- Email: [email protected]
Decontamination of carcasses by vacuum-hot water cleaning and steam pasteurizing during routine operations at a beef packing plant
PII:
SO309-1740(97)00058-2
Meat Science, Vol. 41, No. 3/4, 267-276, 1997 Q 1997 Elsevier Science Ltd All rights reserved. Printed in Great Britain 0309-1740/97 $17.00+0.00
ELSEVIER
Decontamination of Carcasses by Vacuum-hot Water Cleaning and Steam Pasteurizing during Routine Operations at a Reef Packing Plant C. 0. Gill* dz J. Bryant Agriculture and Agri-Food Canada Research Centre, 6000 C and E Trail, Lacombe, Alberta, Canada, T4L 1Wl (Received 4 April 1997; revised version received 14 May 1997; accepted 14 May 1997)
ABSTRACT The microbiological eflects of three operations for cleaning areas on dressed beef carcasses with vacuuming equipment which also applies hot water to the carcass, and of an operation for pasteurizing beef carcass sides with steam, were assessed. All four operations were routine in a commercial carcass dressing process. For each operation, swab samples were obtainedfrom randomly selected carcasses, with a single sample being collectedfrom each carcass, from a site selected at random from those affected by the operation. For the cleaning operations, 25 samples were obtained before and 25 after each operation. For the pasteurizing operation, 50 samples were obtained before and 50 after the operation. In addition, 50 samples were obtained from beef sides after the carcass cooling process which followed the pasteurizing operation. Total aerobic counts, coliforms and Escherichia coli from each sample were enumerated. The cleaning operations generally reduced the log mean numbers of bacteria on treated areas by 5 0.5 and had no discernible effect on the overall microbiological condition of the carcasses emerging from the process. The pasteurizing operation reduced the log mean numbers of total aerobic bacteria on carcasses by about 1, and the log mean numbers of coltforms and E. coli by > 2. The cooling process had no affect on the total counts, but further reduced the log mean numbers of coltforms and E. coli, apparently by about I, to give beef sides from which E. coli were not recovered. 0 1997 Elsevier Science Ltd. All rights reserved
INTRODUCTION Meat inspecting authorities are introducing requirements which are intended to improve control over the contamination of product with pathogenic bacteria during the processing of meat, with particular emphasis on the control of contamination during carcass dressing processes (USDA, 1996). To meet with those requirements, some beef packing plants have installed equipment for decontaminating carcasses by heating the surfaces
*To whom correspondence
should be addressed. Fax: (403) 782-6120; e-mail: [email protected] 267
268
C. 0. Gill, J. Bryant
with hot water or steam. Two types of equipment are being used for such pasteurizing treatments. The first type is composed of a vacuum head fitted with nozzles to deliver hot water and/or steam. That hand held equipment is applied to selected areas of the carcass surface where visible contamination is relatively frequent on the carcasses passing through a process. The second type is equipment in which beef sides are exposed to an atmosphere of steam at above atmospheric pressure, to rapidly raise all the meat surfaces to pasteurizing temperatures. Reports of laboratory studies indicate that treatments with both those types of apparatus can reduce the numbers of bacteria on meat surfaces by about 4 loglo units (Cygnarowicz-Provost et al., 1994; Dorsa et al., 1996). However, in commercial practice, the application of any decontaminating treatment may be constrained by factors such as the speed of the processing line, inconsistent selection by workers of areas for treatment, or the need to minimize the adverse effects of a treatment on the appearance of the product. Consequently, decontaminating treatments, such as the trimming of visible contamination from carcasses, which are highly effective for removing bacteria from meat in experimental circumstances may be ineffective in commercial practice (Hardin et al., 1995; Gill et al., 1996; Reagan et al., 1996). As there have been no reports of the performance in commercial circumstances of equipment for heating the surfaces of beef carcasses with hot water or steam, the microbiological effects of such equipment during their routine use at a beef packing plant were examined.
MATERIALS
AND METHODS
The carcass dressing and cooling processes
The carcass dressing and cooling processes at a beef packing plant which processes 280 carcasses from feedlotted cattle per hr were examined. The dressing process consists of 39 operations (Table 1). The microbiological effects of the vacuum-hot water cleaning operations 14, 17 and 30, the carcass side pasteurizing operation 39, and the following carcass cooling process were assessed. The equipment (Vat-San; Kentmaster, Monrovia, CA) used in operations 14, 17 and 30 consists of a vacuum head which is fitted with a nozzle for the delivery of a continuous stream of hot water onto the surface being treated and which is surrounded by a jacket for delivering steam onto the vacuum head. The temperatures of the water and steam in the lines to the head and the vacuum drawn in the vacuum line are continuously monitored. The equipment is operated with water and steam temperatures > 82°C and with a line vacuum > 175 mm Hg. The equipment (Steam Pasteurizing System. Frigoscandia, Bellevue, WA) used in operation 39 consists of an entrance section, in which air is blown over the beef sides to dry the surfaces after washing; a pasteurizing chamber, which for the treatment is sealed and filled with steam, under pressure and at a temperature about 105°C; and an exit section in which the carcasses are sprayed with cold water. The pasteurizing treatment is applied for 6.5 s. The spray-cooling process for carcasses, which follows operation 39, has been previously described (Gill and Bryant, 1997). Sampling of carcasses
Carcasses were sampled with reference to a grid which specified 126 areas of the surface of a beef carcass side (Fig. 1).
Decontamination of beef carcasses
269
TABLE 1 Operations in the Beef Carcass Dressing Process 1. 2. 3. 4. 5. 6. I. 8. 9. 10. 11. 12. 13. “14. 15. 16. “17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. “30. 31. 32. 33. 34. 35. 36. 31. 38. “39.
Open hide: skin right, rear hock Remove right, rear hoof Skin right butt Hook right, rear leg Skin left, rear hock Remove left, rear hoof Skin left butt Hook left, rear leg Open brisket skin Open tail skin Skin rump Blow hair from rear hocks Skin tail Vacuum-hot water clean between back legs Remove horns, ears and front hooves Skin brisket Vacuum-hot water clean brisket Skin back Remove hide Trim and separate head Split sternum Trim forelegs Free and tie bung Remove head, tie oesophagus Remove viscera Split carcass Change from dressing to main chain hook
Trim butt Trim rump Vacuum-hot water clean anal area Trim brisket Remove tail Remove hanging tender Remove mesenteric fat Remove diaphragm remnants Trim neck Weight Wash Pasteurize carcass sides
“Operations examined in this study.
Samples were obtained from carcasses before and after each of the operations which were assessed, and after the carcass cooling process. Each sample was obtained by swabbing an undelimited area of approximately lOOcm*, as previously described (Gill et al., 1996). Both before and after each of the cleaning operations, a sample was collected from each of 25 carcasses selected at random, from a site randomly selected from those treated during the operation. The sites treated in the cleaning operations were: for operation 14, the projections onto the area between the back legs of sites 8 1,41,62 and 6 1; for operation 17, sites 102,103,104 and 105; and for operation 30, sites 82 and 84. For each operation, on each of five days, five samples were collected before and five after the operation.
270
C. 0. Gill, J. Bryant
Fig. 1. The grid used for the identification
Both before and after ple was collected from selected from the whole the operation, and after
of sites on the surfaces of beef carcass sides.
the pasteurizing operation, and after the chilling process, a sameach of 50 carcasses selected at random, from a site randomly grid, with five samples being collected before the operation, after the cooling process on each of 10 days.
Microbiological analysis The total aerobic counts recovered from each sample were enumerated by a spread plate procedure with a detection level of 1 cfu cme2, and coliforms and E. coli were enumerated by a hydrophobic grid membrane filtration procedure with a detection level of 1 cfu 100cm-2, as previously described (Gill et al., 1996).
271
Decontamination of beef carcasses
All bacterial counts were transformed to log values. Values for the mean log (X) and standard deviation (s) of each set of bacterial counts were calculated on the assumption of a log normal distribution of counts (Brown and Baird-Parker, 1982). In the calculation of Tzand s for sets of E. co/i or coliform counts, log value of -0.5 100 cm-* were assumed for samples in which E. colt’ and/or coliforms were not detected at the level of 1 E. coli or coliform cfu 100 cmp2. A value for the log mean (log A) for each set of counts was calculated from the formula log A = X + In 10 s* 2-l (Kilsby and Pugh, 1981). All calculations were performed with Microsoft Excel Version 4, statistical functions (Microsoft Corp., Redmond, WA, USA).
RESULTS Total aerobic, coliform and E. coli counts were recovered from all the samples obtained both before and after the cleaning operation 14. The values for the logs of the arithmetic means and of the total numbers recovered indicated that the log numbers of total counts and coliforms on the treated area were reduced by IO.5 as a result of the cleaning operation, but that numbers of E. co/i were little affected by the operation (Table 2). Total counts were recovered from all samples, and coliforms and E. coli from most samples obtained both before and after the cleaning operation 17, although the numbers of coliform- and E. co&negative samples were greater after than before the treatment (Table 3). The values for the logs of the arithmetic means and of the total numbers recovered for all three groups of bacteria indicated that the log numbers of total counts, coliforms and E. co/i were each reduced by between 0.5 and 1 as a result of the cleaning treatment, with the largest reduction occurring in the numbers of E. coli. Total counts were recovered from all samples obtained both before and after the cleaning operation 30. Coliforms and E. coli were recovered from majorities of the samples obtained before the operation but from minorities of the samples obtained after the operation (Table 4). Despite that, the values for the logs of the arithmetic means and of TABLE 2
Statistics for Sets of 25 Total Aerobic Counts (cfucme2), Coliform Counts (cfu lOO~rn_~) or Escherichia coli Counts (cfu lOOcm-*) Obtained from the Area of the Beef Carcass Surface between the Back Legs which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 14 before and after the Operation Count
Stage of the operation
Statistics
x
s
log A
N
Before After
3.76 3.47
0.65 0.46
4.25 3.71
5.50” s.17*
Coliform
Before After
3.05 2.30
I .27 1.27
4.89 4.17
S.Olb 4.80b
E. coli
Before After
2.78 2.06
1.17 1-39
4.36 4.28
4.90h 4.84b
Total
Aerobic
X = mean log. s = standard deviation. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm2 or b2500 cm2.
272
C. 0. Gill, J. Bryant
TABLE 3 Statistics for Sets of 25 Total Aerobic Counts (cfu cmP2), Coliform Counts (cfu lOOcm-*) or Escherichia coli Counts (cfu 100 cm-*) Obtained from the Brisket Area of the Beef Carcass Surface which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 17 before and after the Operation Count
Stage of the operation
Statistics
x
s
No.
log A
N
Total Aerobic
Before After
4.03 3.70
0.63 0.40
0 0
4.48 3.88
5.7ga 528”
Coliforrn
Before After
1.68 1.12
0.92 0.84
1 2
2.66 I .93
3.79h 3.20’
E. coli
Before After
1.24 0.61
1.10 0.97
4 7
2.63 1.70
3.71b 3.056
55 = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm* or h2500 cm*.
TABLE 4 Statistics for Sets of 25 Total Aerobic Counts (cfu cm-*), Coliform Counts (cfu lOOcm-*) or Escherichia coli Counts (cfu lOOem-*) Obtained from the Anal Area of the Beef Carcass Surface which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 30 before and after the Operation Count
Stage of the operation
Statistics
x
S
No.
log A
N
Total Aerobic
Before After
2.62 2.19
0.52 0.65
0 0
2.94 2.68
4.22” 4.20”
Coliform
Before After
0.38 0.18
0.79 0.87
7 14
I.10 1.06
2.81b 2.46h
E. coli
Before After
0.21 0.01
0.78 0.76
10 14
0.90 0.68
2.51b 2.29”
X = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm* or b2500 cm2.
the total numbers recovered indicated that the log numbers of all three groups of bacteria were reduced by < 0.5 by the cleaning treatment. Total counts were recovered from all samples both before and after the pasteurizing operation 39 and after the cooling process. The values for the logs of the arithmetic means and of the total numbers recovered indicated that the pasteurizing treatment reduced the
213
Decontamination of beef carcasses
log total counts by about 1, but that the cooling progress has little, if any affect upon the total counts (Table 5). Coliforms and E. coli were recovered from most samples obtained before the pasteurizing operation, but from only minorities of the samples obtained after that operation (Table 5). After the cooling process, the numbers of coliform- and E. co/i-negative samples were increased, with no E. co/i being recovered from any sample. The values for the logs of the arithmetic means and/or of the total numbers recovered indicated that the pasteurizing treatment reduced the log numbers of coliforms and E. coli by > 2, while the cooling process further reduced their log numbers by about 1.
DISCUSSION The operator of vacuum-hot water cleaning equipment applies the apparatus to remove visible contamination. There is only weak correlation between visible and bacterial contamination on beef carcasses (Jericho ef al., 1993). Therefore, the equipment is unlikely to be reliably applied to all sites carrying relatively high bacterial loads. On a high speed line, the operator has approximately 20s to apply the apparatus to a carcass. To obtain the maximum destruction of the natural flora, a pasteurizing treatment with water must raise the temperature of the carcasses surface to at least 80°C for about 10 s (Davey and Smith, 1989). A single operator could then apply an optimum pasteurizing treatment to no more than two areas of a carcass surface, each the size of the head of the steam-vacuuming equipment. In practice, the head is moved back and forth over the carcass
surface,
so that no area is heated continuously
for 10 s.
TABLE 5 Statistics for Sets of 50 Total Aerobic Counts (cfucmP2), Coliform Counts (cfu 100cm-2) or Escherichia cofi Counts (cfu 100cm-2) Obtained from Randomly Selected Sites on Beef Carcasses after the Washing Operation 38, or after the Pasteurizing Operation 39 in the Carcass Dressing Process, or after the Carcass Cooling Process Count
Preceeding operation or process
Statistics
x
S
No.
log A
N
Total Aerobic
Washing Pasteurizing Cooling
2.70 1.49 1.65
0.76 0.81 087
0 0 0
3.36 2.24 2.52
5.23” 4-19” 4.02“
Coliform
Washing Pasteurizing Cooling
1.20
1.03
2 37 47
2.43 -
4.06b 1.69b 0.78b
E. coli
Washing Pasteurizing Cooling
084 -
1.09 -
8 44 50
2.20 -
3.84b l.llb ndh
X = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total number recovered from a25 cm2 or b2500 cm2. - = insullicient data for calculation of the statistic. nd = none detected.
214
C.0. Gill, J. Bryant
With inoculated samples, various decontaminating treatments of meat have been reported to reduce bacterial numbers by several orders of magnitude (Anderson and Marshall, 1989; Hardin et al., 1995; Regan et al., 1996). However, pasteurizing and other treatments will apparently reduce the numbers of the natural flora by about two orders of magnitude at most (Smith and Graham, 1978; Gill et al., 1995). Thus, with the sub-optimal pasteurizing treatment applied during the vacuuming-hot water cleaning of carcasses, reductions in bacterial numbers of substantially less than two orders of magnitude could be expected. In conformance with that expectation, the data from this study indicate that the log numbers of bacteria were generally reduced by about O-5, and by at most 1, as a result of vacuum-hot water cleaning. Moreover, those findings appear to be comparable with the reductions in numbers of natural flora achieved by vacuum-steam or -hot water cleaning in recently reported experimental treatments of commercial carcasses (Kochevar et al., 1997). The differences in the efficacy of the cleaning treatment for removing bacteria from the three sites examined in this study would be anomalous if the reductions were due to the destruction of bacteria by heating, because the treatment was least effective at the anal area, the smallest area subjected to a cleaning operation, the expectation being that in a given, short treatment time higher temperatures for longer periods would be attained in smaller than in larger areas. However, the findings are explicable if the removal of matter containing bacteria in the flows of water and air into the vacuum nozzle, rather than heating, is largely responsible for the removal of bacteria from the carcass. The removal of matter would be most effective in areas when the vacuum head is applied to a relatively flat surface, as to the brisket in operation 17; and least effect in areas where sharply curved surfaces generally prevent contact of the whole face of the vacuum head with the carcass surface, as in cleaning of the anal area in operation 30. The limited reductions in bacterial numbers achieved with vacuum-hot water cleaning, and the small portion of the total surface of a carcass that could in practice be subjected to such cleaning indicate that, at present, such treatments should not be regarded as effective means of improving the microbiological condition of carcasses. That view is supported by the similarity of the findings for the microbiological condition of washed carcasses in this study and in a previous study of the same dressing process, which was completed before the installation of vacuum-hot water cleaning equipment (Gill et al., 1996). However, the equipment is apparently highly effective for the removal of visible contamination, and no less effective than trimming for removing microbial contamination (Kochevar et al., 1997). Its use for the removal of visible contamination would then seem unobjectionable, as there is no indication that use of the equipment would degrade the microbiological condition of carcasses. In contrast, the pasteurizing treatment for beef sides is evidently effective for reducing the numbers of E. coli and other coliforms on carcasses, although less so for reducing the total bacterial load. The surviving flora would likely be enriched for Gram positive species (Gill et al., 1995), but that aspect of the effects of the treatment will require further investigation. The microbiological effects of the pasteurizing treatment appear similar to those of the cooling process on non-pasteurized carcasses, as both reduced the log numbers of total counts by about 1 and the log number of coliforms and E. coli by > 2 (Gill and Bryant, 1997). The cause or causes of the reductions of bacteria numbers during the spray-cooling process are uncertain, although freezing of carcass surfaces is likely to be involved, as that could account for larger reduction in the numbers of Gram negative than of Gram positive species (Lowry and Gill, 1985). With regard to the coliform and E. coli counts, the effect of the cooling process is apparently additive to that of the pasteurizing treatment. However, the cooling process
Decontamination of beef carcasses
275
appears to be without effect on most of the flora which survive the pasteurizing treatment. Again, further investigation of the flora which survives the cooling process is required, as are the implications of the changes in the flora composition for the storage stability of the product (Sheridan et al., 1997). Despite uncertainties about the overall effects of the decontaminating treatments examined in this study, it is apparent that pasteurizing treatments and cooling processes can be operated in commercial practice to produce carcasses from which E. coli cannot be recovered at the level of 1 lOOcm-*. The wide use of pasteurizing treatments and decontaminating cooling processes in conjunction with the control of the numbers of E. co/i deposited on carcasses during dressing could therefore largely eliminate the dressing process as a source of the enteric pathogens associated with meat.
ACKNOWLEDGEMENTS We thank the staff of the beef packing plant involved in this study for the provision of information on plant processes and for assistance with the collection of samples. Funding for this study was provided by the Beef Industry Development Fund of Agriculture and Agri-Food Canada. REFERENCES Anderson, M. E. and Marshall, R. T. (1989) Interaction of concentration and temperature of acetic acid solution on reduction of various species of microorganisms on beef surfaces. Journal of Food Protection 52, 312-315.
Brown, M. H. and Baird-Parker, A. C. (1982) The microbiological examination of meat. In Meat Microbiology, ed. M. H. Brown, pp. 423-520. Applied Science Publishers, London. Cygnarowicz-Provost, M., Whiting, R. C. and Craig Jr., J. C. (1994) Steam surface pasteurization of beef frankfurters. Journal of Food Science 59, l-5. Davey, K. R. and Smith, M. G. (1989) A laboratory evaluation of a novel hot water cabinet for the decontamination of sides of beef. International Journal of Food Science and Technology 24, 305-316.
Dorsa, W. J., Cutter, C. N., Siragusa, G. R. and Koohmaraie, M. (1996) Microbial decontamination of beef and sheep carcasses by steam, hot water spray washes, and a steam-vacuum sanitizer. Journal of Food Protection 59, 127-135.
Gill, C. 0. and Bryant, J. (1997) Assessment of the hygienic performances of two beef carcass cooling processes from product temperature history data or enumeration of bacteria on carcass surfaces. Food Microbiology, in press. Gill, C. O., McGinnis, D. S., Bryant, J. and Chabot, B. (1995) Decontamination of commercial, polished pig carcasses with hot water. Food Microbiology 12, 143-149. Gill, C. O., Badoni, M. and Jones, T. (1996) The hygienic effects of the trimming and washing operations in a beef carcass dressing process. Journal of Food Protection 59, 666-669. Hardin, M. D., Acuff, G. R., Lucia, L. M., Osmar, J. S. and Savell, J. W. (1995) Comparisons of methods for decontamination from beef carcass surfaces. Journal of Food Protection 58,368-374. Jericho, K. W. F., Bradley, J. A., Gannon, V. P. J. and Kozub, G. C. (1993) Visual demerit and microbiological evaluation of beef carcasses: Methodology. Journal of Food Protection 56, 114-119.
Kilsby, D. C. and and Pugh, M. E. (1981) The relevance of the distribution of micro-organisms within batches of food to the control of microbiological hazards from foods. Journal of Applied Bacteriology 51, 345-354.
Kochevar, S. L., Sofos, J. N., Bolin, R. R., Reagan, J. 0. and Smith, G. C. (1997) Steam vacuuming as a pre-evisceration intervention to decontaminate beef carcasses. Journal of Food Protection 60.107-113.
276
C. 0. Gill, J. Bryant
Lowry, P. D. and Gill, C. 0. (1985) Microbiology of frozen meat and meat products. In Microbiology of Frozen Fooa!s, ed. R. K. Robinson, pp. 109-168. Elsevier Applied Science, London. Regan, J. O., Acuff, G. R., Buege, D. R., Buyeck, M. J., Dickson, J. S., Kastner, C. L., Marsden, J. L., Morgan, J. B., Nickelson, R., Smith, G. C. and Sofos, J. N. (1996) Trimming and washing of beef carcasses as a method of improving the microbiological quality of meat. Journal of Food Protection 59, 751-756.
Sheridan, J. J., Doherty, A. M., Allen, P., McDowell, D. A., Blair, I. S. and Harrington, D. (1997) The effect of vacuum and modified atmosphere packaging on the shelf-life of lamb primals, stored at different temperatures. Meat Science 45, 107-l 17. Smith, M. G. and Graham, A. (1978) Destruction of Escherichia coli and salmonellae on mutton carcasses by treatment with hot water. Meat Science 2, 119-128. USDA (1996) US Department of Agriculture, Food Safety and Inspection Service. Pathogen reduction; hazard analysis and critical control point (HACCP) systems; final rule. Federal Register 61, 3880538989.
SO309-1740(97)00058-2
Meat Science, Vol. 41, No. 3/4, 267-276, 1997 Q 1997 Elsevier Science Ltd All rights reserved. Printed in Great Britain 0309-1740/97 $17.00+0.00
ELSEVIER
Decontamination of Carcasses by Vacuum-hot Water Cleaning and Steam Pasteurizing during Routine Operations at a Reef Packing Plant C. 0. Gill* dz J. Bryant Agriculture and Agri-Food Canada Research Centre, 6000 C and E Trail, Lacombe, Alberta, Canada, T4L 1Wl (Received 4 April 1997; revised version received 14 May 1997; accepted 14 May 1997)
ABSTRACT The microbiological eflects of three operations for cleaning areas on dressed beef carcasses with vacuuming equipment which also applies hot water to the carcass, and of an operation for pasteurizing beef carcass sides with steam, were assessed. All four operations were routine in a commercial carcass dressing process. For each operation, swab samples were obtainedfrom randomly selected carcasses, with a single sample being collectedfrom each carcass, from a site selected at random from those affected by the operation. For the cleaning operations, 25 samples were obtained before and 25 after each operation. For the pasteurizing operation, 50 samples were obtained before and 50 after the operation. In addition, 50 samples were obtained from beef sides after the carcass cooling process which followed the pasteurizing operation. Total aerobic counts, coliforms and Escherichia coli from each sample were enumerated. The cleaning operations generally reduced the log mean numbers of bacteria on treated areas by 5 0.5 and had no discernible effect on the overall microbiological condition of the carcasses emerging from the process. The pasteurizing operation reduced the log mean numbers of total aerobic bacteria on carcasses by about 1, and the log mean numbers of coltforms and E. coli by > 2. The cooling process had no affect on the total counts, but further reduced the log mean numbers of coltforms and E. coli, apparently by about I, to give beef sides from which E. coli were not recovered. 0 1997 Elsevier Science Ltd. All rights reserved
INTRODUCTION Meat inspecting authorities are introducing requirements which are intended to improve control over the contamination of product with pathogenic bacteria during the processing of meat, with particular emphasis on the control of contamination during carcass dressing processes (USDA, 1996). To meet with those requirements, some beef packing plants have installed equipment for decontaminating carcasses by heating the surfaces
*To whom correspondence
should be addressed. Fax: (403) 782-6120; e-mail: [email protected] 267
268
C. 0. Gill, J. Bryant
with hot water or steam. Two types of equipment are being used for such pasteurizing treatments. The first type is composed of a vacuum head fitted with nozzles to deliver hot water and/or steam. That hand held equipment is applied to selected areas of the carcass surface where visible contamination is relatively frequent on the carcasses passing through a process. The second type is equipment in which beef sides are exposed to an atmosphere of steam at above atmospheric pressure, to rapidly raise all the meat surfaces to pasteurizing temperatures. Reports of laboratory studies indicate that treatments with both those types of apparatus can reduce the numbers of bacteria on meat surfaces by about 4 loglo units (Cygnarowicz-Provost et al., 1994; Dorsa et al., 1996). However, in commercial practice, the application of any decontaminating treatment may be constrained by factors such as the speed of the processing line, inconsistent selection by workers of areas for treatment, or the need to minimize the adverse effects of a treatment on the appearance of the product. Consequently, decontaminating treatments, such as the trimming of visible contamination from carcasses, which are highly effective for removing bacteria from meat in experimental circumstances may be ineffective in commercial practice (Hardin et al., 1995; Gill et al., 1996; Reagan et al., 1996). As there have been no reports of the performance in commercial circumstances of equipment for heating the surfaces of beef carcasses with hot water or steam, the microbiological effects of such equipment during their routine use at a beef packing plant were examined.
MATERIALS
AND METHODS
The carcass dressing and cooling processes
The carcass dressing and cooling processes at a beef packing plant which processes 280 carcasses from feedlotted cattle per hr were examined. The dressing process consists of 39 operations (Table 1). The microbiological effects of the vacuum-hot water cleaning operations 14, 17 and 30, the carcass side pasteurizing operation 39, and the following carcass cooling process were assessed. The equipment (Vat-San; Kentmaster, Monrovia, CA) used in operations 14, 17 and 30 consists of a vacuum head which is fitted with a nozzle for the delivery of a continuous stream of hot water onto the surface being treated and which is surrounded by a jacket for delivering steam onto the vacuum head. The temperatures of the water and steam in the lines to the head and the vacuum drawn in the vacuum line are continuously monitored. The equipment is operated with water and steam temperatures > 82°C and with a line vacuum > 175 mm Hg. The equipment (Steam Pasteurizing System. Frigoscandia, Bellevue, WA) used in operation 39 consists of an entrance section, in which air is blown over the beef sides to dry the surfaces after washing; a pasteurizing chamber, which for the treatment is sealed and filled with steam, under pressure and at a temperature about 105°C; and an exit section in which the carcasses are sprayed with cold water. The pasteurizing treatment is applied for 6.5 s. The spray-cooling process for carcasses, which follows operation 39, has been previously described (Gill and Bryant, 1997). Sampling of carcasses
Carcasses were sampled with reference to a grid which specified 126 areas of the surface of a beef carcass side (Fig. 1).
Decontamination of beef carcasses
269
TABLE 1 Operations in the Beef Carcass Dressing Process 1. 2. 3. 4. 5. 6. I. 8. 9. 10. 11. 12. 13. “14. 15. 16. “17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. “30. 31. 32. 33. 34. 35. 36. 31. 38. “39.
Open hide: skin right, rear hock Remove right, rear hoof Skin right butt Hook right, rear leg Skin left, rear hock Remove left, rear hoof Skin left butt Hook left, rear leg Open brisket skin Open tail skin Skin rump Blow hair from rear hocks Skin tail Vacuum-hot water clean between back legs Remove horns, ears and front hooves Skin brisket Vacuum-hot water clean brisket Skin back Remove hide Trim and separate head Split sternum Trim forelegs Free and tie bung Remove head, tie oesophagus Remove viscera Split carcass Change from dressing to main chain hook
Trim butt Trim rump Vacuum-hot water clean anal area Trim brisket Remove tail Remove hanging tender Remove mesenteric fat Remove diaphragm remnants Trim neck Weight Wash Pasteurize carcass sides
“Operations examined in this study.
Samples were obtained from carcasses before and after each of the operations which were assessed, and after the carcass cooling process. Each sample was obtained by swabbing an undelimited area of approximately lOOcm*, as previously described (Gill et al., 1996). Both before and after each of the cleaning operations, a sample was collected from each of 25 carcasses selected at random, from a site randomly selected from those treated during the operation. The sites treated in the cleaning operations were: for operation 14, the projections onto the area between the back legs of sites 8 1,41,62 and 6 1; for operation 17, sites 102,103,104 and 105; and for operation 30, sites 82 and 84. For each operation, on each of five days, five samples were collected before and five after the operation.
270
C. 0. Gill, J. Bryant
Fig. 1. The grid used for the identification
Both before and after ple was collected from selected from the whole the operation, and after
of sites on the surfaces of beef carcass sides.
the pasteurizing operation, and after the chilling process, a sameach of 50 carcasses selected at random, from a site randomly grid, with five samples being collected before the operation, after the cooling process on each of 10 days.
Microbiological analysis The total aerobic counts recovered from each sample were enumerated by a spread plate procedure with a detection level of 1 cfu cme2, and coliforms and E. coli were enumerated by a hydrophobic grid membrane filtration procedure with a detection level of 1 cfu 100cm-2, as previously described (Gill et al., 1996).
271
Decontamination of beef carcasses
All bacterial counts were transformed to log values. Values for the mean log (X) and standard deviation (s) of each set of bacterial counts were calculated on the assumption of a log normal distribution of counts (Brown and Baird-Parker, 1982). In the calculation of Tzand s for sets of E. co/i or coliform counts, log value of -0.5 100 cm-* were assumed for samples in which E. colt’ and/or coliforms were not detected at the level of 1 E. coli or coliform cfu 100 cmp2. A value for the log mean (log A) for each set of counts was calculated from the formula log A = X + In 10 s* 2-l (Kilsby and Pugh, 1981). All calculations were performed with Microsoft Excel Version 4, statistical functions (Microsoft Corp., Redmond, WA, USA).
RESULTS Total aerobic, coliform and E. coli counts were recovered from all the samples obtained both before and after the cleaning operation 14. The values for the logs of the arithmetic means and of the total numbers recovered indicated that the log numbers of total counts and coliforms on the treated area were reduced by IO.5 as a result of the cleaning operation, but that numbers of E. co/i were little affected by the operation (Table 2). Total counts were recovered from all samples, and coliforms and E. coli from most samples obtained both before and after the cleaning operation 17, although the numbers of coliform- and E. co&negative samples were greater after than before the treatment (Table 3). The values for the logs of the arithmetic means and of the total numbers recovered for all three groups of bacteria indicated that the log numbers of total counts, coliforms and E. co/i were each reduced by between 0.5 and 1 as a result of the cleaning treatment, with the largest reduction occurring in the numbers of E. coli. Total counts were recovered from all samples obtained both before and after the cleaning operation 30. Coliforms and E. coli were recovered from majorities of the samples obtained before the operation but from minorities of the samples obtained after the operation (Table 4). Despite that, the values for the logs of the arithmetic means and of TABLE 2
Statistics for Sets of 25 Total Aerobic Counts (cfucme2), Coliform Counts (cfu lOO~rn_~) or Escherichia coli Counts (cfu lOOcm-*) Obtained from the Area of the Beef Carcass Surface between the Back Legs which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 14 before and after the Operation Count
Stage of the operation
Statistics
x
s
log A
N
Before After
3.76 3.47
0.65 0.46
4.25 3.71
5.50” s.17*
Coliform
Before After
3.05 2.30
I .27 1.27
4.89 4.17
S.Olb 4.80b
E. coli
Before After
2.78 2.06
1.17 1-39
4.36 4.28
4.90h 4.84b
Total
Aerobic
X = mean log. s = standard deviation. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm2 or b2500 cm2.
272
C. 0. Gill, J. Bryant
TABLE 3 Statistics for Sets of 25 Total Aerobic Counts (cfu cmP2), Coliform Counts (cfu lOOcm-*) or Escherichia coli Counts (cfu 100 cm-*) Obtained from the Brisket Area of the Beef Carcass Surface which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 17 before and after the Operation Count
Stage of the operation
Statistics
x
s
No.
log A
N
Total Aerobic
Before After
4.03 3.70
0.63 0.40
0 0
4.48 3.88
5.7ga 528”
Coliforrn
Before After
1.68 1.12
0.92 0.84
1 2
2.66 I .93
3.79h 3.20’
E. coli
Before After
1.24 0.61
1.10 0.97
4 7
2.63 1.70
3.71b 3.056
55 = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm* or h2500 cm*.
TABLE 4 Statistics for Sets of 25 Total Aerobic Counts (cfu cm-*), Coliform Counts (cfu lOOcm-*) or Escherichia coli Counts (cfu lOOem-*) Obtained from the Anal Area of the Beef Carcass Surface which was Subjected to Vacuum-hot Water Cleaning in the Carcass Dressing Operation 30 before and after the Operation Count
Stage of the operation
Statistics
x
S
No.
log A
N
Total Aerobic
Before After
2.62 2.19
0.52 0.65
0 0
2.94 2.68
4.22” 4.20”
Coliform
Before After
0.38 0.18
0.79 0.87
7 14
I.10 1.06
2.81b 2.46h
E. coli
Before After
0.21 0.01
0.78 0.76
10 14
0.90 0.68
2.51b 2.29”
X = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total numbers recovered from “25 cm* or b2500 cm2.
the total numbers recovered indicated that the log numbers of all three groups of bacteria were reduced by < 0.5 by the cleaning treatment. Total counts were recovered from all samples both before and after the pasteurizing operation 39 and after the cooling process. The values for the logs of the arithmetic means and of the total numbers recovered indicated that the pasteurizing treatment reduced the
213
Decontamination of beef carcasses
log total counts by about 1, but that the cooling progress has little, if any affect upon the total counts (Table 5). Coliforms and E. coli were recovered from most samples obtained before the pasteurizing operation, but from only minorities of the samples obtained after that operation (Table 5). After the cooling process, the numbers of coliform- and E. co/i-negative samples were increased, with no E. co/i being recovered from any sample. The values for the logs of the arithmetic means and/or of the total numbers recovered indicated that the pasteurizing treatment reduced the log numbers of coliforms and E. coli by > 2, while the cooling process further reduced their log numbers by about 1.
DISCUSSION The operator of vacuum-hot water cleaning equipment applies the apparatus to remove visible contamination. There is only weak correlation between visible and bacterial contamination on beef carcasses (Jericho ef al., 1993). Therefore, the equipment is unlikely to be reliably applied to all sites carrying relatively high bacterial loads. On a high speed line, the operator has approximately 20s to apply the apparatus to a carcass. To obtain the maximum destruction of the natural flora, a pasteurizing treatment with water must raise the temperature of the carcasses surface to at least 80°C for about 10 s (Davey and Smith, 1989). A single operator could then apply an optimum pasteurizing treatment to no more than two areas of a carcass surface, each the size of the head of the steam-vacuuming equipment. In practice, the head is moved back and forth over the carcass
surface,
so that no area is heated continuously
for 10 s.
TABLE 5 Statistics for Sets of 50 Total Aerobic Counts (cfucmP2), Coliform Counts (cfu 100cm-2) or Escherichia cofi Counts (cfu 100cm-2) Obtained from Randomly Selected Sites on Beef Carcasses after the Washing Operation 38, or after the Pasteurizing Operation 39 in the Carcass Dressing Process, or after the Carcass Cooling Process Count
Preceeding operation or process
Statistics
x
S
No.
log A
N
Total Aerobic
Washing Pasteurizing Cooling
2.70 1.49 1.65
0.76 0.81 087
0 0 0
3.36 2.24 2.52
5.23” 4-19” 4.02“
Coliform
Washing Pasteurizing Cooling
1.20
1.03
2 37 47
2.43 -
4.06b 1.69b 0.78b
E. coli
Washing Pasteurizing Cooling
084 -
1.09 -
8 44 50
2.20 -
3.84b l.llb ndh
X = mean log. s = standard deviation. No. = number of samples from which bacteria were not recovered. log A = estimated log of the arithmetic mean. N = log of the total number recovered from a25 cm2 or b2500 cm2. - = insullicient data for calculation of the statistic. nd = none detected.
214
C.0. Gill, J. Bryant
With inoculated samples, various decontaminating treatments of meat have been reported to reduce bacterial numbers by several orders of magnitude (Anderson and Marshall, 1989; Hardin et al., 1995; Regan et al., 1996). However, pasteurizing and other treatments will apparently reduce the numbers of the natural flora by about two orders of magnitude at most (Smith and Graham, 1978; Gill et al., 1995). Thus, with the sub-optimal pasteurizing treatment applied during the vacuuming-hot water cleaning of carcasses, reductions in bacterial numbers of substantially less than two orders of magnitude could be expected. In conformance with that expectation, the data from this study indicate that the log numbers of bacteria were generally reduced by about O-5, and by at most 1, as a result of vacuum-hot water cleaning. Moreover, those findings appear to be comparable with the reductions in numbers of natural flora achieved by vacuum-steam or -hot water cleaning in recently reported experimental treatments of commercial carcasses (Kochevar et al., 1997). The differences in the efficacy of the cleaning treatment for removing bacteria from the three sites examined in this study would be anomalous if the reductions were due to the destruction of bacteria by heating, because the treatment was least effective at the anal area, the smallest area subjected to a cleaning operation, the expectation being that in a given, short treatment time higher temperatures for longer periods would be attained in smaller than in larger areas. However, the findings are explicable if the removal of matter containing bacteria in the flows of water and air into the vacuum nozzle, rather than heating, is largely responsible for the removal of bacteria from the carcass. The removal of matter would be most effective in areas when the vacuum head is applied to a relatively flat surface, as to the brisket in operation 17; and least effect in areas where sharply curved surfaces generally prevent contact of the whole face of the vacuum head with the carcass surface, as in cleaning of the anal area in operation 30. The limited reductions in bacterial numbers achieved with vacuum-hot water cleaning, and the small portion of the total surface of a carcass that could in practice be subjected to such cleaning indicate that, at present, such treatments should not be regarded as effective means of improving the microbiological condition of carcasses. That view is supported by the similarity of the findings for the microbiological condition of washed carcasses in this study and in a previous study of the same dressing process, which was completed before the installation of vacuum-hot water cleaning equipment (Gill et al., 1996). However, the equipment is apparently highly effective for the removal of visible contamination, and no less effective than trimming for removing microbial contamination (Kochevar et al., 1997). Its use for the removal of visible contamination would then seem unobjectionable, as there is no indication that use of the equipment would degrade the microbiological condition of carcasses. In contrast, the pasteurizing treatment for beef sides is evidently effective for reducing the numbers of E. coli and other coliforms on carcasses, although less so for reducing the total bacterial load. The surviving flora would likely be enriched for Gram positive species (Gill et al., 1995), but that aspect of the effects of the treatment will require further investigation. The microbiological effects of the pasteurizing treatment appear similar to those of the cooling process on non-pasteurized carcasses, as both reduced the log numbers of total counts by about 1 and the log number of coliforms and E. coli by > 2 (Gill and Bryant, 1997). The cause or causes of the reductions of bacteria numbers during the spray-cooling process are uncertain, although freezing of carcass surfaces is likely to be involved, as that could account for larger reduction in the numbers of Gram negative than of Gram positive species (Lowry and Gill, 1985). With regard to the coliform and E. coli counts, the effect of the cooling process is apparently additive to that of the pasteurizing treatment. However, the cooling process
Decontamination of beef carcasses
275
appears to be without effect on most of the flora which survive the pasteurizing treatment. Again, further investigation of the flora which survives the cooling process is required, as are the implications of the changes in the flora composition for the storage stability of the product (Sheridan et al., 1997). Despite uncertainties about the overall effects of the decontaminating treatments examined in this study, it is apparent that pasteurizing treatments and cooling processes can be operated in commercial practice to produce carcasses from which E. coli cannot be recovered at the level of 1 lOOcm-*. The wide use of pasteurizing treatments and decontaminating cooling processes in conjunction with the control of the numbers of E. co/i deposited on carcasses during dressing could therefore largely eliminate the dressing process as a source of the enteric pathogens associated with meat.
ACKNOWLEDGEMENTS We thank the staff of the beef packing plant involved in this study for the provision of information on plant processes and for assistance with the collection of samples. Funding for this study was provided by the Beef Industry Development Fund of Agriculture and Agri-Food Canada. REFERENCES Anderson, M. E. and Marshall, R. T. (1989) Interaction of concentration and temperature of acetic acid solution on reduction of various species of microorganisms on beef surfaces. Journal of Food Protection 52, 312-315.
Brown, M. H. and Baird-Parker, A. C. (1982) The microbiological examination of meat. In Meat Microbiology, ed. M. H. Brown, pp. 423-520. Applied Science Publishers, London. Cygnarowicz-Provost, M., Whiting, R. C. and Craig Jr., J. C. (1994) Steam surface pasteurization of beef frankfurters. Journal of Food Science 59, l-5. Davey, K. R. and Smith, M. G. (1989) A laboratory evaluation of a novel hot water cabinet for the decontamination of sides of beef. International Journal of Food Science and Technology 24, 305-316.
Dorsa, W. J., Cutter, C. N., Siragusa, G. R. and Koohmaraie, M. (1996) Microbial decontamination of beef and sheep carcasses by steam, hot water spray washes, and a steam-vacuum sanitizer. Journal of Food Protection 59, 127-135.
Gill, C. 0. and Bryant, J. (1997) Assessment of the hygienic performances of two beef carcass cooling processes from product temperature history data or enumeration of bacteria on carcass surfaces. Food Microbiology, in press. Gill, C. O., McGinnis, D. S., Bryant, J. and Chabot, B. (1995) Decontamination of commercial, polished pig carcasses with hot water. Food Microbiology 12, 143-149. Gill, C. O., Badoni, M. and Jones, T. (1996) The hygienic effects of the trimming and washing operations in a beef carcass dressing process. Journal of Food Protection 59, 666-669. Hardin, M. D., Acuff, G. R., Lucia, L. M., Osmar, J. S. and Savell, J. W. (1995) Comparisons of methods for decontamination from beef carcass surfaces. Journal of Food Protection 58,368-374. Jericho, K. W. F., Bradley, J. A., Gannon, V. P. J. and Kozub, G. C. (1993) Visual demerit and microbiological evaluation of beef carcasses: Methodology. Journal of Food Protection 56, 114-119.
Kilsby, D. C. and and Pugh, M. E. (1981) The relevance of the distribution of micro-organisms within batches of food to the control of microbiological hazards from foods. Journal of Applied Bacteriology 51, 345-354.
Kochevar, S. L., Sofos, J. N., Bolin, R. R., Reagan, J. 0. and Smith, G. C. (1997) Steam vacuuming as a pre-evisceration intervention to decontaminate beef carcasses. Journal of Food Protection 60.107-113.
276
C. 0. Gill, J. Bryant
Lowry, P. D. and Gill, C. 0. (1985) Microbiology of frozen meat and meat products. In Microbiology of Frozen Fooa!s, ed. R. K. Robinson, pp. 109-168. Elsevier Applied Science, London. Regan, J. O., Acuff, G. R., Buege, D. R., Buyeck, M. J., Dickson, J. S., Kastner, C. L., Marsden, J. L., Morgan, J. B., Nickelson, R., Smith, G. C. and Sofos, J. N. (1996) Trimming and washing of beef carcasses as a method of improving the microbiological quality of meat. Journal of Food Protection 59, 751-756.
Sheridan, J. J., Doherty, A. M., Allen, P., McDowell, D. A., Blair, I. S. and Harrington, D. (1997) The effect of vacuum and modified atmosphere packaging on the shelf-life of lamb primals, stored at different temperatures. Meat Science 45, 107-l 17. Smith, M. G. and Graham, A. (1978) Destruction of Escherichia coli and salmonellae on mutton carcasses by treatment with hot water. Meat Science 2, 119-128. USDA (1996) US Department of Agriculture, Food Safety and Inspection Service. Pathogen reduction; hazard analysis and critical control point (HACCP) systems; final rule. Federal Register 61, 3880538989.