indicate points,

steps, or procedures that can be applied to food to manage chemical, physical, and/or biological hazards (Katsuyama and Stevenson, 1995). These control points should be applied

when a hazard is significant and reasonably likely to occur. A control point becomes critical

if the “point, step, or procedure at which control can be applied and a food safety hazard can be prevented, eliminated, or reduced to acceptable levels” (Katsuyama and Stevenson, 1995).

Furthermore, FSIS requires meat plants to provide technical documentation of the

effectiveness of CCPs in plant HACCP plans.

In the slaughter process, biological hazards are more significant and reasonably likely

to occur than chemical or physical ones. The final rinse step is usually where a CCP, if

applicable, is applied. Typically, there is not a subsequent step beyond the final wash during

which a CCP can be implemented.

To reduce spoilage and pathogenic bacteria, large and small plants have incorporated

antimicrobial treatments or interventions into their slaughter operations that have been shown

to greatly reduce the levels of any bacteria that may be present (Dickson and Anderson,

1992; Siragusa, 1995; Bolder, 1997; Dorsa, 1997). Employing at least one such treatment

will not, by itself, solve the problem of contamination with pathogenic bacteria. However,

such treatments are a step among many that can reduce the risk of raw product reaching the

consumer with hazardous levels of pathogenic bacteria (USDA-FSIS, 1996a).

FoodNet, the national food borne illness surveillance network that operates under the

Centers for Disease Control and Prevention, has detected significant decreases in the

incidence of food borne illness among the U. S. population from 1996 to 2001. During these

years, the incidences of infection caused by Campylobacter spp., Salmonella spp., and E. coli

O157:H7 decreased by 27%, 15%, and 21%, respectively (CDC, 2002). Network personnel report that the implementation of HACCP in all meat and poultry plants may be responsible for some of the progress. The reduction in Salmonella infections could specifically be attributable to successful compliance with Salmonella performance standards (CDC, 2002).

This may be true of red meat; however, Salmonella incidence in broilers from randomly selected establishments has slightly increased from 11.5% in 2002 to 13.5% in

2004 (USDA-FSIS, 2006). This increase is cause for concern, not alarm, because a prevalence of 13.5% is still lower than the national baseline, which provides a foundation for

Salmonella performance standards in the Pathogen Reduction Act (USDA-FSIS, 2006). As a result, FSIS recently issued a notice about upcoming changes to reporting and use of

Salmonella test results and the targeted enforcement of performance standards in plants with

recurring test failures (USDA-FSIS, 2006). This action mirrors the verification of E. coli

O157:H7 reassessment, which required plants to incorporate at least one CCP in HACCP plans, as applicable, to verify control of E. coli O157:H7 in raw beef products (USDA-FSIS,

2002). FSIS hopes that the recent Salmonella initiative is as successful at controlling the

incidence of Salmonella in meat and poultry as the E. coli O157:H7 reassessment was at

controlling presence in beef (USDA-FSIS, 2006).

CHEMICAL AND PHYSICAL HAZARDS

In general, a food safety hazard can be any agent or property (biological, chemical or physical) that compromises the safety of a food for human consumption (9 CFR 417.1;

NACMCF, 1998). The meat processor and plant employees should be aware of the chemical and physical hazards that are likely to be present in all areas of the meat plant. The HACCP plan must address the hazards that are reasonably likely to occur. Biological hazards are of greatest concern, since they are capable of causing widespread food borne illness affecting a greater proportion of individuals; however, chemical and physical hazards also have been associated with food borne illness and injury. Below is a description of chemical and physical hazards that are likely or known to be of concern in meat processing facilities.

Biological hazards will receive greater emphasis for the remainder of this review.

Chemical Hazards

A wide variety of chemical hazards may be present in any meat processing plant. The chemicals that are most likely to impact beef safety during slaughter originate from incoming cattle, processing aids and ingredients, plant maintenance and plant sanitation (Katsuyama et al., 1999). As a minimum, the HACCP team should consider all of these sources when performing a hazard analysis.

Live animals that enter the plant typically come from reputable feedlots, farms, and other sources. Livestock producers in the U.S. are permitted to use hormones and antibiotics in meat animal production. Growth hormones allow for rapid weight gain while optimizing meat quality (USMEF, 2005b). Antibiotics are approved for use by the Food and Drug

Administration for the prevention, treatment, and maintenance of healthy animals (USMEF,

2005a). However, it is the responsibility of the livestock producer who uses growth promoters, antibiotics, and other chemicals to do so exactly as indicated by the drug manufacturer, in accordance with the label instructions, and to meet the recommended requirements for withdrawal before sending an animal to slaughter. Moreover, FSIS inspectors routinely monitor food animals for drug residues. In 2001, most of the drug residue violations were traced back to the use of animal drugs at levels that exceed the legal limit or failure to allow sufficient withdrawal time prior to slaughter (USDA-FSIS, 2001).

In addition to incoming animals, the plant itself may be the origin of significant chemical hazards. The infrastructure and machinery in meat plants require periodic maintenance to prevent the introduction of these undesirable compounds. Processing equipment, floors, and walls must be cleaned and sanitized as dictated in plant SSOPs.

Employees must also wash their hands, personal protective equipment, knives, and other equipment throughout the day. Furthermore, many plants may employ one or more chemical interventions to reduce biological hazards on beef surfaces and in equipment surfaces during processing. For these reasons, forethought is necessary to prevent contact between product and employees and the substances that could be harmful if consumed (e.g. lubricants; paints; cleaners; sanitizers; and excessive concentrations of antimicrobial compounds, food additives, and processing aids; Katsuyama et al., 1999). The establishment must also ensure that only approved chemicals are used in the facility and that they are used in accordance with label and regulatory recommendations with appropriate written SSOPs. Employee training is essential to prevent misuse and to protect both the employees and the final product.

Incoming non-meat ingredients, in the form of packaging materials, restricted

ingredients, such as nitrates and nitrites, as well as allergens, also may bring chemical

hazards into the production facility. While letters of guarantee from suppliers will help to

ensure that the chemical safety of these ingredients has been addressed prior to receipt into

the facility, it is the responsibility of the establishment to inspect the incoming vehicles and products, provide adequate and appropriate storage, and assure their proper use and labeling.

Physical Hazards

The unintentional introduction of foreign matter into foods also could indicate a physical hazard. Physical hazards associated with meat safety generally are linked to incidents of personal injury and not public health. Katsuyama and Jantschke (1999) have discussed many specific physical hazards that could occur on the slaughter floor and in the fabrication room. At slaughter, incoming animals may enter the facility with hazards such as injection needles. Furthermore, processing equipment that comes in contact with the carcass is usually made of metal or plastic. For example, teeth from a saw or the tip of a knife can break off, become lodged in the carcass tissue, and later cause injury to a consumer

(Katsuyama and Jantschke, 1999). Pieces of plastic may tear or break away from bags, films, or equipment and become attached to or embedded in carcass surfaces, only to be discovered post-processing by a consumer. Wood splinters (from wooden-handled equipment, beams, pallets, etc.) and glass shards (from unprotected lighting, glass covered gauges, thermometers, etc.) also could create physical hazards if these materials are present in areas where carcasses are processed.

Employees also can be a significant source of foreign objects such as meat hooks,

knives, thermometers, pens and pencils, hair, fingernails and jewelry, which can be

hazardous for the final product if employees do not receive proper training or fail to adhere to plant GMPs. Many plants use frocks with inner pockets and snaps instead of buttons to prevent the introduction of foreign objects into edible product. Preventive maintenance of

facilities and equipment by plant maintenance, and constant training and observation of

employees by plant quality assurance and management are necessary to reduce issues

associated with physical hazards.

BIOLOGICAL HAZARDS

Individuals who are most susceptible to contracting food borne disease include the

elderly (65 years or older), children who are five years or younger, pregnant women, and

individuals with weakened immune systems. The Centers for Disease Control and Prevention

estimate that approximately 76 million cases occur each year, in which 5,000 deaths also

occur (CDC, 2003). Moreover, disease incidents from five common food borne pathogens

(Campylobacter spp., nontyphoidal Salmonella , E. coli O157:H7, other enterohemorrhagic E. coli , and Listeria monocytogenes ) incurred an annual public health cost of almost $6.9 billion in 2000 (USDA-ERS, 2004). Food borne illnesses that are traced back to the consumption of fresh meat products are commonly caused by E. coli O157:H7, Salmonella spp., or

Campylobacter spp.

Escherichia coli O157:H7

In the early 1980’s E. coli O157:H7 was regarded as a rare serotype that was

identified initially by the CDC in a 50-year-old woman who suffered from severe abdominal

cramps and bloody diarrhea (Riley et al., 1983). An infectious dose of as few as 10 cells has been documented as causing human illness (Griffin et al., 1994). E. coli O157:H7 illnesses

can be life-threatening in children, the elderly, and other individuals with weak immune

systems.

coli O157:H7 is mediated through shiga or shiga-like toxins, which are also known as verotoxins or verocytotoxins (Wieler et al., 1992). The pathogen can attach to and invade human intestinal epithelia, as well as produce Shiga-toxin, resulting in cellular death. Patients suffering from hemorrhagic colitis often begin by having abdominal cramps and diarrhea that may become profuse and bloody due to the shedding and hemorrhaging of the mucosal lining of the colon (Su and Brandt, 1995). Fever, nausea, and other flu-like symptoms often accompany the colitis. Once hemorrhaging occurs, the bacteria have a direct conduit to the bloodstream, which transport them to other vulnerable tissues such as the kidney and nervous system. Shiga-toxin can damage the endothelial lining of blood vessels. Clots, which may begin to form within capillaries in patients with HUS, block the flow of nutrients to the kidneys and to other organs (Padhye and Doyle, 1992). As the kidneys fail to filter the blood properly, wastes accumulate in the bloodstream (Padhye and Doyle, 1992). In cases of TTP, blood clots can also form in the brain, which may lead to death (Padhye and Doyle, 1992).

The hide and gastrointestinal tract of beef animals have been well established as significant reservoirs for E. coli O157:H7 (Fegan et al., 2005; Arthur et al., 2004). An

examination of five recent publications sets a range of percentages between 11% and 94% for

the prevalence of E. coli O157:H7 on the hides of fed cattle. Elder et al. (2000) detected E.

coli O157:H7 on 11% the hides of feedlot cattle from four Midwestern beef plants.

Additionally, Tutenel et al. (2003) studied the prevalence of E. coli O157 on the hides of

cattle processed in Belgium. Almost 57% of hide swabs taken from the shoulder were positive for the pathogen while 25% of rectal swabs were positive (Tutenel et al., 2003).

Moreover, Barkocy-Gallagher et al. (2004) recovered E. coli O157:H7 from 61% of hide

samples from fed cattle processed in large establishments in the Midwest. Most recently,

Arthur et al. (2004) determined an overall prevalence of E. coli O157:H7 on fed cattle hides of 76%. The range of prevalence extends from 50% to 94% as a result of the six plant visits necessary to collect all of the samples (Arthur et al., 2004). Further, Bosilevac et al. (2004) documented a prevalence of 56% on bovine hides before treatment with water washing and a cetylpyridinium chloride rinse. In these studies, contamination of the hide with E. coli O157 or E. coli O157:H7 was associated with its presence on carcass surfaces (Elder et al., 2000,

Reid et al., 2002, Arthur et al., 2004, Bosilevac et al., 2004, and Fegan et al., 2005).

Many experts have supported the association between E. coli O157:H7 and cattle feces (Chapman et al., 1993; Hancock et al, 1994; Heuvelink et al., 1998b). The survival of this pathogen in excreted feces became the focus of several pre-harvest studies, due to the tremendous potential of E. coli O157:H7 cross-contamination via fecal matter on cattle farms, auction markets, and during live animal transport. The pathogen survived in experimentally inoculated manure at 5°C for up to 70 days under laboratory conditions

(Wang et al., 1996). This pathogen also could be isolated from a minimally disturbed manure pile for 21 months (Kudva et al., 1998). These contributions underscored the significance of bovine manure as an ecological niche for this organism. Recently, Echeverry et al. (2005) have argued that earlier studies of fecal pat sampling have underestimated the prevalence of

E. coli O157:H7. The pathogen is now thought to colonize the rectum in lymphoid tissue that is 5 cm proximal to the anus; so, it may not be evenly distributed within the fecal bulk during each incidence of defecation (Naylor et al., 2003; Echeverry et al., 2005).

Up to 50% of cattle herds and 10% of market cattle harbor the pathogen in England and Wales (Synge and Paiba, 2000). In Britain, 5% of cattle are thought to harbor the organism in the lower gastrointestinal tract (Synge and Paiba, 2000). Van Donkersgoed et al.

(1999) detected verotoxin in 43% of fecal samples from beef cattle processed in Alberta over

one year. Additionally, E. coli O157:H7 was found in 8% of fecal samples from yearling beef animals, culled beef and dairy cattle (Van Donkersgoed et al., 1999).

E. coli O157:H7 also has been retrieved from other red meat sources. A recent study in the US detected the pathogen in 2% of fecal samples (n = 305) from ligated pig colons

(Feder et al., 2003). In addition to fecal samples, carcass swabs from almost 200 pork carcasses in France also were sampled post-exsanguination and pre- and post-chilling for the presence of E. coli O157:H7 (Bouvet et al., 2002). The pathogen was recovered from the feces in 31% of the carcasses, while 46% of swabs tested positive after exsanguination, thereby suggesting that cross contamination may have occurred at some time between animal commingling in holding pens and bleeding. Subsequent washing and chilling of pork carcasses reduced the prevalence to 16% and 15%, respectively (Bouvet et al., 2002). A less recent study isolated E. coli O157:H7 from 1.5% of retail pork samples from grocery stores in Wisconsin and Alberta (Doyle and Schoeni, 1987). However, the prevalence of this pathogen on retail pork in the late 1980s has likely been reduced since the advent of improved carcass decontamination techniques and pathogen reduction legislation (CDC,

2002).

In another study, E. coli O157 (H7-negative) was identified in 1% of bob veal calf feces from California, Washington, and Wisconsin (Martin et al., 1994). However, the analytical tests that were used are less sensitive than most currently used methods. The true incidence of E. coli O157:H7 in bob veal feces may have been underestimated. Most recently, E. coli O157:H7 was detected in carcass swabs and ileal contents of special-fed veal

(10%) and bob veal (7%) processed in the Northeastern US (Flowers, 2002).

In addition, ground baby beef samples were collected from grocery stores in Croatia

and tested for the presence of E. coli O157:H7 (Uhitil et al., 2001). To clarify, baby beef is red meat that comes from beef animals that are too old to be considered veal and too young to be considered market beef. None of the Croatian samples contained E. coli O157:H7

(Uhitil et al., 2001). These investigations are among the very few that have examined veal and young beef for microorganisms of significance to food safety. Additionally, a much higher incidence of E. coli O157:H7 has been reported in calves than in mature cattle.

Heuvelink et al. (1998a) reported the prevalence of E. coli O157 in calves 4 to 12 months old

(21%) to be higher than any other age group. In a study to ascertain the prevalence of E. coli

O157:H7 in Michigan dairy herds, a greater proportion of calves were found to harbor the pathogen than dairy cattle older than 2 years of age (Wells et al., 1991).

Salmonella spp.

Salmonella spp. have been implicated in meat borne illnesses for several decades.

To cause illness in humans, Salmonella cells must be ingested. Common symptoms that are associated with Salmonella infection include nausea, vomiting, diarrhea, and severe abdominal pain (Jay, 2000). In some salmonellosis cases, severe dehydration can lead to death of infected individuals.

As few as 100 cells of S. Eastbourne can cause illness, while the average infectious

5 6 dose of other Salmonella strains can be 10 to 10 CFU/ml (Jay, 2000). At the other extreme, an individual would have to ingest approximately one to ten billion cells of a less virulent strain, such as S. Pullorum, before the onset of illness (Jay, 2000). Nevertheless, the strains with a low infective dose most commonly cause human illness. A low infective dose is

especially problematic in foods that have a high fat content and abundant buffering capacity,

like meat products, because viable cells of Salmonella that have attached to these foods are better protected during gastric digestion (Humphrey, 2000).

Among more than 2,300 Salmonella serotypes, S. Typhimurium and S. Enteriditis are

estimated to cause 43.5% of the human cases of salmonellosis (CDC, 1998). Of these two

species, S. Typhimurium poses the most critical hazard to red meat safety because it is the serotype most frequently isolated from beef animals (15.7%) and market hogs (20.4%;

Sarwari et al., 2001). This pathogen frequently has been isolated from the fecal matter and rumen contents of beef animals (Gay et al., 1994; McEvoy et al., 2003). Cattle of all ages may be susceptible to Salmonella infection by the fecal-oral route when contaminated feedstuffs, water, or pasture are consumed on the farm (Wray and Davies, 2000).

Salmonella also is problematic in veal calves. Neonatal calves that do not receive adequate colostrum tend to have weaker immune systems and, thus, are prone to diarrhea, which is often caused by Salmonella , various types of E. coli , or one of many viruses and parasites (McDonough et al., 1994). Veal producers must be cautious in the treatment of calf diarrhea because the threat of drug residues in meat renders some medications impractical for use in animals that are ready for market at early ages. In another study, Edel et al. (1970) isolated 16 different serotypes of Salmonella spp. from calves (n = 1,000) in slaughterhouses

in The Netherlands. S. Typhimurium comprised 52% of the strains that were isolated from the infected calves. Cecal contents (28.6%) and feces (28.6%) were also positive for

Salmonella . It was suggested that most of the calves were infected by contact with contaminated feces and hair. Most recently, S. Newport was isolated from 4% of bob veal

carcasses in the Northeastern US with a total Salmonella prevalence determined at nearly

12% in bob veal (Flowers, 2002).

It is widely accepted that the contamination of red meat carcasses with Salmonella

occurs via fecal contamination and that the severity of contamination is largely due to the

extent of infection in the live animal and the handling of carcasses by slaughter employees

(Humphrey, 2000). To illustrate this point, S. Typhimurium was isolated from 70% of lamb

carcasses (n = 10) immediately following the final wash in a Spanish plant with poor

hygienic conditions, but not from lamb carcasses (n = 20) sampled in two other plants

considered to have more hygienic practices (Sierra et al., 1995). In another lamb study,

carcasses from six plants that underwent 24 h chilling before sampling demonstrated a

Salmonella prevalence of 1.5% (Duffy et al., 2001). All of these lamb carcasses were steam

vacuumed, and approximately half of them were treated with an acetic or lactic acid rinse prior to chilling.

Despite efforts to remove Salmonella from meat surfaces during slaughter, carcasses

may still contain small populations, which may be embedded in surface tissues and crevices

or deposited on surfaces during fabrication and further processing. An extensive study of

retail meats from Washington, D. C. grocery stores demonstrated that Salmonella still is present on 3.3% of fresh pork and 1.9% of fresh beef (Zhao et al., 2001). Moreover,

Salmonella spp. have been isolated from veal. Buchanan and Bevan (1939) reported an

outbreak of S. Typhimurium on a passenger train in South Africa that was traced to roasted

veal. Anderson and others (1961) reported a series of incidents of food borne salmonellosis

in 90 people from the handling and consumption of veal and calf products in Southeast

England.

Campylobacter spp.

Campylobacter spp. are one of the most common causes of food borne illness. Due to

the difficulty in isolating these organisms, Campylobacter spp. has only recently been identified as a major meat borne pathogen. In fact, Campylobacter surpassed Salmonella as

the most frequent cause or reported food borne illness in 1997. (CDC, 1997). Many cases of

campylobacteriosis have been linked to the cross contamination of prepared foods with raw

or undercooked foods and with infected food handlers (Doyle, 2004). Outbreaks of food borne illness associated with Campylobacter spp. are generally associated with drinking raw milk or unchlorinated water, mishandling during preparation of raw poultry or consumption of undercooked poultry or poultry products (Gill and Harris, 1982; NACMCF, 1998;

Altekruse and Swerdlow, 2002).

Disease symptoms are typically less severe than illnesses caused by E. coli O157:H7 or Salmonella spp. although a victim of campylobacteriosis would likely show symptoms that are very similar to salmonellosis (Skirrow and Blaser, 2000; Shallow et al., 2001). An infective dose of 500 C. jejuni organisms can cause disease in humans (Robinson, 1981).

Also, C. jejuni and C. coli are the two species that are most often linked to illness from the consumption of muscle foods (Jacobs-Reitsma, 2000). Campylobacteriosis is diagnosed by fecal culture following the onset of severe gastrointestinal upset including abdominal cramps, diarrhea and vomiting. More severe secondary manifestations of Campylobacter infection include Reiter’s syndrome, which is often characterized by chronic joint pain or arthritis for several weeks or even a year following infection, and Guillain-Barré syndrome (GBS), which can cause acute neuromuscular paralysis, long-term disability or death in humans (Stern et al., 1992; Skirrow and Blaser, 2000). Of all known Campylobacter spp., C. jejuni has been

implicated most often in cases of GBS, which develop within one to three weeks of infection

(Nachamkin et al., 2000).

When compared with other Gram-negative bacteria, Campylobacter spp. are

considered fragile organisms and, thus, easier to eradicate. When exposed to an environment

outside the host animal, the survival and growth of Campylobacter spp. can be inhibited by

atmospheric oxygen concentrations, competition with other microflora for nutrients and

suboptimal temperatures (> 30°C; Doyle, 2004; Stern et al., 1992). Furthermore, routine

sanitation of slaughter equipment and facilities easily prevents cross-contamination of

carcass surfaces with Campylobacter (Gill and Harris, 1982).

Compared with E. coli O157:H7 and Salmonella spp., there are fewer data that

describe the prevalence of Campylobacter spp. in red meat animals. Even though data are

insufficient to support performance standards for Campylobacter spp. in the meat and poultry

industry, this organism may be a likely target of future regulation (Lammerding et al., 1988).

Given existing information, it is apparent that pathogenic Campylobacter are common in the

feces and gastrointestinal tracts of red-blooded animals.

Approximately 10% of fecal samples (n = 935) from calves, fat cattle, and cows in

Switzerland contained C. jejuni (Al-Saigh et al., 2004). In Japan, C. jejuni was isolated from

the feces 50% of Japanese black cattle and 31% of Holstein cattle at one slaughterhouse,

whereas C. coli was isolated from the feces of 82.5% of the market hogs at the same plant

(Ono and Yamamoto, 1999). However, in the same study, none of the beef or pork sampled

from retail stores in the same geographical area yielded positive results for Campylobacter .

In a study of fresh meats from markets in Washington, D. C., C. jejuni and C. coli were

isolated from 1% of retail beef and pork and from 3% of retail pork, although the origin of

these organisms was not traced (Zhao et al., 2001).

Gill and Harris (1982) frequently isolated C. jejuni from the fecal samples of unweaned calves but determined that the pathogen may be present, though less prevalent in mature sheep and cattle. When present on calf carcasses, C. jejuni were present in populations no greater than one to ten organisms per cm 2 (Gill and Harris, 1982).

Campylobacter also have been isolated from lamb intestines at slaughter. After collecting the intestinal contents of 360 lambs from a processing facility in central England, 91.7% of lambs were determined to carry Campylobacter in the small intestine (Stanley et al, 1998).

Stanley et al. (1998) concluded that the incidence of Campylobacter in lambs previously had been underestimated because prior studies usually examined fecal matter from the colon and not from the more proximal sections of the intestinal tract. Given the high incidence of

Campylobacter in the gastrointestinal tract, it is of utmost importance to prevent its rupture during evisceration.

On the other hand, Campylobacter spp. were not found on lamb carcasses (n = 30) that were sampled in Spain; however, 30% of these carcasses did contain S. Typhimurium

(Sierra et al., 1995). Relative to the study by Stanley et al. (1998), a much lower incidence of

Campylobacter was detected in feces (7.4%) and on carcasses (0.7%) processed in

Yorkshire, England (Chapman et al., 2001). While lambs may commonly harbor

Campylobacter in the small intestine, good manufacturing practices and carcass hygiene may be the key to preventing contamination of lamb carcass surfaces.

Bacterial attachment to meat surfaces

It is clear that pathogenic microorganisms can create a hazard to human health. The pathogens that may be present in fecal matter, mud, hair, and other foreign matter may not cause harm to humans unless they become attached to meat surfaces. In order to develop and evaluate effective carcass decontamination strategies, it is useful to understand the mechanism of bacterial attachment to meat surfaces.

When physical contact is made between unattached bacteria and a meat surface (particularly, collagen fibers), bacteria are likely to adhere (Notermans, 1979;

Fratamico et al., 1996). Often, the bacteria readily attach themselves to meat surfaces within one minute of exposure (Fratamico et al., 1996). Based on one study, exposure time beyond one minute does not appreciably enhance the number of bacteria that attach to meat surfaces (Fratamico et al., 1996). Fratamico et al. (1996) reported that the amount of bacteria that do attach to meat surfaces is in direct proportion with the magnitude of the inoculum. However, Benito et al. (1997) did not find any correlation between the number of bacteria that attached to beef surfaces and bacterial hydrophobicity, which suggests that the hydrophobicity of the cell membrane determines how strongly bacteria adhere to meat surfaces.

In a study of bacterial hydrophobicity and attachment strength, Benito et al. (1997) examined a variety of spoilage and pathogenic bacteria that are associated with meat products. Clostridium perfringens , Staphylococcus aureus , and Yersinia enterocolitica possessed the greatest strength of attachment, while E. coli and S. Cholerasuis showed

moderate strength (Benito et al., 1997). E. coli and S. Cholerasuis are also moderately less

hydrophobic than these other bacterial species. To counteract the hydrophobicity of bacterial

cell membranes, Tween 20, a surfactant, was added to a 2% lactic acid solution (Calicioglu et

al., 2002). Since Tween 20 is ampiphilic, it can reduce surface tension between hydrophobic

and hydrophilic substances. Compared to water rinsing alone, 1.1 and 1.4 log CFU/cm 2

greater reductions in E. coli O157:H7 populations were achieved when 2% lactic acid and

2% lactic acid mixed with 0.5% sodium benzoate, respectively, were sprayed on beef

surfaces following pretreatment with 5% Tween 20 (Calicioglu et al., 2002). However, beef

samples were chilled at 4°C for 2 days before final bacterial counts were made; so, it is

difficult to separate the combined effect that chilling and residual lactic acid and/or sodium benzoate had on the reduction of E. coli O157:H7 during 2 d of storage.

Similarly, the treatment of preevisceration beef surfaces with a water wash reduced

the likelihood that meat surfaces were contaminated with E. coli O157:H7 and Salmonella

spp. than unwashed meat surfaces by 0.7 log CFU (Dickson, 1995). Beef surfaces that were

washed prior to evisceration had significantly lower surface free energy readings than

unwashed tissue, which is the likely explanation for the difference in bacterial attachment

(Dickson, 1995). The water film on the washed tissue may have reduced the affinity for

attachment by bacteria with strongly hydrophobic cell membranes. Furthermore, bacteria are

more difficult to remove from lean tissue surfaces than adipose tissue surfaces (Cutter and

Siragusa, 1994). When inoculated meat pieces were treated with organic acid rinses at

various concentrations, treated lean tissue retained populations of E. coli O157:H7 and

Pseudomonas fluorescens at 1 log 10 or 2 log 10 , respectively, greater than adipose surfaces

(Cutter and Siragusa, 1994).

Although a water film on washed carcass surfaces may prevent the additional

attachment of bacteria during slaughter, one might also consider the potential disadvantage of

a dilution effect when an antimicrobial rinse is subsequently applied. While no studies to date

have investigated such a dilution effect, it is reasonable to speculate that antimicrobial

effectiveness of chemical mists or rinses is diminished to a certain extent when applied to a

carcass immediately after washing. After bacteria have been physically removed by a water

wash or, perhaps, even injured or killed by hot water washing, it may be prudent to allow

carcasses to drip momentarily before sanitizing with an antimicrobial rinse to allow the rinse

to perform at intended strength.

Another phenomenon to take into account is the increased heat tolerance of bacterial

cells that become attached to meat surfaces. When the heat tolerance (58°C) of S.

Typhimurium was determined in an aqueous suspension or attached to ground pork, 43% of the cells that adhered to pork tissue survived after three minutes of heat treatment while only

3% of the cells heated in nutrient broth could be recovered (Humphrey et al., 1997). It is unclear whether physiological changes occur in the Salmonella cells to enhance their heat tolerance or whether the meat pieces provided insulation from the heat for the attached cells.

Nevertheless, the augmentation of heat tolerance and the strong affinity for adherence to meat surfaces are two qualities that may challenge efforts to decontaminate carcasses.

ANTIMICROBIAL INTERVENTIONS

Water washing

In very small meat plants, spray washing employs the use of a hand-held hose with a spray gun to deliver water to carcass surfaces for a specified time and within a range of temperature and pressure. Hand-held hoses are inexpensive and easy to maneuver.

Essentially, the employee directs the water stream at the carcass surface to wash away gross contamination.

Spray washing in larger plants can also be accomplished with a washing cabinet.

Washing cabinets are made of stainless steel and are equipped with either fixed or rotating nozzles to apply the wash uniformly over the carcass, as well as within the body cavity. For red meat processing, these cabinets may extend from floor to ceiling, and generally range from six to fifteen feet long. In some processing facilities, meat carcasses also will undergo spray washing as described above after evisceration and immediately before refrigeration.

The cost of a spray washing cabinet ranges from $30,000 to $100,000, depending upon the cabinet manufacturer and design specifications. Given the cost and space requirements for this equipment, washing cabinets are uncommon in very small meat establishments.

Water washing can take place at any stage during slaughter. Large and small plants have adequate manpower to assign workers to washing stations at various phases of slaughter, whereas very small plants may delay water washing until the final wash step.

Dickson (1995) explored pre-evisceration washing as a means of reducing bacterial attachment to beef surfaces and demonstrated that the affinity for attachment by aerobes and

Enterobacteriaceae to the carcass surface was lowered by 0.7 log CFU/cm 2 as compared with control carcasses.

Carcass washing can be performed with cold, warm, or hot water. Of these three temperature ranges, a hot water wash is considered the most effective at eliminating bacteria due to mechanical removal combined with high temperature (Bolton et al., 2002). According to Gill et al. (1999), one can expect a 1 to 2 log reduction in aerobic plate counts on carcass surfaces treated with hot water washing (85°C) for at least 10 s, while coliforms and generic

E. coli are generally reduced by 2 log with minimal lean discoloration. In a similar study,

Gill et al. (1998) channeled pork and lamb carcasses through a washing cabinet that applied a hot water wash at 85°C for 10 s, and observed paralleled success at reducing bacterial populations while maintaining satisfactory muscle color. Consequently, the application of this hot water wash in a washing cabinet (pressure and flow rate were not reported) for 20 s was observed to unacceptably discolor lean surfaces of beef carcasses (Gill et al., 1999).

Besides prolonged exposure time, the presence of excessive protein and other solids in the water, which accumulate in the washing apparatus from recirculation, also causes discoloration of carcass surfaces; but, does not contribute to the bacterial load of washed carcass surfaces (Gill and Bryant, 2000). Frequent replacement of the fouled water diminishes this particular discoloration effect (Gill and Bryant, 2000). Gill and Bryant (2000) also assert that discoloration by hot water pasteurization is alleviated by air chilling, given that all exposed lean surfaces experience desiccation and drying and develop a dark red color. Moreover, there was no obvious difference between the appearance of treated and untreated carcasses that have undergone air chilling (Gill and Bryant, 2000).

In addition, Castillo et al. (1998b) applied a hot water wash (24 psi, 95°C, 14.4L/min,

5 s, distance from sample = 12.5 cm) with a hand-held garden sprayer to beef surfaces

inoculated with a fecal slurry containing E. coli O157:H7 and S. Typhimurium and observed

log reductions of 3.7 and 3.8, respectively. APC and coliform counts were also reduced by

2.9 and 3.3 log, in that order (Castillo et al., 1998b). The 2.9 log reduction in APC by

2 Castillo et al. (1998b) is 0.9 log CFU/cm higher than that expected by Gill et al. (1999),

which could be explained by the higher water temperature (95°C).

Although both hot and cold water effect the mechanical removal of bacteria, hot water is considered a more lethal treatment on account of high temperature. The lethality of a hot water wash is attributed to irreparable damage to the cell membrane and vital cytoplasmic components, such as ribosomes and enzymes (Smith and Palumbo, 1982). In the presence of excessive heat, most proteins denature causing leakage of the cell membrane and halting metabolic activity. Hence, cell death, or the inability to recover, occurs.

At the other end of the spectrum, a cold water wash can also eradicate bacteria by physical force. However, it is considered a much less effective decontaminant when compared to hot water washing (Bolton et al., 2002). In Germany, cold water washing (12°C) of lamb carcasses immediately following evisceration was deemed ineffective since bacterial loads on ventral carcass surfaces, which were adjacent to the cut made during evisceration, did not change appreciably (Ellerbroek et al., 1993).

An effective washing treatment should rid the carcass of most gross contamination.

However, washing sometimes causes the redeposition of contaminants from one area of the carcass to another, instead of completely eliminating it (Sheridan, 1998). For example, when the dorsal region of lamb carcasses was swabbed before and after cold water washing,

lumbar swabs after washing showed higher aerobic plate counts than before washing

(Ellerbroek et al., 1993). Overall, water washing is employed as an important carcass

cleaning step, which sets the stage for sanitizer effectiveness.

Rinsing with antimicrobial compounds

Organic acids (e.g. lactic, acetic and citric), chlorinated compounds or other generally recognized as safe (GRAS) compounds are usually applied to carcass surfaces after water washing. These acids also can be applied using washing cabinets, similar to the ones used for water washing in larger plants. A less expensive way for smaller processors to apply antimicrobial solutions to carcasses is with a hand-held spray tank, much like the garden sprayers commonly available at hardware stores or home improvement centers. Overall, these rinses can eliminate microscopic contaminants that remain after the final water wash step. More importantly, some antimicrobial rinse solutions produce an inhospitable environment for microbes that remain on carcass surfaces after water washing and for those that may come in contact with the surface further down the slaughter line. The prolonged contact of the antimicrobial, either before evisceration, after evisceration, or during the refrigeration process has been demonstrated to improve both the shelf life and safety of the meat (Barkate et al., 1993; Cutter and Siragusa, 1994; Hardin et al., 1995; Reagan et al.,

1996; Davey and Smith, 1989; Dickson, 1995; Dormedy et al., 2000). Various chemical compounds used as carcass sanitizers and their efficacy are discussed below.

Lactic, acetic and citric acids are naturally occurring compounds in many foods. FSIS permits the use of organic acids as solutions up to 2.5% for carcass washing prior to chilling

(USDA-FSIS, 2005). Lactic, acetic, and citric acids are the organic acids that are often used

as carcass rinses, with lactic and acetic acids most commonly in use (Cherrington et al.,

1991; Doores, 1993; Siragusa, 1995). When applied as a processing aid, within the approved

limits for species and product type, there is no labeling requirement (USDA-FSIS, 2005).

The complete inactivation of harmful organisms on beef carcass surfaces cannot be

guaranteed, although substantial evidence exists that acid rinsing is an effective means of

reducing bacterial populations on carcass surfaces. Hence, acid rinsing can provide

reductions as a measure of control, but not elimination, in HACCP plans for slaughter,

fabrication and other fresh meat applications (Dormedy et al., 2000).

The mechanism of the growth inhibition of common meat spoilage bacteria by

organic acids can be attributed to more than the effect of lowering pH alone (Ouattara et al.

1997). At pH levels other than established acid dissociation constants (pKa), organic acids do

not fully dissociate in aqueous solutions (Doores, 1993). The dissolved organic acid does

lower pH, forcing bacteria to expend energy to actively export the excess protons.

Furthermore, undissociated organic acid molecules are able to enter the cell and then

dissociate in the cytoplasm where pH is slightly higher than outside the cell (Ouattara et al.,

1997). This shift creates a greater influx of unwanted protons in the cytoplasm and causes the bacterium to work even harder to get rid of them. Eventually, the cell becomes energy- deficient, leading to injury or death (Ouattara et al., 1997).

Lactic acid

In 1780, Carl Wilhelm Scheele, a Swedish chemist, first discovered lactic acid in sour milk (Holten, 1971a). J. Berzelius, another chemist from Sweden, supported Scheele’s finding by isolating lactic acid from ox meat, blood, and fresh milk (Holten et al., 1971a).

During the mid-1800s, the lactic acid found in meat and in fermented foods was thought to be of two different forms. The lactic acid found in meat was provisionally known as paralactic or sarcolactic acid until 1869 when J. Wislicenus determined that the different

forms [L (+) and D (-)] are lactic acid stereoisomers (Holten et al., 1971a).

Since Charles Avery established the first commercial plant for lactic acid

manufacture in 1881, the production scheme has remained conceptually the same; lactic acid bacteria ferment carbohydrate and the by-product is purified. Specifically, sucrose or glucose

is fermented by homolactic lactic acid bacteria in the presence of lime and water to create

calcium lactate (Holten et al., 1971c; PURAC, 2000). Sulfuric acid is added to calcium

lactate to produce lactic acid and hydrous calcium sulfate (gypsum; PURAC, 2000). Next the

crude lactic acid is purified and concentrated for use in foods, health and beauty aids or for

industrial applications (Holten, et al., 1971b).

Lactic acid is the most commonly used organic acid for carcass decontamination

(Sheridan, 1998). van Netten et al. (1994) recommend its use at 2%, which optimizes

antimicrobial efficacy without sacrificing quality, or discoloring of the lean surfaces. Hardin

et al. (1995) measured the efficacy of water washing followed by organic acid (2% lactic acid

or 2% acetic acid) rinsing for the removal of E. coli O157:H7 and S. Typhimurium from beef

carcass surfaces. Rinsing with 2% lactic acid (pH = 2.2, 40 psi, 55°C, 200 ml for 11 s,

distance from sample = 80 cm) eliminated significantly greater numbers of E. coli O157:H7

at more carcass locations than 2% acetic acid, with no disparity in the removal of S.

Typhimurium (Hardin et al., 1995).

In a challenge study of pork surfaces, van Netten et al. (1995) washed bellies,

inoculated with lactic acid-habituated S. Typhimurium, with cold (11°C) or hot (55°C) lactic

acid at 2% or 5% concentrations for 30, 60, 90, and 120 s. For this study, hot lactic acid at

2% applied for 30 s was the mildest combination of physical attributes at which bacterial

elimination (approximately 1 log) could be detected. The authors reported that lean

discoloration after both hot acid rinses was considered unacceptable by a panel of meat

inspectors, who were employed by a plant in The Netherlands. However, the addition of

0.015% nicotinic acid and/or 0.05% ascorbic acid to the lactic acid solutions improved visual

appearance scores by 20% (van Netten et al., 1995).

Acetic acid

Since 10,000 B.C., acetic acid has been utilized by humans and, until the 19 th century, was made available mainly from the oxidation of ethyl alcohol in wine (Partin and Heise,

1993). Modern commercial production of acetic acid relies heavily upon the chemical processes of acetaldehyde oxidation, liquid phase hydrocarbon oxidation, and methanol carbonylation (Fanning, 1993). Because methanol carbonylation is the lowest cost method of large-scale acetic acid production, it is the most widely used method (Fanning, 1993).

Acetic acid has found practical application as a carcass decontaminant. Bell et al.,

(1997) applied a 1% acetic acid spray (40 psi; 25°C, 3.8L/min, 15 s spray followed by 90 s rest followed by 15 s spray; distance from sample = 22 cm) to beef surfaces experimentally inoculated with pathogen models of E. coli and Salmonella and achieved 2.5 log 10 and 3.5

log 10 reductions, respectively, with slight discoloration. Meanwhile, the effectiveness of rinsing with 2% acetic acid at removing E. coli O157:H7 from beef surfaces was compared to rinsing with tap water from 30 to 70°C at 10°C increments (80 psi, 4.2 L/min, 15 s, distance from sample = 17 cm; Cutter et al., 1997). Although the different temperatures had no effect

on the reduction of E. coli O157:H7, 2% acetic acid delivered a >4.3 log CFU/cm 2 reduction

while tap water removed <2.7 log CFU/cm 2 (Cutter et al., 1997). In short, use of acetic acid

at 2% resulted in greater removal of E. coli from beef surfaces than 1% acetic acid. It should

also be noted that the spray pressure and flow rate used by Cutter et al. (1997) was greater

than that used by Bell et al. (1997). The greater pressure and flow rate of the 2% acid rinse

could have interacted with the higher acid concentration to synergistically eliminate more bacteria from beef surfaces.

Citric acid

Citric acid, an erstwhile popular product of Pfizer Inc., was primarily imported from

Italy until World War I disrupted Italian citrus production (Pfizer Inc., 2005). James Currie

secretly developed a large scale fermentation process for Pfizer Inc. to produce citric acid. In

modern industrial production, over 99% of citric acid is manufactured by fermentation

(Crueger and Crueger, 1989). Typically, this process utilizes the fermentative capability of

Aspergillus niger to convert molasses into citric acid, which is then separated and purified

(Noyes, 1969). Mutants of A. niger are preferentially used to inhibit the production of by- products, such as oxalic acid and gluconic acid (Crueger and Crueger, 1989).

Citric acid has been studied as a meat surface decontaminant to a lesser extent than

lactic and acetic acid. Cutter and Siragusa (1994) observed no significant differences in the

reduction of E. coli O157:H7 and Pseudomonas fluorescens , a common meat spoilage

organism, among citric, acetic and lactic acid sprays applied at 1, 3 and 5% concentrations.

In that study, bacterial elimination increased along with acid concentration and bacteria were

more effectively removed from adipose tissue than lean tissue surfaces (Cutter and Siragusa,

1994). Also, 0.25% citric acid was added to mixtures of acetic (1 or 2%), lactic (1 or 2%) and

ascorbic (0.1%) acids and its antimicrobial effectiveness compared with 3% acetic or lactic

acid to rinse beef surfaces inoculated with a fecal slurry of E. coli , S. Typhimurium and

Enterobacteriaceae (Anderson et al., 1992). When applied at 45 or 70°C, the acid mixture with twice as much lactic as acetic acid was the most effective of all acid treatments at reducing microbial load. According to Siragusa (1995), citric and ascorbic acids at low concentrations provide a chelating effect, which improves the antimicrobial effectiveness of acetic and lactic acid applied to meat surfaces as a mixture. However, the greater antimicrobial efficacy could also be attributed to the additive effect of having four organic acids in one solution (Nazer et al., 2005). Although the mixed acid treatment used by

Anderson et al. (1992) contained small quantities of four acids, the rinse solution as a whole contained 3.35% organic acid.

Ouattara et al. (1997) investigated the effect of various organic acids (including acetic, lactic, and citric acids) on pure cultures of common meat spoilage bacteria. Acetic and lactic acids were able to inhibit growth of most of the organisms studied at minimum concentrations of 0.1 to 0.75% (w/v; Ouattara et al., 1997). Citric acid was considered less effective because higher minimum inhibitory concentrations (0.2 to >1.0% w/v) were necessary for the inactivation of spoilage bacteria. However, when these three organic acids were applied to beef carcass tissue in a pilot scale carcass washer (60 psi, 28°C, 4.2 L/min, distance from sample = 17.8 cm, chain speed = 14 m/min, spray nozzle oscillation speed = 80 cycles/min) at different concentrations (1, 3, or 5%), antimicrobial efficacy against E. coli

O157:H7 and Pseudomonas fluorescens was significantly influenced by acid concentration rather than acid type (Cutter and Siragusa, 1994).

Peroxyacetic acid

Peroxyacids, such as peroxyacetic acid, also can be used in antimicrobial rinses for

carcasses and variety meats. In this class of oxidizers, two or more of the following

compounds are mixed together before direct application to meats: peroxyacetic acid; octanoic

acid; acetic acid; hydrogen peroxide; peroxyoctanoic acid; and 1-hydroxyethylidene-1, 1-

diphosphonic acid (also known as HEDP; 21 CFR 173.370). Peroxyacetic acid, by itself, is

considered a strong oxidizer that is capable of denaturing microbial proteins and disrupting

the bacterial cell membrane (Cords and Dychdala, 1993).

When used as a 0.02% solution, peroxyacetic acid reduced E. coli O157:H7 from beef

carcass tissue and beef short plates by 1.4 and 1.0 log CFU/cm 2, whereas water washing produced comparable reductions of 1.2 and 0.3 CFU/cm 2, in that order (Ransom et al., 2003).

Gill and Badoni (2004) also tested the effectiveness of a 0.02% peroxyacetic acid and 4% lactic acid on chilled beef carcass surfaces. Overall, the peroxyacetic acid rinse provided a

Mixtures of peroxyacids that have been optimized for use on beef carcasses are commercially available. The plant employee can dilute the mixture with water to achieve a peroxyacetic acid concentration ≤ 220 ppm and a hydrogen peroxide concentration of ≤ 75 ppm. When used at various concentrations (200, 600, and 1000 ppm) and temperatures (45 and 55°C) to eliminate E. coli O157:H7 and S. Typhimurium from chilled or hot beef carcass surfaces, peroxyacetic acid achieved minor reductions, when compared with 2% or 4% lactic acid rinses (King et al., 2005). As a sanitizer of meat grinding equipment that had been

experimentally inoculated with E. coli O157:H7, 0.2% peroxyacetic acid reduced the presence of the pathogen by at least 2.6 log CFU/g (Farrell et al., 1998).

Chlorinated compounds

Chlorine-based compounds that may be used for microbial decontamination include

sodium hypochlorite (bleach), chlorine dioxide and acidified sodium chlorite. The common

mode of action of these chlorinated compounds is the ability to oxidize components of the

microbial cell membrane, DNA, and amino acids, resulting in the disruption of energy

generation and protein synthesis (Scatina et al., 1985; Cords and Dychdala, 1993).

Chlorinated compounds are the most commonly used sanitizing agent in the food industry because of cost and ease of use (Cords and Dychdala, 1993). However, there has been

increased interest in non-chlorine sanitizers because chlorine can yield disinfection by- products, such as chloroform (an example of a trihalomethane), that are considered harmful

to human heath (Richardson et al., 1998). Residual chlorine in plant wastewater can react

with organic materials in other wastewaters to generate trihalomethanes, which can enter the

drinking water supply (Cords and Dychdala, 1993).

Sodium hypochlorite

Sodium hypochlorite, or household bleach, is an inexpensive, broad-spectrum

sanitizer. Fluctuations in pH, temperature, organic load can vary the quantity of chlorine

available for disinfecting poultry chiller water (Tsai et al., 1992). An earlier report of various

sodium hypochlorite rinses (0.02 to 0.025%; 50 or 200 psi; 2, 15, or 30 s application time;

1.7, 3.4 or 6.8 L/min) on beef plates demonstrated that reduction of APC is most effective

(0.6 to 0.7 log CFU) when the rinse is applied at high pressure, high flow rate, and for the

longest application time studied (Marshall et al., 1977). When applied to beef carcasses in a processing plant, a 0.02% sodium hypochlorite solution effected a 0.3 log CFU/cm 2 greater

reduction in APC than untreated carcasses (Stevenson et al., 1978). In both of these studies,

the pH of sodium hypochlorite was adjusted to 6.0 or 6.5, respectively, with acetic acid,

which converts hypochlorite to hypochlorous acid. It is thought that acidification of chlorine

solutions improves antimicrobial effectiveness due to the increased dissociation of the

chlorinated compound (Cords and Dychdala, 1993).

More recently, a four-year study of chiller water from one commercial poultry plant

indicated that a chlorine concentration in the range of 150 to 250 ppm was necessary to

consistently attain a 2-log reduction in aerobic plate counts (APC) (Tsai et al., 1992).

Furthermore, in samples of chiller water within the lower range of chlorine concentrations (0

to 100 ppm), APC reductions were minimal (Tsai et al., 1992). When used experimentally as

a meat surface decontaminant at 800 ppm, sodium hypochlorite yielded a 0.7 log CFU/cm 2 greater reduction of E. coli O157:H7 than a control water wash (Cutter and Siragusa, 1994).

Although sodium hypochlorite can provide an antimicrobial benefit, it may not be sufficient to reduce potentially harmful bacteria to acceptable levels.

Acidified sodium chlorite

Currently, acidified sodium chlorite (ASC) is permitted for use as a processing aid to decontaminate beef carcasses, parts, and organs, at a range of 500 to 1,200 ppm (21 CFR

173.325). Preparation of ASC in a meat plant can be relatively convenient when powdered sodium chlorite is acidified with a GRAS acid (e.g., citric, acetic, or phosphoric) such that

the resulting pH is 2.5 to 2.9 (21 CFR 173.325). Automated equipment that consistently generates ASC can also be installed in the meat plant. Due to the strong oxidative ability of

ASC and solution instability (i.e., volatility of chlorine dioxide), it must prepared on-site and used immediately.

Recently, Gill and Badoni (2004) applied 1,200 ppm ASC activated with 2% citric acid (final solution ph ≤ 3) to chilled beef carcass surfaces and obtained an average reduction of 0.8 log CFU/cm 2 in aerobic plate counts greater than a water rinse (temperature not reported). Acidified sodium chlorite can also be used in combination with another treatment to achieve greater lethality. For example, Castillo et al. (1999) achieved a 4.5 to 4.6 log reduction of E. coli O157:H7 and S. Typhimurium by spray washing (mean pH = 2.62, 69 kPa, 22.4 to 24.7°C, 140 ml for 10 s) beef carcass tissue with water, followed by spraying

1,200 ppm ASC activated with citric acid. When activated with phosphoric acid and applied under the same physical conditions, the log reduction in bacterial populations by ASC was

3.8 to 3.9 log, while water washing (manual wash: 69 kPa, 1.5 L for 90 s; followed by automated wash: 5 L applied at 1.72 MPa increasing to 2.76 MPa for 9 s) alone afforded a

2.3 log reduction (Castillo et al., 1999). Kemp et al. (2000) speculate that citric acid activation of sodium chlorite may impart greater effectiveness than activation by phosphoric acid because citric acid is thought to chelate components of the cell membrane and cause it to rupture.

Chlorine dioxide

Chlorine dioxide is another chemical alternative for sanitizing red meat carcass surfaces. Used extensively in poultry chiller water, chlorine dioxide retains more oxidizing

capacity than chlorine when organic matter is present (Villareal et al., 1990). Unlike many

chlorine compounds currently in use, chlorine dioxide is not converted into chloramines,

which are toxic to fish, or trihalomethanes, which are harmful to humans at sufficient levels

(Cords and Dychdala, 1993). Chlorine dioxide also maintains antimicrobial effectiveness

over a broad pH range.

Foschino et al. (1998) examined the effectiveness of chlorine dioxide against a

nonpathogenic strain of E. coli , which was suspended in physiological saline with 0.1% tryptone, and reported a 5 log reduction after exposure to 1.4 ppm solution for 30 s.

However, when used as a spray wash (520 kPa, 16°C, 10 s) to decontaminate beef carcass surfaces at 0 to 20 ppm, chlorine dioxide was no more effective than water washing (Cutter and Dorsa, 1995). The abundance of organic matter (lean, fat, as well as microbes) on beef carcass surfaces may overpower the oxidizing capacity of chlorine dioxide to inactivate microorganisms as dissociated chlorine loses efficacy once bound to organic matter

(Siragusa, 1995). Furthermore, the porosity of surfaces may hamper the effectiveness of chlorine dioxide. Although a 5 log reduction in E. coli could be achieved on a smooth, stainless steel surface treated with 7 ppm chlorine dioxide for 6 min, the same concentration applied for 8 min to PVC floor tiles only achieved a reduction of 2 log (Foschino et al.,

1998). Overall, chlorinated compounds are not as popular as other beef carcass sanitizers

(Siragusa, 1995).

Aqueous ozone

Ozone has a long history of use as an antimicrobial for drinking water, given that the first drinking water treatment plant to rely on ozone for disinfection was built in The

Netherlands in 1893 (Brink et al., 1991). The gentle bubbling of ozone gas into water

generates ozonated water, or aqueous ozone, which can then be used to disinfect foods and

food handling surfaces. Ozone is more soluble and less likely to decompose in cold water

than hot water (Kim et al., 1999). Essentially, the mode of action is the release of a third

oxygen atom from an unstable ozone molecule. The singular oxygen atoms make contact

with organic and inorganic debris and rapidly oxidize them. In fact, ozone is considered to be

a stronger oxidizing agent than chlorine dioxide, hypochlorous acid, and chlorine gas

(Manley and Niegowski, 1967). The remaining oxygen gas remains dissolved in the water or

enters the atmosphere. A very detailed review of ozone chemistry is presented by Bablon et

al. (1991).

In the bacterium, ozone is thought to cause lethality by the oxidation of cell

components. It is generally agreed that ozone first oxidizes glycoproteins and lipoproteins in

the outer cell membrane, which may lead to leakage of the cytoplasm. Inside the cell, ozone

nonspecifically oxidizes proteins resulting in their degradation, flocculation, or reduced

ability to function (Kim et al., 1999; Güzel-Seydim et al., 2004b). Furthermore, ozone can

damage DNA and RNA by modifying nitrogen bases, particularly thymine, which usually

causes death due to the inability of the cell to synthesize new proteins (Kim et al., 1999).

Reagan and associates (1996) tested the efficacy of aqueous ozone, hydrogen peroxide, water washing, and knife trimming on experimentally contaminated beef carcasses.

Aqueous ozone (0.3 to 2.3 ppm), hydrogen peroxide (5%) and water washing (74 to 88°C) all

resulted in an approximate 1 log reduction of microbial populations (Reagan et al., 1996).

The results of a more recent study agree with those of Reagan and others (1996). Excised

carcass surfaces were treated with 95 ppm aqueous ozone for 30 s or washed with water

(28°C) for 9 s with very little difference in effectiveness (Castillo et al., 2003). The

effectiveness of aqueous ozone as a carcass decontaminant may be stifled by the

overwhelming presence of organic material on a carcass surface, since ozone can oxidize

components of bacterial cell as favorably as lean or adipose tissue (Castillo et al., 2003).

Güzel-Seydim et al. (2004a) demonstrated that suspensions of locust bean gum, sodium

caseinate and whipping cream actually protected E. coli and Staphylococcus aureus inocula

against ozonation. Yang and Chen (1979) also demonstrated that ozone (2.5%) in the presence of egg albumin (0.05 or 1.0%) reduced antimicrobial effectiveness by

approximately 4 log CFU/ml when compared to the control. On the other hand, its use as a

food sanitizer may be desirable to some processors because the rapid decomposition of ozone

leaves no residue on foods (Güzel-Seydim et al., 2004b).

Poultry processors also can use ozone in chiller water. Yang and Chen (1979)

demonstrated a 7.1 or 3.6 log CFU/ml reduction in total bacteria when spoiled or fresh poultry slurries (rinsate of spoiled gizzards or ground poultry meat diluted with distilled

water, respectively) were treated with 19 ppm aqueous ozone for 4 min (3.2 L/min). Sheldon

and Brown (1986) also experimented with aqueous ozone for chilling poultry carcasses.

Among populations of coliforms, fecal coliforms, APC and psychrotrophic plate counts,

aqueous ozone (3.0 to 4.5 ppm, 7°C, 45 min) yielded a slight antimicrobial advantage over

tap water (7°C, 45 min) in the range of 0.06 to 0.52 log CFU/ml or MPN/ml while

Salmonella were reduced on the order of 0.44 MPN/100 ml (Sheldon and Brown, 1986).

While aqueous ozone may not provide the same level of disinfection on carcass

surfaces as other interventions, some plants apply it to food contact surfaces and equipment

as a sanitizer after cleaning. Use of a portable ozone generator may be the most feasible way

to use aqueous ozone in a very small meat plant due to ease of operation, low equipment

maintenance requirements, and convenience. For plants that decide to use aqueous ozone for

the dual purposes of plant sanitation and carcass sanitation, aqueous ozone can be used on all

meat products in accordance with good manufacturing practices (USDA-FSIS, 2005).

Steam vacuuming

Another technological intervention used in the meat slaughter process is the steam vacuum sanitizer. The steam vacuum sanitizer may employ a stream of hot water (70-80 °C)

to moisten and heat the carcass surface; a vacuum to remove the moisture along with visible

and microscopic contamination from the carcass surface; and steam to sanitize the stainless

steel head. All material vacuumed from the surface is collected into a large container and

emptied regularly. Processors can position steam vacuum sanitizers throughout the plant to

ensure removal of visible contaminants at many steps throughout the slaughter process.

Studies have indicated that steam vacuuming can effectively reduce levels of microbial

contamination (Dorsa et al., 1996; Dorsa et al., 1997; Kochevar et al., 1997; Castillo et al.,

1999). For instance, steam vacuuming applied to beef carcass surfaces for 6 s achieved 2.7 to

2.8 log CFU/cm 2 in APC, Enterobacteriaceae , coliforms, and generic E. coli (Castillo et al.,

1999). A steam vacuum sanitizer costs from $15,000 to $20,000 per station and while it is used primarily by large or small beef processing establishments, some very small plants have also installed them.

Steam pasteurization

Steam pasteurization employs a multi-step process involving water removal, heating, and cooling of meat carcass surfaces within an enclosed cabinet. Similar to the spray washing cabinets described previously, these enclosed, stainless steel systems are 14 to 36 feet long and extend from the ceiling to the floor. After evisceration and before refrigeration, carcasses may undergo spray washing to apply hot water (76 to 96°C) or an antimicrobial; pass through high velocity fans (air velocity = 119 km/h, air volume = 170 m 3/min, drying time = 20 s) to

remove residual moisture from the surface and enter the steam pasteurization cabinet (Nutsch

et al., 1997; Phebus et al, 1997). Within the cabinet, steam (95 °C) is applied directly to the

carcass surface for several seconds to obtain a carcass surface temperature of approximately

80 °C. The heated carcass travels through the cabinet until it is sprayed with cold water to bring the surface temperature down prior to refrigeration. The high temperature is

detrimental to pathogenic microorganisms by affecting 4.22 to 4.85 log reductions in Listeria

monocytogenes Scott A, E. coli O157:H7, and Salmonella Typhimurium (Phebus et al.,

1997). Gill and Bryant (1997) demonstrated a 1.0 log reduction in total aerobic counts and

2.4 and 2.7 log reductions in coliforms and E. coli , respectively, on beef carcasses that were

steam pasteurized in a commercial plant. Moreover, steam pasteurization does not adversely

affect the color or texture of the meat surface (Gill and Bryant, 1997; Phebus et al., 1997).

The cost of steam pasteurization systems ranges from $400,000 for small plants to over one

million dollars per unit for large plants.

Chemical dehairing

Some preliminary studies have elucidated the potential of chemical dehairing as an effective means of pathogen control especially because the hide is known to be a major source of fecal contamination. A patented chemical dehairing method is summarized by

Schnell and others (Clayton and Bowling, 1989; 1995). The stunned beef animal is sent down the rail to a washing cabinet to be pre-rinsed with water (120 psi, 40°C, 23 s), sprayed with

10% sodium sulfide (50 psi, 25°C, 16 s) followed by a 90 s dwell before a second spray treatment with 10% sodium sulfide (80 psi, 25°C, 16 s) and additional dwell of 60 s. The carcass is transported further along the cabinet undergoing a water rinse (300 psi, 40.5°C, 50 s) and then a 3% hydrogen peroxide rinse (50 psi, 17 s), which neutralizes sodium sulfide.

Next, a third water rinse (120 psi, 40.5°C, 23 s) followed by a second rinse with 3% hydrogen peroxide (50 psi, 17 s) are applied to the hide surface followed by two more water washes, one at 120 psi, 40.5°C, and for 23 s followed by a fresh water rinse at 120 psi for 23 s. Upon completion of the dehairing step, the carcass is exsanguinated and subjected to the remainder of the processes associated with primary processing.

Nou et al. (2003) were able to achieve reductions in APC and Enterobacteriaceae by

2.0 and 1.8 log, respectively, greater on dehaired carcasses than conventionally processed carcasses. In these experiments, carcasses were dehaired by a proprietary chemical procedure that is similar to the one tested by Schnell and others (1995). Although dehairing appeared to improve carcass hygiene, this chemical process improved hide hygiene only slightly. Of 240 dehaired carcasses, 67% were positive for E. coli O157:H7 after dehairing while a significantly higher proportion (88%, α = 0.05) of the conventionally processed carcasses were positive for the pathogen (Nou et al., 2003). On the other hand, Schnell et al. (1995)

demonstrate that dehaired carcasses had somewhat greater coliform counts (0.3 log) than

non-dehaired carcasses, which does not corroborate the advantage of dehairing.

To date, few, if any, plants are employing this technology. It has been speculated that

the processing of dehaired carcasses in a plant environment strictly dedicated to dehairing

would improve hide hygiene because, theoretically, there would be fewer microbial aerosols

in the air and processing equipment would stay cleaner (Schnell et al., 1995; Sofos et al.,

1998). Additional research is necessary to optimize the processing of chemically dehaired

carcasses. There may also be cost barriers to adopting this technology associated with buying

equipment, expanding the processing floor, and treating additional wastewater.

Knife trimming

Knife trimming is frequently performed in plants for the physical removal of fecal matter, ingesta, and milk that visually contaminate carcass surfaces. When the antimicrobial effectiveness of knife trimming is compared to water washing and organic acid rinsing, knife trimming was equally as effective as water washing; but, overall it was not as effective as water washing combined with organic acid rinsing (Hardin et al., 1995). In a comparison between knife trimming followed by water washing and water washing alone, no significant differences were detected in the reduction of E. coli or total plate count (Gorman et al.,

1995). In contrast, Prasai et al. (1995) determined that knife trimming followed by a water wash is more advantageous than knife trimming alone or water washing alone. Washing alone can spread contaminants from one area of the carcass to another, whereas knife trimming alone removes visible debris but fails to eliminate contaminants on untrimmed surfaces (Prasai et al., 1995). In combination, knife trimming removes gross contamination

from selected surfaces and thorough water washing makes physical contact with all areas of

the carcass to detach and eliminate microbes.

Antimicrobial interventions used in combination

Antimicrobial carcass interventions that consist of two or more surface treatments are generally more effective than water washing or chemical rinsing alone. The efficacy of these multi-step approaches can be explained by the hurdle concept (Leistner, 1992). When microorganisms are exposed to a sequence of treatments, the first step inhibits the proliferation of surface microflora by physical removal, cell injury or cell death. The subsequent treatments can then target the microbes that linger on beef surfaces to further enhance safety (Leistner, 1992).

For example, the efficacy of pre-evisceration washing (345 kPa, 21-54°C, 5.6 s), acetic acid rinsing (2%, 207 kPa, 38-54°C, 5.6 s), warm water washing (2069 kPa, 21-54°C,

20 s), and hot water washing (207 kPa, 80°C, 5.6 s) were performed individually or as sequential spray treatments (Graves Delmore et al., 1998). The same spraying apparatus (9.5

L/min, distance from sample = 25 cm) was used for all treatments. The authors reported a range of log reductions of APC and E. coli of 1.2 to 2.2 log CFU/cm 2 and 1.1 to 1.8 log

CFU/cm 2, correspondingly, when treatments were performed individually. However, the sequential application of four treatments lowered bacterial populations by 2.8 to 4.3 log

CFU/cm 2 (Graves Delmore et al., 1998).

Castillo et al. (1998b) reported that warm water washing (35°C) or knife trimming in combination with a hot water rinse (95°C) and/or 2% lactic acid rinse on experimentally inoculated and freshly slaughtered beef carcass tissue yielded reductions in E. coli O157:H7,

S. Typhimurium, coliforms, and generic E. coli on the order of 4.0 to >4.9 log CFU/cm 2.

Performing two or more interventions sequentially provided at least a 1.4 log CFU/cm 2 advantage in bacterial reductions than performing a singular cleaning step, i.e. warm water washing or knife trimming (Castillo et al., 1998b). Because water washing and knife trimming can inadvertently redistribute microorganisms to other areas of the carcass, diligent implementation of additional interventions should help reduce levels of potentially harmful bacteria (Castillo et al., 1998b).

Physical attributes of interventions

Pordesimo et al. (2002) provide a compelling argument that engineering variables, or

the physical attributes, of intervention treatments should receive more attention in future

meat safety research. A review of the literature revealed a lack of standardization of process

engineering variables (temperature, pressure, volume, exposure time, distance between food

and spray origin, spray pattern and spray nozzle type, number and orientation) among the

numerous studies that have documented the efficacy of water washing and the application of

antimicrobial solutions (Pordesimo et al., 2002). Furthermore, there is a need to identify

interactions, or the synergistic effects of combining spray wash engineering variables, so that

carcass decontamination procedures can optimally harness the kinetic energy of spray

droplets (Pordesimo et al., 2002).

The pressure at which a water wash or chemical rinse is applied to a beef carcass

surface can influence antimicrobial efficacy. According to Gorman et al. (1995), spray

washes applied at 200, 300 or 400 psi were more effective at removing fecal matter and bacteria on beef adipose surfaces than a lower pressure wash (40 psi). Studies have indicated

the need to demarcate optimal pressure ranges for spray washing since pressure that is too

low inadequately removes microbes from meat surfaces, and pressure that is too high can

actually drive surface microbes deep into carcass tissues (Pordesimo et al., 2002).

Temperature is a critical factor in the effectiveness of water washing as hot and warm

water are generally more lethal treatments than cold water. In fact, washing beef carcasses

with cold water was deemed ineffective at reducing aerobic plate counts at various carcass

locations and appeared to transfer contaminants from the hindquarter down to the forequarter

(Bell, 1997). Furthermore, Dorsa et al. (1997) assert that water washes at 70°C or greater are

very effective at eliminating E. coli , Salmonella spp., and background microflora.

Future/potential interventions

Current and future research to improve beef safety will continue to explore innovative

methods and evaluate novel compounds for carcass decontamination. Recent studies have

suggested that irradiation, ultraviolet light, or sodium metasilicate have the potential to

develop into carcass interventions for microbiological safety.

In December 1999, the USDA finalized the regulations that govern the irradiation of

meat and poultry products (USDA-FSIS, 1999b). According to this final ruling, meat plants

establish their own guidelines for irradiation dosage (not to exceed 4.5 kGy for refrigerated product or 7 kGy for frozen product) and the minimum level of pathogen reduction (USDA-

FSIS, 1999b). Irradiation has been evaluated for its effectiveness to reduce the prevalence of

E. coli O157:H7 on beef surfaces. Recently, cutaneous trunci pieces were inoculated with E.

coli O157:H7 and exposed to E-beam irradiation at 1.0 kGy with 15 mm of penetration

(Arthur et al., 2005). The authors report a 2.6 to 2.9 log reduction of E. coli O157:H7 when

beef surfaces contained a low-level inoculum and a 5.7 to 6.6 log reduction for high-

inoculum beef surfaces (Arthur et al., 2005). Additional studies could evaluate the

antimicrobial effectiveness of E-beam irradiation on carcass surfaces at other stages during primary processing and fabrication.

In many beef safety experiments, ultraviolet light is used to sanitize excised beef surfaces under laboratory conditions, prior to inoculation and/or antimicrobial treatments

(Cutter and Siragusa, 1994; Dorsa et al., 1996; Lim and Mustapha, 2004). Ultraviolet light is classified as radiant energy and it is exposure to radiant energy that accomplishes food irradiation (USDA-FSIS, 1999a). This technology could be a practical way to improve the microbiological safety of beef because carcass contamination is confined to the surface while the internal tissues are considered sterile. Ultraviolet light chambers could be installed along the rail in beef processing plants to rapidly inactivate surface contaminants from beef carcasses at various stages of slaughter or before fabrication.

Sodium metasilicate has shown promise as an effective means of controlling E. coli

O157:H7. This compound is GRAS and currently may be used on beef carcasses as a 4 ± 2% solution (USDA-FSIS, 2005). Preliminary in vitro data demonstrate complete (>99.99%) inhibition of three strains of E. coli O157:H7 after 5 to 10 s of exposure to 0.6% sodium metasilicate (Weber et al., 2004). Further research is needed to test antimicrobial effectiveness of this alkali on meat surfaces and perhaps on carcasses in a plant setting.

Though unclear, the mode of inhibition is thought to be similar to that of trisodium phosphate, another alkaline disinfectant. The combination of high pH and the ability of anions to bind metal ions interfere with bacterial metabolism as well as damage bacterial cell walls and membranes (Weber et al., 2004). Other compounds currently used in the

processing and manufacture of other food commodities may crossover to the beef industry

for use as antimicrobial agents during slaughter.

CONCLUSIONS

In summary, there are numerous chemical, physical, and biological hazards that pose

a threat to red meat safety during slaughter. The development and implementation of a good

HACCP plan provides the backbone for controlling these hazards. Slaughter workers should be vigilant while performing processing tasks and be attentive to personal hygiene. Because

carcass surfaces are unavoidably contaminated during primary processing, meat plants

employ one or more antimicrobial interventions to successfully eliminate harmful

microorganisms. The sum of these factors should keep the red meat supply safe during the

important steps that take place during slaughter.

REFERENCES

Aberle, E. D., Forrest, J. C., Gerrard, D. E., and Mills, E. W. 2001. Conversion of muscle

to meat and development of meat quality. Ch. 5 in Principles of meat science 4 th ed.

p.83-107. Kendall/Hunt Publishing Company, Dubuque, IA.

Al-Saigh, H., Zweifel, C., Blanco, J., Blanco, J. E., Blanco, M., Usera, M. A., and

Stephan, R. 2004. Fecal shedding of Escherichia coli O157:H7, Salmonella , and

Campylobacter in Swiss cattle at slaughter. J. Food Protect. 67:679-684.

Altekruse, S, F. and Swerdlow, D. L. 2002. Campylobacter jejuni and related organisms. Ch.

6 In: Foodborne Diseases, 2nd ed. D. O. Cliver and H. P. Riemann (Eds.) p. 103-112.

Academic Press, San Diego, CA.

Anderson, E. S., Galbraith, N. S., and Taylor, C. E. D. 1961. An outbreak of human

infection due to Salmonella Typhimurium phage-type 20a associated with infection

in calves. Lancet 277:854-858.

Anderson, M. E., Marshall, R. T., and Dickson, J. S. 1992. Efficacies of acetic, lactic and two

mixed acids in reducing numbers of bacteria on surfaces of lean meat. J. Food Safety

12:139-147.

Arthur, T. M., Bosilevac, J. M., Nou, X., Shackelford, S. D., Wheeler, T. L., Kent, M. P.,

Jaroni, D., Pauling, B., Allen, D. M., and Koohmaraie, M. 2004. Escherichia coli

O157 prevalence and enumeration of aerobic bacteria, Enterobacteriaceae , and

Escherichia coli O157 at various steps in commercial beef processing plants. J.

Food Protect. 67:658-665.

Arthur, T. M., Wheeler, T. L., Shackelford, S. D., Bosilevac, J. M., Nou, X., and

Koohmaraie, M. 2005. Effects of low-dose, low-penetration electron beam

irradiation of chilled beef carcass surface cuts on Escherichia coli O157:H7 and

meat quality. J. Food Protect. 68:666-672.

Bablon, G., Bellamy, W. D., Bourbigot, M., Daniel, F. B., Doré, M., Erb, F., Gordon, G.,

Langlais, B., Laplanche, A., Legube, B., Martin, G., Masschelein, W. J., Pacey, G.,

Reckhow, D. A., and Ventresque, C. 1991. Fundamental Aspects. Ch. 2 In: Ozone in

water treatment: application and engineering, B. Langlais, D. A. Reckhow, and D. R.

Brink, (Eds.) p. 11-132. American Water Works Association Research Foundation,

Denver, CO and Lewis Publishers, Inc., Chelsea, MI

Barkate, M. L., Acuff, G. R., Lucia, L. M., and Hale, D. S. 1993. Hot water

decontamination of beef carcasses for reduction of initial bacterial numbers. Meat

Sci. 35, 397-401.

Barkocy-Gallagher, G. A., Arthur, T. M., Rivera-Betancourt, M., Nou, X., Shackelford, S.

D., Wheeler, T. L., and Koohmaraie, M. 2003. Seasonal prevalence of shiga-toxin

producing Escherichia coli , including O157:H7 and non-O157 serotypes, and

Salmonella in commercial beef processing plants. J. Food Prot. 66:1978-1986.

Bell, K. Y., Cutter, C. N., and Sumner, S. S. 1997. Reduction of food borne micro-

organisms on beef carcass tissue using acetic acid, sodium bicarbonate, and

hydrogen peroxide spray washes. Food Microbiol. 14:439-448.

Benito, Y., Pin, C., Luisa Marin, M., Luisa Garcia, M., Dolores Selgas, M., and Casas, C.

1997. Cell surface hydrophobicity and attachment of pathogenic and spoilage

bacteria to meat surfaces. Meat Sci. 45:419-425.

Bolder, N. M. 1997. Decontamination of meat and poultry carcasses. Trends Food Sci.

Technol. 8:221-227.

Bolton, D. J., Pearce, R. A., Sheridan, J. J., Blair, I. S., McDowell, D. A., and Harrington,

D. 2002. Washing and chilling as critical control points in pork slaughter hazard

analysis and critical control point (HACCP) systems. J. Appl. Microbiol. 92:893-

902.

Bosilevac, J. M., Arthur, T. M., Wheeler, T. L., Shackelford, S. D., Rossman, M., Reagan, J.

O., and Koohmaraie, M. 2004. Prevalence of Escherichia coli O157 and levels of

aerobic bacteria and Enterobacteriaceae are reduced hen hides are washed and

treated with cetylpyridinium chloride at a commercial beef processing plant. J. Food

Protect. 67:646-650.

Bouvet, J., Montet, M. P., Rossel, R., Le Roux, A., Bavai, C., Ray-Gueniot, S., Mazuy, C.,

Atrache, V., and Vernozy-Rozand, C. 2002. Effects of slaughter processes on pig

carcass contamination by verotoxin-producing Escherichia coli and E. coli O157:H7.

Int. J. Food Microbiol. 77:99-108.

Brink, D. R., Langlais, B., and Reckhow, D. A. 1991. Introduction. Ch. 1 In: Ozone in

water treatment: application and engineering, B. Langlais, D. A. Reckhow, and D. R.

Brink, (Eds.) p. 1-10. American Water Works Association Research Foundation,

Denver, CO and Lewis Publishers, Inc., Chelsea, MI

Buchanan, G. and Bevan, C. de V. 1939. Record of a food-poisoning outbreak: The role

played by ice. S. Afr. J. Med. Sci. 4:111-116.

Calicioglu, M., Kaspar, C. W., Buege, D. R., and Luchansky, J. B. 2002. Effectiveness of

spraying with Tween 20 and lactic acid in decontaminating inoculated

Escherichia coli O157:H7 and indigenous Escherichia coli biotype I on beef. J.

Food Protect. 65:26-32.

Castillo, A., Lucia, L. M., Goodson, K. J., Savell, J. W., and Acuff, G. R. 1998a.

Comparison of water wash, trimming, and combined hot water and lactic acid

treatments for reducing bacteria of fecal origin on beef carcasses. J. Food Protect.

61:823-828.

Castillo, A., Lucia, L. M., Goodson, K. J., Savell, J. W., and Acuff, G. R. 1998b. Use of hot

water for beef carcass decontamination. J. Food Protect. 61:19-25.

Castillo, A., L. M. Lucia, K. J. Goodson, J. W. Savell, and G. R. Acuff. 1999.

Decontamination of beef carcass surface tissue by steam vacuuming alone and

combined with hot water and lactic acid sprays. J. Food Protect. 62:146-151.

Castillo, A., McKenzie, K. S., Lucia, L. M., and Acuff, G. R. 2003. Ozone treatment for

reduction of Escherichia coli O157:H7 and Salmonella serotype Typhimurium on

beef carcass surfaces. J. Food Protect. 66:775-779.

CDC, 1998. 1997 Final FoodNet surveillance report. Foodborne Diseases Active

Surveillance Network. Centers for Disease Control and Prevention, Atlanta, GA.

CDC, 1998. 1998 Annual Summary: Table 1. The 20 most frequently reported Salmonella

serotypes from human sources reported to CDC in 1998 and from nonhuman sources

reported to CDC and USDA in 1997. PHLIS Surveillance data.

http://www.cdc.gov/ncidod/dbmd/phlisdata/salmonella.htm, accessed: February 16,

2002.

CDC, 2002. Preliminary FoodNet data on the incidence of foodborne illnesses – selected

sites, United States, 2001. MMWR 51:325-329.

CDC, 2003. Infectious disease information. December 5, 2003.

http://www.cdc.gov/ncidod/diseases/food/index.htm, accessed October 10, 2005.

Chapman, P. A., Cerdán Malo, A. T., Ellin, M., Ashton, R., and Harkin, M. A. 2001.

Escherichia coli O157 in cattle and sheep at slaughter, on beef and lamb carcasses

and in raw beef and lamb products in South Yorkshire, UK. Int. J. Food Microbiol.

64:139-150.

Cherrington, C. A., Hinton, M., Mead, G. C., and Chopra, I. 1991. Organic acids: chemistry,

antibacterial activity and practical applications. Adv. Microb. Physiol. 32:87-108.

Clayton, R. P. and Bowling, R. A., inventors; Monfort of Colorado, Inc., assignee.

Animal slaughtering chemical treatment and method. US Patent 4,862,557. 1989

Sep 5

Cords, B. R. and Dychdala, G. R.1993. Sanitizers: halogens, surface-active agents, and

peroxides. Ch. 14 in Antimicrobials in foods , 2 nd ed. P. M. Davidson and A. L.

Barmen (Eds.), p. 469-537. Marcel Dekker, Inc., New York.

Crueger, W. and Crueger, A. 1989. Organic acids. Ch. 8 in: Biotechnology: a textbook of

industrial microbiology , 2 nd ed. T. D. Brock (Ed.), p. 134-149. Sinauer

Associates, Inc., Sunderland, MA. Translation of: Lehrbuch der angewandten

Mikrobiologie.

Cutter, C. N. and Dorsa, W. J. 1995. Chlorine dioxide spray washes for reducing fecal

contamination on beef. J. Food Protect. 58:1294-1296.

Cutter, C. N., Dorsa, W. J., and Siragusa, G. R. 1997. Parameters affecting the efficacy of

spray washes against Escherichia coli O157:H7 and fecal contamination on beef. J.

Food Protect. 60:614-618.

Cutter, C. N. and G. R. Siragusa. 1994. Efficacy of organic acids against Escherichia coli

O157:H7 attached to beef carcass tissue using a pilot scale model carcass washer. J.

Food Protect. 57:97-103.

Cutter, C. N. and Siragusa, G. R. 1994. Application of chlorine to reduce populations of

Escherichia coli on beef. J. Food Safety 15:67-75.

Davey, K. R., and Smith, M. G. 1989. A laboratory evaluation of a novel hot water

cabinet for the decontamination of sides of beef. Int. J. Food Sci. 24, 305-316.

Delazari, I., Iaria, S. T., Riemann, H., Cliver, D. O., Jothikumar, N. 1998. Removal of

Escherichia coli O157:H7 from surface tissues of beef carcasses inoculated with wet

and dry manure. J. Food Protect. 61:1265-1268.

Dickson, J. S. and M. E. Anderson. 1992. Microbiological decontamination of food

animal carcasses by washing and sanitizing systems: a review. J. Food Protect.

55:133-140.

Dickson, J. S. 1995. Susceptibility of preevisceration washed beef carcasses to

contamination by Escherichia coli O157:H7 and salmonellae. J. Food Protect. 58:

1065-1068.

Doores, S. 1993. Organic acids. Chapter 4 in Antimicrobials in foods , 2 nd ed. Davidson, P. M.

and Branen, A. L. (Eds.) p. 95-136. Marcel Dekker, Inc. New York.

Dormedy, E. S., M. M. Brashears, C. N. Cutter, D. E. Burson. 2000. Validation of acid

washes as critical control points in hazard analysis and critical control point

systems. J. Food Protect. 63:1676–1680.

Dorsa, W. J., C. N. Cutter, G. R. Siragusa, and M. Koohmaraie. 1996. Microbial

decontamination of beef and sheep carcasses by steam, hot water spray washes, and a

steam-vacuum sanitizer. J. Food Protect. 59:127-135.

Dorsa, W. J., C. N. Cutter, and G. R. Siragusa. 1997. Effects of steam-vacuuming and hot

water spray wash on the microflora of refrigerated beef carcass surface tissue

inoculated with Escherichia coli O157:H7, Listeria innocua , and Clostridium

sporogenes . J. Food Protect. 60:619-624.

Dorsa, W. J. 1997. New and established carcass decontamination procedures commonly

used in the beef-processing industry. 60: 1146-1151.

Doyle, M. P. 2004. Campylobacter jejuni and other species. In: Bacteria Associated with

food borne diseases. IFT Scientific Status Summary,

http://members.ift.org/NR/rdonlyres/3DEA7A91-DF48-42CE-B195

06B01C14E273/0/bacteria.pdf, accessed August 17, 2005.

Doyle, M. P. and Schoeni, J. L. 1987. Isolation of Escherichia coli O157:H7 from retail

fresh meats and poultry. Appl. Environ. Microbiol. 53:2394-2396.

Duffy, E. A., Belk. K. E., Sofos, J. N., Bellinger, G. R., Pape, A., and Smith, G. C. 2001.

Extent of microbial contamination in United States pork retail products. J. Food

Protect. 64:172-178.

Echeverry, A., Loneragan, G. H., Wager, B. A., and Brashears, M. M. 2005. Effect of

intensity of fecal pat sampling on estimates of Escherichia coli O157 prevalence.

Amer. J. Vet. Res. 66:2023-2027.

Edel, W., Guinée, P. A. M., and Kampelmacher, E. H. 1970. Salmonella infection in

fattening calves after slaughter. Zbl. Vet. Med. B. 17:479-484.

Elder, R. O., Keen, J. E., Siragusa, G. R., Barkocy-Gallagher, G. A., Koohmaraie, M., and

Laegreid, W. W. 2000. Correlation of enterohemorrhagic Escherichia coli O157

prevalence in feces, hides, and carcasses of beef cattle during processing. Proc. NAS

97:2999-3003.

Ellerbroek, L. L., Wegener, J. F., and Arndt, G. 1993. Does spray washing of lamb

carcasses alter bacterial surface contamination? J. Food Protect. 56:432-436.

Fanning, A. T. 1993. Ethylene- and acetylene-based processes. Ch. 2 in: Acetic acid and its

derivatives . p. 15-25. V. H. Agreda and J. R. Zoeller (Eds.) Series: Chemical

industries; v. 4 Marcel Dekker Inc., New York.

Farrell, B. L., Ronner, A. B., and Wong, A. C. L. 1998. Attachment of Escherichia coli

O157:H7 in ground beef to meat grinders and survival after sanitation with

chlorine and peroxyacetic acid. J. Food Protect. 61:817-822.

FDA, 1992. Food borne Pathogenic Microorganisms and Natural Toxins Handbook (Bad

Bug Book). Center for Food Safety & Applied Nutrition

http://www.cfsan.fda.gov/~mow/chap6.html, accessed: August 8, 2005.

Feder, I., Wallace, F. M., Gray, J. T., Fratamico, P., Fedorka-Cray, P. J., Pearce, R. A.,

Call, J. E., Perrine, R., and Luchansky, J. B. 2003. Isolation of Escherichia coli

O157:H7 from intact colon fecal samples of swine. Emerging Infect. Dis. 9:380-

383.

Fegan, N., Higgs, G., Vanderlinde, P., and Desmarchelier, P. 2005. An investigation of

Escherichia coli O157 contamination of cattle during slaughter at an abattoir. J.

Food Protect. 68:451-457.

Flowers, S. L. 2002. Prevalence of Escherichia coli O157:H7 and Salmonella spp. in

special-fed and bob veal in the Northeastern United States. MS Thesis, The

Pennsylvania State University, University Park, PA.

Foschino, R., Nervegna, I., Motta, A., and Galli, A. 1998. Bactericidal activity of chlorine

dioxide against Escherichia coli in water and on hard surfaces. J. Food Protect.

61:668-672.

Fratamico, P. M., Schultz, F. J., Benedict, R. C., Buchana, R. L., and Cooke, P. H. 1996.

Factors influencing attachment of Escherichia coli O157:H7 to beef tissues and

removal using selected sanitizing rinses. J. Food Protect. 59:453-459.

Gay, J. M., Rice, D. H., and Steiger, J. H. 1994. Prevalence of fecal Salmonella shedding by

cull dairy cattle marketed in Washington state. J. Food Protect. 57:195-197.

Gill, C. O., and Badoni, M. 2004. Effects of peroxyacetic acid, acidified sodium chlorite or

lactic acid solutions on the microflora of chilled beef carcasses. Int. J. Food

Microbiol. 91:43-50.

Gill, C. O., and J. Bryant. 1997. Decontamination of carcasses by vacuum-hot water

cleaning and steam pasteurizing during routine operations at a beef packing plant.

Meat Sci. 47: 267-276.

Gill, C. O., and Bryant, J. 2000. The effects on product of a hot water pasteurizing

treatment applied routinely in a commercial beef carcass dressing process. Food

Microbiol. 17:495-504.

Gill, C. O., Bryant, J., and Bedard, D. 1999. The effects of hot water pasteurizing treatments

on the appearances and microbiological conditions of beef carcass sides. Food

Microbiol. 16:281-289.

Gill, C. O. and Harris, L. M. 1982a. Contamination of red-meat carcasses by

Campylobacter fetus subsp. jejuni . Appl. Environ. Microbiol. 43:977-980.

Gill, C. O., Jones, T., and Badoni, M. 1998. The effects of hot water pasteurizing treatments

on the microbiological conditions and appearances of pig and sheep carcasses. Food

Res. Internat. 31:273-278.

Gorman, B. M., Morgan, B. J., Sofos, J. N., and Smith, G. C. 1995. Microbiological and

visual effects of trimming and/or spray washing for removal of fecal material

from beef. J. Food Protect. 58:984-989.

Graves Delmore, L. R., Sofos, J. N., Schmidt, G. R., and Smith, G. C. 1998.

Decontamination of inoculated beef with sequential spraying treatments. J. Food

Sci. 63:890-893.

Griffin, P. M., Bell, B. P., Cieslak, P. R., Tuttle, J., Barrett, T. J., Doyle, M. P.,

McNamara, A. M., Shefer, A. M. and Wells, J. G. 1994. Large outbreak of

Escherichia coli O157:H7 infections in the Western United States: the big picture.

In Recent advances in verocytotoxins-producing Escherichia coli infections, M. A.

Karmali and A. G. Goglio (Eds.) pgs. 7-12. Elsevier Science, Amsterdam, The

Netherlands.

Griffin, P. M., and Tauxe, R. V. 1991. Epidemiology of infections caused by Escherichia coli

O157:H7, other enterohemorrhagic E. coli , and the associated hemolytic uremic

syndrome. Epidemiol. Rev. 13:60-98.

Güzel-Seydim, Z., Bever, Jr., P. I., and Greene, A. K. 2004a. Efficacy of ozone to reduce

bacterial populations in the presence of food components. Food Microbiol. 21:

475-479.

Güzel-Seydim, Z., Greene, A. K., and Seydim, A. C. 2004b. Use of ozone in the food

industry. Lebensm.-Wiss. u.Technol. 37:453-460.

Hancock, D. D., Besser, T. E., Kinsel, M. L., Tarr, P. I., Rice, D. H., and Paros, M. G.

1994. The prevalence of Escherichia coli O157:H7 in dairy and beef cattle in

Washington state. Epidemiol. Infect. 113:199-207.

Hardin, M. D., Acuff, G. R., Lucia, L. M., Oman, J. S., and Savell, J. W. 1995.

Comparison of methods for decontamination from beef carcass surfaces. J. Food

Protect. 58:368-374.

Heuvelink, A. E., van den Biggelaar, F. L. A. M., de Boer, E., Herbes, R. G., Melchers, W. J.

G., Huis In ‘T Veld, J. H. J., and Monnens, L. A. H. 1998a. Isolation and

characterization of verocytotoxin-producing Escherichia coli O157 strains from

Dutch cattle and sheep. J Clin. Microbiol. 36:878-882.

Heuvelink, A. E., van den Biggelaar, F. L. A. M., Zwartkruis-Nahuis, J. T. M., Herbes, R.

G., Huyben, R., Nagelkerke, N., Melchers, W. J. G., Monnens, L. A. H., and de

Boer, E. 1998b. Occurrence of verocytotoxin-producing Escherichia coli O157 on

Dutch dairy farms. J. Clin. Microbiol. 36:3480-3487.

Holten, C. H., Müller, A., and Rehbinder, D. 1971a. History. Chpt. 1 in Lactic acid:

Properties and chemistry of lactic acid and derivatives . p. 3-6. Verlag Chemie

GmbH, Weinheim/Bergstr.

Holten, C. H., Müller, A., and Rehbinder, D. 1971b. Physical chemistry. Chpt. 5 in Lactic

acid: Properties and chemistry of lactic acid and derivatives . p. 62-86. Verlag

Chemie GmbH, Weinheim/Bergstr.

Holten, C. H., Müller, A., and Rehbinder, D. 1971c. Biochemistry. Chpt. 17 in Lactic

acid: Properties and chemistry of lactic acid and derivatives . p. 412-460. Verlag

Chemie GmbH, Weinheim/Bergstr.

Humphrey, T. 2000. Public-health aspects of Salmonella infection. Chpt. 15 in: Salmonella

in domestic animals. Pgs. 245-263. C. Wray and A. Wray (Eds.). CAB International ,

New York.

Humphrey, T. J., Wilde, S. J., and Rowbury, R. J. 1997. Heat tolerance of Salmonella

Typhimurium DT104 isolates attached to muscle tissue. Letters Appl. Microbiol.

25:265-268.

Jacobs-Reitsma, W. 2000. Campylobacter in the food supply. Chpt. 24 in Campylobacter ,

2nd ed. I. Nachamkin and M. J. Blaser (Eds.) pgs. 467-481. American Society for

Microbiology, Washington, D. C.

Jay, J. M. 2000. Foodborne gastroenteritis caused by Salmonella and Shigella . In: Modern

Food Microbiology, 6 th ed. Aspen Publishers, Inc. Gaithersburg, MD. p. 511-530.

Katsuyama, A. M., Jantschke, M. and Gombas, D. E. 1999. Chemical hazards and

controls. Chpt. 6 In: Establishing hazard analysis and critical control point

programs: a workshop manual , 3 rd ed. K. E. Stevenson and D. T. Bernard, (Eds.) p.

53 - 61. The Food Processors Institute, Washington, D. C.

Katsuyama, A. M. and Jantschke, M. 1999. Physical hazards and controls. Chpt. 7 In:

Establishing hazard analysis and critical control point programs: a workshop

manual , 3 rd ed. K. E. Stevenson and D. T. Bernard, (Eds.) p. 63 - 66. The Food

Processors Institute, Washington, D. C.

Katsuyama, A. M. and Humm, B. J. 1995. The regulations of HACCP to CGMPs and

Sanitation. Chpt 13. In: Establishing hazard analysis and critical control point

programs: a workshop manual , 2 nd ed. K. E. Stevenson and D. T. Bernard (Eds.), p.

13-1 to 13-7. The Food Processors Institute, Washington, D. C.

Katsuyama A. M. and Stevenson, K. E. 1995. Hazard analysis and identification of

critical control points. Chpt 8. In: Establishing hazard analysis and critical

control point programs: a workshop manual , 2 nd ed. K. E. Stevenson and D. T.

Bernard (Eds.), p. 8-1 to 8-14. The Food Processors Institute, Washington, D. C.

Kemp, G. K., Aldrich, M. L., and Waldroup, A. L. 2000. Acidified sodium chlorite

antimicrobial treatment of broiler carcasses. J. Food Protect. 63:1087-1092.

Kim, J., Yousef, A. E., and Dave, S. 1999. Application of ozone for enhancing the

microbiological safety and quality of foods: a review. J. Food Protect. 62:1071-

1087.

King, D. A., Lucia, L. M., Castillo, A., Acuff, G. R., Harris, K. B., and Savell, J. W.

2005. Evaluation of peroxyacetic acid as a post-chilling intervention for control of

Escherichia coli O157:H7 and Salmonella Typhimurium on beef carcass surfaces.

Meat Sci. 69:401-407.

Kochevar, S. L., J. N. Sofos, R. R. Bolin, J. O. Reagan, and G. C. Smith. 1997. Steam

vacuuming as a pre-evisceration intervention to decontaminate beef carcasses. J.

Food Protect. 60:107-113.

Kudva, I. T., Blanch, K., and Hovde, C. J. 1998. Analysis of Escherichia coli O157:H7

survival in ovine or bovine manure and manure slurry. Appl. Env. Microbiol.

64:3166-3174.

Lammerding, A. M., Garcia, M. M., Mann, E. D., Robinson, Y., Dorward, W. J., Truscott, R.

B., and Tittiger, F. 1988. Prevalence of Salmonella and thermophilic Campylobacter

in fresh pork, beef, veal and poultry in Canada. J. Food Prot. 51:47-52.

Leistner, L. 1992. Food preservation by combined methods. Food Res. Internat. 25:151-

158.

Lim, K. and Mustapha, A. 2004. Effects of cetylpyridinium chloride, acidified sodium

chlorite, and potassium sorbate on populations of Escherichia coli O157:H7,

Listeria monocytogenes , and Staphylococcus aureus on fresh beef. J. Food

Protect. 67:310-315.

Manley, T. C. and Niegowski, S. J. 1967. Ozone. In Encyclopedia of chemical technology

(Vol. 14, 2nd ed., pgs. 410-432). Wiley, New York.

Marshall, R. T., Anderson, M. E., Naumann, H. D., and Stringer, W. C. 1977.

Experiments in sanitizing beef with sodium hypochlorite. J. Food Protect. 4:246-

249.

Martin, D. R., Uhler, P. M., Okrend, A. J. G., Chiu, J. Y. 1994. Testing of bob calf fecal

swabs for the presence of Escherichia coli O157:H7. J. Food Protect. 57:70-72.

McDonough, S. P., Stull, C. L., and Osburn, B. I. 1994. Enteric pathogens in intensively

reared veal calves. Am. J. Vet. Res. 55:1516-1520.

McEvoy, J. M., Doherty, A. M., Sheridan, J. J., Blair, I. S., and McDowell, D. A., 2003.

The prevalence of Salmonella spp. in bovine faecal, rumen, and carcass samples at a

commercial abattoir. J. Appl. Microbiol. 94:693-700.

Nachamkin, I., Allos, B. M., and Ho, T. W. 2000. Campylobacter jejuni infection and the

association with Guillain-Barré syndrome. Chpt. 8 in Campylobacter , 2 nd ed. I.

Nachamkin and M. J. Blaser (Eds.). pgs. 155-175. American Society for

Microbiology, Washington D. C.

NACMCF. 1998. Hazard analysis and critical control point principles and application

guidelines. National Advisory Committee on Microbiological Criteria for Foods. J.

Food Protect. 61:762-775.

Naylor, S. W., Low, J. C., Besser, T. E., Mahajan, A., Gunn, G. J., Pearce, M. C.,

McKendrick, I. J., Smith, D. G. E., and Gally, D. L. 2003. Lymphoid follicle-

dense mucosa at the terminal rectum is the principal site of colonization of

enterohemorrhagic Escherichia coli O157:H7 in the bovine host. Infect. Immun.

71:1505-1512.

Nazer, A. I., Kobilinsky, A., Tholozan, J. L., and Dubois-Brissonet, F. 2005.

Combinations of food antimicrobials at low levels to inhibit the growth of Salmonella

sv. Typhimurium: a synergistic effect? Food Microbiol. 22:391-398.

Notermans, S. 1979. Attachment of bacteria to meat surfaces. Antonie van Leeuwenhoek,

45:324-325.

Nou, X., Rivera-Betancourt, M., Bosilevac, J. M., Wheeler, T. L., Shackelford, S. D.,

Gwartney, B. L., Reagan, J. O., and Koohmaraie, M. 2003. Effect of chemical

dehairing on the prevalence of Escherichia coli O157:H7 and the levels of aerobic

bacteria and Enterobacteriaceae on carcasses in a commercial beef processing

plant. J. Food Protect. 66:2005-2009.

Noyes, R. 1969. Processing. Chpt. 1 in Citric acid production processes. P. 1-69. Chemical

Process Review No. 37. Noyes Development Corporation, Park Ridge, NJ.

Nutsch, A. L., Phebus, R. K., Riemann, M. J., Schafer, D. E., Boyer, Jr., J. E., Wilson, R. C.,

Leising, J. D., and Kastner, C. L. 1997. Evaluation of a steam pasteurization process

in a commercial beef processing facility. J. Food Protect. 60:485-492.

Ono, K. and Yamamoto, K. 1999. Contamination of meat with Campylobacter jejuni in

Saitama, Japan. Int. J. Food Microbiol. 47:211-219.

Ouattara, B., Simard, R. E., Holley, R. A., Piette, G. J.-P., and Bégin, A. 1997.

Inhibitory effect of organic acids upon meat spoilage bacteria. J. Food Protect.

60:246-253.

Padhye, N. V. and Doyle, M. P. 1992. Escherichia coli O157:H7: epidemiology,

pathogenesis, and methods for detection in food. J. Food Protect. 55:555-565.

Partin, L. R. and Heise, W. H. 1993. Bioderived acetic acid. Chpt. 1 in: Acetic acid and its

derivatives . p. 3-13. V. H. Agreda and J. R. Zoeller (Eds.) Series: Chemical

industries; v. 4 Marcel Dekker Inc., New York.

Pfizer, Inc. 2005. 1919 – Mass production of lactic acid.

http://www.pfizer.com/pfizer/history/1919.jsp accessed October 25, 2005

Phebus, R. K., Nutsch, A. L., Schafer, D. E., Wilson, R. C., Riemann, M. J., Leising, J. D.,

Kastner, C. L., Wolf, J. R., and Prasai, R. K. 1997. Comparison of steam

pasteurization and other methods for reduction of pathogens on freshly slaughtered

beef surfaces. J. Food Protect. 60:476-484.

Pordesimo, L. O., Wilkerson, E. G., Womac, A. R., and Cutter, C. N. 2002. Process

engineering variables in the spray washing of meat and produce. J. Food Protect.

65:222-237.

Prasai, R. K., Phebus, R. K., Garcia Zepeda, C. M., Kastner, C. L., Boyle, A. E., and

Fung, D. Y. C., 1995. Effectiveness of trimming and/or washing on

microbiological quality of beef carcasses. J. Food Protect. 58:1114-1117.

PURAC. 2000. Education Kit (Food Acids). Published February 2000 accessed April 1,

2005. http://www.purac.com 27 p. PURAC, Lincolnshire, IL.

Ransom, J. R., Belk, K. E., Sofos, J. N., Stopforth, J. D., Scanga, J. A., Smith, G. C.,

2003. Comparison of intervention technologies for reducing Escherichia coli

O157:H7 on beef cuts and trimmings. Food Protect. Trends. 23:24-34.

Reagan, J. O., G. R. Acuff, D. R. Buege, M. J. Buyck, J. S. Dickson, C. L. Kastner, J. L.

Marsden, J. B. Morgan, R. Nickelson, II, G. C. Smith, and J. N. Sofos. 1996.

Trimming and washing of beef carcasses as a method of improving the

microbiological quality of meat. J. Food Protect. 59: 751-756.

Reid, C. A., Small, A., Avery, S. M., and Buncic, A. S. 2002. Presence of food-borne

pathogens on cattle hides. Food Control 13:411-415.

Richardson, S. D., Thruston, Jr., A. D., Caughran, T. V., Collette, T. W., Patterson, K. S.,

and Lykins, Jr., B. W. 1998. Chemical by-products of chlorine and alternative

disinfectants. Food Technol. 52:58-61.

Riley, L. W., Remis, R. S., Helgerson, S. D., McGee, H. B., Wells, J. G., Davis, B. R.,

Hebert, R. J., Olcott, E. S., Johnson, L. M., Hargrett, N. T., Blake, P. A., and

Cohen, M. L. 1983. Hemorrhagic colitis associated with a rare Escherichia coli

serotype. New Eng. J. Med. 308:681-685.

Robinson, D. A. 1981. Infective dose of Campylobacter jejuni in milk. Brit. Med. J.

282:1584.

Sarwari, A. R., Magder, L. S., Levine, P., McNamara, A., Knower, S., Armstrong, G. L.,

Etzel, R., Hollingsworth, J., and Morris, Jr., J. G. 2001. Serotype distribution of

Salmonella isolates from food animals after slaughter differs from that of isolates

found in humans. J. Infect. Dis. 183:1295-1299.

Scatina, J., Abdel-Rahman, M. S., and Goldman, E. 1985. The inhibitory effect of

Alcide ®, an antimicrobial drug, on protein synthesis in Escherichia coli . J. Appl.

Toxicol. 5:388-394.

Schnell, T. D., Sofos, J. H., Littlefield, V. G., Morgan, J. B., Gorman, B. M., Clayton, R. P.,

and Smith, G. C. 1995. Effects of postexsanguination dehairing on the microbial load

and visual cleanliness of beef carcasses. J. Food Protect. 58:1297-1302.

Shallow, S., M. Samuel, A. McNees, G. Rothrock, D. Vugia, T. Fiorentino, R. Marcus, S.

Hurd, P. Mshar, Q. Phan, M. Cartter, J. Hadler, M. Farley, W. Baughman, S. Segler,

S. Lance-Parker, W. MacKenzie, K. McCombs, P. Blake, J. G. Morris, M. Hawkins,

J. Roche, K. Smith, J. Besser, E. Swanson, S. Stenzel, C. Medus, K. Moore, S.

Zansky, J. Hibbs, D. Morse, P. Smith, M. Cassidy, T. McGivern, B. Shiferaw, P.

Cieslak, M. Kohn, T. Jones, A. Craig, and W. Moore. 2001. Preliminary FoodNet

Data on the Incidence of Food borne Illnesses - Selected Sites, United States, 2000.

Morb. Mortal. Week. 50:241-246.

Sheldon, B. W. and Brown, A. L. 1986. Efficacy of ozone as a disinfectant for poultry

carcasses and chill water. J. Food Sci. 51:305-309.

Sheridan, J. J. 1998. Sources of contamination during slaughter and measures for control. J.

Food Safety 18:321-339.

Sierra, M., Gonzalez-Fandos, E., García-López, M., Garcia Fernandez, M. C., and Prieto, M.

1995. Prevalence of Salmonella , Yersinia , Aeromonas , Campylobacter , and

cold-growing Escherichia coli on freshly dresses lamb carcasses. J. Food Protect.

58:1183-1185.

Siragusa, G. R. 1995. The effectiveness of carcass decontamination systems for controlling

the presence of pathogens on the surfaces of meat animal carcasses. J. Food Safety

15:229-238.

Skirrow M. B. and Blaser, M. J. 2000. Clinical aspects of Campylobacter infection. Chpt.

4 in Campylobacter , 2 nd ed. I. Nachamkin and M. J. Blaser (Eds.), p. 69-88.

American Society for Microbiology, Washington, D. C.

Smith, J. L. and Palumbo, S. A. 1982. Microbial injury reviewed for the sanitarian. Dairy

Food Sanit. 2:57-63.

Sofos, J. N. and Smith, G. C. 1998. Nonacid meat decontamination technologies: model

studies and commercial applications. Int. J. Food Microbiol. 44:171-188.

Stanley, K. N., Wallace, J. S., Currie, J. E., Diggle, P. J., and Jones, K. 1998. Seasonal

variation of thermophilic campylobacters in lambs at slaughter. J. Appl. Microbiol.

84:1111-1116.

Stern, N. J., Line, J. E., and Chen, H. 1992. Campylobacter. Chpt. 31 in Compendium of

methods for the microbiological examination of foods, 3rd ed. C. Vanderzant and D.

F. Splittstoesser (Eds.), p. 301-310. American Public Health Association,

Washington, D. C.

Stevenson, K. E., Merkel, R. A., and Lee, H. C. 1978. Effects of chilling rate, carcass

fatness and chlorine spray on microbiological quality and case-life of beef. J.

Food Sci. 43:849-852.

Su, C. and Brandt, L. J. 1995. Escherichia coli O157:H7 infection in humans. Ann.

Intern. Med. 123:698-714.

Synge, B. and Paiba, G. 2000. Verocytotoxin-producing E. coli O157. Vet. Rec. 147:27.

Tsai, L., Schade, J. E., and Molyneux, B. T. 1992. Chlorination of poultry chiller

water: chlorine demand and disinfection efficiency. Poultry Sci. 71:188-196.

Tutenel, A. V., Pierard, D., Van Hoof, J., and De Zutter, L. Molecular characterization of

Escherichia coli O157 contamination routes in a cattle slaughterhouse. J. Food

Protect. 66:1564-1569.

Uhitil, S., Jakšić, S., Petrak, T., and Botka-Petrak, K. 2001. Presence of Escherichia coli

O157:H7 in ground beef and ground baby beef meat. J. Food Protect. 64:862-864.

USDA-ERS, 2004. Economics of foodborne disease: feature.

http://www.ers.usda.gov/briefing/FoodborneDisease/features.htm, Economic

Research Service, United States Department of Agriculture, Washington, D. C.,

accessed October 10, 2005.

USDA-FSIS, 1996a. Use of knife trimming and vacuuming of beef carcasses with hot

water or steam; other beef carcass intervention systems. FSIS Directive 6350.1

USDA-FSIS, 1996b. Notice of policy change; achieving the zero tolerance performance

standard for beef carcasses by knife trimming and vacuuming with hot water or

steam; use of acceptable carcass interventions for reducing carcass contamination

without prior agency approval. Federal Register 61, no. 66 (4 April 1996).

USDA-FSIS, 1996c. Pathogen reduction act; hazard analysis and critical control point

(HACCP) systems; final rule. Federal Register 61, no. 144 (25 July 1996).

USDA-FSIS, 1999a. USDA issues final rule on meat and poultry irradiation. Food Safety

and Inspection Service. http://www.fsis.usda.gov/OA/background/irrad_final.htm,

accessed August 17, 2005.

USDA-FSIS, 1999b. Irradiation of meat food products; final rule. Federal Register 64, no.

246 (23 December 1999).

USDA-FSIS, 2001. 2001 National Residue Program Data- Red Book.

http://www.fsis.usda.gov/Science/2001_National_Residue_Data_Red_

Book/index.asp, accessed August 17, 2005.

USDA-FSIS, 2002. Instructions for verifying E. coli O157:H7 reassessment. Notice 44-02, (4

November 2002).

USDA-FSIS, 2004. Prohibition of the use of specified risk materials for human food and

requirements for the disposition of non-ambulatory disabled cattle; meat produced

by advanced meat/bone separation machinery and meat recovery (AMR) systems;

prohibition of the use of certain stunning devices used to immobilize cattle during

slaughter; bovine spongiform encephalopathy surveillance program; interim final

rules and notice. Federal Register 69, no. 7 (12 January 2004).

USDA-FSIS, 2005. Safe and suitable ingredients used in the production of meat and

poultry products. FSIS Directive 7120.1, Amendment 4.

USDA-FSIS, 2006. Salmonella verification sample result reporting: agency policy and use in

public health protection. Notice, Federal Register 71, No. 38 (27 February 27, 2006)

USMEF. 2005a. Backgrounder Hormone Use. United States Meat Export Federation.

http://www.usmef.org/TradeLibrary/assets/12233/horbackgrounderfile/Backgroun

der%20-%20Hormones.pdf, United States Meat Export Federation, Washington,

D. C. accessed September 15, 2005.

USMEF. 2005b. Backgrounder Antibiotics.

http://www.usmef.org/TradeLibrary/assets/12233/antfactsheetfile/Factsheet%20-

%20Antibiotics.pdf, accessed September 15, 2005.

Van Donkersgoed, J., Graham, T., and Gannon, V. 1999. The prevalence of verotoxin,

Escherichia coli O157:H7, and Salmonella in the feces and rumen of cattle at

processing. Can. Vet. J. 40:332-338. van Netten, P., Huis in ‘t Veld, J. H. J., and Mossel, D. A. A. 1994. The immediate

bactericidal effect of lactic acid on meat-borne pathogens. J. Appl. Bacteriol.

77:490-496. van Netten, P., Mossel, D. A. A., and Huis in ‘t Veld, J. H. J. 1995. Lactic acid

decontamination of fresh pork carcasses: a pilot plant study. Int. J. Food Microbiol.

25:1-9.

Villareal, M. E., Baker, R. C., and Regenstein, J. M. 1990. The incidence of Salmonella on

poultry carcasses following the use of slow release chlorine dioxide (Alcide). J. Food

Protect. 53:465-467.

Wang, G., Zhao, T., and Doyle, M. P. 1996. Fate of enterohemorhagic Escherichia coli

O157:H7 in bovine feces. Appl. Environ. Microbiol. 62:2567-2570.

Weber, G. H., O’Brien, J. K., and Bender, F. G. 2004. Control of Escherichia coli

O157:H7 with sodium metasilicate. J. Food Protect. 67:1501-1506.

Wells, J. G., Shipman, L. D., Greene, K. D., Sowers, E. G., Green, J. H., Cameron, D. N.,

Downes, F. P., Martin, M. L., Griffin, P. M., Ostroff, S. M., Potter, M. E., Tauxe, R.

V., and Wachsmuth, I. K. 1991. Isolation of Escherichia coli serotype O157:H7 and

other shiga-like-toxin-producing E. coli from dairy cattle. J. Clin. Microbiol. 29:985-

989.

Wieler, L. H., Bauerfeind, R., and Baljer, G. 1992. Characterization of shiga-like toxin

producing Escherichia coli (SLTEC) isolated from calves with and without

diarrhoea. Zbl. Bakt. 276:243-253.

Wray, C. and Davies, R. H. 2000. Salmonella infections in cattle. Chpt. 10 in: Salmonella

in domestic animals. P. 169-190. C. Wray and A. Wray (Eds.). CAB

International , New York.

Yang, P. P. W. and Chen, T. C. 1978. Stability of ozone and its germicidal properties on

poultry meat microorganisms in liquid phase. J. Food Sci. 44:501-504.

Zhao, C., Ge, B., de Villena, J., Sudler, R., Yeh, E., Zhao, S., White, D. G., Wagner, D., and

Meng, J. 2001. Prevalence of Campylobacter spp., Escherichia coli , and Salmonella

serovars in retail chicken, turkey, pork, and beef from the greater Washington, D. C.,

area. Appl. Environ. Microbiol. 67:5431-5436.

CHAPTER THREE

PRELIMINARY SURVEY

INTRODUCTION

Primary processing, or slaughter, of red meat animals in very small plants (10 or

fewer employees, average annual revenue < $2.5 million) is achieved through a universal

series of steps with minor differences in processing equipment and techniques, as well as the

application of antimicrobial interventions. Regardless of the methods used to decontaminate

carcasses prior to chilling, carcasses must be free of visible fecal contamination and steps

must be taken to reduce populations of potentially harmful bacteria on carcass surfaces

(USDA-FSIS, 1996a). When used to reduce a biological hazard on carcass surfaces, critical

control points must be incorporated in plant Hazard Analysis Critical Control Point

(HACCP) plans and their antimicrobial effectiveness supported by valid, scientific

documentation (USDA-FSIS, 1996b).

There is ample scientific documentation to support the various antimicrobial

interventions used in meat plants. Water washing alone is capable of reducing aerobic plate

counts (APC) consistently by 1 log CFU/cm 2 as the kinetic energy of the water physically detaches bacteria from carcass surfaces (Siragusa, 1995; Sheridan, 1998). Meanwhile, hot water washing typically yields a 3 log reduction in bacterial populations (Siragusa, 1995).

Siragusa (1995) also generalizes that the rinsing of meat surfaces with food grade organic acids elicits a 1 to 2 log reduction in bacterial populations. Steam vacuuming is commonplace in large meat plants, although some smaller plants have also adopted this method. Essentially, steam vacuuming can be used to remove visible debris and fecal contamination and is at least as efficacious as knife trimming (Kochevar et al., 1997). One of the advantages over knife trimming is that steam vacuumed carcass surfaces remain intact

resulting in less trimmed waste (Sheridan, 1998). Moreover, plants with sufficient capital and

workspace can install washing cabinets for the controlled application of water washes and

antimicrobial rinses at different temperatures and pressures. Steam pasteurization also can provide effective elimination of potentially harmful bacteria with a rapid, repeatable pressurized steam treatment. Carcasses that undergo programmed steam pasteurization

treatment can be cleared of coliforms and generic E. coli by approximately 2 log and APC by

>1 log CFU/cm 2 in a commercial setting (Nutsch et al., 1997; Sofos and Smith, 1998).

Even so, some of these interventions (spray washing in cabinets or steam pasteurization chambers) may be too costly or physically impractical to implement in very small plants. Before scientific information can be generated to validate the HACCP slaughter plans of very small meat plants, it is necessary to obtain an accurate description of existing carcass decontamination interventions and assess the training methods for new or improved interventions. Because these data do not presently exist for very small meat plants, a survey tool was designed to investigate these needs. Analysis of survey results should provide guidance for the research and development of effective carcass decontamination techniques that can be easily implemented in very small meat establishments.

METHODS

Very small meat and poultry plants (n = 298) under federal inspection in

Pennsylvania, Texas, and Washington were mailed a written survey, which was developed in

cooperation with the Penn State Office for Outreach Marketing, Research, and Planning with

Dr. Naana Nti as the researcher. The questionnaire was used to determine the types of rinses

and washes used on carcass surfaces immediately prior to chilling (see Appendix A).

Questions in the first half of the survey were directed at the identification of antimicrobials

used [hot water, organic acids, and other compounds generally recognized as safe (GRAS)]

and how these antimicrobials were applied to carcass surfaces (equipment, pressure,

application time, etc.). The second half of the survey asked the processors for their preference of educational materials (CD-ROM, video, website, etc.) and their preferred

method of receiving technical training, such as in-plant demonstrations, for the

implementation of carcass antimicrobial interventions.

Prior to participation in the survey, plants were informed that the names of plants or plant personnel would not be reported to any regulatory official or used in any publications related to the project. Furthermore, the maintenance of plant confidentiality was of the utmost importance. Although plant employees were assured of confidentiality, most of them did not participate in the survey.

After receiving a small number of paper surveys in the mail, a second effort was made to collect survey information via telephone or in-plant visit. Before the telephone survey, very small establishments in Pennsylvania were pre-screened according to the name of each company. Based on the names of some companies (nonspecifically, XYZ Wholesale

and Purveying or Family Custom Butchering), it was apparent that these plants were not

slaughter facilities or were not operating under federal inspection. Thus, these plants were

not contacted by telephone. During telephone surveys, processors also were asked if they processed poultry, meat or both as this question was not part of the written survey. Survey

responses collected by mail or telephone from very small meat and poultry plants in

Pennsylvania were summarized by the Penn State Office for Outreach Marketing, Research

and Planning. Specifically, the survey responses were tabulated for each question and were

expressed as a percentage of responses or they were summarized descriptively, as

appropriate. Additionally, survey data were collected by making personal visits to very small plants following a very low response rate in Washington and Idaho.

RESULTS

Survey information was collected from a total of 49 very small establishments in

Pennsylvania for a response rate of 16% (Nti, 2002). Fifty-five responses among all very

small establishments, whether involved in red meat slaughter under federal inspection or not,

were needed to establish a 90% confidence level (10% margin of error; Nti, 2002). Only the

surveys from Pennsylvania meat plants were available for analysis by the Office for Outreach

Marketing, Research and Planning. This response rate accounted for the surveys sent to 298

very small establishments in Pennsylvania whether or not these plants operate under federal

inspection or perform slaughter. Four of these plants process poultry only and the other 45 plants process red meat only (market beef, dairy beef, suckling pigs, market hogs, sows,

market lambs, bob veal, and/or special-fed veal). Some Pennsylvania respondents elected not

to complete the entire survey.

No surveys were returned from plants in Texas. The reasons for survey non-response

in Texas are unknown. In Washington and Idaho, a very small number of surveys were

returned by mail. To collect more survey data, a research technician from Washington State

University made personal visits to very small plants in Washington and Idaho to collect

responses from 24 processors. The surveys collected from Washington and Idaho contained

responses only to questions 1 through 7 and 11 (see Appendix A). These data were used to

identify the antimicrobial compounds used during slaughter in Washington and Idaho and

describe the method of their application. These additional survey responses are reported

alongside the results from Pennsylvania plants.

In postscript, there are only 78 very small establishments in Pennsylvania that

slaughter red meat under federal inspection (J. Behney; personal communication on May 17,

2006; FSIS Office of Field Operations, District 60, Philadelphia, PA). Furthermore, only 24 plants in both Idaho and Washington fit these same criteria (R. Jones; personal

communication on May 17, 2006; FSIS Office of Field Operations, District 15, Denver, CO).

These new figures indicate that the survey response rates for Pennsylvania (63%) and the

Northwest (100%) are much more favorable than previously thought.

At the time of this survey, none of the very small meat and poultry plants was using

steam pasteurization to reduce microbial contaminants on carcass surfaces (Table 1). More

than half of the plants utilized knife trimming (58%) and cold water washing (56%). Warm

water (23%) and hot water washes (17%) were used in fewer plants in Pennsylvania. In

Washington and Idaho, cold water and warm water washes were more common than hot

water washes and only 6 plants (18%) implemented knife trimming (Table 21). Only seven plants in Pennsylvania reported using antimicrobial compounds and one plant was using a

steam vacuum. Although this question was multiple choice, several plants volunteered

additional information. Of the organic acids used to wash carcasses, lactic acid and citric acid

were each used by at least one plant. All four poultry plants regularly used sodium

hypochlorite in chill tanks, for manual water washing, and for washing carcasses in

evisceration cabinets. One survey respondent commented on the use of spray chilling as an

intervention, which was applied after slaughter. Only one plant in the Northwest reported

using an organic acid rinse while two plants use sodium hypochlorite (Table 21).

Very small plants in Pennsylvania that were applying antimicrobial compounds to

carcasses were also asked to report the concentration of these compounds (Q-8., Appendix

A). The information collected from this question was strictly qualitative, not multiple choice.

The majority of the plants that provided a description of antimicrobial usage were the four poultry plants. One poultry plant mentioned that approximately 50 ppm sodium hypochlorite

was added to the processing floor water supply to achieve approximately 25 to 50 ppm free

chlorine in the final wash water. When organic acids were used, citric acid was applied at

approximately 2.5% and lactic acid at 2%. Another plant used 60% sodium lactate. In

Northwestern plants, 2% acetic acid and sodium hypochlorite (0.39 and 0.11% solutions)

were the only antimicrobial compounds reported to be in use (Table 21).

Answers to the second question comprise a broad range of application times for

slaughter interventions. In general, the majority of plants in Pennsylvania and the Northwest

(44 and 52%, respectively) were administering some type of intervention to carcass surfaces

for longer than 30 seconds (Tables 2 and 22). Specifically, one plant in Pennsylvania

reported spending 15 minutes on each carcass to perform both knife trimming and water

washing while another plant washed each carcass with cold water for three minutes followed by a three-minute hot water wash. Many of the other plants in Pennsylvania that applied

interventions for more than 30 seconds reported intervals of one, two, or five minutes per

carcass or that the application time varied depending on the amount of visible contamination

on carcasses. In contrast, 35% of plants in Pennsylvania applied some interventions (organic

acids or chlorine) for five to ten seconds.

Most very small plants in Pennsylvania use a hand-held hose (87%) to apply water

washes (Table 3). Pennsylvania plants that apply chemical rinses often use a hand-held spray

apparatus (15%). None of the plants surveyed used a continuous or automatic spray cabinet

for washing carcasses. However, one poultry plant reported washing carcasses in an

evisceration cabinet with a hand-held hose and water containing 20 ppm chlorine. Another plant specified the use of a water pump, not to be confused with a pressure washer, which

was connected to a hand-held hose in order to deliver water to carcass surfaces. Also, several

Pennsylvania plants responded that knives were used to carry out a knife trimming

intervention. In Washington and Idaho, most (73%) plants also used a hand-held hose for

water washing while 23% used hand-held sprayers to apply water washing or antimicrobial

rinses (Table 23). One other plant reported that employee hands were used to scrub carcass

surfaces during washing (Table 23).

Slaughter interventions were usually applied prior to chilling (96%) in Pennsylvania plants (Table 4), whereas Northwestern plants were equally as likely to intervene after

evisceration (33%) or before chilling (33%; Table 24). Nevertheless, interventions were

applied in some plants at processing steps other than the final wash. Some plants in

Pennsylvania replied that washing and/or trimming took place at multiple steps during

slaughter, such as before (15%) and after evisceration (9%) and after debunging (2%). Small percentages of Northwestern plants also applied interventions during other processing stages: before evisceration (12%), after debunging (18%), or at the commencement of chilling (3%;

Table 24).

In Tables 5 and 25, most or all survey participants in Pennsylvania or the Northwest

replied that a washing cabinet was not used (88 and 100%, respectively); hence, there was no

need to determine whether these plants used fixed or rotating nozzles. However, there was

one positive response each for use of a fixed nozzle and a rotating nozzle. It is unclear how

these nozzles were being used in Pennsylvania.

The survey participants also reported a range of application pressures used to deliver

hot water or organic acid to carcass surfaces (Table 6). As conveyed previously, all

respondents to this question would likely be using a hand-held hose or sprayer instead of a

washing cabinet to apply hot water or organic acid solutions. Most (59%) of the

Pennsylvania respondents did not know the washing or spraying system pressure in their plants. Almost 10% of Pennsylvania plants apply hot water or organic acid in the range of 51 to 100 psi, whereas 14% apply rinses at 25 to 50 psi. Conversely, plants in the Northwest were more familiar with application pressures. While only 6% did not know application pressures, 22% were applying treatments at 25 to 50 psi, an additional 22% employed 51 to

100 psi, and 6% applied washes or rinses at 151 to 200 psi (Table 26).

Application distances between carcass surfaces and the water hose or garden sprayer nozzle were estimated (Table 7). Most processors in Pennsylvania (52%) hold the hose or sprayer nozzle more than 12 inches away from the carcass surface when applying water or chemical rinses. In comparison, other processors attempt to maintain an application distance of 4 to 6 inches (17%), 7 to 9 inches (7%), or 10 to 12 inches (14%). The application distances employed by very small processors in Pennsylvania who do not use hot water or organic acids (10%) on carcass surfaces are not accounted for in this survey. In contrast, most Northwestern plants were not in the habit of applying organic acids or hot water (54%;

Table 27). Those that applied these types of treatments most often (29%) employed an application distance of 10 to 12 inches (Table 27).

Although some processors (9%, n = 23) applied more than one antimicrobial compound or other intervention to carcass surfaces, most (91%, n = 23) did not (Table 8).

One poultry processor added sulfuric acid to chill tanks containing 30 ppm chlorine to

maintain bath pH between 6 and 6.5 (Q-10, Appendix A). Moreover, one meat plant steam

vacuums carcasses before a cold water wash, which is followed immediately by a citric acid

rinse (Q-10, Appendix A). None of the Northwestern meat plants applied more than one

antimicrobial compound to carcass surfaces.

Twenty processors in Pennsylvania reported monitoring antimicrobial interventions by measuring temperature and/or concentration (Table 9). Use of a digital or dial

thermometer to monitor temperature was the most commonly (50%) used method in very

small meat plants. Chlorine test kits, pH tape, and pH meters were used by poultry plants.

The concentration of sodium hypochlorite solutions was monitored with a chlorine test kit

(HACH Company, Loveland, CO) in two poultry plants. One of these two plants used a pH

meter (Hanna Instruments, Woonsocket, RI). A third poultry plant outsourced the monitoring

of chlorine concentration to a contract sanitation company, which used pH tape and a

chlorine test kit. Other monitoring methods included measuring the mass or volume of

antimicrobial compounds before adding them to a known quantity of water. In the Northwest,

only one processor (6%) reported that temperature was monitored (Table 28). None of the

other Northwestern processors used a monitoring method.

Very small processors also were asked if they would be willing to participate in a

microbial sampling study of carcass surfaces in their plants. Processors were told that

carcasses would be sampled after all antimicrobial treatments were performed. Over half

(59%) of the processors were interested (Table 10). Moreover, almost 90% of processors

were interested in receiving training about new slaughter interventions (Table 11).

As soon as scientific information could be generated to validate antimicrobial

treatments for carcasses during slaughter, very small plants would receive training for the

new methods. To facilitate planning, processors were asked about their preferences for instructional methods, access to the Internet and CD-ROM, and when and where to hold training sessions. Most processors (81%) indicated a strong interest in receiving paper-based training materials (brochures, booklets, etc.; Table 12). Others strongly favored instructional methods like videotapes (64%), web-based/Internet training, and on-site training sessions at the workplace (46%). On the other hand, 33% of processors were not interested in the audiotape format (Table 12). The majority of survey participants, who answered questions regarding computer use, said that Internet access (Table 13) was available at home (79%) and at work (55%) and that they were more likely to have a CD-ROM player (Table 14) at home

(86%) than at work (50%). Also, the most suitable times of the year for very small plant employees to attend training seminars are spring (37%) and summer (40%; Table 15). While there did not appear to be any outstanding days or times of day to schedule a group training session, Mondays (29%) and Tuesdays (35%) were generally the most fitting (Table 16).

Near the end of the survey, four demographic questions were asked. Most processors had completed high school (60%) while 25% graduated from college (Table 17).

Additionally, 98% of the respondents work in a supervisory capacity as plant managers, owners, presidents, or foremen (Table 18). Almost 60% of respondents had 20 to 39 years of experience in the meat and poultry industry (Table 19). Lastly, 64% of plants employed fewer than 10 people, whereas two of the plants employed 100 or more people.

A few processors provided additional comments at the end of the survey and are summarized as follows (Q-20, Appendix A). There was interest in learning about low-cost interventions for very small plants. Several processors expressed concern about maintaining the confidentiality of plant and employee identities, if selected to participate in the microbial

sampling study. Also, many of the processors belong to the American Association of Meat

Processors (AAMP) and it was suggested that the results of this project be shared with members during an AAMP convention.

DISCUSSION

This preliminary survey was conducted in spring and summer of 2002. The Pathogen

Reduction Act required all very small establishments to have implemented HACCP by

January 25, 2000, whereas large and small plants had to begin operating under HACCP in

1998 and 1999, respectively (USDA-FSIS, 1996b). Hence, the stressful memory of

implementing HACCP for the first time may have remained vivid in the minds of these processors.

The survey also generated a list of very small meat plants that may be willing to allow

university personnel to sample carcasses for the presence of pathogens during subsequent

stages of this research project. When asked to participate in a telephone survey, some plants

expressed strong concern regarding the potential isolation of E. coli O157:H7 from carcasses

and were worried about the consequences of taking a remedial action, such as a product

recall. Others were simply not interested in participating in a research study that was

associated with HACCP. Furthermore, the completion of a survey requires time and effort,

two resources than are perceived to cost money to a business, which may explain why some processors were reticent on the telephone. These reactions are likely to be some of the

reasons behind survey non-response.

In survey research, non-respondents are a significant source of error, which biases

survey results (Fowler, 1993). Nti (2002) reported that the survey may not accurately

describe the range of interventions used in and training needs of very small meat plants, because the responses were not sufficient to achieve a 90% confidence level. However, more

recent information confirms that the response rates in Pennsylvania and the Northwest

accounted for 63% and 100%, respectively, of very small plants Hence, the survey responses

do adequately represent the types and characteristics of carcass decontamination treatments

in very small plants and the actual training needs of plant employees.

There is a predominance of techniques to remove carcass contamination in very small plants. Overall, plant employees are interested in learning more about carcass decontamination alternatives. This interest may be best satisfied by developing and distributing training materials (perhaps a video and accompanying brochure) rather than a organizing a group training session in a university facility.

From another angle, most of the participating establishments process red meat, not poultry. This is not surprising because the vast majority of poultry processing plants in the

U.S. employs 400 or more workers (Ollinger et al., 2000). In the vertically integrated poultry industry, hatcheries, growing facilities, and processing plants are often owned by large corporations that can take advantage of economies of scale. Since the 1960s, vertical integration in the poultry industry has led to the near disappearance of small and very small poultry processing plants (Ollinger et al., 2000). Given this finding, all subsequent research objectives are focused on the study of antimicrobial treatments in the context of very small plants that process red meat, rather than poultry.

The outcomes of this survey might also suggest that some very small meat processors have not been approaching carcass decontamination strategies in a scientific manner. Almost

60% of respondents in Pennsylvania did not know the operating pressure of the spraying or washing equipment used on the slaughter floor. Furthermore, poultry processors, who responded to the survey, were more likely to monitor solution concentrations than meat processors. Only 10 out of 45 very small meat processors (22%; Table 9) reported using a

thermometer to monitor the temperature of water washes in Pennsylvania while 1 out of 18

used a thermometer in the Northwest (6%; Table 28). According to Tables 1 and 21, 27 meat processors (56%) in Pennsylvania and 10 in the Northwest (29%) reported using a cold water

wash on carcasses. Because cold water is not considered a lethal antimicrobial treatment,

these processors may not have been inclined to measure cold water temperatures (Bolton et

al., 2002). According to Pordesimo et al. (2002), process engineering variables (i.e. pressure,

time spent washing, distance between the carcass and hose nozzle, temperature, antimicrobial

concentration, etc.) associated with meat decontamination strategies should be optimized to

achieve maximum effectiveness and efficiency.

It is clear that very small meat and poultry establishments have adopted a diverse

assortment of carcass decontamination techniques. When used specifically to reduce

microbial populations to safe levels, these techniques may be seamlessly incorporated into plant HACCP plans, provided that there is a scientific justification for doing so. According to

this study, almost all very small meat plants employ water washing during the final rinse step

and most of these interventions are applied with hand-held hoses or portable sprayers. In a

subsequent study, the antimicrobial effectiveness of water washing was studied under various

lengths of time at different pressures, application distances, and temperatures followed by

different drip, or dwell, times. Furthermore, lactic acid, acetic acid, and sodium hypochlorite

were already being used in very small meat plants at the time of this survey. Additional

antimicrobial compounds need to be identified as potential candidates for use in very small

meat plants.

In subsequent research, red meat carcasses were sampled soon after the completion of

this survey to establish a microbiological baseline for very small meat plants. Next, an

effective multi-step treatment was identified from water washing and antimicrobial rinsing experiments performed in a laboratory setting. After this multi-step carcass treatment was implemented in very small meat plants, a second microbiological baseline was established. A comparison of the two baselines provides validation of the antimicrobial effectiveness of the treatment. At the completion of this project, scientific documentation of a multi-step antimicrobial intervention will be available to very small meat plants for use in slaughter

HACCP plans.

REFERENCES

Bolton, D. J., Pearce, R. A., Sheridan, J. J., Blair, I. S., McDowell, D. A., and Harrington,

D. 2002. Washing and chilling as critical control points in pork slaughter hazard

analysis and critical control point (HACCP) systems. J. Appl. Microbiol. 92:893-

902.

Fowler, F. J. 1993. Survey research methods, 2 nd ed. Sage Publications, Inc., Newbury

Park, CA.

Kochevar, S. L., Sofos, J. N., Bolin, R. R., Reagan, J. O., and Smith, G. C. 1997. Steam

vacuuming as a pre-evisceration intervention to decontaminate beef carcasses. J.

Food Protect. 60:107-113.

Nti, N. 2002. Interventions for very small meat and poultry establishments: summary

report. Penn State Office for Outreach Marketing, Research and Planning, State

College, PA.

Nutsch, A. L., Phebus, R. K., Riemann, M. J., Schafer, D. E., Boyer, Jr., J. E., Wilson, R. C.,

Leising, J. D., and Kastner, C. L. 1997. Evaluation of a steam pasteurization process

in a commercial beef processing facility. J. Food Protect. 60:485-492.

Ollinger, M., MacDonald, J. and Madison, M. 2000. Technological change and economies of

scale in U.S. poultry slaughter. Discussion paper CES 00-05. Center for Economic

Studies, U.S. Bureau of the Census, Washington, D. C. 41 pp.

Pordesimo, L. O., Wilkerson, E. G., Womac, A. R., and Cutter, C. N. 2002. Process

engineering variables in the spray washing of meat and produce. J. Food Protect.

65:222-237.

Sheridan, J. J. 1998. Sources of contamination during slaughter and measures for control. J.

Food Safety 18:321-339.

Siragusa, G. R. 1995. The effectiveness of carcass decontamination systems for controlling

the presence of pathogens on the surfaces of meat animal carcasses. J. Food Safety

15:229-238.

Sofos, J. N. and Smith, G. C. 1998. Nonacid meat decontamination technologies: model

studies and commercial applications. Int. J. Food Microbiol. 44:171-188.

USDA-FSIS, 1996a. Notice of policy change; achieving the zero tolerance performance

standard for beef carcasses by knife trimming and vacuuming with hot water or

steam; use of acceptable carcass interventions for reducing carcass contamination

without prior agency approval. Federal Register 61, no. 66 (4 April 1996).

USDA-FSIS, 1996b. Pathogen reduction act; hazard analysis and critical control point

(HACCP) systems; final rule. Federal Register 61, no. 144 (25 July 1996).

Table 1. Slaughter interventions used to reduce microbial contaminants on carcass surfaces in very small meat establishments in Pennsylvania (n = 48) a

Interventions Number of responses Percent of responses b

Trimming 28 58.3

Cold water wash 27 56.3

Warm water wash 11 22.9

Hot water wash 8 16.7

Organic acid wash 4 8.3

Chlorine 3 6.3

Steam vacuum 1 2.1

Steam pasteurization 0 0.0

Other c 3 6.3

a n = number of survey participants who answered question 1 (Appendix A) b Percentages add up to more than 100% due to multiple responses. c Several plants volunteered additional information regarding interventions: spray chilling after slaughter, 150°F pressure wash as needed, lactic acid, or sodium hypochlorite in chill tanks and water washes.

All data from Pennsylvania plants were analyzed and tabulated by Dr. Naana Nti (2002). Table titles and footnotes have been modified from the original version.

Table 2. Approximate application time of slaughter intervention to carcass surfaces in very small meat establishments in Pennsylvania (n = 23) a

Length of time Number of responses Percent of responses b

5-10 seconds 8 34.8

10-20 seconds 5 21.7

20-30 seconds 3 13.0

More than 30 seconds 10 43.5

Other c 3 13.0

a n = number of survey participants who answered question 2 (Appendix A) b Percentages add up to more than 100% due to multiple responses. c Some respondents said knife trimming was performed throughout slaughter.

100

Table 3. Types of equipment used to apply slaughter interventions in very small meat establishments in Pennsylvania (n = 45) a

Application methods Number of responses Percent of responses b

Hand held hose 34 87.2

Hand held sprayer 6 15.4

Continuous spray cabinet 0 0.0

Automatic cabinet 0 0.0

Other c 5 12.8

a n = number of survey participants who answered question 3 (Appendix A) b Percentages add up to more than 100% due to multiple responses. c Other equipment that were mentioned include: knives for trimming, chill tank or tub with paddles, washing a carcass in a cabinet with 20 ppm chlorine, and use of a pressure pump (not pressure washer) for water washing.

101

Table 4. Timing of the application of slaughter interventions to carcass surfaces in very small meat establishments in Pennsylvania (n = 47) a

Timing of slaughter intervention Number of responses Percent of responses b

Before the chiller/hot box 45 95.7

Before evisceration 7 14.9

After evisceration 4 8.5

After debunging 1 2.1

Other 5 10.6

a n = number of survey participants who answered question 4 (Appendix A) b Percentages add up to more than 100% due to multiple responses.

102

Table 5. Type of nozzle used in washing cabinets to apply hot water or organic acids in very small meat establishments in Pennsylvania (n = 17) a

Nozzle type Number of responses Percent of responses

Fixed 1 5.9

Rotating 1 5.9

Do not use cabinet 15 88.2

a n = number of survey participants who answered question 5 (Appendix A)

103

Table 6. Application pressure of hand-held garden sprayer or spray cabinet used to apply hot water or organic acids to carcass surfaces in very small meat establishments in Pennsylvania (n = 44) a

Application pressure Number of responses Percent of responses

25-50 psi 6 13.6

51-100 psi 4 9.1

Do not know 26 59.1

Do not use 8 18.2

a n = number of survey participants who answered question 6 (Appendix A)

104

Table 7. Distance between hand-held hose or hand-held sprayer and carcass surfaces during application of hot water or organic acids in very small meat establishments in Pennsylvania (n = 42) a

Distance Number of responses Percent of responses

4-6 inches 7 16.7

7-9 inches 3 7.1

10-12 inches 6 14.3

Greater than 12 inches 22 52.4

Do not use 4 9.5

a n = number of survey participants who answered question 7 (Appendix A)

105

Table 8. Proportion of very small meat establishments in Pennsylvania that apply more than one antimicrobial compound to carcass surfaces as a slaughter intervention (n = 23) a

More than one compound is used Number of responses Percent of responses

Yes 2 8.7

No 21 91.3

a n = number of survey participants who answered question 9 (Appendix A)

106

Table 9. Method for monitoring the concentration of antimicrobial compounds used for slaughter interventions in very small meat establishments in Pennsylvania (n = 20) a

Method Number of responses Percent of responses b

Temperature monitor 10 50.0

Chlorine test kit 3 15.0 pH meter 1 5.0 pH tape 1 5.0

Other 2 10.0

Do not use antimicrobials 8 40.0

a n = number of survey participants who answered question 11 (Appendix A) b Percentages add up to more than 100% due to multiple responses.

107

Table 10. Interest in participating in a microbial sample study for evaluation of slaughter interventions in very small meat establishments in Pennsylvania (n = 39) a

Interest in microbial study Number of responses Percent of responses

Yes 23 59.0

No 16 41.0

a n = number of survey participants who answered question 12 (Appendix A)

108

Table 11. Interest in training for new slaughter interventions in very small meat establishments in Pennsylvania (n = 45) a

Interest in training Number of responses Percent of responses

Yes 40 88.9

No 5 11.1

a n = number of survey participants who answered question 13 (Appendix A)

109

Table 12. Interest in various types of instructional methods by processors in Pennsylvania

Instructional method Very Somewhat Not Do not Mean a Nb interested interested interested know (3) (2) (1) Brochures/frequently asked 25 (81%) c 6 (19%) 0 (0%) 0 (0%) 2.81 31 questions/booklets

Videotapes 16 (64%) 6 (24%) 2 (8%) 1 (4%) 2.58 25

On-site (at workplace) 13 (46%) 10 (36%) 4 (14%) 1 (4%) 2.33 28

Web-based/Internet 7 (47%) 4 (27%) 3 (20%) 1 (7%) 2.29 15

At a county extension office 12 (38%) 16 (50%) 2 (6%) 2 (6%) 2.33 32

On a university campus 8 (26%) 17 (55%) 4 (13%) 2 (6%) 2.14 31

Correspondence/independent 5 (50%) 3 (30%) 1 (10%) 1 (10%) 2.44 10 learning (print-based)

Audiotapes 5 (33%) 4 (27%) 5 (33%) 1 (7%) 2.00 15

Video conferencing 2 (20%) 4 (40%) 2 (20%) 2 (20%) 2.00 10

Satellite downlink 2 (18%) 4 (36%) 2 (18%) 3 (27%) 2.00 11

CD-ROM 8 (40%) 8 (40%) 3 (15%) 1 (5%) 2.26 20

a Mean = [(number of very interested responses × 3) + (number of somewhat interested responses × 2) + (number of not interested responses × 1)] ÷ N b N = number of survey respondents who demonstrated a level of interest for each instructional method; question 14 (Appendix A) c Number of responses followed by percent of responses in parentheses

110

Table 13. Availability of Internet access to survey participants in Pennsylvania for educational purposes (n = 29) a

Location Number of responses b Percent of responses c

Home 23 79.3

Work 16 55.2

No access 0 0.0

Other location 0 0.0

a n = number of survey participants who answered question 15 (Appendix A) b Respondents were given the options of “yes” or “no”. This column indicates the number of “yes” responses. c Percentages add up to more than 100% due to multiple responses.

111

Table 14. Availability of CD-ROM access to survey participants in Pennsylvania for educational purposes (n = 22) a

Location Number of responses b Percent of responses c

Home 19 86.4

Work 11 50.0

No access 1 4.5

Other d 0 0.0

a n = number of survey participants who answered question 16 (Appendix A) b Respondents were given the options of “yes” or “no”. This column indicates the number of “yes” responses. c Percentages add up to more than 100% due to multiple responses. d No other locations were specified.

112

Table 15. Most suitable time of year for employees of very small meat plant establishments in Pennsylvania to attend a training program (n = 35) a

Time of year Number of responses Percent of responses b

Summer 14 40.0

Spring 13 37.1

Winter 6 17.1

Fall 2 5.7

No preference 6 17.1

a n = number of survey participants who answered question 17 (Appendix A) b Percentages add up to more than 100% due to multiple responses.

113

Table 16. Best times and days of the week for employees of very small meat establishments in Pennsylvania to attend training programs (n = 23) a

Time Mon. Tues. Wed. Thurs. Fri. Sat. Sun.

Morning 24% b 24% 18% 12% 0% 6% 6%

Afternoon 29% 29% 29% 24% 12% 12% 12%

Evening 12% 18% 12% 12% 0% 6% 0%

All day 29% 35% 12% 18% 6% 18% 12%

a n = number of survey participants who answered question 18 (Appendix A) b Respondents checked all applicable days and times. The percentage of responses is shown for each day and time.

114

Table 17. Highest level of education completed by processors in Pennsylvania (n = 40) a

Education level Number of responses Percent of responses

Less than high school 1 2.5

High school 24 60.0

Trade school 2 5.0

Some college 3 7.5

College graduate 10 25.0

a n = number of survey participants who answered question 19, part I (Appendix A)

115

Table 18. Job title of survey participants in Pennsylvania (n = 40) a

Job title Number of responses Percent of responses

Manager 17 42.5

Owner 16 40.0

President 3 7.5

Foreman 3 7.5

Worker 1 2.5

a n = number of survey participants who answered question 19, part II (Appendix A)

116

Table 19. Number of years worked in meat and/or poultry processing industry by survey participants in Pennsylvania (n = 39) a

Years Number of responses Percent of responses

Less than 10 6 15.4

10-19 3 7.7

20-29 14 35.9

30-39 9 23.1

40-49 6 15.4

50-59 1 2.6

a n = number of survey participants who answered question 19, part III (Appendix A)

117

Table 20. Number of employees who work in surveyed establishments in Pennsylvania (n = 39) a

Number of employees Number of responses Percent of responses

Fewer than 10 25 64.1

10-24 7 17.9

25-49 1 2.6

50-74 2 5.1

75-99 2 5.1

100 or more 2 5.1

a n = number of survey participants who answered question 19, part IV (Appendix A)

118

Table 21. Slaughter interventions used to reduce microbial contaminants on carcass surfaces in very small meat establishments in Washington and Idaho (n = 24)a

Interventions Number of responses Percent of responses

Trimming 6 17.6

Cold water wash 10 29.4

Warm water wash 10 29.4

Hot water wash 5 14.7

Organic acid wash 1 2.9

Chlorine 2 5.9

Steam vacuum 0 0.0

Steam pasteurization 0 0.0

Other b 0 0.0

a n = number of survey participants who answered question 1 (Appendix A) b Several plants volunteered additional information regarding interventions. Sodium hypochlorite was used at 0.39% or 0.11% based on preparation directions provided by processors. Also, a 2% acetic acid solution was used in one plant.

119

Table 22. Approximate application time of slaughter intervention to carcass surfaces in very small meat establishments in Washington and Idaho (n = 21)a

Length of time Number of responses Percent of responses

5-10 seconds 5 23.8

10-20 seconds 4 19.0

20-30 seconds 1 4.8

More than 30 seconds 11 52.4

Other b 0 0.0

a n = number of survey participants who answered question 2 (Appendix A) b Some respondents said knife trimming was performed throughout slaughter.

120

Table 23. Types of equipment used to apply slaughter interventions in very small meat establishments in Washington and Idaho (n = 22)a

Application methods Number of responses Percent of responses

Hand held hose 16 72.7

Hand held sprayer 5 22.7

Continuous spray cabinet 0 0.0

Automatic cabinet 0 0.0

Other b 1 4.5

a n = number of survey participants who answered question 3 (Appendix A) b One other method of application included scrubbing by hand

121

Table 24. Timing of the application of slaughter interventions to carcass surfaces in very small meat establishments in Washington and Idaho (n = 24)a

Timing of slaughter intervention Number of responses Percent of responses b

Before the chiller/hot box 11 33.3

Before evisceration 4 12.1

After evisceration 11 33.3

After debunging 6 18.2

Other c 1 3.0

a n = number of survey participants who answered question 4 (Appendix A) b Percentages add up to more than 100% due to multiple responses. c One plant reported applying an intervention in a drip cooler.

122

Table 25. Type of nozzle used in washing cabinets to apply hot water or organic acids in very small meat establishments in Washington and Idaho (n = 24)a

Nozzle type Number of responses Percent of responses

Fixed 0 0.0

Rotating 0 0.0

Do not use cabinet 24 100.0

a n = number of survey participants who answered question 5 (Appendix A)

123

Table 26. Application pressure of hand-held garden sprayer or spray cabinet used to apply hot water or organic acids to carcass surfaces in very small meat establishments (n = 18)a

Application pressure Number of responses Percent of responses

25-50 psi 4 22.2

51-100 psi 4 22.2

151-200 psi 1 5.6

Do not know 1 5.6

Do not use 8 44.4

a n = number of survey participants who answered question 6 (Appendix A)

124

Distance Number of responses Percent of responses

1-3 inches 1 4.2

4-6 inches 1 4.2

7-9 inches 1 4.2

10-12 inches 7 29.2

Greater than 12 inches 1 4.2

Do not use 13 54.2

a n = number of survey participants who answered question 7 (Appendix A)

125

Table 28. Method for monitoring the concentration of antimicrobial compounds used for slaughter interventions in very small meat establishments in Washington and Idaho (n = 18)a

Method Number of responses Percent of responses

Temperature monitor 1 5.6

Chlorine test kit 0 0.0 pH meter 0 0.0 pH tape 0 0.0

Other 0 0.0

Do not use antimicrobials 17 94.4

a n = number of survey participants who answered question 11 (Appendix A)

126

CHAPTER FOUR

INVESTIGATION OF WATER WASHES ON THE ELIMINATION OF

MEAT BORNE PATHOGENS FROM INOCULATED BEEF SURFACES

127

INTRODUCTION

As part of HACCP, FSIS requires the removal of all visible fecal contamination, ingesta, or milk from carcasses, by knife trimming or vacuuming with hot water or steam

(USDA-FSIS, 1996b). Despite the use of sanitary dressing procedures to reduce or eliminate contamination during slaughter and processing, pathogenic bacteria still may be present on red meat carcasses (USDA-FSIS, 1996a). Therefore, antimicrobial treatments or interventions are recommended by USDA to reduce the prevalence of pathogens in raw products. Some plants decide to integrate antimicrobial interventions into their HACCP plans as control points, CPs, or critical control points, or CCPs. Both CPs and CCPs indicate points, steps, or procedures that can be applied to food to manage chemical, physical, and/or biological hazards (Katsuyama and Stevenson, 1995). These control points should be applied when a hazard is significant and reasonably likely to occur. A control point becomes critical if the “point, step, or procedure at which control can be applied and a food safety hazard can be prevented, eliminated, or reduced to acceptable levels” (Katsuyama and Stevenson, 1995).

Furthermore, FSIS requires meat plants to provide technical documentation of the effectiveness of CCPs in plant HACCP plans.

In 2002, meat establishments that produce raw beef had to reassess their operations to determine whether E. coli O157:H7 is a hazard reasonably likely to occur. If so, the establishment was required to incorporate at least one CCP in HACCP plans, as applicable, to verify control of E. coli O157:H7 in raw beef (USDA-FSIS, 2002). According to preliminary survey data (Chapter Three), the final rinse step is usually when a CP or CCP is applied in very small meat plants. During slaughter, there is not usually a subsequent step

128

beyond the final wash in which a CCP can be implemented to effectively control food safety hazards.

Types of interventions

Several antimicrobial treatments, or interventions, are shown to substantially reduce bacteria levels on red meat carcasses and are approved for use to meet HACCP requirements.

The most common carcass decontamination strategies include water washing, knife

trimming, chemical treatments, and moist heat under pressure or vacuum. When considering

which interventions to implement, plants of all sizes should consider the expenses of purchase, maintenance, and daily operation (fixed costs); available space and technical

expertise; and documented antimicrobial effectiveness. Some interventions, which are well-

suited for large plants, are difficult or impossible to implement in small or very small plants.

Selection of antimicrobial treatments

In 2002, very small meat plants in Pennsylvania were asked to complete a survey questionnaire to identify carcass decontamination treatments used during slaughter (Chapter

Three). According to survey results, knife trimming (58%) and cold water washing (56%) were the most common methods to remove fecal matter and other debris from carcasses. In most plants, the application of water washes and/or chemical rinses was performed manually

(e.g. hand-held hoses, portable spray tanks, 87%) instead of with automated equipment (e.g. cabinet washers, 0%). Manual water washes were performed in very small plants under a wide range of temperatures, pressures, application times and application distances. Although most plants use cold water to wash carcasses, a small number of plants reported using warm

129

and/or hot water to wash carcass surfaces. The antimicrobial effectiveness of cold, warm, and hot washing can be investigated under laboratory conditions and the best performing one selected for use in an in-plant study.

Water washing

In very small meat plants, spray washing employs the use of a hand-held hose with a spray gun to deliver water to carcass surfaces for a specified time and within a range of temperatures and pressures. Hand-held hoses are inexpensive and easy to maneuver.

Essentially, the employee directs the water stream at the carcass surface to wash away gross contamination.

Dickson (1995) explored pre-evisceration washing as a means of reducing bacterial attachment to beef surfaces and demonstrated that the affinity for attachment by aerobes and

Enterobacteriaceae to the carcass surface was lowered by 0.7 log CFU/cm 2 as compared with control carcasses.

Carcass washing can be performed with cold, warm, or hot water. Of these three temperature ranges, a hot water wash is considered the most effective at eliminating bacteria due to mechanical removal combined with high temperature (Bolton et al., 2002). At the other end of the spectrum, a cold water wash also can eradicate bacteria by physical force.

However, it is considered a much less effective decontaminant when compared to hot water washing (Bolton et al., 2002). In Germany, cold water washing (12°C) of lamb carcasses

130

immediately following evisceration was deemed ineffective since bacterial loads on ventral carcass surfaces, which were adjacent to the cut made during evisceration, did not change appreciably (Ellerbroek et al., 1993).

Physical attributes of interventions

Physical attributes associated with the application of antimicrobial spray treatments

often do not receive sufficient consideration by food processors and researchers (Pordesimo

et al. 2002). A review of the literature revealed a lack of standardization of process

engineering variables (temperature, pressure, volume, exposure time, distance between food

and spray origin, spray pattern and spray nozzle type, number and orientation) among the

numerous studies that have documented the efficacy of water washing and the application of

antimicrobial solutions (Pordesimo et al., 2002). Pressure and temperature are two physical

attributes that easily be monitored by a plant employee and can have significant impact on

the efficacy of spray treatments.

Studies have indicated the need to demarcate optimal pressure ranges for spray

washing since pressure that is too low inadequately removes microbes from meat surfaces,

and pressure that is too high can actually drive surface microbes deep into carcass tissues

(Pordesimo et al., 2002). According to Gorman et al. (1995), spray washes applied at 200,

300 or 400 psi were more effective at removing fecal matter and bacteria on beef adipose

surfaces than a 40 psi wash. Temperature is also a critical factor in the effectiveness of water

washing as hot and warm water are generally more lethal treatments than cold water. In fact,

washing beef carcasses with cold water was deemed ineffective at reducing aerobic plate

counts at various carcass locations and appeared to transfer contaminants from the

131

hindquarter down to the forequarter (Bell, 1997). Furthermore, Dorsa et al. (1997) assert that water washes at 70°C or greater are very effective at eliminating E. coli , Salmonella spp., and background microflora.

RESEARCH OBJECTIVES

In this study, a series of water washing experiments is performed under laboratory controlled conditions. First, cold water washing is conducted manually at various pressures, application distances, application times, and drip times as a pilot study. The objective of this first study is to determine which combination of physical attributes is the most effective at removing bacteria from experimentally inoculated beef surfaces. Secondly, water washing is performed with cold, warm, and hot water while incorporating the physical attributes that were considered to be the most effective during the first study.

132

METHODS

Preparation of fecal slurry

Feces were collected within 5 min of defecation from non-diarrheic beef cattle at the

Penn State Beef and Sheep Center on the first day of each experiment. After donning sterile latex gloves, several handfuls of a fresh fecal pat were placed into a sterile whirl-pak bag

(15.2 cm × 22.9 cm, NASCO, Fort Atkinson, WI) making sure to collect only the feces and not the soil or debris from the floor or ground. The bagged feces were placed in a styrofoam shipping cooler without coolant and transported immediately to the laboratory for slurry preparation.

Pure cultures of E. coli O157:H7 (ATCC 43889; Penn State Gastroenteric Disease

Center 93-0133) and Salmonella Typhimurium (ATCC 13311, ATCC 14028), were incubated statically overnight (18 – 24 h) in Tryptic soy broth [TSB, Becton Dickinson

Company (BD), Sparks, MD] to obtain approximately 1 X 10 9 CFU/ml. Campylobacter jejuni (ATCC 33559), and C. coli (ATCC 33560) were incubated microaerophilically for 44

– 48 h at 42 ± 2°C in Brucella broth (BB, BD) to obtain the same cell concentration.

Pathogens were considered to be in the stationary growth phase, which was determined previously by generating growth curves (Appendix 2). Each incubated broth was thoroughly

vortexed, aseptically transferred to sterile 250 ml bottles, and centrifuged (5,000 rpm for 5

min, Model RC-5B Plus, Sorvall ®, Thermo Electron Corporation, Asheville, NC). After resuspending each cell pellet in 10 ml TSB or BB, 2 ml of each pathogen was aseptically transferred to 94 g feces to which was added 94 ml Butterfield’s

133

phosphate diluent (BPD*). This mixture was manually homogenized in a sterile stomacher bag for 1 min. Subsequently, 15 ml of fecal slurry was transferred to a sterile test tube and stored at ambient temperature for 1 to 3 h prior to inoculation of beef surfaces.

Inoculation of beef plates with fecal slurry

Vacuum-packaged, boneless beef plates were obtained from a very small meat processor in Pennsylvania and stored at 4°C before use in inoculation challenge experiments.

After removal of vacuum packaging, plates were placed on a sanitized cutting board and excess subcutaneous fat and intercostal meat trimmed away with a sterile scalpel. Trimmed plates were cut into 30 cm by 15 cm sections, which were randomly assigned to treatments.

Each treatment was replicated in triplicate. A 5 cm X 5 cm sanitized, stainless steel template was applied to the lateral side of each plate section and three 25 cm 2 squares were outlined on each section with sterile cotton swabs dipped in edible carcass marking ink (GL #31

Perma-brite purple, Koch Supplies Inc., Kansas City, MO). All three outlines were horizontally aligned 3 to 5 cm apart. The marked plate sections were placed on sanitized trays and exposed to ultraviolet light in a biological safety hood for 15 min (Cutter and

______* Butterfield’s Phosphate Diluent was prepared by adding 1.25 ml stock solution to 1 L distilled water and autoclaving for 15 minutes at 121°C. Stock solution contained 34.0 g KH 2PO 4 (Fisher Scientific Co., Fair

Lawn, NJ) dissolved in 500.0 ml distilled water, which was adjusted to pH 7.2 with approximately 175 ml 1 N

NaOH (VWR International, West Chester, PA). Additional distilled water was added for a final volume of 1 L stock solution (Murano and Hudnall, 2001 Compendium Ch. 63 Media, Reagents, and Stains pgs. 601-648).

134

Siragusa, 1994). A 0.1 ml aliquot of fecal slurry was uniformly dispensed within each marked square. Inoculated plate sections were left undisturbed in the biological safety hood for 15 minutes to allow bacterial attachment to meat surfaces. During pilot washing trials, laminar air flow inside the hood desiccated the inoculated meat surfaces to an undesirable extent. To minimize this desiccation during UV sterilization, inoculation, and bacterial attachment, the laminar air flow was turned off.

Each plate section was fixed on stainless steel tines, which had been welded to a stainless steel stage. The tines were arranged in a rectangular formation (two rows by three columns) on a slightly acute angle to prevent dislodging of the meat piece during treatment.

Prior to the application of any treatment to inoculated plate sections, one marked square (left, center, or right) was randomly selected as a control square. The tissue in this square was excised approximately 1 cm deep with a sterile scalpel and forceps and placed in a sterile stomacher bag. Excised tissue was diluted with 100 ml BPD. The remaining plate section was left attached to the stage, which was then hung vertically on a stainless steel stand and placed inside of a benchtop, stainless steel washing cabinet for exposure to a water wash,

chemical rinse, or combined treatment.

Treatment of inoculated plates with cold water washes

Experimental cold water washes (15 ± 2°C) were performed once under combinations of the following four variables: water application pressure of 10 or 30 psi; distance between hose nozzle and meat surface no greater than 5 ± 1 cm, 30.5 ± 1 cm, or 61 ± 1 cm (2, 12, or

24 in.); application time of 2 s, 10 s, or 20 s; drip or dwell time for 5 min or 20 min. The various water washes were applied to each plate section using a spray washing gun [12.5

135

gpm (47.3 L/min), 150 psi max., 200°F (93.3°C) max., 3/4” I.D., aluminum body, built-in thermometer, Part no. 33505K68, McMaster-Carr, Aurora, OH] fitted with a pressure gauge

(1/4” NPT, 100 PSI Model No. RG-2, Part no. 05003A500, Water Ace Pump Co., Ashland,

OH), which was connected to a packinghouse hose [50 ft. (15.2 m), 500 psi max., 200°F

(93.3°C) max., 3/4” I.D., Part no. 3538 2050, KOCH Supplies, North Kansas City, MO)].

The hose was connected to a faucet in the laboratory to supply tap water for cold water washing. The distance between the spray nozzle and meat surface was measured with a steel carpenter’s square (61 cm body, 41 cm tongue, Stanley Tools Product Group, New Britain,

CT). Following the dwell period, the stainless steel stage was carefully removed from the washing cabinet so that the treated 25 cm 2 squares could be excised approximately 1 cm deep

with a sterile scalpel and forceps. Each excised piece was placed in a sterile stomacher bag

and diluted with 100 ml BPD, followed by microbiological methods for the enumeration and

detection of bacterial populations (see pg. 153).

Treatment of inoculated plates with cold, warm or hot water washes

In addition to water temperature, the importance of drip time and distance were

examined further in subsequent washing treatments of inoculated beef plates. This second

tier of experiments investigated the effectiveness of water washing with cold (15 ± 2°C),

warm (54 ± 2°C), or hot (77 ± 2°C) water, which was applied at 30 psi for 20 s from a

distance of 5.1 cm or 30.5 cm, and allowed to drip for 5 or 20 min. Cold and warm tap water

were supplied through the same faucet, hose, and spray gun. For hot water washes, water was

heated in a water bath (99 ± 1°C; RTE-221, NESLAB Instruments, Inc., Newington, NH)

and pumped (1/3 HP, 1725 RPM, 115 V, Catalog No. H139, GE Industrial Systems, Ft.

136

Wayne, IN) directly out of the bath reservoir through a braided hose (980 PVC, 3/8” I.D. X

9/16 O. D., 3/32” wall thickness, part no. 63012-483, VWR International, West Chester, PA), which was connected to the same spray gun that was used for cold and warm water washes.

In a separate experiment, self-adhesive temperature monitoring strips (MINI Eight-Position

101-8-110F, Palmer Wahl Instruments, Asheville, NC) were affixed to lean tissue surfaces to detect the maximum change in temperature on beef plates while warm (54 ± 2°C) or hot water (77 ± 2°C) washes are directly applied for 20 s. Following the 5 min dwell period, the stainless steel stage was carefully removed from the washing cabinet so that the treated 25 cm 2 squares could be excised approximately 1 cm deep with a sterile scalpel and forceps.

Each excised piece was placed in a sterile stomacher bag and diluted with 100 ml BPD, followed by microbiological methods for the enumeration and detection of bacterial populations.

Enumeration/isolation of pathogens and hygiene indicators

Following the administration of water washes, chemical rinses, or a combination

treatment, all excised meat pieces in stomacher bags were stomached at 260 rpm for 2 min

(Seward Stomacher, Tekmar Co., Cincinnati, OH) and used to enumerate pathogens ( E. coli

O157:H7, S. Typhimurium, Campylobacter spp.) and hygiene indicators (mesophilic aerobic plate count, total coliforms, and generic E. coli ). Although other Salmonella spp. may have been naturally present in bovine feces, these counts were reported as S. Typhimurium, which

was likely to be the most predominant serotype. Homogenized excision samples (HES) also

were analyzed for the presence of pathogens at low levels, which may not be detected by the

enumeration protocol. Excised squares that were randomly selected as the control were

137

enumerated for pre-treatment bacterial populations. The remaining two squares of each treated plate section were used to enumerate post-treatment bacterial populations.

HES were spiral plated (Autoplate ® 4000, Spiral Biotech, Norwood, MA) in duplicate on Sorbitol MacConkey agar (CT-SMAC, BD) supplemented with cefixime (1 mg/L) and potassium tellurite (5 mg/L, CT supplement, Invitrogen), XLD agar (BD), and modified cefoperazone Campylobacter agar (mCCDA, REMEL) for the enumeration of populations of

E. coli O157:H7, S. Typhimurium, and Campylobacter spp., respectively. CT-SMAC and

XLD plates were incubated at 37 °C for 24 hours. mCCDA Plates were incubated microaerophilically (BBL™ CampyPak Plus Microaerophilic System Envelopes with

Palladium Catalyst, BD) for 48 h at 42 °C. Bacterial colonies were enumerated automatically

(Q-Count ®, Spiral Biotech).

Detection of pathogens at low levels

E. coli O157:H7

A 1 ml aliquot of HES was aseptically transferred to 9 ml GN broth (BD). One loopful each of E. coli O157:H7 (ATCC 43889) and E. coli O157:H7 (PSU Gastroenteric

Disease Center 93-0133) were also transferred into GN broth as a positive control. Tubes were incubated without agitation for 6 h at 37 °C. A secondary enrichment step,

immunomagnetic separation with Dynabeads anti-E.coli O157 (Invitrogen Corporation), was performed after primary enrichment in GN broth according to manufacturer instructions. If

no colonies were detected on spiral-plated CT-SMAC plates, enriched beads were streaked

for isolation on duplicate CT-SMAC and incubated at 37 °C for 24 h. If typical colonies were

138

observed, they were tested for latex agglutination ( RIM ® E. coli O157:H7 latex test kit,

REMEL) as directed.

Salmonella spp.

HES (1 ml) was transferred to 9 ml Lactose broth (BD) and was gently vortexed prior to static incubation at 37 °C for 24 hours as a primary enrichment for Salmonella spp. S.

Typhimurium (ATCC 14028) and S. Typhimurium (ATCC 13311) were used as positive

controls. If spiral-plated XLD exhibited no visible colonies of S. Typhimurium after 24 h of

incubation, 1 ml enriched Lactose broth was transferred to 9 ml each Selenite Cystine and

Tetrathionate broths and incubated for 18-24 h at 37°C with caps loosened. Each selective broth was streaked for isolation to duplicate XLD plates and incubated overnight (37°C).

Presumptive colonies typical of Salmonella spp. were confirmed by latex agglutination

(Oxoid Salmonella Latex Kit, Unipath, Hampshire, UK).

Campylobacter spp.

On the same day of rinsing experiments, 1 ml of HES was aseptically transferred to 9

ml Bolton broth (REMEL Inc., Lenexa, KS), which had been supplemented with 5 ml

Bolton broth selective supplement (REMEL Inc.) per 500 ml and 50ml/L laked horse blood

(HemoStat Laboratories, Dixon, CA). One loopful each of C. jejuni (ATCC 33560) and C.

coli (ATCC 33559) also were placed into Bolton broth tubes as positive controls. Tubes were incubated microaerophilically with CampyPak Plus packets 4 h at 37 °C. Then, the anaerobic jar was transferred to a 42 °C incubator for an additional 44 h of enrichment. Selectively enriched samples were streaked for isolation in duplicate onto mCCDA and incubated for 48 139

h at 42°C under microaerophilic conditions. Each incubated plate was inspected for positive colonies using the control plates for visual comparison. Plates showing no growth were incubated for another 24 to 48 h. Colonies that appeared to be positive for Campylobacter spp. were confirmed by the Oxoid Campylobacter DrySpot Latex Test kit (REMEL) as directed.

Enumeration of hygiene indicators

Serial dilutions were performed on a 1-ml aliquot of homogenate in BPD and then plated in duplicate on Aerobic Plate Count (APC) and E. coli/coliform Petrifilm™ (3M™,

St. Paul, MN) followed by incubation at 35 °C for 48 h according to manufacturer specifications. Mesophilic APC, generic E. coli (EC), and coliforms (CF) were counted manually according to the AOAC method described by the manufacturer or automatically

(3M™ Petrifilm™ Plate Reader, 3M™).

Statistical analysis of cold water washes

Bacterial populations were determined before and after cold water washing and compared by two-sample t-test to determine whether treated beef plates differed from the control (MINITAB ®, Release 14.1, State College, PA). Data underwent a log (X + 1)

2 transformation so that bacterial counts could be reported as log 10 CFU/cm . Post-washing populations were subtracted from pre-treatment populations to calculate mean log reductions

(Microsoft Office EXCEL 2003, Redmond, WA). Each mean log reduction or bacterial count

is reported with the standard error of the mean (SEM) to provide a measure of the variation

of the mean as it relates to the entire population (MINITAB ®). Also, physical variables

140

associated with the cold water washing of meat surfaces with a hand-held hose and spray gun were analyzed statistically to determine which combinations of water pressure, application distance, application time, and drip time provided the greatest antimicrobial effectiveness.

These variables were analyzed using the General Linear Models (GLM) procedure of

ANOVA (MINITAB ®). GLM output provided identical SEMs for mean log reductions because the error is pooled across factor levels and sample sizes are the same (Nichols,

1998). Pressure, distance, or, application time, and drip time were specified as fixed and crossed factors in each model, which was generated for each pathogen and hygiene indicator.

Tukey’s multiple comparison method was used to separate mean log reductions. Factor levels of washing variables that were considered to be the most effective by GLM were applied in a subsequent study of cold, warm, and hot water washes. All statistical analyses were conducted at the 0.05 significance level.

Statistical analysis of cold, warm, and hot water washes

Bacterial counts following water washing at three temperatures were calculated as described previously for cold water washing. Water temperatures, application distances, and drip times were analyzed by the GLM procedure of MINITAB ® with means separated by

Tukey’s method to determine which least squares means (factor levels) were best attributed

to the effective reduction of bacterial populations. Pre- and post-treatment bacterial

2 ® populations (log 10 CFU/cm ) also were compared by two-sample t-tests (MINITAB ). All

statistical analyses were performed at a significance level of 0.05. When the best

combination of physical factors was identified, it was incorporated into subsequent chemical

rinsing treatments and combination treatments.

141

RESULTS

Cold water washes

According to t-test results, every bacterial population, except for Campylobacter spp.,

underwent a significant decrease as a result of being treated with a cold water wash (Tables 1

through 6). For Campylobacter spp., the water washes applied at 30 psi, 30.5 cm distance, 2 s wash time and 20 min drip time did not lower appreciably post-treatment populations (Table

2). A 10 psi cold water wash is more effective than a 30 psi wash at reducing Campylobacter spp. and generic E. coli (Tables 2 and 5). Conversely, cold water delivered at 30 psi significantly lowered APC by 0.7 log CFU/cm 2 more than a 10 psi wash (Table 4). There were no important differences in the reductions of S. Typhimurium, E. coli O157:H7, or coliforms between the two wash pressures (Tables 1, 3, and 6). Reasoning that most plants are likely to use tap water at a pressure that is somewhat greater than 10 psi for washing carcasses, further experiments employed a water pressure of 30 psi. Also, 30 psi was the most intense water pressure that could be generated in the laboratory using a hand-held sprayer.

The results of application distance (between the spray gun nozzle and meat surface) do not provide any meaningful trends. The ANOVAs for the reduction of Campylobacter spp., E. coli O157:H7, and APC point out that a 30.5 cm application distance is the least effective numerically, but that it is one of the most effective distances (2.07 log CFU/cm 2)

for the removal of generic E. coli (Tables 2, 3, 5, and 6). For S. Typhimurium., E. coli

O157:H7, and APC, there were no statistically important differences among the three spray distances (Tables 1, 3, and 4). 142

Cold water washing data were analyzed three times by the GLM procedure of

ANOVA. The decisions to select or exclude spray washing variables for further analysis were based solely on this first, flawed data analysis. The first analysis demonstrated the 61 cm spray distance clearly to be the least effective across most of the bacterial populations

(Table 14). Hence, it was deselected for further analysis in cold, warm, and hot water washes.

Subsequent cold, warm, and hot water wash experiments were performed at application distances of 5.1 and 30.5 cm based on the first analysis of cold water washing data.

Errors in the first data analysis were recognized after the completion of cold, warm, and hot water washing experiments. The second analysis of the data accounted for the dilution effect of placing a 25 cm 2 excision of beef tissue into 100 ml BPD (i.e., raw bacterial

counts were multiplied by 4) with GLM performed for post-treatment populations (Table 15).

This second analysis was still incorrect because the response variable was mean log

reductions, not populations. Finally, the third analysis of the data accounted for the dilution

of samples being enumerated for pathogens and hygiene indicators and included mean log

reductions, instead of population counts. Unfortunately, the third and most accurate analysis

of the data casts greater doubt about the relative effectiveness of spray washes than the first

data analysis. Among all four physical factors of washing, the greatest disparity between the

two statistical outputs is distance (Tables 1 through 6 and 14). Perhaps all three distances

should have been included in the cold, water, and hot water wash experiments instead of the

two shorter distances (Tables 1 through 6).

The effect of application time on bacterial elimination also was measured. A 20 s

application time provided the greatest reductions among bacterial populations (Tables 1, 2, 3,

5 and 6) and was statistically significant. Though not significant, a 20 s cold water wash

143

provided the most effective numerical reduction of mesophilic APC (Table 4). Given these results, all subsequent water washing treatments in this study were applied for 20 s.

Finally, the effect of drip time on the recovery of pathogens and hygiene indicators was researched. For most populations, a 20 min drip time provided significantly greater reductions on the order of 0.21 to 0.47 log CFU/cm 2 (Tables 1, 2, 4, 5, and 6). Also, a 20 min drip time was associated with numerically greater reductions in S. Typhimurium and E. coli

O157:H7, but it was not statistically important (Tables 1 and 3). In the first, flawed analysis of drip time data, log reductions after 5 min and 20 min of drip appeared less divergent with substantial differences only in counts of E. coli O157:H7 and generic E. coli (Table 14).

Based on the first analysis of data, both drip times were investigated further in studies of cold, warm, and hot water washes.

Cold, warm, and hot water washes

Two-sample t-tests demonstrate that all bacterial populations were significantly diminished to some extent by being washed with cold, warm, or hot water from 5.1 or 30.5 cm spray distances and after 5 or 20 min of drip time (Tables 7 through 12). Washing with hot water was undoubtedly the most effective way to reduce all six bacterial populations with log reductions ranging from 2.73 to 5.05 CFU/cm 2 (Tables 7 through 12). Between warm and cold water washes, warm water treatments results in numerically greater reductions for all bacterial populations except for E. coli O157:H7. While cold water removed 0.31 log

CFU/cm 2 more E. coli O157:H7 than warm water, there was no statistical significance between any of the log reductions for cold and warm water.

144

Drip time and application distance were further investigated in the application of cold, warm, and hot washes. A 5 min drip time resulted in substantially greater reductions of

S. Typhimurium (1.0 log) and E. coli O157:H7 (0.74 log) than a 20 min drip (Tables 7 and

9). A 5 min drip also was associated with a slight increase in log reductions of 0.25 to 0.39 when compared with 20 min (Tables 8, 10, 11, and 12). In addition, there were no significant differences between the two application distances for any pathogen or hygiene indicator

(Tables 7 through 12). Numerically, there was no greater than a 0.66 log difference between

5.1 cm and 30.5 cm spray distances (Table 10). Consequently, all subsequent washing and rinsing experiments were performed with a 5 min drip.

The surface temperatures of beef plates were monitored during treatment with warm water (54 ± 2°C) or hot water (77 ± 2°C). The mean surface temperature of lean beef tissue during a warm water wash was 53.1°C and 73.9°C during hot water washing (Table 13).

Adipose tissue surfaces were not monitored in this study as fecal slurries and washing treatments were applied only to lean surfaces.

145

DISCUSSION

Cold water washes were conducted at various pressures and application distances as well as for various application and drip times using a hand-held sprayer. There was very little difference between washing beef plates at 10 psi or 30 psi. Even though 10 psi washes were slightly more effective, additional water washing experiments were performed at 30 psi because it better represents the water pressure that is likely to exist in meat packing plants than 10 psi. Many of the spray guns sold in a popular meat processing supply catalog are rated to 150 psi, which suggests that the expected water pressure during use is greater than 30 psi (Koch Supplies Inc., North Kansas City, MO). According to Pordesimo et al. (2002), increasing spray pressure also increases bacterial removal. In this study, the use of a wider range of application pressures (i.e. 10 to 100 psi) may have been necessary to detect a more prominent eliminatory effect. Actually, water pressure in the range of 100 to 300 psi is considered optimal for washing red meats (De Zuniga et al., 1991). Nevertheless, it was not feasible to perform water washes at pressures greater than 30 psi in this study.

Results from cold water washing alone did not demonstrate noteworthy differences among spray distances, except for Campylobacter spp. In this study, the mean log reductions

of Campylobacter spp. do not decrease in order of increasing spray distances (Table 2). In particular, the 1.04 CFU/cm 2 log reduction associated with a 30.5 cm distance was expected

to be more than the 1.59 CFU/cm 2 log reduction at 61 cm. This disparity can be explained by the low level of Campylobacter species that initially were present on the untreated beef plate.

According to Dorsa et al. (1996), the application of an antimicrobial carcass intervention to a meat surface that has been experimentally inoculated with a high (7.6 log CFU/cm 2) dose of 146

E. coli O157:H7 results in a high (2.1 CFU/cm 2) log reduction. Conversely, the application of

the same intervention to a meat surface with a low (5.5 log CFU/cm 2) dose inoculum resulted in a lower (2.1 CFU/cm 2) log reduction of E. coli O157:H7 (Dorsa et al., 1996). In the current study, one might expect a larger log reduction in Campylobacter spp. following water

2 washing at 30.5 cm if the initial inoculum had contained as many log 10 CFU/cm

Campylobacter spp. as the other inocula.

Nevertheless, spray distance did not seem to be associated with significant reductions in bacterial populations when performed at cold, warm, or hot temperatures. Pordesimo et al.

(2002) assert that spray washing efficiency is dependent on contact force, which is a function of the distance from a spray nozzle (spray origin) to the target food surface. As the distance increases, the contact force decreases because the water loses momentum as it travels to the target food surface. The range of distances in this study (5.1 to 61 cm) may have not been broad enough to identify differences in contact force. Hence, a very small processor can achieve bacterial reductions at 5.1 cm that are probably comparable to reductions achieved from a spray distance of 61 cm. In the context of the survey results (Chapter Three), the very small processors who apply cold water washes from a distance of 4 to 24 inches (10.2 to 61 cm) are likely to realize bacterial reductions between 0.88 and 2.07 log CFU/cm 2.

With respect to water temperature, one could use either cold or warm water to achieve similar log reductions in bacterial counts (Tables 7-12) although hot water washing eliminated significantly greater magnitudes of pathogens, APC, coliforms, and generic E. coli . Even though hot water was the most effective of the water washes, several very small

147

meat processors have remarked that they will not use hot water to wash carcasses because it discolors meat surfaces (Anonymous, personal communication*). According to these processors, surface lean discoloration is an economic issue associated with trim loss and customer unacceptance.

According to Gill et al. (1999), one can expect a 1 to 2 log reduction in aerobic plate counts on carcass surfaces treated with hot water washing (85°C) for at least 10 s, while coliforms and generic E. coli are generally reduced by 2 log with minimal lean discoloration.

Consequently, the application of this hot water wash in a washing cabinet (pressure and flow rate were not reported) for 20 s was found to unacceptably discolor lean surfaces of beef carcasses (Gill et al., 1999). However, Gill and Bryant (2000) assert that discoloration by hot water pasteurization is alleviated by air chilling since exposed lean surfaces undergo desiccation and develop a dark red color, whether treated with hot water or not. Lean discoloration also was observed on beef carcass surfaces following a 2 min hot water wash

(95°C; Barkate et al., 1993). After 24 h of chilling, normal lean color returned to previously discolored surfaces (Barkate et al., 1993).

______

* These remarks are from very small meat processors who participated in the survey via telephone and/or in- plant carcass sampling studies. All participating processors were assured that their identities would not be

disclosed. 148

In water washing studies to reduce bacterial loads on meat surfaces, water temperature often receives more attention than other spray washing variables, such as distance and application pressure (Pordesimo et al., 2002). Castillo et al. (1998) applied a hot water rinse (24 psi, 95°C, 14.4L/min, 5 s, distance from sample = 12.5 cm) with a hand-held garden sprayer to beef surfaces inoculated with a fecal slurry containing E. coli O157:H7 and

S. Typhimurium and observed log reductions of 3.7 and 3.8, respectively. The results of

Castillo et al. (1998) agree with the current study in that log reductions of 3.88 and 4.1 were achieved for E. coli O157:H7 and S. Typhimurium, respectively. In both studies, spray wash engineering variables also were similar.

APC and coliform counts also were reduced by 2.9 and 3.3 log, in that order (Castillo

2 et al., 1998). The 2.9 log reduction in APC by Castillo et al. (1998) was 0.9 log CFU/cm higher than that expected by Gill et al. (1999), which could be explained by the higher water temperature (95°C). As hot water washing relates to the current study, APC and coliforms were reduced by 4.0 and 5.1 log. The discrepancy between these results and those of Castillo et al. (1998) are not easily explained by the difference in temperature. Hot water in the current study was at 77°C and applied for 15 s longer at a slightly higher pressure (30 psi).

Perhaps the longer exposure time to hot water in this study caused the additional log reduction.

Furthermore, the removal of non-pathogenic E. coli by hot water washing (74°C) was compared with warm water (35°C; Cabedo et al., 1996). Hot water eliminated 4.17 log

CFU/cm 2 E. coli , while warm water yielded a 3.52 log reduction (Cabedo et al., 1996). The

temperatures for hot and warm water employed by Cabedo et al. (1996) were slightly lower

than those used in the current study in which hot water (77°C) yielded a comparable 4.33 log 149

decline in generic E. coli (Table 11). Warm water (54°C) provided a 2.43 log reduction

(Table 11), which was approximately 1 log less effective than that achieved by Cabedo et al.,

(1996). Cabedo et al. (1996) applied water washes with an automated spraying cabinet at 300 psi for 12 s, whereas the current study conducted water washes manually at 30 psi for 20 s.

The difference in application pressures may account for the disparity in E. coli elimination between the two warm water washes. However, the two hot water washes yielded similar log reductions, which suggests that hot water can be an effective decontaminant regardless of pressure. It is possible that the high water temperature provided the greater lethality than mechanical removal, as a function of water pressure.

Drip time following a water wash also was investigated. Based on cold, warm, and hot water wash data, beef plates underwent slightly greater reductions in populations of S.

Typhimurium and E. coli O157:H7 following a 5 min drip than a 20 min drip. Allowing meat pieces to drip for a longer period of time (20 min) in a benchtop washing cabinet resulted in slightly higher populations of pathogens and hygiene indicators. Given these data, it is not clear whether exposure to the open air for a longer period of time could be the cause of the slightly higher bacterial counts. In practice, it might be prudent to place carcasses in a chilling environment within 5 min of performing the final water wash or other intervention.

Any bacteria remaining on treated carcass surfaces would be quickly subjected to the stress of chilling after being subjected to displacement by water washing (preferably hot) and desiccation during a 5 min drip period.

In afterthought, it is evident that an experimental control was omitted from these investigations of drip time. Bacterial reductions relative to a 5 or 20 min drip time were enumerated. However, it would have been useful to incorporate a 0 min drip time as a

150

control. Hence, there is no foundation against which the effects of a 5 or 20 min drip can be measured. While this study suggests that a 5 min drip is slightly more effective than a 20 min drip, it is unknown whether the effectiveness of a 5 min drip differs from not providing a drip period at all. Prior to this realization, the recommendations that were formulated for the water washing of carcasses in very small meat plants include a 5 min drip time.

Although an experimental control for drip time was omitted in this study, one can speculate that drip time is essential under typical conditions of primary processing. Allowing carcass surfaces to drip for a few minutes after water washing facilitates contact between any bacteria remaining on the carcass surface and a subsequently applied antimicrobial rinse.

Furthermore, excess water film on a carcass surface can dilute the antimicrobial rinse, which follows water washing, thereby reducing the ultimate antimicrobial effectiveness of the rinse.

These findings are used to generate recommendations for the water washing of red meat carcasses in very small meat establishments. Subsequent studies investigate the antimicrobial effectiveness of various chemical rinses and the combination of water washing with chemical rinsing at eliminating potentially harmful bacteria from meat surfaces.

151

REFERENCES

Barkate, M. L., Acuff, G. R., Lucia, L. M., and Hale, D. S. 1993. Hot water

decontamination of beef carcasses for reduction of initial bacterial numbers. Meat

Sci. 35, 397-401.

Bell, K. Y., Cutter, C. N., and Sumner, S. S. 1997. Reduction of food borne micro-

organisms on beef carcass tissue using acetic acid, sodium bicarbonate, and

hydrogen peroxide spray washes. Food Microbiol. 14:439-448.

Bolton, D. J., Pearce, R. A., Sheridan, J. J., Blair, I. S., McDowell, D. A., and Harrington,

D. 2002. Washing and chilling as critical control points in pork slaughter hazard

analysis and critical control point (HACCP) systems. J. Appl. Microbiol. 92:893-

902.

Cabedo, L., Sofos, J. N., and Smith, G. C. 1996. Removal of bacteria from beef tissue by

spray washing after different times of exposure to fecal material. J. Food Protect.

59:1284-1287.

Castillo, A., Lucia, L. M., Goodson, K. J., Savell, J. W., and Acuff, G. R. 1998. Use of hot

water for beef carcass decontamination. J. Food Protect. 61:19-25.

Cutter, C. N. and G. R. Siragusa. 1994. Efficacy of organic acids against Escherichia coli

O157:H7 attached to beef carcass tissue using a pilot scale model carcass washer. J.

Food Protect. 57:97-103.

De Zuniga, A. G., Anderson, M. E., Marshall, R. T., and Iannotti, E. L. 1991. A model

system for studying the penetration of microorganisms into meat. J. Food Protect.

54:256-258.

152

Dickson, J. S. 1995. Susceptibility of preevisceration washed beef carcasses to

contamination by Escherichia coli O157:H7 and salmonellae. J. Food Protect. 58:

1065-1068.

Dorsa, W. J., Cutter, C. N., and Siragusa, G. R. 1996. Effectiveness of a steam-vacuum

sanitizer for reducing Escherichia coli O157:H7 inoculated to beef carcass surface

tissue. Letters Appl. Microbiol. 23:61-63.

Dorsa, W. J., C. N. Cutter, and G. R. Siragusa. 1997. Effects of steam-vacuuming and hot

water spray wash on the microflora of refrigerated beef carcass surface tissue

inoculated with Escherichia coli O157:H7, Listeria innocua , and Clostridium

sporogenes . J. Food Protect. 60:619-624.

Ellerbroek, L. L., Wegener, J. F., and Arndt, G. 1993. Does spray washing of lamb

carcasses alter bacterial surface contamination? J. Food Protect. 56:432-436.

Gill, C. O., and Bryant, J. 2000. The effects on product of a hot water pasteurizing

treatment applied routinely in a commercial beef carcass dressing process. Food

Microbiol. 17:495-504.

Gill, C. O., Bryant, J., and Bedard, D. 1999. The effects of hot water pasteurizing treatments

on the appearances and microbiological conditions of beef carcass sides. Food

Microbiol. 16:281-289.

Gorman, B. M., Morgan, B. J., Sofos, J. N., and Smith, G. C. 1995. Microbiological and

visual effects of trimming and/or spray washing for removal of fecal material

from beef. J. Food Protect. 58:984-989.

Katsuyama A. M. and Stevenson, K. E. 1995. Hazard analysis and identification of

critical control points. Ch. 8. In: Establishing hazard analysis and critical control

153

point programs: a workshop manual , 2 nd ed. K. E. Stevenson and D. T. Bernard

(Eds.), p. 8-1 to 8-14. The Food Processors Institute, Washington, D. C.

Nichols, D. 1998. Re: How is std. error calculated for est. marginal means in

GLMRepeated Measures? electronic mail, 9 April 1998. Available online:

http://www.listserv.uga.edu/cgi-bin/wa?A2=ind9804&L=spssx-

l&F=&S=&P=14812

Pordesimo, L. O., Wilkerson, E. G., Womac, A. R., and Cutter, C. N. 2002. Process

engineering variables in the spray washing of meat and produce. J. Food Protect.

65:222-237.

USDA-FSIS, 1996a. FSIS Directive 6350.1: Use of knife trimming and vacuuming of

beef carcasses with hot water or steam; other beef carcass intervention systems.

USDA-FSIS, 1996b. Notice of policy change; achieving the zero tolerance performance

standard for beef carcasses by knife trimming and vacuuming with hot water or

steam; use of acceptable carcass interventions for reducing carcass contamination

without prior agency approval. Federal Register 61, no. 66 (4 April 1996).

USDA-FSIS, 2002. Instructions for verifying E. coli O157:H7 reassessment. Notice 44-02,

11/04/02 Food Safety and Inspection Service, Washington, D. C.

154

2 Table 1. Mean populations and reductions of Salmonella Typhimurium (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 5.17 ± 0.12 3.75 ± 0.12 0.000 1.42 ± 0.09 1

30 psi 4.96 ± 0.22 3.88 ± 0.11 0.000 1.20 ± 0.09 1

p-value d 0.108

Distance 5.1 cm 4.69 ± 0.11 3.41 ± 0.08 0.000 1.28 ± 0.12 1

30.5 cm 5.15 ± 0.04 3.73 ± 0.11 0.000 1.42 ± 0.12 1

61 cm 5.35 ± 0.33 4.30 ± 0.16 0.012 1.23 ± 0.12 1

p-value 0.496

Appl. time e 2 s 4.61 ± 0.25 3.90 ± 0.12 0.020 0.71 ± 0.12 3

10 s 5.25 ± 0.19 3.91 ± 0.17 0.000 1.34 ± 0.12 2

20 s 5.33 ± 0.15 3.63 ± 0.13 0.000 1.87 ± 0.12 1

p-value 0.000

Drip time 5 min 5.10 ± 0.19 3.90 ± 0.13 0.000 1.20 ± 0.09 1

20 min 5.03 ± 0.17 3.72 ± 0.10 0.000 1.42 ± 0.09 1

p-value 0.100

a Mean populations before and after treatment within analyzed by two-sample t-test (MINITAB ®, α = 0.05); sample size by factor (before, after): Pressure (18, 36), Distance (12, 24), Appl. time (12, 24), and Drip time (18, 36) b p-value from two-sample t-test c Least squares means in column group (antimicrobial and water rinses) sharing the same superscripts are not significant by GLM procedure of ANOVA; means separated by Fisher’s LSD (MINITAB ®, α = 0.05) d p-value of one-way ANOVA e application time

155

2 Table 2. Mean populations and reductions of Campylobacter spp. (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 4.78 ± 0.06 2.89 ± 0.08 0.000 1.89 ± 0.07 1

30 psi 2.96 ± 0.51 1.85 ± 0.23 0.061 1.15 ± 0.07 2

p-value d 0.000

Distance 5.1 cm 4.74 ± 0.07 2.81 ± 0.08 0.000 1.93 ± 0.08 1

30.5 cm 2.40 ± 0.73 1.36 ± 0.29 0.206 1.04 ± 0.08 3

61 cm 4.45 ± 0.09 2.94 ± 0.12 0.000 1.59 ± 0.08 2

p-value 0.000

Appl. time e 2 sec 3.86 ± 0.53 2.67 ± 0.26 0.060 1.20 ± 0.08 2

10 sec 3.86 ± 0.53 2.29 ± 0.24 0.016 1.57 ± 0.08 1

20 sec 3.86 ± 0.53 2.17 ± 0.21 0.009 1.79 ± 0.08 1

p-value 0.000

Drip time 5 min 3.86 ± 0.42 2.48 ± 0.21 0.007 1.38 ± 0.07 2

20 min 3.87 ± 0.43 2.26 ± 0.18 0.002 1.66 ± 0.07 1

p-value 0.006

156

2 Table 3. Mean populations and reductions of Escherichia coli O157:H7 (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 5.00 ± 0.14 3.41 ± 0.14 0.000 1.59 ± 0.081

30 psi 5.11 ± 0.15 3.65 ± 0.10 0.000 1.55 ± 0.08 1

p-value d 0.736

Distance 5.1 cm 4.72 ± 0.09 3.08 ± 0.13 0.000 1.64 ± 0.10 1

30.5 cm 4.85 ± 0.10 3.44 ± 0.14 0.000 1.41 ± 0.10 1

61 cm 5.60 ± 0.20 4.06 ± 0.10 0.000 1.67 ± 0.10 1

p-value 0.117

Appl. time e 2 sec 4.88 ± 0.15 3.83 ± 0.11 0.000 1.05 ± 0.10 3

10 sec 5.11 ± 0.20 3.57 ± 0.18 0.000 1.54 ± 0.10 2

20 sec 5.18 ± 0.18 3.19 ± 0.13 0.000 2.13 ± 0.10 1

p-value 0.000

Drip time 5 min 5.10 ± 0.14 3.57 ± 0.13 0.000 1.53 ± 0.08 1

20 min 5.01 ± 0.16 3.49 ± 0.11 0.000 1.62 ± 0.08 1

p-value 0.443

157

2 Table 4. Mean mesophilic aerobic plate counts and reductions (log 10 CFU/cm ± SE) on inoculated beef plates following treatment with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 8.09 ± 0.17 7.15 ± 0.18 0.000 0.93 ± 0.12 2

30 psi 7.54 ± 0.13 6.06 ± 0.08 0.000 1.63 ± 0.12 1

p-value d 0.000

Distance 5.1 cm 7.90 ± 0.10 6.72 ± 0.16 0.000 1.18 ± 0.14 2

30.5 cm 7.62 ± 0.14 6.74 ± 0.23 0.003 0.88 ± 0.14 2

61 cm 7.92 ± 0.30 6.36 ± 0.22 0.000 1.79 ± 0.14 1

p-value 0.000

Appl. time e 2 sec 7.51 ± 0.20 6.28 ± 0.15 0.000 1.22 ± 0.14 1

10 sec 7.89 ± 0.23 6.71 ± 0.22 0.001 1.18 ± 0.14 1

20 sec 8.03 ± 0.15 6.83 ± 0.23 0.000 1.44 ± 0.14 1

p-value 0.393

Drip time 5 min 7.87 ± 0.15 6.82 ± 0.16 0.000 1.05 ± 0.12 2

20 min 7.75 ± 0.17 6.39 ± 0.17 0.000 1.52 ± 0.12 1

p-value 0.008

158

2 Table 5. Mean populations and reductions of generic E. coli (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 6.25 ± 0.23 4.14 ± 0.22 0.000 2.11 ± 0.06 1

30 psi 6.35 ± 0.23 4.63 ± 0.11 0.000 1.72 ± 0.06 2

p-value d 0.000

Distance 5.1 cm 6.48 ± 0.17 4.51 ± 0.13 0.000 1.96 ± 0.07 1,2

30.5 cm 5.83 ± 0.14 3.76 ± 0.16 0.000 2.07 ± 0.07 1

61 cm 6.60 ± 0.40 4.89 ± 0.27 0.002 1.72 ± 0.07 2

p-value 0.005

Appl. time e 2 sec 5.81 ± 0.29 4.39 ± 0.19 0.001 1.42 ± 0.07 3

10 sec 6.40 ± 0.29 4.39 ± 0.27 0.000 2.02 ± 0.07 2

20 sec 6.69 ± 0.20 4.37 ± 0.19 0.000 2.32 ± 0.07 1

p-value 0.000

Drip time 5 min 6.42 ± 0.22 4.61 ± 0.20 0.000 1.81 ± 0.06 2

20 min 6.18 ± 0.23 4.16 ± 0.14 0.000 2.02 ± 0.06 1

p-value 0.018

159

2 Table 6. Mean populations and reductions of coliforms (log 10 CFU/cm ± SE) on inoculated beef plates treated with a cold water wash (15°C) under various pressures, spray distances, application times, and drip times

Factor Factor level Before a After a p-value b Reduction c

Pressure 10 psi 7.26 ± 0.20 5.32 ± 0.20 0.000 1.94 ± 0.11 1

30 psi 7.08 ± 0.16 5.35 ± 0.09 0.000 1.88 ± 0.11 1

p-value d 0.709

Distance 5.1 cm 7.23 ± 0.11 5.23 ± 0.15 0.000 2.01 ± 0.14 1

30.5 cm 6.83 ± 0.08 4.86 ± 0.14 0.000 1.97 ± 0.14 1

61 cm 7.45 ± 0.34 5.91 ± 0.22 0.001 1.75 ± 0.14 1

p-value 0.387

Appl. time e 2 sec 6.88 ± 0.21 5.63 ± 0.14 0.000 1.25 ± 0.14 3

10 sec 7.27 ± 0.26 5.36 ± 0.22 0.000 1.91 ± 0.14 2

20 sec 7.38 ± 0.16 5.01 ± 0.19 0.000 2.58 ± 0.14 1

p-value 0.000

Drip time 5 min 7.27 ± 0.18 5.54 ± 0.17 0.000 1.73 ± 0.11 2

20 min 7.08 ± 0.18 5.13 ± 0.13 0.000 2.09 ± 0.11 1

p-value 0.033

160

2 Table 7. Mean populations and reductions (log 10 CFU/cm ± SE) of Salmonella Typhimurium on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances

Factor Factor level Before a After a p-value b Reduction c

Temperature Cold 4.46 ± 0.21 3.38 ± 0.09 0.000 1.08 ± 1.89 2

Warm 4.39 ± 0.17 3.16 ± 0.08 0.000 1.23 ± 1.89 2

Hot 5.19 ± 0.17 1.11 ± 0.22 0.000 4.08 ± 1.89 1

p-value d 0.000

Drip time 5 min 5.17 ± 0.15 2.54 ± 0.20 0.000 2.63 ± 0.15 1

20 min 4.20 ± 0.09 2.56 ± 0.22 0.000 1.63 ± 0.15 2

p-value 0.000

Distance 5.1 cm 4.69 ± 0.15 2.51 ± 0.19 0.000 2.18 ± 0.15 1

30.5 cm 4.67 ± 0.19 2.59 ± 0.22 0.000 2.08 ± 0.15 1

p-value 0.632

a Populations means before and after treatment within the same row analyzed by two-sample t-test (MINITAB®, α = 0.05); sample size by factor (before, after): Temperature (12, 24), Drip time (18, 36), and Distance (18, 36) b p-value of two-sample t-test c Data subjected to GLM procedure of ANOVA with means separated by Tukey method; least squares means within the same factor and column sharing the same superscripts do not differ (MINITAB®, α = 0.05). d p-value of ANOVA

161

2 Table 8. Mean populations and reductions (log 10 CFU/cm ± SE) of Campylobacter spp. on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances

Factor Factor level Before a After a p-value b Reduction c

Temperature Cold 2.34 ± 0.34 1.20 ± 0.17 0.027 1.14 ± 0.27 2

Warm 3.88 ± 0.45 2.08 ± 0.30 0.003 1.80 ± 0.27 1,2

Hot 3.33 ± 0.29 0.60 ± 0.00 0.000 2.73 ± 0.27 1

p-value d 0.001

Drip time 5 min 2.86 ± 0.30 0.78 ± 0.10 0.000 2.08 ± 0.22 1

20 min 3.51 ± 0.39 1.81 ± 0.22 0.001 1.70 ± 0.22 1

p-value 0.238

Distance 5.1 cm 2.82 ± 0.38 1.10 ± 0.17 0.000 1.71 ± 0.22 1

30.5 cm 3.55 ± 0.30 1.48 ± 0.20 0.000 2.07 ± 0.22 1

p-value 0.268

162

2 Table 9. Mean populations and reductions (log 10 CFU/cm ± SE) of Escherichia coli O157:H7 on inoculated beef plates following a cold (15°C), warm (54°C), or hot (77°C) water wash (30 psi, 20 s application time) with varying drip times and spray distances

Factor Factor level Before a After a p-value b Reduction c

Temperature Cold 4.67 ± 0.11 3.26 ± 0.08 0.000 1.41 ± 0.14 2

Warm 3.96 ± 0.08 2.87 ± 0.10 0.000 1.10 ± 0.14 2

Hot 4.78 ± 0.12 0.98 ± 0.16 0.000 3.79 ± 0.14 1

p-value d 0.000

Drip time 5 min 4.75 ± 0.09 2.28 ± 0.20 0.000 2.47 ± 0.12 1

20 min 4.19 ± 0.11 2.45 ± 0.19 0.000 1.73 ± 0.12 2

p-value 0.000

Distance 5.1 cm 4.54 ± 0.12 2.35 ± 0.18 0.000 2.19 ± 0.12 1

30.5 cm 4.40 ± 0.12 2.39 ± 0.20 0.000 2.01 ± 0.12 1

p-value 0.287

163

Factor Factor level Beforea After a p-value b Reduction c

Temperature Cold 7.07 ± 0.16 5.69 ± 0.12 0.000 1.38 ± 0.33 2

Warm 8.47 ± 0.30 6.55 ± 0.10 0.000 1.92 ± 0.33 2

Hot 8.12 ± 0.13 4.14 ± 0.31 0.000 3.99 ± 0.33 1

p-value d 0.000

Drip time 5 min 7.98 ± 0.21 5.40 ± 0.19 0.000 2.58 ± 0.27 1

20 min 7.80 ± 0.23 5.52 ± 0.27 0.000 2.28 ± 0.27 1

p-value 0.434

Distance 5.1 cm 8.01 ± 0.25 5.25 ± 0.23 0.000 2.76 ± 0.27 1

30.5 cm 7.76 ± 0.18 5.66 ± 0.24 0.000 2.10 ± 0.27 1

p-value 0.100

164

Factor Factor level Before a After a p-value b Reduction c

Temperature Cold 4.25 ± 0.10 2.20 ± 0.10 0.000 2.06 ± 0.28 2

Warm 5.79 ± 0.40 3.37 ± 0.38 0.000 2.43 ± 0.28 2

Hot 5.36 ± 0.47 1.03 ± 0.19 0.000 4.33 ± 0.28 1

p-value d 0.000

Drip time 5 min 4.92 ± 0.23 1.79 ± 0.15 0.000 3.13 ± 0.23 1

20 min 5.35 ± 0.41 2.61 ± 0.32 0.000 2.74 ± 0.23 1

p-value 0.249

Distance 5.1 cm 5.05 ± 0.32 2.09 ± 0.23 0.000 2.95 ± 0.23 1

30.5 cm 5.22 ± 0.35 2.30 ± 0.28 0.000 2.92 ± 0.23 1

p-value 0.923

165

Factor Factor level Before a After a p-value b Reduction c

Temperature Cold 6.38± 0.15 4.38 ± 0.11 0.000 2.00 ± 0.26 2

Warm 7.36 ± 0.25 5.13 ± 0.22 0.000 2.23 ± 0.26 2

Hot 7.18 ± 0.18 2.13 ± 0.29 0.000 5.05 ± 0.26 1

p-value d 0.000

Drip time 5 min 6.90 ± 0.10 3.69 ± 0.24 0.000 3.22 ± 0.21 1

20 min 7.04 ± 0.25 4.07 ± 0.31 0.000 2.97 ± 0.21 1

p-value 0.424

Distance 5.1 cm 7.03 ± 0.18 3.88 ± 0.25 0.000 3.16 ± 0.21 1

30.5 cm 6.91 ± 0.20 3.88 ± 0.31 0.000 3.03 ± 0.21 1

p-value 0.673

166

Table 13. Mean a temperature of lean tissue surfaces of beef plates manually washed with warm (54°C) or hot (77°C) water in a benchtop washing cabinet

Water wash type Mean surface temperature (°C) b

Warm water 53.1

Hot water 73.9

a N = 5 b Self-adhesive temperature monitoring strips (110 to 180°F range, MINI Eight-Position 101- 8-110F Temp-Plates ®, Palmer Wahl Instrumentation Group) were attached to lean tissue surfaces.

167

Table 14. Mean log reductions of pathogens and hygiene indicators on inoculated beef plates following cold water (15°C) washing (first, flawed analysis of cold water washing data)

Factor Factor Salmonella E. coli Campylobacter Coliforms Generic E. coli Mesophilic

level Typhimurium a O157:H7 spp. APC

Pressure 10 psi 1.39 1 1.87 1 1.62 1 1.94 1 2.11 1 0.93 1

30 psi 1.05 2 1.15 2 1.50 1 1.74 2 1.72 2 1.47 2

Distance 5.1 cm 1.24 1,2 1.95 1 1.68 1 2.01 1 1.96 1,2 1.18 1

30.5 cm 1.35 1 1.00 3 1.41 2 1.97 1 2.08 1 0.88 2

61.0 cm 1.07 2 1.59 2 1.59 1 1.54 2 1.72 2 1.55 3

Application 2 s 0.66 1 1.18 1 1.04 1 1.25 1 1.42 1 1.22 1 time 10 s 1.33 2 1.60 2 1.57 2 1.91 2 2.02 2 1.18 1

20 s 1.67 3 1.75 3 2.07 3 2.37 3 2.32 3 1.21 1

168

Drip time 5 min 1.18 1 1.40 1 1.56 1 1.73 1 1.81 1 1.05 1

20 min 1.26 1 1.63 2 1.56 1 1.95 1 2.02 2 1.36 1

a Least squares means in column within factor sharing the same superscripts are not significant by GLM procedure of ANOVA; means separated by Fisher’s LSD (MINITAB ®, α = 0.05)

169

2 a, b Table 15. Mean populations (log 10 CFU/cm ) of bacterial populations on inoculated beef plates following cold water (15°C) washing under a variety of pressures, spray distances, application times, and drip times (second analysis of data)

Factor Factor level Salmonella Campylobacter Escherichia Mesophilic Generic Coliforms

spp. spp. coli O157:H7 APC E. coli

Pressure 10 psi 5.14 ± 0.09 1 4.98 ± 0.10 1 4.77 ± 0.04 1 7.25 ± 0.14 1 6.25 ± 0.16 1 6.66 ± 0.52 1

30 psi 4.62 ± 0.27 2 5.06 ± 0.12 2 2.94 ± 0.35 2 7.07 ± 0.11 2 6.33 ± 0.16 2 7.51 ± 0.09 2

p-value 0.002 0.000 0.000 0.000 0.000 0.009

Distance 5.1 cm (2”) 4.49 ± 0.21 1 4.70 ± 0.07 1 4.73 ± 0.05 1 7.23 ± 0.07 1 6.47 ± 0.12 1 6.87 ± 0.55 1

30.5 cm (12”) 5.12 ± 0.03 2 4.80 ± 0.08 2 2.40± 0.50 2 6.82 ± 0.06 2 5.81 ± 0.10 2 6.49 ± 0.52 1

61 cm (24”) 5.03 ± 0.38 2 5.56 ± 0.15 3 4.42 ± 0.07 3 7.42 ± 0.24 3 6.58 ± 0.28 3 7.89 ± 0.21 2

p-value 0.003 0.000 0.000 0.000 0.000 0.002

170

Application 2 sec 4.10 ± 0.36 1 4.81 ± 0.111 3.83 ± 0.37 1 6.84 ± 0.15 1 5.78 ± 0.20 1 7.47 ± 0.14 1 time 10 sec 5.22 ± 0.13 2 5.07 ± 0.15 2 3.85 ± 0.36 2 7.26 ± 0.18 2 6.40 ± 0.20 2 7.18 ± 0.48 1

20 sec 5.31 ± 0.11 2 5.17 ± 0.13 3 3.87 ± 0.36 3 7.37 ± 0.11 3 6.69 ± 0.14 3 6.61 ± 0.62 1

p-value 0.000 0.000 0.000 0.000 0.000 0.082

Drip time 5 min 4.86 ± 0.23 1 5.07 ± 0.10 1 3.84 ± 0.29 1 7.25 ± 0.12 1 6.41 ± 0.15 1 6.72 ± 0.47 1

20 min 4.90 ± 0.18 1 4.97± 0.12 2 3.86 ± 0.30 2 7.06 ± 0.13 2 6.17 ± 0.17 2 7.45 ± 0.24 2

p-value 0.791 0.000 0.000 0.000 0.000 0.025

a Data subjected to GLM procedure of ANOVA with means separated by Tukey method (95.0% confidence interval) b Means within the same factor and column sharing the same superscripts do not differ (α = 0.05).

171

CHAPTER FIVE

INVESTIGATION OF CHEMICAL RINSES ON THE ELIMINATION OF

MEAT BORNE PATHOGENS FROM INOCULATED BEEF SURFACES

172

INTRODUCTION

In 2002, very small meat plants in Pennsylvania were asked to complete a survey questionnaire to identify carcass decontamination methods used during slaughter (Chapter

Three). A small number of plants had implemented decontamination treatments such as steam vacuuming (2%) or the application of organic acids (lactic or citric, 4%), or sodium hypochlorite (6%) to carcass surfaces. Also, nearly all very small processors (96%) reported that some sort of carcass decontamination procedure is implemented immediately prior to chilling.

Besides lactic acid, citric acid, and sodium hypochlorite, the following antimicrobial compounds have been identified as candidates for implementation in very small meat establishments: acidified sodium chlorite, chlorine dioxide, acetic acid, peroxyacetic acid, and aqueous ozone. In the laboratory, the antimicrobial effectiveness of these compounds plus water washing under controlled conditions can be assessed and the best performing ones selected for use in an in-plant validation study. The antimicrobial effectiveness of all eight compounds and of water washing has been reviewed extensively in Chapter Two and will be summarized as follows.

Rinsing with antimicrobial compounds

Organic acids (e.g. acetic, lactic, and citric), chlorinated compounds or other generally recognized as safe (GRAS) compounds are usually applied to carcass surfaces after water washing. An inexpensive way for very small processors to apply antimicrobial solutions to carcasses is with a hand-held spray tank, much like the garden sprayers

173

commonly available at hardware stores or home improvement centers. Overall, these rinses can eliminate microscopic contaminants that remain after the final water wash step. More importantly, some antimicrobial rinse solutions produce an inhospitable environment for microbes that remain on carcass surfaces after water washing and for those that may come in contact with the surface further down the slaughter line. The prolonged contact of the antimicrobial either before evisceration, after evisceration, or during the refrigeration process has been demonstrated to improve both the shelf life and safety of the meat (Barkate et al.,

1993; Cutter and Siragusa, 1994; Hardin et al., 1995; Reagan et al., 1996; Davey and Smith,

1989; Dickson, 1995; and Dormedy et al., 2000).

Lactic, citric, and acetic acids

Lactic, citric and acetic acids are naturally occurring compounds in many foods

(Doores, 1993). FSIS permits the use of organic acids as solutions up to 2.5% for carcass washing prior to chilling (USDA-FSIS, 2005). Other than specifying a maximum concentration of 2.5%, FSIS does not identify a target temperature or pH range for organic acids used explicitly as processing aid (USDA-FSIS, 2005). When applied as a processing aid, within the approved limits for species and product type, there is no labeling requirement

(USDA-FSIS, 2005). Lactic, acetic, and citric acids are the organic acids that are often used as carcass rinses, with lactic and acetic acids most commonly in use (Cherrington et al.,

1991; Doores, 1993; Sheridan, 1998; Siragusa, 1995).

The complete inactivation of harmful organisms on beef carcass surfaces cannot be guaranteed, although substantial evidence exists that acid rinsing is an effective means of reducing bacterial populations on meat surfaces. For example, Castillo et al. (1998a) report a

174

4.6 to > 4.9 log CFU/cm 2 reduction in numerous bacterial populations ( E. coli O157:H7, S.

Typhimurium, generic E. coli , etc.) by spraying with 2% lactic acid (55°C, 40 psi, 11 s) alone. Acetic acid also has found practical application as a carcass decontaminant. Bell et al.,

log 10 reductions, respectively, with slight discoloration. Furthermore, Cutter and Rivera-

Betancourt (2000) observed approximately 2 log CFU/cm 2 reductions in E. coli O157:H7 and

S. Typhimurium immediately after spraying inoculated beef tissue with 2% lactic or acetic acid (125 psi, 35°C, 15 s). Hence, acid rinsing can provide reductions as a measure of control, but not complete elimination, in HACCP plans for slaughter, fabrication and other fresh meat applications (Dormedy et al., 2000).

Peroxyacetic acid

Peroxyacids, such as peroxyacetic acid, also can be used as antimicrobial rinses for carcasses and variety meats. In this class of oxidizers, two or more of the following compounds are mixed together before direct application to meats: peroxyacetic acid; octanoic acid; acetic acid; hydrogen peroxide; peroxyoctanoic acid; and 1-hydroxyethylidene-1, 1- diphosphonic acid (also known as HEDP; 21 CFR 173.370). Mixtures of peroxyacids that have been optimized for use on beef carcasses are commercially available and at a recommended temperature range of 38 to 49°C and peroxyacid concentration of 180 to 200 ppm (ECOLAB Inc., 2003). The plant employee can dilute the mixture with water to achieve

175

a peroxyacetic acid concentration ≤ 220 ppm and a hydrogen peroxide concentration of ≤ 75 ppm, the maximum acceptable level for use on beef carcasses (USDA-FSIS, 2005).

When used as a 0.02% solution, peroxyacetic acid reduced E. coli O157:H7 from beef

carcass tissue and beef short plates by 1.4 and 1.0 log CFU/cm 2, whereas water washing produced comparable reductions of 1.2 and 0.3 CFU/cm 2, in that order (Ransom et al., 2003).

Gill and Badoni (2004) also tested the effectiveness of a 0.02% peroxyacetic acid and 4% lactic acid on chilled beef carcass surfaces. Overall, the peroxyacetic acid rinse provided a

0.5 log CFU/cm 2 greater reduction in aerobic plate counts than a water rinse (temperature not reported), while 4% lactic acid provided at least a 1 log CFU/cm 2 advantage over peroxyacetic acid (Gill and Badoni, 2004). Nevertheless, the use of lactic acid at 4% concentration is not approved for carcass surfaces.

Sodium hypochlorite

Sodium hypochlorite, or household bleach, is an inexpensive, broad-spectrum sanitizer. Fluctuations in pH, temperature, organic load can vary the quantity of chlorine available for disinfecting poultry chiller water (Tsai et al., 1992). A four-year study of chiller water from one commercial poultry plant indicated that a chlorine (2 to 4 °C, pH = 7.4 to 7.8) concentration in the range of 150 to 250 ppm was necessary to consistently attain a 2-log reduction in aerobic plate counts (APC; Tsai et al., 1992). Furthermore, in samples of chiller water within the low-dose spectrum of chlorine concentrations (0 to 80 ppm), APC reductions ranged from 0.1 to 0.5 log CFU/ml (Tsai et al., 1992).

When applied to beef carcasses in a processing plant, a 0.2% sodium hypochlorite solution effected a 0.3 log CFU/cm 2 greater reduction in APC than untreated carcasses

176

(Stevenson et al., 1978). An earlier report of various sodium hypochlorite rinses (0.2 to

0.25%; 50 or 200 psi; 2, 15, or 30 s application time; 1.7, 3.4 or 6.8 L/min) on beef plates demonstrated that reduction of APC is most effective (0.6 to 0.7 log CFU) when the rinse is applied at high pressure, high flow rate, and the for longest application time studied

(Marshall et al., 1977). In both of these studies, the pH of sodium hypochlorite was adjusted to 6.5 or 6.0, respectively, with acetic acid, which converts hypochlorite to hypochlorous acid. It is thought that acidification of chlorine solutions improves antimicrobial effectiveness due to the increased dissociation of the chlorinated compound (Cords and Dychdala, 1993).

Acidified sodium chlorite

Currently, acidified sodium chlorite (ASC) is permitted for use as a processing aid to decontaminate beef carcasses, parts, and organs, at a range of 500 to 1,200 ppm (21 CFR

173.325). Preparation of ASC in a meat plant can be relatively convenient when powdered sodium chlorite is acidified with a GRAS acid (e.g., citric, acetic, or phosphoric) such that the resulting pH is 2.5 to 2.9 (21 CFR 173.325). Automated equipment that consistently generates ASC can also be installed in the meat plant. Due to the strong oxidative ability of

ASC and solution instability (i.e., volatility of chlorine dioxide), it must prepared on-site prior to use. Recently, Gill and Badoni (2004) applied 1,600 ppm ASC to chilled beef carcass

2 surfaces and recovered 0.8 log 10 CFU/cm less for aerobic plate counts than with a water rinse. No experimental control was reported and, thus, no log CFU/cm 2 reductions in plate counts were calculated (Gill and Badoni, 2004).

Acidified sodium chlorite also can be used in combination with another treatment to achieve greater lethality. For example, Castillo et al. (1999) achieved a 4.5 to 4.6 log

177

CFU/cm 2 reduction of E. coli O157:H7 and S. Typhimurium by spray washing (mean pH =

2.62, 69 kPa, 22.4 to 24.7°C, 140 ml for 10 s) beef carcass tissue with water, followed by spraying 1,200 ppm ASC activated with citric acid. When activated with phosphoric acid and applied under the same physical conditions, the log reduction in bacterial populations by

ASC was 3.8 to 3.9 log CFU/cm 2, while water washing (manual wash: 69 kPa, 1.5 L for 90 s;

followed by automated wash: 5 L applied at 1.72 MPa increasing to 2.76 MPa for 9 s) alone

afforded a 2.3 log CFU/cm 2 reduction (Castillo et al., 1999). Kemp et al. (2000) speculate

that citric acid activation of sodium chlorite may impart greater effectiveness than activation by phosphoric acid because citric acid is thought to chelate components of the cell membrane

and cause it to rupture.

Chlorine dioxide

Chlorine dioxide is another option for sanitizing red meat carcass surfaces. Used extensively in poultry chiller water, chlorine dioxide retains more oxidizing capacity than chlorine when organic matter is present (Villareal et al., 1990). Unlike many chlorine compounds currently in use, chlorine dioxide is not converted into chloramines, which are toxic to fish, or trihalomethanes, which are harmful to humans at sufficient levels (Cords and

Dychdala, 1993). Chlorine dioxide also maintains antimicrobial effectiveness over a broad pH range.

Foschino et al. (1998) examined the effectiveness of chlorine dioxide against a

nonpathogenic strain of E. coli , which was suspended in physiological saline with 0.1%

tryptone, and reported a 5 log CFU/ml reduction after exposure to 1.4 ppm solution for 30 s.

However, when used as a spray wash (520 kPa, 16°C, 10 s) to decontaminate beef carcass

178

surfaces at 0 to 20 ppm, chlorine dioxide was no more effective than water washing (Cutter and Dorsa, 1995). The abundance of organic matter (lean, fat, as well as microbes) on beef carcass surfaces may overpower the oxidizing capacity of chlorine dioxide to inactivate microorganisms as dissociated chlorine loses efficacy once bound to organic matter

(Siragusa, 1995).

Aqueous ozone

Ozone has a long history of use as an antimicrobial for drinking water, given that the first drinking water treatment plant to rely on ozone for disinfection was built in The

Netherlands in 1893 (Brink et al., 1991). The gentle bubbling of ozone gas into water generates aqueous ozone, which can then be used to disinfect foods and food handling surfaces. Ozone is more soluble and less likely to decompose in cold water than hot water

(Kim et al., 1999). Essentially, the mode of action is the release of a third oxygen atom from an unstable ozone molecule. The singular oxygen atoms make contact with organic and inorganic debris and rapidly oxidize them.

Reagan and associates (1996) tested the efficacy of aqueous ozone, hydrogen peroxide water washing, and knife trimming on experimentally contaminated beef carcasses.

Ozone (0.3 to 2.3 ppm), hydrogen peroxide (5%) and water washing (74 to 88°C) all resulted in an approximate 1 log CFU/cm 2 reduction of microbial populations (Reagan et al., 1996).

The results of a more recent study agree with those of Reagan and associates (1996). Excised carcass surfaces were treated with 95 ppm aqueous ozone for 30 s or washed with water

(28°C) for 9 s with very little difference in effectiveness (Castillo et al., 2003). The effectiveness of ozone as a carcass decontaminant may be stifled by the overwhelming

179

presence of organic material on a carcass surface, since ozone can oxidize components of bacterial cell as favorably as lean or adipose tissue (Castillo et al., 2003). Güzel-Seydim et al.

(2004a) demonstrated that suspensions of locust bean gum, sodium caseinate and whipping cream actually protected E. coli and Staphylococcus aureus inocula against ozonation. Yang and Chen (1979) also demonstrated that ozone (2.5%) in the presence of egg albumin (0.05 or 1.0%) reduced antimicrobial effectiveness by approximately 4 log CFU/ml when compared to the control. On the other hand, its use as a food sanitizer may be desirable to some processors because the rapid decomposition of ozone leaves no residue on foods

(Güzel-Seydim et al., 2004b).

Types of interventions

Besides antimicrobial effectiveness, it is important to consider costs and worker safety, which may influence the feasible implementation of these decontamination methods in a very small meat plant. Chlorinated compounds are generally inexpensive and widely used in the food industry to sanitize food surfaces and equipment (Cords and Dychdala,

1993). Unstable or volatile compounds, such as aqueous ozone and chlorine dioxide, can be mechanically generated in the plant for immediate use. The procurement of generating equipment usually requires a sizeable capital investment and space in the plant for storage or permanent installation. Reagents and consumable accessories (filters, hoses, etc.) that may be needed to continuously operate these systems also could incur additional costs to the processor. Also, processors who purchase chemical generators (e.g., ozone or chlorine dioxide generators) may want to consider having a secondary method of carcass decontamination readily available in the event of mechanical failure. Freshly slaughtered

180

carcasses need to be treated with an antimicrobial intervention whether or not a chemical generator is working properly.

Worker safety also plays a role in the selection of an antimicrobial carcass treatment.

Several of the chemicals listed above must be used in a well-ventilated area to prevent or minimize irritation to the eyes and respiratory tract of employees during use. In very small meat plants, it may be too expensive or physically impractical to install adequate ventilation for the safe use of chlorinated or acid rinses. Moreover, proper personal protective equipment

(gloves, goggles, rainsuit or apron) should be provided to workers who must handle concentrated chemicals to prepare carcass rinse solutions. In cramped spaces, a worker may need to wear a respirator when handling peroxyacetic acid or chlorine dioxide. When rinsing carcasses, spraying equipment often produces spray mist, which should be contained, if possible. The spray mist can carry aerosolized chemicals through the air and can be deposited on the skin or inadvertently inhaled by workers, who are not properly outfitted. Also, is it prudent to have an emergency plan in case of a chemical spill or worker injury.

181

RESEARCH OBJECTIVES

In this study, a series of rinsing experiments is performed under laboratory controlled conditions to investigate the bactericidal effectiveness of eight antimicrobial compounds. The chemical rinse that provides the greatest reductions in bacterial populations combined with other feasibility attributes (cost, worker safety, space) is selected for use in combination treatments. Furthermore, antimicrobial rinses are used at concentrations that observe or exceed, when possible, maximum usage limits set by the Food Safety and Inspection Service

(USDA-FSIS, 2005). An auxiliary purpose of this study is to determine whether residues of selected antimicrobial rinses exceed maximum usage limits when used at an elevated concentration as no studies have investigated this concept yet.

182