Swine carcass microbiological evaluation and hazard analysis and critical control points (HACCP) in a slaughterhouse in Minas Gerais, Brazil

Arruda Pinto, PS.

Departamento de Veterinária, Universidade Federal de Viçosa, Viçosa, Brazil.

ABSTRACT

We evaluated swine carcass surface contamination by aerobic plate count (APC), total coliforms (TC), fecal coliforms (FC) and Escherichia coli (EC) and identified microbiological hazards as well as and critical control points (CCPs) in different steps of the slaughter process through risk quantification. We collected 120 surface swab samples in swine carcasses at a slaughterhouse plant. Samples were collected after carcass scalding/dehairing (point A), before evisceration (point B), after evisceration and carcass splitting (point C) and after 24 hours of cooling (point D). Point A presented significantly higher contamination than points B, C and D for aerobic total bacteria (5.3 logCFU/cm²), TC (1.7 logNMP/cm²), FC (1.3 logNMP/cm²) and EC (0.6 logNMP/cm²). Risk of contamination at point A was also higher than at the other sampling points, with calculated odds ratios of 16.4, 17.9, 20.0 and 13.5, for the above listed microbiological indicators, respectively. The results indicate that point A, the scalding/dehairing step, constitutes an import critical control point in the swine slaughter process.

Key words:  Aerobic plate count, coliforms, Escherichia coli, swine carcasses, HACCP.

Departamento de Veterinária, Universidade Federal de Viçosa, Viçosa, Brazil.

INTRODUCTION

Contamination by gastric juices and feces is a highly targeted aspect of swine slaughtering. The Hazard Analysis and Critical Control Point (HACCP) system currently permits minimizing swine carcass microbial contamination, thereby reducing risk of pathogen-contaminated meat. The slaughter line can be monitored at Critical Control Points (CCPs) to assure that microbiological hazards are prevented, reduced or eliminated (Declan et al., 1999).

Since pork meat is inevitably contaminated during slaughter by intestinal contents, as well as by dirty equipment and even the meat handlers (Carr et al., 1998), process steps identified as potential CCPs should be monitored to control microbial contamination, especially those steps in which improper control results in public health risks and economic loss.

The main operations involved in obtaining high quality, hygienic swine carcasses include such diverse phases as scalding, dehairing, singeing, polishing, evisceration, washing and cooling. There is an elevated risk of swine carcass surface contamination in the slaughterhouse by deteriorating and pathogenic bacteria, including Salmonella sp., Campylobacter sp., Yersinia sp. and E. coli (Hansson, 2001). The slaughter process includes operations that reduce the number of bacteria but does not include any step able to completely eliminate microbial contamination (Rivas et al., 2000).

Yu et al. (1999) investigated different swine slaughter operations in order to establish CCPs and found that singeing, washing and cooling the carcasses led to an accentuated fall in coliform and aerobic total counts while the polishing and evisceration steps resulted in increased contamination. These latter two steps were thus characterized as important CCPs. Borch et al. (1996) and Berends et al. (1997) observed that besides evisceration, the scalding and dehairing steps are operations in which contaminating bacterial counts often increase and should therefore also be considered as CCPs.

Although microbiological evaluation of swine carcasses in the slaughter process has been the object of many experimental studies and reviews over the past few years in Europe, Asia and the USA (Berends et al., 1997; Miller et al., 1997; Korsak et al., 1998; Yu et al., 1999; Rivas et al., 2000; Hansson, 2001; Rho et al., 2001), few accounts of scientific research of this nature in Brazil exist. The few existing studies analyzed lymph nodes of carcasses and fecal samples (Costa et al., 1972; Zebral & Freitas, 1974; Langenegger et al., 1983). Only one recent study of this nature was performed on swine carcass surfaces, but this was limited to the final slaughter phases (Lopes & Oliveira, 2002). There exists a lack of knowledge on conditions in which pork meat is currently obtained in Brazil.

Given the economic importance of pork meat for Brazil, as exemplified by the exceptional 47% increase in exports in 2001 (Abipecs, 2002), and the need for sanitary control of meat to maintain this economic activity, the objectives of this study were to identify microbiological hazards associated with swine carcasses in different stages of the slaughter process and identify swine slaughter process critical control points through risk quantification, using a Brazilian industry as example.

MATERIAL AND METHODS

Sampling and Experimental Design.

This research was developed in a pork slaughterhouse plant located in the state of Minas Gerais, Brazil. The establishment has a capacity for slaughtering 2000 swine/day, but operates with at 600 swine/day, with an average velocity of 200 swine/hour.

One hundred and twenty carcasses were analyzed in four different steps of the slaughter line flow sheet (Figure 1): immediately after the scalding/dehairing (point A) immediately before evisceration (point B), after evisceration and splitting the carcasses (Point C) and after 24 hours of cooling (point D).

In each slaughter step, 30 swine carcasses were randomly sampled by the surface swab method. A total area of 140 cm2 was analyzed on each carcass, subdivided into seven areas of 20 cm2, defined as follows: the external regions of the ham (2), shoulder (2), rib (2) and belly mid-region (1), when the carcass had not yet been eviscerated. After opening the carcass, the external belly region was replaced by the internal mid-region of one of the last ribs. This material was placed on ice and taken immediately to the laboratory for microbiological analyses.

Figure 1. Slaughter fluxogram

Microbiological analyses

On the day collected, 140 mL buffered peptone water  (0.1 %) was added to each Wirl-Parker bag containing the swab samples which were then mixed for one minute in a peristaltic homogenizer (Stomacher), to prepare the homogenates for microbiological analyses.

Aerobic plate count (APC) was carried out in duplicate for each dilution according to Stevenson & Segner (1992).

Total (TC) and fecal (FC) coliform and E. coli (EC) most probable numbers (MPN) were determined according to Hitchins et al. (1992).

Statistical analysis.

Averages and standard deviations of APC as well as TC, FC and EC MPN for each sampling point were calculated by the program BioEstat (Ayres et al., 2000). Differences between means at the different control points of the slaughter steps were analyzed using the Newman-Keuls test (p < 0.05).

Risk of contamination by the microbiological indicators studied at the different points sampling slaughter were quantified by calculating proportional probabilities by the Odds Ratios (OR) test, with results expressed in a two-by-two table, and statistical significance tested by the Chi-squared (X2) test, with a 95% confidence interval. This analysis was carried out using the program EpiInfo, version 6.04b (WHO, 1997).

Since no microbiological standards exist for fresh meat under current Brazilian legislation, the reference limits (critical limit) used to calculate the OR for the slaughterhouse were defined as the average counts of the 120 samples analyzed for each microbiological indicator. The critical limits established were, therefore: 4.02 logCFU/cm² for APC, and 1.06, 0.66 and 0.20 logMPN/cm2 for TC, FC and EC, respectively.

Samples presenting above average counts were considered positive and those with below average counts as negative in constructing the two-by-two table. Risk was identified when the OR was greater than one and statistically significant (Berends et al., 1998; Carr et al., 1998).

Results and discussion

APC and MPN of TC, FC and EC in the samples collected at point A were significantly greater than in the samples collected at points B,C and D, except for TC at point D (Tables 1 and 2). Although no samples were taken before point A, the high contamination at this point seems to be associated with the dehairing process, in which fecal matter transfer to the carcass through contaminated hairs is common. The scalding tank in the slaughterhouse is a batch process without water renewal during slaughter activities, which increases the risk of cross contamination of carcasses. Similar levels of aerobic total bacteria contamination after swine carcass dehairing were reported by Gill & Bryant (1992), who found 4.0 logCFU/cm² and Rivas et al. (2000), who reported 4.3 logCFU/cm².

Table 1. Aerobic Plate Count (APC) in swine carcasses in different slaughter steps (n=120).

*Average + Standard deviation

Average followed by the same letters do not differ significantly (Newman-Keuls test, p<0.05).

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

The microbiological quality of the scalding tank water can lead to high microbial content in the carcasses after the scalding/dehairing step. Lopes & Oliveira (2002) detected increasing numbers of aerobic total bacteria in the scalding water during pork meat processing in a batch tank process, with counts of 2.55, 2.90 and 3.90 logCFU/ml, at the beginning, middle and end of the process, respectively

Table 2. Most Probable Number (MPN) of Total (CT) and Fecal (CF) Coliforms and Escherichia coli (EC) in swinr carcasses (n=120), in different slaughter steps.

*Average + Standard deviation

Average followed by the same letters do not differ significantly (Newman-Keuls test, p<0.05).

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

Fecal contamination at point A may also be associated with the dehairing process. Gill & Bryant (1993) found elevated E. coli counts (4.3 - 5.0 logCFU/g) in dehairer waste, in scalding water (3.0 - 3.7 logCFU/ml) and in dehaired carcasses (1.9 - 3.6 logCFU/cm2). Borch et al. (1996) reported that E. coli counts on the surface of dehaired swine carcasses could reach 3.8 logCFU/cm2.

The decrease in APC of about 1.4 log units (Table 1) at point B may be due to the process performed at the slaughterhouse (Figure 1), since after scalding/dehairing the carcasses passed through a mechanized dry polishing, followed by mechanized singeing under a shower, manual scraping and a chlorinated water (1.2 to 2 ppm) wash before entering the clean slaughter zone, that included sampling points B and C.

Similar results for APC were obtained by Yu et al. (1999) who found a decrease of approximately 1.5 log cycles after singeing swine carcasses. Rivas et al. (2000) also found an aerobic total bacteria average of 3.7 logCFU/cm² in the operations between swine carcass dehairing and polishing, and attributed this decrease to singeing. These results indicate that singeing is the crucial step for this accentuated decrease, although this phase of the slaughter process was not analyzed separately in the present study. On the other hand, Borch et al. (1996) related an increase of 3.0 to 3.8 logCFU/cm² in APC in pre-evisceration steps, and attributed this increase to the utensils used.

Coliform MPN also decreased significantly at point B (Table 2). Troeger (1993) also found that singeing reduced carcass E. coli surface contamination by up to 2 logCFU/cm², while Yu et al. (1999) observed that fecal coliform decreased from 1.4 logCFU/cm² before singeing to 0.4 logCFU/cm² immediately after this step.

The reduction in EC MPN from point A to point B (0.56 to 0.05 logCFU/cm2, Table 2) agrees with the decrease reported by Rivas et al. (2000), from 0.45 logCFU/cm² after dehairing to 0.05 logCFU/cm² after singeing.

An insignificant increase in APC was found between points B and C (3.4 to 3.6 logCFU/cm², Table 1), in the clean zone of the slaughterhouse investigated. Rivas et al. (2000) also found no significant increase in the levels of APC in swine carcasses after evisceration (3.5 logCFU/cm²), or in the final stages of processing (3.8 logCFU/cm²), before cooling. Rho et al. (2001) also reported APC of about 3.0 logCFU/cm2 after evisceration and through cooling. We therefore concluded that evisceration protocols used at the study site do not favor increased contamination and the microbiological counts found were probably influenced mainly by initial contamination.

Berends et al. (1997) emphasize that singeing is the last slaughter step that can lead to a decrease in the number of bacteria in swine carcasses and that final carcass contamination is due to the polishing and evisceration procedures. 

APC in samples from point C (3.6 logCFU/cm²) are comparable to those found by Hansson (2001), who reported counts of 3.4 logCFU/cm2 after the evisceration process.

The evisceration step did not lead to an increase in EC, in contrast to results of Rivas et al. (2000), who observed an increase to 1.1 log/cm², after this slaughter step. These authors comment that low levels of contamination at this stage of the slaughter process reflect Good Manufacturing Practices (GMP) adopted by the slaughterhouses under investigation. These practices include the technique of rectal occlusion prior to evisceration, the separation of clean and dirty zones and consequently less worker movement between these zones and washing the knives between carcasses. All these GMP procedures are also adopted at the slaughterhouse in the present study and can explain the low contamination levels observed.

APC results after cooling, at point D (3.8 logCFU/cm², Table 2), where not significantly higher than those in the carcasses before cooling. Gill et al. (2000) found average APC of 1.9 to 3.8 logCFU/cm² in swine carcasses and also concluded that there was no substantial difference in counts in carcasses before and after cooling. Rho et al. (2001) also observed that APC in swine carcasses after evisceration and through cooling remain at about 3.0 logCFU/cm² in some slaughterhouses.

TC, FC and EC were also detected in samples collected in carcasses cooled for 24 hours (point D), with a significant increase in TC (Table 2). These results disagree with those presented by Yu et al. (1999), who observed that a coliform count of 0.9 logCFU/cm² before cooling that decreased to 0.5 logCFU/cm² after 24 hours cooling. Carr et al. (1998) and Gill et al. (2000) also affirm that the swine carcass-cooling step reduces the number of coliforms and E. coli.

TC, FC and EC contamination at point D was not reduced even with the lactic acid decontamination step, indicating that this is an important control point. In contrast to these results, Biemuller et al. (1973) found that application of acetic acid in swine carcasses before chilling reduced total bacterial contamination. These results found in the present study suggest that coliforms, principally E. coli, are among the enteric pathogenic microorganisms resistant to organic acids described by Smulders & Greer, (1998).

It is also possible that hygiene measures efficiency and temperature control in the cooling room were neglected, leading to favorable conditions for bacterial multiplication, since the surface swine carcass is always contaminated with a variety of microorganisms and a poorly controlled cooling temperature would favor rapid proliferation of pathogenic and deteriorating bacteria on hot carcass surfaces (Gill & Jones, 1992).

The presence of higher TC MPN at point D may also indicate recontamination through handlers, utensils, or between carcasses, since prior to this slaughter point the carcasses pass through final inspection, removal of the head, feet and tail and washing with water that can spread bacteria on the carcass instead of eliminating them, as was discussed by Charlebois et al. (1991) in bovine carcasses.

To evaluate and quantify microbiological risk, we adopted as base category the slaughter point presenting the lowest percentage of positive results, that we arbitrarily identified as OR = 1.0. This was point B for aerobic total bacteria and point C for coliforms.

Point A therefore presented the greatest contamination risk for the different microbiological indicators analyzed, with OR values of 16.4 for APC, 17.9 for TC, 20.0 for FC and 13.5 for EC (Tables 3, 4, 5 and 6). This greater contamination risk was statistically significant. No statistical difference was found for calculated risks at points B, C and D.

Table 3. “Odds Ratio” (OR) of the swine carcass with aerobic plate count above the limit of 4,02 logFCU/cm².

*Statistically significant (p<0,05). (X2)=Chi-squared. IC95%=95% confidence interval

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

Table 4. “Odds Ratio” (OR) of the swine carcass with total coliforms above the limit of 1,06 logNMP/cm².

*Statistically significant (p<0,05). (X2)=Chi-squared. IC95%=95% confidence interval

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

Table 5. “Odds Ratio” (OR) of the swine carcass with fecal coliforms above the limit of 0,06 logNMP/cm².

*Statistically significant (p<0,05). (X2)=Chi-squared. IC95%=95% confidence interval

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

Table 6. “Odds Ratio” (OR) of the swine carcass with Escherichia coli above the limit of 0,20 logNMP/cm².

*Statistically significant (p<0,05). (X2)=Chi-squared. IC95%=95% confidence interval

A=Immediatelly after scalding/dehairing; B=Immdiatelly prior to evisceration; C= After evisceration and cutting of carcasses; D= After 24 hours of cooling

Few studies have examined and quantified risk of contamination by aerobic total bacteria and coliforms in swine carcasses. Berends et al. (1997) quantified an OR for carcass contamination by enteric bacteria of 6.6 and 10.9 in dirty polishing equipment and in poorly conducted evisceration, respectively.

Some differences between results observed in this study and those reported in the literature with regard to variations in contamination in different swine slaughter process segments are justified by the variation in levels of operational, personal, equipment and installation hygiene at each establishment and demonstrate the complexity and heterogeneity of slaughter activities. Therefore different CCPs may be identified in each slaughterhouse, depending on the general slaughter structure available.

The relatively good control aerobic total bacteria and coliforms at points B, C and D may be attributed in part to structural peculiarities of the slaughterhouse evaluated in the present study. This establishment started up activities about two years ago, and therefore is a new industry with fully functioning, unworn equipment and operates a short daily processing period (three hours of slaughter), conditions that do not favor pathogen multiplication of pathogens up to the carcass-cooling step. It should be emphasized that the cleaning phases, principally in the dirty zone, are semi-automatic and broad, permitting good carcass decontamination.

A well-defined separation of the dirty and clean zones exists in the slaughterhouse under evaluation, which may also contribute to low number of environmental and carcass microbial contaminants. It has already been proven that a strong correlation exists between aerobic total bacteria contamination of air and carcasses (Rahkio & Korkeala, 1996).

CONCLUSIONS

The results of the microbiological analyses of the slaughterhouse plant studied permit us to conclude the following. The combined scalding/dehairing phase, sampling point A, presented the highest levels of APC, TC, FC and EC contamination, confirming its importance as a critical control point that should be monitored to avoid increasing microbial counts in the following slaughter steps. The risk of swine carcass contamination was highest after scalding/dehairing (point A). The risks, quantified as OR values, were 16, 17, 20 and 13 times greater, for contamination by APC, TC, FC and EC, respectively than for the base category. No significant differences in risk of contamination by APC, TC, FC and EC were found at the points immediately before evisceration (point B), after evisceration and carcass splitting (point C) and after 24 hours of cooling (point D) were found. We identified the cooling step at point D as a critical control point requiring monitoring since, even with the use of lactic acid before this step, APC, TC, FC and EC MPN increased, although this increase was not statistically significant. The presence of fecal coliforms at almost constant levels in steps B, C and D should be continue to be monitored, even when at low levels, since this group indicates the presence of pathogenic microorganisms of fecal origin

REFERENCES

1. Associação Brasileira da Indústria Produtora e Exportadora de Carne Suína-Abipecs: Relatório Anual Abipecs 2002, 2003.        [ Links ]

2.  Ayres M, Ayres Júnior M, Ayres DL, Santos AS. BioEstat 2.0: Aplicações estatísticas nas áreas das ciências biológicas e médicas. Belém/Brasília: Sociedade Civil Mamirauá / CNPq. 272p, 2000.        [ Links ]

3. Berends BR, Knapen F, van Snijders JMA, Mossel DAA. Identification and quantification of risk factors regarding Salmonella spp. on pork carcasses. Int J Food Microbiol 1997; 36(2/3): 199-206.        [ Links ]

4. Berends BR, Knapen F, van Mossel DAA, Burt SA, Snijders JMA. Salmonella spp. on pork at cutting plants and at the retail level and the influence of particular risk factors. Int J Food Microbiol 1998; 44(3): 207-17.        [ Links ]

5. Biemuller GW, Carpenter JA, Reynolds AE. Reduction of bacteria on pork carcasses. J Food Sci 1973; 38(2): 261-3.        [ Links ]

6. Borch E, Nesbakken T, Christensen H. Hazard identification in swine slaughter with respect to foodborne bacteria. Int J Food Microbiol 1996; 30(1/2): 9-25.        [ Links ]

7. Carr MA, Thompson LD, Miller MF, Ramsey CB, Kaster CS. Chilling and trimming effects on the microbial populations of pork carcasses. J Food Prot 1998; 61(4): 487-9.        [ Links ]

8. Charlebois R, Trudel R, Messier S. Surface contamination of beef carcasses by fecal coliforms. J Food Prot 1991; 54(12): 950-6.        [ Links ]

9. Costa GA, Hofer E, Costa MDM, Silva JAH, Santos JV, Doria JD. Isolation of Salmonella from lymph nodes of pigs slaughtered at the abattoir of Salvador, BA. Mem Inst Oswaldo Cruz 1972; 70: 417-31.        [ Links ]

10. Declan J, Bolton AH, Oser GJ, Cocoma S, Palumbo A, Miller AJ. Integrating HACCP & TQM reductions pork carcass contamination. Food Technol 1999; 53(4): 40-3.        [ Links ]

11. Gill CO, Bryant J. The contamination of pork with spoilage bacteria during commercial dressing, chilling and cutting of pig carcasses. Int J Food Microbiol 1992;  16(1): 51-62.        [ Links ]

12. Gill CO, Bryant J. The presence of Escherichia coli, Salmonella and Campylobacter in pig carcass dehairing equipment. Food Microbiol 1993; 10(4): 337-44.        [ Links ]

13. Gill CO, Dussault F, Holley RA, Houde A, Jones T, Rheault N, Rosales A, Quessy S. Evaluation of the hygienic performances of the processes for cleaning, dressing and cooling pig carcasses at eight packing plants. Int J Food Microbiol 2000; 58(1,2): 65-72.        [ Links ]

14. Gill CO, Jones T. Assessment of the hygienic efficiencies of two commercial processes for cooling pig carcass. Food Microbiol 1992; 9(4): 335-43.        [ Links ]

15. Hansson IB. Microbiological meat quality in high-and low-capacity slaughterhouse in Sweden. J Food Prot 2001; 64(6): 820-5.        [ Links ]

16. Hitchins, AD, Hartman, PA, Todd, ECD: Coliforms - Escherichia coli and its toxins. En: Vanderzant, C., Splittstoesser, D.F. Compendium of methods for the microbiological examination of foods. 3. ed. Washington: American Public Health Association (APHA), p. 325-370, 1992.        [ Links ]

17. Korsak N, Daube G, Ghafir Y, Chahed A, Jolly S, Vindevogel H. An efficient sampling technique used to detect four foodborne pathogens on pork and beef carcasses in nine belgian abattoirs. J Food Prot 1998; 61(5): 535-41.        [ Links ]

18. Langenegger CH, Alfinito J, Langenegger J. Salmonelas isoladas de suínos de abate no estado do Pará. Pes Vet Bras 1983; 3(3): 91-4.        [ Links ]

19. Lopes CMM, Oliveira CAF. Avaliação da contaminação microbiana superficial de carcaças, em diferentes etapas do abate de bovinos e suínos. Revista Higiene Alimentar 2002; 16(92/93): 71-5.        [ Links ]

20. Miller MF, Carr MA, Bawcom DB, Ramsey CB, Thompson LD. Microbiology of pork carcasses from pigs with differing origins and feed withdraw times. J Food Prot 1997; 60(3): 242-5.        [ Links ]

21. Rahkio TM, Korkeala HJ. Microbiological contamination of carcasses related to hygiene practice and facilities on slaughtering lines. Acta Vet Scand 1996; 37(3): 219-28.        [ Links ]

22. Rivas T, Vizcaíno JA, Herrera FJ. Microbial contamination of carcasses and equipment from an Iberian pig slaughterhouse. J Food Prot 2000; 63(12): 1670-5.        [ Links ]

23. Rho M, Chung M, Lee J, Park J. Monitoring of microbial hazards at farms, slaughterhouses, and processing lines of swine in Korea. J Food Prot 2001; 64(9): 1388-91.        [ Links ]

24. Smulders FJM, Greer GG. Integrating microbial decontamination with organic acids in HACCP programmes for muscle foods: prospects and controversies. Int J Food Microbiol 1998; 44: 149-69.        [ Links ]

25. Stevenson KE, Segner WP. Mesophilic aerobic sporeformes. In: Vanderzant, C., Splittstoesser, D.F. Compendium of methods for the microbiological examination of foods. 3. ed. Washington: American Public Health Association (APHA), p. 365-274, 1992.        [ Links ]

26. Troeger K. Scalding and dehairing technology: influence on the bacterial count of pig carcasses. Fleischwirtsch 1993; 73: 1157-60.        [ Links ]

27. World Health Organization-WHO. World processing database and statistics and program for public health (EpiInfo). Versão 6.04b. Genebra: WHO, 1997.        [ Links ]

28. Yu SL, Bolton D, Laubach C, Kline P, Oser A, Palumbo SA. Effect of dehairing operations on microbiological quality of swine carcasses. J Food Prot 1999; 62(12): 1478-81.        [ Links ]

29. Zebral AA, Freitas CA. The occurrence of Salmonella in lymph nodes of seemingly normal swine slaughtered at abattoir of Santa Cruz, Rio de Janeiro. Mem Inst Oswaldo Cruz 1974; 62: 223-36.        [ Links ]