Use of Hot Water for Beef Carcass Decontamination

Transcription

1 19 Journal of Food Protection, Vol. 61, No.1, 1998, Pages Copyright, International Association of Milk. Food and Environmental Sanitarians Use of Hot Water for Beef Carcass Decontamination A. CASTILLO, L. M. LUCIA, K. J. GOODSON, J. W. SAVELL, AND G. R. ACUFF* Institute of Food Science and Engineering, Center for Food Safety, Department of Animal Science, Texas A&M University, College Station, Texas , USA MS 97-15: Received 31 January 1997/Accepted 4 April 1997 ABSTRACT Hot water treatment of beef carcass surfaces for reduction of Escherichia coli 0157:H7, Salmonella typhimurium, and various indicator organisms was studied using a model carcass spray cabinet. Paired hot carcass surface regions with different external fat characteristics (inside round, outside round, brisket, flank, and clod) were removed from carcasses immediately after the slaughter and dressing process. All cuts were inoculated with bovine feces containing 10 6 /g each of rifampicin-resistant E. coli 0157:H7 and S. typhimurium, or with uninoculated bovine feces. Surfaces then were exposed to a carcass water wash or a water wash followed by hot water spray (95 C). Counts of rifampicin-resistant Salmonella and E. coli or aerobic plate count (APC) and coliform counts were conducted before and after each treatment. All treatments significantly reduced levels of pathogens from the initial inoculation level of 5.0 logro CFU/cm 2. Treatments including hot water sprays provided mean reductions of initial counts for E. coli 0157:H7 and S. typhimurium of 3.7 and 3.8 log, APC reductions of 2.9 log, and coliform and thermotolerant coliform count reductions of 3.3 log. The efficacy of hot water treatments was affected by the carcass surface region, but not by delaying the treatment (30 min) after contaminating the surface. Verification of efficacy of hot water interventions used as critical control points in a hazard analysis critical control point (HACCP) system may be possible using coliform counts. Various reports have indicated that application of hot water treatments can be an effective measure for reducing microbiological contamination of beef carcasses. In an early report on hot water decontamination Patterson (13) reported that beef carcasses treated with a steam and hot water spray (80 to 96 C) for 2 min contained significantly lower bacterial numbers than untreated carcasses. Some discoloration on the carcass surface occurred initially, but normal color returned after cooling for 24 h. A volume of 18.9 liters of water was sprayed on each carcass; however, the temperature increase obtained at the carcass surface in this study was not described. According to Smith and Graham (15), pouring hot water (80 C) on beef and lamb samples for los destroyed more than 99% of Escherichia coli and Salmonella inoculated at levels of 6.5 loglcm 2 The surface tissues of the beef and mutton were not permanently discolored by this treatment, but the authors reported discoloration when water at 90 C for 120 s was used. In a laboratory evaluation of a hot water cabinet, Davey and Smith (4) obtained E. coli reductions of 2.98 log for artificially contaminated beef carcass sides treated with hot water raising the carcass surface temperature to 83.soC for 20 s. Kelly et al. (12) reported that lamb carcasses sprayed with hot water at temperatures above 80 C caused significant decreases (>1.0 log) in aerobic plate counts (APC). In more recent studies where hot water treatments were evaluated, Dorsa et al. (7) and Gorman et al. (8) obtained reductions in coliform or E. coli counts of approximately 3.0 log. Recently, in our laboratory, Barkate et al. (1) sprayed * Author for correspondence. Tel: ; Fax: ; gacuff@tamu.edu areas of hot beef carcass surfaces using 95 C water with the objective of raising the carcass surface temperature to 82 C for approximately 10 s. Significant reductions in bacterial contamination on the surface of carcasses were obtained. Slight discoloration of carcasses occurred immediately after spraying with hot water, but the discoloration was temporary, and treated and untreated areas of carcasses could not be distinguished after 24 h. Problems encountered in applying hot water included obtaining a water spray that would adequately raise the surface temperature of the carcass to a bactericidal level. The volume of the spray and the size of the water droplets were found to have a profound effect on the temperature of the water after leaving the spray nozzle and before contacting the carcass surface. In this study a type of nozzle which addressed the limitations reported by Barkate et al. (1) was used to spray hot water onto different hot carcass surface regions. The objectives of this study included providing a model for determining the effectiveness of hot water treatment for reduction of enteric pathogens (E. coli 0157:H7 and S. typhimurium) originating from fecal contamination of hot carcass surfaces and examining the possibility of utilizing traditional indicator organisms for verification of the efficacy of hot water sprays as a critical control point in a HACCP system. MATERIALS AND METHODS Media. A selective, differential medium (lactose-sulfitephenol red-rifampicin agar, LSPR) was developed to simultaneously enumerate rifampicin-resistant marker pathogens inoculated onto carcass surfaces. The medium consisted of the following ingredients per liter: tryptic soy agar (TSA, Difco, Detroit, MI) 40 g, yeast extract (Difco) 3 g, beef extract (Difco) 3 g, lactose (EMI

2 20 CASTILLO, LUCIA, GOODSON, SAVELL, AND ACUFF J. Food Prot., Vol. 61, No.1 Industries, Inc., Gibbstown, NJ) 5 g, sodium sulfite (MCB Reagents, Cincinnati, OH) 2.5 g, ferrous sulfate (MCB Reagents) 0.3 g, phenol red (Fisher Scientific, Fair Lawn, NJ) 25 mg, cycloheximide (Sigma Chemical, St. Louis, MO) 0.1 g, and rifampicin (Sigma Chemical) 0.1 g. Phenol red was dissolved in 2 ml 0.1 N NaOH before adding to the medium. The medium without rifampicin was autoclaved at 121 C for 15 min and cooled to 50 C. Rifampicin was dissolved in 5 ml methanol, filter-sterilized, and added to the sterile medium prior to pouring into petri plates. Prepoured plates were dried at 25 C overnight before use. The Rifampicin-resistant E. coli produced yellow colonies on the medium, whereas the rifampicin-resistant Salmonella developed colonies with a black center surrounded by a pink halo. APC, coliform, and thermotolerant coliform counts were conducted using APC or E. coli Petrifilm plates OM, St. Paul, MN). Bacterial cultures. Rifampicin-resistant mutants derived from parent strains of S. typhimurium ATCC and E. coli 0157:H7 (from ground beef implicated in an outbreak in Washington, 1993; provided by P. 1. Tarr, Children's Hospital and Medical Center, Seattle, WA) were used to inoculate beef carcass surfaces to be treated in this study. Growth curves, heat resistance, and acid sensitivity of the mutant strains were determined to be virtually indistinguishable from the parent strains (9, 10). RifaJIlpicinresistant cultures were produced according to the method published by Kaspar and Tamplin (11). The selected mutants were maintained on TSA slants at 4 C. Before inoculation into feces, these cultures were transferred into tryptic soy broth (TSB, Difco) and incubated at 35 C for 12 to 14 h. Rifampicin resistance was then confirmed by streaking TSB cultures onto LSPR plates and incubating at 35 C for 24 h. Characteristic colonies were inoculated into TSB, incubated at 35 C for 12 h, and then used for the fecal inoculum as described below. Inoculum preparation. Each slaughter day, feces were collected from a randomly selected cow at the Texas A&M University Dairy Center immediately after defecation. Feces were hand-kneaded for I min in a stomacher bag and then dispensed in 10-g portions to individual stomacher bags. Twelve-hour cultures of rifampicin-resistant S. typhimurium and rifampicin-resistant E. coli 0157:H7 were transferred together (0.1 ml of each) into tubes containing 9 ml of sterile 0.1 % peptone water and then mixed using a vortexing mixer. The contents of each tube were poured into a stomacher bag containing 10 g of feces and then hand-kneaded for I min to mix. The prepared inoculum (containing each pathogen at a concentration of ca. 6 log CFU/g) was used within 4 h after preparation and was kept at room temperature (23 to 24 C) during the experiments. An identical set of bags containing feces not inoculated with pathogens was prepared with 9 ml sterile 0.1 % peptone water and mixed in the same manner as the inoculated bags. The uninoculated feces were used to evaluate the effect of treatments on traditional indicator organisms. Carcass selection and inoculation. Fed steers or heifers typical ofthose entering the U.S. meat supply were selected for use in the study. Each slaughter day, three cattle were transported to the Texas A&M University Rosenthal Meat Science and Technology Center (RMSTC), slaughtered, and dressed in the university abattoir following USDA-FSIS-regulated commercial procedures. Paired beef inside rounds, outside rounds, briskets, flanks, and clods were separated from the remainder of the carcass just subsequent to carcass splitting. These particular carcass surface regions were selected for use in this study since they are located in areas where fecal contamination is likely to occur (3). In addition, it was theorized that differences in fat surface characteristics from these areas on the carcass might affect contamination removal. Carcasses were not washed or decontaminated in any manner before removing carcass surface regions for use in this study. The cuts were transported from the slaughter floor in insulated containers to an adjacent isolated treatment area where they were contaminated with the prepared feces and subjected to different decontamination treatments. Each carcass surface region was contaminated by spreading either inoculated or uninoculated feces over a 400-cm 2 area using a sterile stainless steel spatula. The 400-cm 2 area was delineated on the carcass surface region using purple edible ink. Mean counts of S. typhimurium and E. coli 0157:H7 on the carcass surface regions immediately after inoculation were 5.0 and 5.2 log CFU!cm 2, respectively. Mean initial counts on carcass surface regions after contamination with uninoculated feces were 6.1 log CFU/cm 2 for APC and 5.3 log CFU/cm 2 for both coliforms and thermotolerant coliforms. Application of treatments. All feces-contaminated (with or without marker pathogens) carcass surface regions were exposed to one of the four following treatments: (I) warm carcass wash 5 min after contamination with feces, (2) warm carcass wash 20 to 30 min after contamination with feces, (3) warm carcass wash followed by treatment with hot water 5 min after contamination with feces, or (4) warm carcass wash followed by treatment with hot water 20 to 30 min after contamination with feces. The warm carcass wash consisted of a two-step procedure which included an initial low-pressure manual wash to remove gross fecal matter followed by a high pressure wash in an automated spray cabinet (Chad Co., Lenexa, KS). In the initial hand wash, the carcass surface region to be treated was hung in the cabinet in the same orientation as it would be positioned on an intact carcass and 1.5 liters of water (ca. 25 C) was sprayed at 10 psi (69 kpa) for 90 s using a hand-held, noncorrosive, polyethylene compressed-air sprayer (10.56 I, Universal-Gerwin, Saranac, MI). An automated high-pressure wash was then applied consisting of spraying 5 liters potable water at 35 C for 9 s, starting at an initial pressure of 250 psi (1.72 MPa) for 4 s and gradually increasing to 400 psi (2.76 MPa) within 2 s, maintaining this pressure for 3 s to complete a total treatment time of 9 s. A description of the automated cabinet and spraying system used in this study has been provided (9). Hot water treatment was applied at 24 psi (165 kpa) for 5 s using a flat spray nozzle (HY4USS5050, Spraying Systems, Wheaton, IL) from a distance of 12.5 em. This spray nozzle delivered a 46 continuous fan of water from an orifice with an equivalent diameter of 43.7 mm, providing 14.4 liters/min at 24 psi (165 kpa). Water was heated to 97 C in a constant temperature water bath (Magni Whirl, Blue M, Blue Island, IL) and pumped (B-472, Magnetek, St. Louis, MO) through insulated tubing (1.5-m flexible Nalgene 0.93-cm-diameter reinforced PVC tube) to the nozzle. In preliminary experiments, when the temperature of the water at the nozzle was 95 C, the carcass surface temperature during the treatment would be raised to 82 to 85 C in I to 2 s. Additional temperature measurements were obtained during the study to verify the preliminary data. Temperatures were measured with a Tegam 871 digital thermometer (ElL Instruments, Inc. Sparks, MD) connected to a type K thermocouple. Sampling and microbiological analysis. Three lo-cm 2 samples were excised from randomly selected areas of each carcass surface region to measure the background microflora (30 cm 2 total area) before inoculation. In addition, three lo-cm 2 samples were randomly excised from the 400-cm 2 inoculated area after inoculation and again after applying the different treatments. To measure the spreading of the inoculated organisms to areas outside the inoculated area as a result of the treatment (runoff), and their

3 J. Food Prot., Vol. 61, No. 1 BEEF DECONTAMINATION BY HOT WATER 21 reduction by the treatments, three 1O-cm 2 samples were obtained, again at random, from areas closely surrounding each inoculated area before and after treatments. All microbiological samples were obtained using a sterile stainless steel 3.6-cm-diameter borer to initially cut an outline (approximately 2 to 3 mm deep) for a 1O-cm 2 surface area, followed by excising the surface sample (taking 2 mm or less of underlying tissue) from the area using a sterile scalpel and forceps. Each group of three samples were compo sited in a stomacher bag to which 100 ml of sterile 0.1 % peptone water was added before examination. Counts of rifampicin-resistant E. coli OIS7:H7 and S. typhimurium were determined by plating appropriate dilutions of the compo sited samples onto plates of the selective, differential agar described above and incubating at 37 C for 24 h. For the samples obtained from carcass surface regions contaminated with uninoculated feces, counts of coliforms, thermotolerant coliforms, and APCs were determined by plating appropriate dilutions of the compo sited sample onto E. coli Petrifilm plates incubated at 3SoC and 44.SoC for 24 h (total and thermotolerant coliforms, respectively) and APC Petri film plates incubated at 37 C for 48 h. Visual evaluation. Treated carcass surface regions were evaluated visually following a previously established procedure (9), grading the remaining visible fecal contamination from 0 to S, where a score of 0 indicated no fecal contamination and a score of S indicated total retention of fecal contamination. Statistical analysis. Count data were transformed logarithmically before comparison of means by analysis of variance or general linear model procedures. When significant differences were observed (P < O.OS), separation of means was accomplished by using Duncan's multiple range test (14). Means from other sets of data were compared by t test. RESULTS AND DISCUSSION All treatments significantly reduced levels of pathogens from fecal contamination of carcass surface regions compared to the initial inoculation level (ca. 5.0 log CFU/cm 2 ). In all but one instance, no differences were observed (P > 0.05) in the effects of immediate and delayed (up to 30 min) treatment on bacterial reduction (data not shown). For all carcass surface regions, reductions of 1.9 to 2.7 log in counts of the marker pathogens were obtained by the carcass wash and additional reductions of 0.7 to 2.2 log were observed after spraying with hot water (Table 1). Overall, treatments which included hot water sprays reduced the numbers of S. typhimurium and E. coli 0157:H7 by 2.7 to 4.3 log on various carcass surface regions, and these reductions were significantly greater than those obtained by water wash alone. Reductions of marker pathogens by hot water treatment were consistent on all carcass surface regions with the exception of the inside round region. This region displayed consistently smaller reduction of pathogens, including a significantly smaller reduction (P < 0.05) of S. typhimurium (Table 1). Difficulty encountered in decontamination of the inside round area also was noted by Hardin et a1. (9) when treating carcass surface regions with organic acids. These investigators reported that the inside round region contained a substantial amount of exposed lean tissue and a pronounced collar of fat at the edge of the lean, possibly allowing for fecal material (and bacteria) to become imbedded in the juncture of fat and lean and between muscle bundles of the lean surface. After cleaning the inoculated fecal contamination from carcass surface regions using water only, recovered counts of S. typhimurium and E. coli 0157:H7 were consistently in the range of 0.7 to 1.8 log/cm 2 on surfaces adjacent to but outside the 400-cm 2 inoculation area (Table 2). These data confirmed that some spreading of bacterial contamination is probable when areas of visible fecal contamination are washed with water. However, after treatment of the surfaces with hot water, counts of both pathogens were significantly reduced to levels very close to, and often below, the minimum detection level of 0.5 log/cm 2 It is apparent, therefore, that a water wash of fecal contamination on TABLE I. Log reductions in populations of S. typhimurium and E. coli 0157:H7 recovered from within 400-cm 2 contaminated areas of carcass surface regions as affected by treatment and type of surface Log reduction a Microorganism S. typhimurium E. coli OIS7:H7 Treatment b (I) (0) (B) (F) (C) means C Water wash 2.0A d 2.6A 2.3A 2.0A 1.9A OBFIC Water wash + hot water 2.7B 4.3B 3.8B 3.9B 4.1B OCFBI Difference e I.S Water wash 2.1A 2.7A 1.7A 1.9A 2.0A OICFB Water wash + hot water 2.9B 4.0B 3.9B 3.8B 4.1B COBFI Difference' a Log reduction = (log CFU/cm 2 before treatment) - (log CFU/cm 2 after treatment). Log CFU/cm 2 before treatment for S. typhimurium was S.Oand for E. coli OIS7:H7 was S.2. b Water: I.S-liter hand wash (90 s, 10 psi) followed by S-liter automated cabinet wash (9 s, 2S0 to 400 psi), 3S C. Hot water: 9SoC water at 24 psi for S s using a flat spray nozzle (HV4USSSOSO,Spraying Systems) from a distance of 12.S cm. C Means within rows underlined by a common line are not significantly different (P > O.OS). d Means in columns for each pathogen with same letter (A or B) are not significantly different (P > O.OS)., Difference = (log reduction by water wash + hot water) - (log reduction by water wash).

4 22 CASTILLO, LUCIA, GOODSON, SAVELL, AND ACUFF J. Food Prot., VoL 61, No.1 TABLE 2. Mean populations (log CFU/cm 2 ) ofs. typhimurium and E. coli 0157:H7 recoveredfrom outside 400-cm 2 contaminated areas of carcass surface regions as affected by treatment and type of surface Mean population (log CFU/cm 2 ) Microorganism S. typhimurium E. coli 0l57:H7 Treatment" (I) (0) (B) (F) (C) means b Water wash 1.4A C 0.7A 1.4A 1.5A 1.5A CFBIO (O)d (50) (0) (0) (0) Water wash + hot water <0.5B <0.5A 0.5B 0.5B LOA CFBOI (100) (83) (67) (50) (50) Water wash 1.6A LOA 1.8A 1.8A 1.7A FBCIO (0) (17) (0) (0) (17) Water wash + hot water <0.5B <0.5B 0.6B 0.7B 0.6B FCBIO (67) (67) (67) (33) (33) a Water: 1.5-liter hand wash (90 s, 10 psi) followed by 5-liter automated cabinet wash (9 s, 250 to 400 psi), 35 C. Hot water: 95 C water at 24 psi for 5 s using a flat spray nozzle (HlJ4USS5050, Spraying Systems) from a distance of 12.5 em. b Means within rows underlined by a common line are not significantly different (P > 0.05). C Means in columns for each pathogen with same letter (A or B) are not significantly different (P> 0.05). d Values in parentheses denote the percentage of samples below the minimum detection level of 0.5 log CFU/cm 2 carcass surfaces may spread any pathogens present over the surface of the carcass; however, sanitizing with hot water is capable of eliminating most of that contamination, even when initial levels were as high as those used in this study (5.0 log/cm 2 ). Similar results also were reported by Hardin et al. (9) when sanitizing washed carcass surfaces with organic acids. No differences in counts after water wash or hot water treatments were observed for the different carcass surface regions studied. As with the pathogens, all treatments reduced levels of indicator organisms from fecal contamination of carcass surface regions (Table 3). No significant differences were observed in the effects of the treatments on the reductions of individual types of indicators or in the effects of immediate versus delayed (up to 30 min) treatment on the reductions in APC, total coliforms, or thermotolerant coliforms. A reduction of 1.3 to 2.3 log in counts of the indicators was obtained by the water wash and an additional reduction of 0.5 to 2.3 log was observed after spraying with hot water. Overall, treatments which included hot water sprays reduced indicator organisms by 2.3 to 4.0 log (APC by 2.3 to 3.4 log; coliforms by 2.6 to 4.0 log; and thermotolerant coliforms by TABLE 3. Log reductions in APC, calif arms and thermotolerant calif arms recovered from within 400-cm 2 contaminated areas of carcass surface regions as affected by treatment and type of surface Log reduction" Microorganism APC Total coliforms Thermotolerant coliforms Treatment b (I) (0) (B) (F) (C) means c Water wash 1.8A d 1.8A 1.3A 2.lA 1.8A FCIOB Water wash + hot water 2.3A 3.4B 2.9B 3.3B 2.4B OFBCI Difference' Water wash 2.0A 1.6A 1.4A 2.2A 1.8A FlCOB Water wash + hot water 2.6A 3.8B 3.4B 4.0B 2.6A FOBCI Difference' Water wash 2.0A 1.6A 1.3A 2.3A 1.9A FICOB Water wash + hot water 2.7A 3.9B 3.2B 4.0B 2.7A FOBCI Difference e a Log reduction = (log CFU/cm 2 before treatment) - (log CFU/cm 2 after treatment). Log CFU/cm 2 before treatment for APC was 6.1, and 5.3 for both total coliforms and thermotolerant coliforms. b Water: 1.5-liter hand wash (90 s, 10 psi) followed by 5-liter automated cabinet wash (9 s, 250 to 400 psi), 35 C. Hot water: 95 C water at. 24 psi for 5 s using a flat spray nozzle (HlJ4USS5050, Spraying Systems) from a distance of 12.5 em. e Means within rows underlined by a common line are not significantly different (P > 0.05). d Means in columns for each indicator with same letter (A or B) are not significantly different (P > 0.05). e Difference = (log reduction by water wash + hot water) - (log reduction by water wash).

5 J. Food Prot., Vol. 61, No. I BEEF DECONTAMINATION BY HOT WATER 23 TABLE 4. Mean APC, coliform, and thermotolerant coliform counts (log CFU/cm 2 ) recoveredfrom outside 400-cm 2 contaminated areas of carcass suiface regions as affected by treatment and type of suiface Mean count (log CFU/cm 2 ) Microorganism Treatment" (I) (0) (B) (F) (C) means b APC Water wash 3.6A c 3.0A 3.2A 3.1A 3.6A ICBFO (O)d (0) (0) (0) (0) -- Water wash + hot water 2.5B 1.5B 1.9B 2.8A 2.2B FICBO (0) (17) (0) (0) (0) Coliforms Water wash 2.0A l.la 1.6A 1.8A 2.1A CIFBO (0) (17) (0) (0) (0) Water wash + hot water 1.2B <O.5B 0.5B 0.8B 0.8B ICFBO (0) (100) (50) (33) (50) Thermotolerant coliforms Water wash 1.8A 0.9A 1.7A 1.6A 2.0A CIBFO (0) (50) (0) (0) (0) Water wash + hot water 0.9A <O.5A <O.5B 0.8B l.lb CIFBO (33) (100) (83) (33) (33) "Water: 1.5-liter hand wash (90 s, 10 psi) followed by 5-liter automated cabinet wash (9 s, 250 to 400 psi), 35 C. Hot water: 95 C water at 24 psi for 5 s using a flat spray nozzle (HY4USS5050, Spraying Systems) from a distance of 12.5 em. b Means within rows underlined by a common line are not significantly different (P > 0.05). c Means in columns for each indicator with same letter (A or B) are not significantly different (P > 0.05). d Values in parentheses denote the percentage of samples below the minimum detection level of 0.5 log CFU/cm to 4.0 log) on various carcass surface regions. As observed with the inoculated pathogens, consistently smaller reductions in indicator organisms were found on the inside round region. However, smaller reductions were seen for all indicator organisms in the clod region as well. The surface of the clod is similar to that of the inside round, where the lean surface may allow for bacteria to attach or penetrate into the crevices of the muscle surface. In a pattern similar to that seen for the inoculated pathogens, APCs, coliforms, and thermotolerant coliforms were recovered consistently from surface areas adjacent to but outside the 400-cm 2 inoculation area after cleaning the fecal contamination from carcass surface regions using water only (Table 4). Following treatment of the surfaces with hot water, log reductions in APCs ranged from 0.3 to 1.5 for the different carcass surface regions; however, counts ranging from 1.5 to 2.8 log CFU/cm 2 remained on the carcass surface regions. Counts of both coliforms and thermotolerant coliforms were reduced, regardless of the type of carcass surface region studied, to levels close to or below the minimum detection level (0.5 log/cm 2 ) after treatment with hot water. In most instances, total and thermotolerant coliform counts were significantly smaller than those obtained after water wash alone. With the exception of two samples from the outside round region, no sample from the carcass surface regions treated only by water wash had counts of coliforms or thermotolerant coliforms below the minimum detection level of 0.5 log/ cm 2, whereas about 50% of the carcass surface regions had counts of these indicators below the minimum detection level after applying the hot water treatment. These data confirmed that a similar pattern of spreading bacterial contamination also occurs with the indicator organisms when areas of visible fecal contamination are washed with water. The pattern observed for recovery of coliforms and thermotolerant coliforms is consistent with that demonstrated with both S. typhimurium and E. coli 0157:H7. It is likely, therefore, that critical control point (CCP) verification of the ability of a hot water carcass treatment CCP to reduce these pathogens could be accomplished by obtaining coliform counts before and after treatment. Log reductions of coliforms (35 or 44.5 C) by application of hot water to carcass surfaces would likely approximate the expected log reduction in S. typhimurium or E. coli 0157:H7, if present. In general, the visual scores established to grade the remaining visible fecal contamination on the carcass surface regions after the treatments showed no significant differences among the various treatments, as well as no effect of the carcass surface region on the final score (Table 5). Therefore, all treatments were equally successful in that no treatment was demonstrated to have an advantage in producing carcass surfaces with an absence of visible feces. Hot water treatments effectively reduced enteric pathogens on beef carcass surfaces. The reductions in pathogen numbers obtained in this study were greater than the reductions reported by Davey and Smith (4) for E. coli on carcass sides that were treated in a hot water cabinet. No differences in log reductions were observed when the treatment was delayed after contamination, even though delaying the treatment would supposedly allow for bacteria to attach to the meat surfaces. Butler et al. (2) reported that attachment of E. coli and other similar microorganisms to the surface of different types of meat increased during the first 1 to 20 min of contact with the attachment surface, with little further attachment after this time. Bacterial attachment is still a process that is not fully understood. Dickson and

6 24 CASTILLO, LUCIA, GOODSON, SAVELL, AND ACUFF J. Food Prot., Vol. 61, No. I TABLE 5. Mean visual scores of carcass regions contaminated with pathogen-inoculated or uninoculatedfecal material Mean visual score" Carcass region b (n = 6) Order of Treatment (treatment no.) IR (I) OR (0) BK(B) FL(F) CL(C) means C Delayed water wash (1) IBOCF Delayed water wash + hot water spray (2) FBCIO Immediate water wash (3) FIB CO Immediate water wash + hot water spray (4) FBCIO Order of means d "The visual score ranged from 0 (no visible fecal contamination remaining after treatment) to 5 (total contamination remaining after treatment). b IR: Inside round; OR: outside round; BK: brisket; FL: flank; CL: clod. e Mean values within a row underscored by a common line do not differ significantly (P > 0.05). d Treatment numbers within the same carcass region underscored by a common line do not differ significantly (P > 0.05). Anderson (5) described bacterial attachment on meat surfaces as consisting of an initial reversible attachment followed by a permanent attachment to the surface, resulting in entrapment of the cells in the connective tissue fibers. Therefore, some protection against decontamination could be available to bacteria attached to meat surfaces. However, results of this study indicate that hot water treatments may be effective for reducing attached or unattached bacteria. The inoculation menstruum has been shown to affect the attachment rate as well. Dickson and MacNeil (6) reported that S. typhimurium and Listeria monocytogenes attached to beef carcass surfaces at a higher rate when the inoculum had been diluted in phosphate buffer compared to cow manure. Results of the present study also show that an automated carcass water wash alone is not effective in eliminating fecal contamination of beef carcass surfaces. Removal of any visible fecal material from beef carcass surfaces during slaughter and dressing is required by USDA-FSIS regulations. Currently, this removal of contamination is accomplished by steam-vacuuming the affected area or trimming the affected surface from the carcass. However, bacteria of fecal origin are not necessarily confined to areas of visible fecal material contamination, and treating only the visibly affected areas has not been shown to have great benefit over whole-carcass wash procedures in reducing the possible bacterial contamination. Therefore, a water wash may be an important component of the overall contamination reduction process in carcass processing. Another concern associated with washing procedures for removal of feces include spreading of contamination to previously uncontaminated areas through carriage of bacteria in liquid runoff. In this study it was evident that the water wash process dispersed the microorganisms to areas outside the 400-cm 2 inoculated area. This problem was contained by treatment with hot water. The reduction in bacterial counts by hot water sprays was obvious; however, these reductions may have been due to a combined effect of washing and heat inactivation of the microorganisms. Hot water decontamination of carcasses could have major advantages over the use of chemicals for decontamination, including near elimination of pathogenic vegetative bacteria such as Salmonella and E. coli OI57:H7. In addition, the use of a heat treatment in carcass processing might allow for greater use of simple, less expensive laboratory tests for bacterial indicators, such as coliforms, to verify bacterial control at CCPs in slaughter HACCP programs. ACKNOWLEDGMENTS This is a technical article from the Texas Agricultural Experiment Station. The authors wish to acknowledge the financial support of the Beef Industry Council of the National Live Stock and Meat Board. Appreciation is extended to Chad Co., for building and supplying the decontamination cabinet. REFERENCES 1. Barkate, M. L., G. R. Acuff, L. M. Lucia, and D. S. Hale Hot water decontamination of beef carcasses for reduction of initial bacterial numbers. Meat Sci. 35: Butler, J. L., J. C. Stewart, C. Vanderzant, Z. L. Carpenter, and G. C. Smith Attachment of microorganisms to pork skin and surfaces of beef and lamb carcasses. J. Food Prot. 42: Charlebois, R., R. Trudel, and S. Messier Surface contamination of beef carcasses by fecal coliforms. J. Food Prot. 54: Davey, K. R., and M. G. Smith A laboratory evaluation of a novel hot water cabinet for the decontamination of sides of beef. Int. J. Food Sci. Technol. 24: Dickson, J. S., and M. E. Anderson Microbiological decontamination of food animal carcasses by washing and sanitizing systems. A review. J. Food Prot. 55: Dickson, J. S., and M. D. MacNeil Contamination of beef tissue surfaces by cattle manure inoculated with Salmonella typhimurium and Listeria monocytogenes. J. Food Prot. 54: Dorsa, W. J., C. N. Cutter, G. R. Siragusa, and M. Koohrnaraie Microbial decontamination of beef and sheep carcasses by steam, hot water spray washes, and a stearn-vacuum sanitizer. J. Food Prot. 59: Gorman, B. M., J. N. Sofos, J. B. Morgan, G. R. Schmidt, and G. C. Smith Evaluation of hand-trimming, various sanitizing agents, and hot water spray-washing as decontamination interventions for beef brisket adipose tissue. J. Food Prot. 58: Hardin, M. D., G. R. Acuff, L. M. Lucia, J. S. Oman, and J. W. Savell Comparison of methods for contamination removal from beef carcass surfaces. J. Food Prot. 58: Jackson, T. J., M. D. Hardin, and G. R. Acuff Heat resistance of Escherichia coli 0157:H7 in a nutrient medium and in ground beef patties as influenced by storage and holding temperatures. J. Food Prot. 59: II. Kaspar, C. W., and M. L. Tamplin Effects of temperature and salinity on the survival of Vibrio vulnificus in seawater and shellfish. Appl. Environ. Microbiol. 59:

7 J. Food Prot., Vol. 61, No.1 BEEF DECONTAMINATION BY HOT WATER Kelly, C.A., J. F. Dempster, and A. J. McLoughlin The effect of temperature, pressure and chlorine concentration of spray washing water on numbers of bacteria on lamb carcases. J. Appl. Bacteriol. 51: Patterson, J. T Hygiene in meat processing plants 4. Hot-water washing of carcasses. Rec. Agric. Res. Minist. Agric. N.!. 18: SAS Institute User's guide: statistics. SAS Institute Inc., Cary, NC. 15. Smith, M. G., and A. Graham Destruction of Escherichia coli and salmonellae on mutton carcasses by treatment with hot water. Meat Sci. 2: