Janet E L Corry

The prevalence of campylobacters and arcobacters in broiler chickens.

Chicken carcasses from a supermarket and from a poultry abattoir were examined using methods designed to isolate as many strains of campylobacters and related organisms as possible. Strains of arcobacter, but no campylobacters, were isolated from every carcass after enrichment. Campylobacter jejuni subsp. jejuni was isolated from all carcasses examined by direct plating and other Campylobacter‐like strains were isolated from nine out of 15 abattoir carcasses by direct plating but not after enrichment. Only the Camp. jejuni subsp. jejuni strains could be identified to species level using a readily available identification scheme and/or a commercial identification kit. Examination of caecal contents from the 15 abattoir poultry yielded Camp. jejuni subsp. jejuni and Campylobacter‐like strains from 15 and eight by direct plating, and from six and nine after enrichment, respectively. Four sites in the intestine of the abattoir birds (60 samples) were examined for arcobacters and only one strain was isolated. This indicates that arcobacters are probably not normal inhabitants of the poultry intestine. Poultry is a rich source of other campylobacteria besides the thermophilic Campylobacter spp.

Sources of salmonella on broiler carcasses during transportation and processing: modes of contamination and methods of control

Aims: The prevalence and types of salmonella in broiler chickens during transportation and during slaughter and dressing were studied. This was part of a comprehensive investigation of salmonellas in two UK poultry companies, which aimed to find the origins and mechanisms of salmonella contamination.

Diversity and prevalence of Arcobacter spp. in broiler chickens

Ninety‐nine strains of Arcobacter spp., isolated from 10 chicken carcasses purchased from a supermarket and 15 chicken carcasses collected from a poultry abattoir, were speciated using a variety of phenotypic identification methods. All were tested using API Campy test strips and the 16‐test (Preston) identification scheme developed for campylobacters. Fifty strains were selected for examination using a more comprehensive probabilistic identification scheme, and the identity of representative strains confirmed by protein profiling using SDS‐PAGE. All 25 carcasses yielded Arcobacter butzleri. Three supermarket and 10 abattoir carcasses also carried A. cryaerophilus, and two abattoir carcasses carried A. skirrowii. The API Campy scheme proved unsatisfactory for identifying these strains: only 20 of 99 strains were accurately identified, all of which were A. cryaerophilus, the only Arcobacter sp. included in the database. Moreover, 76 of 99 strains were misidentified. The 16‐test scheme identified all the arcobacter strains as A. cryaerophilus, since neither A. butzleri nor A. skirrowii had been described when the scheme was developed. The computer‐assisted probabilistic scheme succeeded in identifying all but one strain, the identity of which was clarified by the use of SDS‐PAGE. To our knowledge this is the first time that arcobacters other than A. butzleri have been reported in poultry meat or any other food of animal origin. Their high prevalence in poultry products may be of significance to public health.

The isolation and prevalence of campylobacters from dairy cattle using a variety of methods

Faecal samples from 94 dairy cows and 42 calves in three different herds were examined by a variety of techniques for campylobacters. Cefoperazone amphotericin teicoplanin (CAT) agar, modified cefoperazone charcoal deoxycholate agar (mCCDA), Karmali agar, and membrane filtration onto blood agar, were used with and without enrichment in CAT broth. Seventy‐nine percent of cattle in herd A carried campylobacters, compared with 40% and 37·5% of cattle in herds B and C, respectively. Most animals carried only one species of Campylobacter. Campylobacter hyointestinalis was isolated most frequently (32% animals positive) with Camp. fetus subsp. fetus and Camp. jejuni subsp. jejuni detected in 11% and 7% of animals, respectively. In addition, a novel biotype of Camp. sputorum was isolated from 60% of 47 cows tested in herd A. Direct plating detected only two of the total of 40 animals positive for campylobacter. Enrichment in CAT broth before membrane filtration onto blood agar or CAT agar were the most successful methods of plating. Campylobacter sputorum was isolated from CAT agar and blood agar but not from mCCDA or Karmali agar. Karmali agar incubated at 30 °C was especially effective for isolating Camp. fetus subsp. fetus.

Hygiene aspects of modern poultry chilling

An evaluation was made of six commercial poultry chilling systems in relation to factors affecting microbial-contamination of carcasses. These systems included water immersion chilling, air chilling and air chilling with evaporative cooling using water sprays. Samples of neck skin and body cavity were taken from carcasses, together with samples from the chilling environment. These were examined for total aerobic mesophilic microbes and counts of presumptive coliform bacteria and Pseudomonas spp. at specific points in the chilling process. Physical measurements included surface and deep-muscle temperatures of carcasses, water temperatures and chlorine concentrations in the immersion system and air speed and temperature during air chilling. The results obtained for water immersion chilling confirmed previous experience that the washing effect reduces microbial contamination of carcasses, although initially the numbers of pseudomonads tended to increase. The air chillers varied in design and mode of operation, but had little overall effect on microbial contamination of the skin. When a completely dry process was used, microbial numbers were reduced approximately ten-fold in the body cavity. However, the use of water sprays tended to increase contamination of the cavity, while relatively heavy spraying using non-chlorinated water, resulted in a substantial increase in the numbers of pseudomonads.

Evaluation of a new arcobacter enrichment medium and comparison with two media developed for enrichment of Campylobacter spp.

The productivity of an arcobacter enrichment medium (AM), newly developed by Oxoid was compared with two campylobacter enrichment media (Preston broth (Oxoid) and LabM broth), with arcobacter basal medium (ABM) as control. Twenty strains of Arcobacter and Campylobacter spp. were tested for growth, with target inocula of < 4 cfu per ml of medium. Incubation was carried out aerobically for 48 h with tightly closed caps at 25 degrees C for arcobacters and 37 degrees C for campylobacters. After incubation the numbers of cfu in the broths were counted by surface-plating on blood agar. None of the Campylobacter spp. grew in the complete AM, and only one grew (very poorly) in the arcobacter basal medium. However, AM supported good growth of all three species of Arcobacter (A. butzleri, A. skirrowii and A. cryaerophilus) which have been associated with human and animal disease. The one strain of A. nitrofigilis tested failed to grow in any of the media except poorly in ABM. None of the Arcobacter spp. grew in Preston Broth, but nine grew in LabM broth, although productivity was poor compared to AM. None of the Campylobacter spp. grew in AM and only one grew (very poorly) in ABM.

Salmonella, Campylobacter and Escherichia coli 0157:H7 decontamination techniques for the future

Raw meat, particularly poultry meat, remains an important, and probably the major source of human infection with campylobacters and salmonellas. In spite of decades of effort it has so far proved extremely difficult to raise food animals free of these pathogens. For the foreseeable future, therefore, the most effective approach must be to decontaminate the final raw product. In this way numbers of these pathogens entering kitchens and commercial food processing premises will be reduced substantially, and hence opportunities for cross-contamination onto ready-to-eat foods or for survival during cooking or other processes will be much lower. The ideal method of decontamination will have the following attributes: it will not change appearance, smell, taste or nutritional properties; it will leave no residues; it will pose no threat to the environment; it will encounter no objections from consumers or legislators; it will be cheap and convenient to apply; it will improve the shelf life by inactivating spoilage organisms as well as pathogens. Various techniques will be listed and their potential assessed (see Table 1).

Wide variety of bioserotypes of enteropathogenic Yersinia in tonsils of English pigs at slaughter

The tonsils of 630 pigs from 45 English farms using three different rearing methods (Assured British Pigs, Open Management and Organic) were examined between 2003 and 2005 in order to investigate if the low incidence of human yersiniosis could be attributed to a low prevalence of enteropathogenic Yersinia among English pigs. In addition, different isolation methods were compared, possible differences in prevalence among pigs were studied, as well as the prevalence of different bioserotypes of enteropathogenic Yersinia. A high prevalence and a wide diversity of bioserotypes of enteropathogenic Yersinia compared to other European countries were observed. The prevalence of pathogenic Yersinia enterocolitica was 44% and of Yersinia pseudotuberculosis 18%. Overall, 60% of pigs carried enteropathogenic Yersinia. Y. pseudotuberculosis was detected on 78% of farms and Y. enterocolitica on 69%. The most common bioserotypes of Y. enterocolitica were 2/O:9 (33%) and 2/O:5 (26%), and of Y. pseudotuberculosis 2/O:3 (34%), 1/O:1 (26%) and 1/O:4 (24%). Cold enrichment gave the highest isolation rate for both species. Y. enterocolitica was more prevalent (P<0.001) and Y. pseudotuberculosis less prevalent (P<0.05) in winter than in summer in Eastern England. Y. enterocolitica was more common in Eastern England and in assured British pigs, whereas Y. pseudotuberculosis was more common in Western England and in organic pigs. Y. pseudotuberculosis 1/O:1 was predominant (P<0.05) in Western England. Types 1/O:4 (P<0.05) and 2/O:3 (P<0.001) predominated in Eastern England. The high prevalence of Y. enterocolitica bioserotypes 2/O:9 and 2/O:5 found in this study suggests that English pigs are an important reservoir of these bioserotypes whereas in other European countries bioserotype 4/O:3 predominates.

Prevalence of campylobacters and arcobacters in ducks at the abattoir

Ten duck carcasses, five from each of two different flocks, and four pairs of pooled duck caecal contents, each pair from a separate flock, were examined by a variety of techniques for arcobacters and campylobacters. Campylobacter coli, C. jejuni ssp. jejuni, C. upsaliensis, Arcobacter cryaerophilus and A. butzleri were isolated from duck caecal contents. Campylobacter coli, C. jejuni ssp. jejuni, A. cryaerophilus, A. butzleri and A. skirrowii were isolated from carcasses. The most effective methods for isolating these bacteria from carcasses involved selective enrichment in campylobacter enrichment broth, containing a cefoperazone, amphotericin, teicoplanin supplement, followed by plating onto modified charcoal cefoperazone deoxycholate agar (mCCDA), or plating onto non‐selective blood agar after filtration through a 0·65 μm pore size cellulose acetate filter. In contrast, recovery from caecal contents was most effective by direct plating onto mCCDA. API test strips performed poorly, failing to identify A. skirrowii or A. butzleri (which are not included in the scheme), or even many common campylobacters. The Preston biochemical characterization scheme was more helpful, though it did not distinguish between Arcobacter species. The species of most isolates of campylobacter, identified using the Preston scheme, was confirmed by the use of SDS‐PAGE of whole cell proteins and this technique was also used successfully to speciate arcobacters.

Microbial contamination of food refrigeration equipment

Refrigeration systems in chilled rooms in 15 plants processing a variety of foods were studied. These included plants processing raw meat and salads, Chinese ready meals, dairy products, slicing and packing of cooked meats and catering establishments. An initial survey of total numbers of microbes at a total of 891 sites on evaporators, drip trays and chilled room walls was followed up with a more detailed examination of 336 sites with high counts, selecting for Listeria spp., coliforms, enterococci, Staphylococcus aureus and Bacillus cereus. Temperatures (particularly air on and air off, maximum and near defrost heaters) relative humidity, airflow, layout and cleaning regimes were surveyed. In general, no correlation could be found between any of the physical measurements and the numbers and types of bacteria detected. Maximum mean temperatures in the chilled rooms varied from -1 to +16.9 °C and few chilled units were regularly cleaned. Twenty five percent of sites examined had more than 105 colony-forming units per cm2, although, very few pathogens or faecal indicator bacteria were detected. Listeria spp. were not found and coliforms were found only once, in low numbers. Low numbers of S. aureus or B. cereus were present in 9 of the 15 plants, B. cereus was found on evaporators and associated drip trays in two catering plants and two plants processing cooked meat. Enterococci and S. aureus were found most frequently in a raw red meat slaughterhouse (always in low numbers). In general, microbial contamination was lower in rooms where wrapped rather than unwrapped products were stored. The type of product also affected the degree of contamination, with raw red meat and poultry or dry ingredients giving highest counts, and raw vegetables and cooked products lowest. The work demonstrated that bacteria were present on evaporator cooling coils in all factory cold rooms visited. Although evaporator-cleaning procedures were carried out in some factories as part of routine maintenance these were not shown to be effective at maintaining low levels of bacteria on evaporators. To maintain evaporator hygiene it is suggested that more regular cleaning procedures, possibly by means of automated cleansing systems, should be considered.