Microbiological sampling of poultry carcass portions by excision, rinsing, or swabbing

Groups of 25 skin-on thighs or skin-on, skinned, or tumbled breast portions of broiler chicken carcasses were sampled by excision of skin or muscle tissue, rinsing, or swabbing. Counts of total aerobic bacteria, coliforms, and Escherichia coli were recorded for each sample. For all types of carcass portions, the mean log counts and the total log counts obtained for each group of bacteria by excision or rinsing mostly differed by <0.5 log unit. However, the counts obtained by swabbing were generally >0.5 log unit lower than the smaller of the values obtained by the other two sampling methods.     

Microbiological sampling of carcasses by excision or swabbing

Groups of 25 carcasses were obtained by random selection of carcasses at the end of each of eight commercial processes for the dressing or cooling of carcasses. Samples were collected from six groups of pig or beef carcasses by excision or swabbing with sponge, gauze, or cotton wool, with one sample obtained by each of the four methods from a separate, randomly selected site on each carcass. Total aerobic counts, coliforms, and Escherichia coli from each sample were enumerated. Values for the mean log10, log10 mean, and log10 total numbers recovered were calculated for each set of total aerobic counts. Those statistics indicated that the numbers of bacteria recovered by excision or swabbing with sponge or gauze were similar, while the numbers recovered by swabbing with cotton wool were at the lower end of or below the range of the numbers recovered by the other methods. The numbers of coliforms or E. coli recovered from carcasses by sampling areas up to 100 cm2 were too few for the estimation of log mean numbers. Sampling of two groups of carcasses by swabbing with gauze indicated that each 10-fold increase in the area sampled, from 10 to 1,000 cm2, approximately doubled the number of samples from which coliforms or E. coli were recovered. Sampling of six groups of carcasses from one process indicated that the sizes of swabs and volumes of diluent used for processing swabs did not have to be increased proportionally to the area of carcass surface sampled to recover numbers of E. coli proportional to the sampled area. It therefore appears that carcass sampling techniques can be varied widely without compromising the recovery of bacteria, and that the relative efficiencies with which bacteria are recovered by different techniques can be assessed by sampling each carcass in a group of 25 by each of the methods to be compared.     

Assessment of the microbiological conditions of red-meat carcasses from bacterial counts recovered by sampling via excision or swabbing with cotton wool

Samples from 240 carcasses were collected from four animal species (porcine, ovine, bovine and equine). Two samples were taken from each carcass, one using the excision method (EX) and the other the wet–dry swabbing method (SW). Eight areas from each carcass were sampled. Most of the samples obtained by SW revealed total aerobic viable counts (TVC) levels of between 3.1 and 4.0 log CFU cm⁻², while most of the values corresponding to excision were located between 4.1 and 5.0 log CFU cm⁻². Moreover, Enterobacteriaceae (EC) counts were only detected above 3.0 log CFU cm⁻² in 0.85% of the carcasses when the samples were collected by swabbing, while the excision method revealed that 13.75% of the carcasses presented EC greater than 3.0 log CFU cm⁻². TVC and EC by EX revealed statistically significant differences compared to SW, while no significant linear relationship was found between carcass surface bacterial counts obtained by SW and EX.     

Experimental comparison of excision and swabbing microbiological sampling methods for carcasses

Bovine sides, ovine carcasses, and porcine carcasses were individually inoculated by dipping in various suspensions of a marker organism (Escherichia coli K-12 or Pseudomonas fluorescens), alone or in combination with two meat-derived bacterial strains, and were sampled by two standard methods: cotton wet-dry swabbing and excision. The samples were examined for bacterial counts on plate count agar (PCA plate counts) and on violet red brilliant green agar (VRBGA plate counts) by standard International Organization for Standardization methods. Average bacterial recoveries by swabbing, expressed as a percentage of the appropriate recoveries achieved by excision, varied widely (2 to 100%). Several factors that potentially contributed to relatively low and highly variable bacterial recoveries obtained by swabbing were investigated in separate experiments. Neither the difference in size of the swabbed area (10, 50, or 100 cm2 on beef carcasses) nor the difference in time of swabbing (20 or 60 min after inoculation of pig carcasses) had a significant effect on the swabbing recoveries of the marker organism used. In an experiment with swabs preinoculated with the marker organism and then used for carcass swabbing, on average, 12% of total bacterial load was transferred inversely (i.e., from the swab to the carcass during the standard swabbing procedure). In another experiment, on average, 14% of total bacterial load was not released from the swab into the diluent during standard swab homogenization. Use of custom-made swabs with abrasive butts, around which metal pieces of pan scourers were wound, markedly increased PCA plate count recoveries from noninoculated lamb carcasses at commercial abattoirs compared with cotton swabs. In spite of the observed inferiority of the cotton wet-dry swabbing method compared with the excision method for bacterial recovery, the former is clearly preferred by the meat industry because it does not damage the carcass. Therefore, further large-scale evaluation of the two carcass sampling methods has been undertaken under commercial conditions and reported separately.     

Formation of biogenic amines by Gram-negative bacteria isolated from poultry skin

The aim of this study is to explore production of biogenic amines (histamine, tyramine, putrescine, cadaverine, agmatine, spermine and spermidine) by 88 Gram-negative bacteria isolated from poultry skin (41 isolates of family Enterobacteriaceae, 21 isolates of genus Aeromonas, 16 isolates of genus Pseudomonas, and 10 isolates of other Gram-negative rods). Ion-exchange chromatography was employed to analyse the above mentioned amines. Enterobacteria were found to be the largest producers of amines with proved presence of tyramine, agmatine, putrescine, and cadaverine in cultivation broth after incubation of bacteria. Putrescine and cadaverine were the most abundant products. Presence of at least two biogenic amines, i.e. mainly concurrent presence of putrescine and cadaverine, was revealed in 19 enterobacteria strains. Eleven isolates classified into Aeromonas genus produced putrescine and five of them also formed cadaverine. The other observed biogenic amines (histamine, spermine and spermidine) were not found among tested isolates. Production of biogenic amines by any Pseudomonas family isolates was not proved.     

Microbiological sampling of meat cuts and manufacturing beef by excision or swabbing

Groups of 25 beef or pork loin primal cuts or of pieces of stored or not stored manufacturing beef were sampled by excision and by swabbing with cotton wool, sponge, and gauze. Total aerobic counts, coliforms, and Escherichia coli from each sample were enumerated. Values for the mean log10, log10 mean, and/or the log10 total numbers recovered were calculated for each set of 25 bacterial counts. Those statistics indicated that, for product sampled without storage, swabbing with cotton wool or sponge recovered about 30%, and swabbing with gauze recovered about 10% of the bacteria recovered by excision sampling; but that for product sampled after storage, swabbing with cotton wool or sponge recovered about 50% and swabbing with gauze recovered about 15% of the bacteria recovered by excision sampling. However, the incidences of samples positive for coliforms and E. coli were less for stored than for nonstored product with all methods of sampling. The findings indicate that the conditions of meat surfaces, the handling of product, and the state of the microflora might all affect the numbers of bacteria recovered by any sampling technique. Thus, the relationship between the numbers recovered by excision or any selected swabbing technique may differ for different types of noncomminuted, raw meat product.     

Tracking spoilage bacteria in commercial poultry processing and refrigerated storage of poultry carcasses

Four trials were conducted to examine the effect of commercial processing and refrigerated storage on spoilage bacteria in the native microflora of broiler carcasses. Prescalded, picked, eviscerated, and chilled carcasses were obtained from a commercial processing facility, and psychrotrophs in the bacterial flora were enumerated on Iron Agar, Pseudomonas Agar, and STAA Agar. The size of the population of spoilage bacteria on processed carcasses stored at 4 degrees C for 7, 10, or 14 days was also determined. Bacterial isolates were identified and dendrograms of the fatty acid profiles of the isolates were prepared to determine the degree of relatedness of the isolates. Findings indicated that although some processing steps increased the level of carcass contamination by selected bacteria, the number of spoilage bacteria recovered from processed carcasses was significantly (P< or = 0.05) less than the number of bacteria recovered from carcasses entering the processing line. Acinetobacter and Aeromonas spp. were the primary isolates recovered from carcasses taken from the processing line. During refrigerated storage, there was a significant (P < or =0.05) increase in the population of bacteria on the carcasses, and Pseudomonas spp. were the predominant bacteria recovered from these carcasses. Dendrograms of the fatty acid profiles of the isolates indicated that bacterial cross-contamination of carcasses occurs during all stages of processing and that some bacteria can survive processing and proliferate on carcasses during refrigerated storage. Furthermore, cross-contamination was detected between carcasses processed on different days at the same facility. Findings indicate that although poultry processing decreases carcass contamination by psychrotrophic spoilage bacteria, significant levels of bacterial cross-contamination occur during processing, and bacteria that survive processing may multiply on the carcasses during refrigerated storage.     

Identification of lactic acid bacteria isolated from poultry carcasses by high-resolution melting (HRM) analysis

Lactic acid bacteria (LAB) are a complex group of Gram-positive bacteria belonging to different genera with common morphological, metabolic and physiological characteristics. Their classification was initially based on biochemical methods, but nowadays molecular methods are usually applied for the identification of LAB. Herein, real-time PCR assay coupled with high-resolution melting (HRM) analysis was developed for identifying and distinguishing LAB isolates coming from poultry carcasses. The 16S rRNA gene from these isolates was amplified using primers that annealed to conserved regions. The melting curve analysis of the amplicons classified all isolates into ten LAB species and generated ten distinct HRM curve profiles. The results from HRM analysis were compared to those produced by API 50 CH biochemical microkits and ribosomal DNA sequencing, suggesting the superiority of HRM against API. In conclusion, HRM was proven to be a fast, reliable and cost-effective method for identification of LAB isolates. HRM analysis could be used in order to reduce the time needed for the identification assay and the cost of sequencing the entire group of LAB isolates. The melting profile of known LAB species could be used as a reference for the rapid identification of unknown LAB isolates without the need of sequencing.     

An in vitro system for the comparison of excision and wet-dry swabbing for microbiological sampling of beef carcasses

An in vitro system for the comparison of wet-dry swabbing and surface tissue excision was developed to ascertain whether the commonly accepted statement of the advantage (in terms of bacterial recovery) of the tissue excision method is also legitimate when different kinds of bacteria are used. A total of 1,770 sections (2.5 by 10 cm) of bovine skin were individually inoculated on the subcutaneous fat side by spreading various suspensions of marker organisms (nalidixic acid-resistant Escherichia coli, vancomycin-resistant Enterococcus faecalis, and methicillin-resistant Staphylococcus aureus) at different concentrations and sampled by two standard methods: cotton wet-dry swabbing and excision. Most counts from cuts sampled by excision were significantly (P < 0.05) higher than the wet-dry swabs; however, no differences were observed between the control and the sampling method when sections were inoculated with bacterial solutions at a concentration of 10(3) CFU/ml and sampled by excision. For sections inoculated with bacterial solutions at a concentration of 10(3) CFU/ml, counts given as log CFU/25 cm2 ranged from 1.97 (S. aureus sampled by wet-dry swab) to 3.06 (S. aureus sampled by excision). For sections inoculated at a concentration of 10(4), counts given as log CFU/25 cm(2) ranged from 2.15 (E. faecalis sampled by wet-dry swab) to 3.19 (S. aureus sampled by excision). For sections inoculated at 10(5), counts given as log CFU/25 cm(2) ranged from 2.94 (E. faecalis, wet-dry swab) to 3.98 (S. aureus, excision), and for sections inoculated at 106, counts given as log CFU/25 cm(2) ranged from 3.53 (E. coli, wet-dry swab) to 4.69 (S. aureus, excision). The proposed system, which enabled a considerable amount of samples to be analyzed under controlled experimental conditions and a large number of data to be generated in a short time, demonstrated among the tested microorganisms that whereas the excision method recovered the highest number of bacteria, control means were always (with the exception of an inoculum of 10(3)/ml) significantly higher than means from either of the sampling methods. Our results indicate that particular attention should be paid to the diverse microflora that can contaminate carcasses in a given slaughterhouse and that it is not appropriate to generalize by saying that the destructive method is the reference technique for the bacteriological sampling of carcasses in slaughterhouses, especially when the contamination is higher than 10(3) CFU/25 cm(2).     

A comparison of wet-dry swabbing and excision sampling methods for microbiological testing of bovine, porcine, and ovine carcasses at red meat slaughterhouses

A comparison of wet-dry swabbing and surface tissue excision of carcasses by coring was undertaken. Samples from 1,352 bovine, 188 ovine, and 176 porcine carcasses were collected from 70 separate visits to commercial slaughterhouses operating under normal conditions. The mean total aerobic viable bacterial counts (TVCs) for all species sampled by excision was 5.36 log units, which was significantly greater than the 4.35 log units measured for swabbing. Poorly correlated linear relationships between swab- and excision-derived bacterial numbers from near-adjacent carcasses were observed for all three animal species. R2 values for least squares regressions for bovine, ovine, and porcine carcasses were 0.09, 0.27, and 0.21, respectively. The reasons why it was not possible to calculate a factor that allowed the interconversion of bacterial numbers between samples collected by each sampling method were investigated. Uncertainty associated with laboratory analyses was a contributing factor because the geometric relative standard deviations measured for TVCs were 0.174 and 0.414 for excision and swabbing, respectively. Uneven distribution of bacteria at identical sampling sites on near-adjacent carcasses on processing lines was also a contributory factor. The implications of these findings for process control verification were investigated by intensive sampling for 13 weeks in three commercial slaughterhouses. As many as 4 log units of difference in TVCs were observed in duplicate samples collected within a narrow timeframe from near-adjacent carcasses on the processing line. We conclude that it might not be appropriate to institute corrective actions in slaughterhouses on the basis of a single week's test results.     

Microbiological sampling of swine carcasses: a comparison of data obtained by swabbing with medical gauze and data collected routinely by excision at Swedish abattoirs

Swab sample data from a 13-month microbiological baseline study of swine carcasses at Swedish abattoirs were combined with excision sample data collected routinely at five abattoirs. The aim was to compare the numbers of total aerobic counts, Enterobacteriaceae, and Escherichia coli, recovered by swabbing four carcass sites with gauze (total area 400 cm2) with those obtained by excision at equivalent sites (total area 20 cm2). The results are considered in relation to the process hygiene criteria that are stated in Commission Regulation (EC) No 2073/2005. These criteria apply only to destructive sampling of total aerobic counts and Enterobacteriaceae, but alternative sampling schemes, as well as alternative indicator organisms such as E. coli, are allowed if equivalent guarantees of food safety can be provided. Swab sampling resulted in higher mean log numbers of total aerobic counts at four of the five abattoirs, compared with excision, and lower or equal standard deviations at all abattoirs. The percentage of swab and excision samples positive for Enterobacteriaceae at the different abattoirs ranged from 68 to 100% and 15 to 24%, respectively. Similarly, the percentages of swab samples that were positive for E. coli were higher than the percentages of positive excision samples (range 52 to 84% and 3 to 14%, respectively). Due to the low percentage of positive excision results, the mean log numbers of Enterobacteriaceae and E. coli were only compared at two and one abattoirs, respectively, using log probability regression to substitute censored observations. Higher mean log numbers of Enterobacteriaceae were recovered by swabbing compared with excision at one abattoir, whereas the numbers of Enterobacteriaceae and E. coli did not differ significantly between sampling methods at one abattoir. This study suggests that the same process hygiene criteria as those stipulated for excision can be used for swabbing with gauze without compromising food safety. For monitoring of low numbers of Enterobacteriaceae and E. coli, like those found on swine carcasses at Swedish abattoirs, the results also show that swabbing of a relatively large area is superior to excision of a smaller area.     

Effect of dry air or immersion chilling on recovery of bacteria from broiler carcasses

A study was conducted to investigate the effect of chilling method (air or immersion) on concentration and prevalence of Escherichia coli, coliforms, Campylobacter, and Salmonella recovered from broiler chicken carcasses. For each of four replications, 60 broilers were inoculated orally and intracloacally with 1 ml of a suspension containing Campylobacter at approximately 10(8) cells per ml. After 1 day, broilers were inoculated with 1 ml of a suspension containing Salmonella at approximately 10(8) cells per ml. Broilers were processed, and carcasses were cooled with dry air (3.5 m/s at -1.1 degrees C for 150 min) or by immersion chilling in ice water (0.6 degrees C for 50 min). Concentrations of E. coli, coliforms, Campylobacter, and Salmonella recovered from prechill carcasses averaged 3.5, 3.7, 3.4, and 1.4 log CFU/ml of rinse, respectively. Overall, both chilling methods significantly reduced bacterial concentrations on the carcasses, and no difference in concentrations of bacteria was observed between the two chilling methods (P < 0.05). Both chilling methods reduced E. coli and coliforms by 0.9 to 1.0 log CFU/ml. Air and immersion chilling reduced Campylobacter by 1.4 and 1.0 log CFU/ml and reduced Salmonella by 1.0 and 0.6 log CFU/ml, respectively. Chilling method had no effect on the prevalence of Campylobacter and Salmonella recovered from carcasses. These results demonstrate that air- and immersion-chilled carcasses without chemical intervention are microbiologically comparable, and a 90% reduction in concentrations of E. coli, coliforms, and Campylobacter can be obtained by chilling.     

Attachment of bacteria to beef from steam-pasteurized carcasses

The extent to which a bacterial cocktail containing equal numbers of Pseudomonas fragi NCTC 10689, Listeria monocytogenes BL5/2, Salmonella Typhimurium LT2, and Escherichia coli JM 109 attached to loin surface cuts (7 by 5 cm) derived from steam-pasteurized beef carcasses has been evaluated. The extent of attachment was categorized as loosely attached (removed by rinsing), firmly attached (released by stomaching), and irreversibly bound. No significant difference (P > 0.10) in the attachment of bacteria to steam-pasteurized carcasses was found compared with control loin samples that had received no treatment. No significant difference (P > 0.05) was also found in the attachment strength between the different bacterial species tested. Most bacteria inoculated onto the loin cuts were reversibly bound, since they had been removed by rinsing and stomaching. The irreversible attachment of bacteria to loin cuts was found to vary significantly (P < 0.01) among the different carcass sets used but was independent of whether the carcass had undergone steam pasteurization treatment. Use of a bioluminescent strain of E. coli showed that cells bound preferentially to cut edges and convoluted areas on the loin surface and could not be removed by rinsing. The possible mechanisms of bacterial attachment and the suitability of steam pasteurization to remove contamination incurred during slaughter are discussed.     

Class 1 and class 2 integrons in poultry carcasses from broiler house and poultry processing environments

Integrons have been identified as major genetic contributors to the dissemination of antimicrobial resistance in bacteria. The objective of this study was to examine the prevalence of integrons in poultry processing at the broiler house and in processing plants. Class 1 and class 2 integrons were found throughout the processing environment. Of the two classes of integrons, class 1 was the most prevalent in all processing areas. The levels of both classes of integrons decreased from the farm to the processing plant. Within the chiller tank in the processing plant, the persistence of these sequences appears to be related to the free chlorine concentration of the chiller tank water. The variable regions of the amplified integrons showed size diversity (from 680 to 2,000 bp), suggesting diversity in types of antibiotic-resistance-coding gene cassettes. The presence of the class 1 and class 2 integrons in the chlorinated chiller tank suggests that these sequences are capable of withstanding this critical step in the reduction of microbial loads on poultry carcasses. The persistence of the integron gene sequences on the farm and throughout processing highlights the stability of these transmissible antibiotic-resistance-coding nucleotide sequences and their potential role as reservoirs of antibiotic-resistance-coding genetic elements within the poultry rearing and processing environments.     

Diversity of Campylobacter isolates from retail poultry carcasses and from humans as demonstrated by pulsed-field gel electrophoresis

Campylobacter spp. are a major contaminant of poultry. Eating undercooked chicken and handling raw poultry have been identified as risk factors for campylobacteriosis in humans. Previous studies have found Campylobacter spp. on 90% of poultry carcasses. In the present study, pulsed-field gel electrophoresis (PFGE) was used to assess the genetic diversity of strains on retail poultry carcasses. PFGE patterns of isolates from campylobacteriosis cases were compared to those from the poultry isolates. Over a 1-year study period (March 2000 through February 2001), whole fresh young chickens (n = 72) were obtained from three retail outlets in an urban community in the south-central United States. Campylobacter spp. were isolated from 82% of these carcasses. Strains (n = 70) were defined on the basis of their PFGE pattern. Sixty-seven percent of the carcasses from which Campylobacter spp. were isolated were contaminated with more than one PFGE-distinguishable strain. During the 1-year study period, most of the PFGE patterns (59%) were limited to isolates obtained from a single carcass. Forty-one percent of the PFGE-distinguishable strains were recovered from more than one carcass. Ninety-seven percent of the carcasses contaminated with the same strain were purchased at the same time from the same store. To examine the degree of genetic stability, four strains were followed in vitro over an estimated 1,000 doublings. The PFGE pattern of one of these isolates underwent minor changes during in vitro growth. The data indicate extensive variability in the PFGE patterns of Campylobacter spp. isolated from humans and from poultry carcasses. In spite of difficulties caused by such diversity and the fact that some carcasses are contaminated with more than one strain, the pattern variation provides a useful method for linking a particular strain to its source.     

Decontamination of poultry carcasses using steam or hot water in combination with rapid cooling, chilling or freezing of carcass surfaces

Reconstituted infant formula has been implicated in outbreaks of Enterobacter sakazakii infections, causing high mortality and serious sequelae. Current prevention methods appear to be insufficient to ensure that such foods are free of E. sakazakii. In this study, the usefulness of bacteriophages for biocontrol of E. sakazakii was investigated. Of a total of six new E. sakazakii phages isolated from sewage and UV irradiated cultures, two were selected for further study by electron microscopy, DNA restriction analysis and SDS-PAGE of structural proteins. Purified phages were used to control bacterial growth in broth medium and reconstituted infant formula. Both phages effectively prevented development of E. sakazakii in formula at various temperatures (12, 24 and 37 degrees C), the efficiency of which was dependent upon intrinsic lysis properties and the applied phage concentration. We conclude that application of specific bacteriophages may provide a means for efficient prevention of E. sakazakii infection through reconstituted infant formula.     

Comparison of destructively and rinsing gained samples to determine TVC of pig carcasses by bioluminescence

A rinsing-technique with a spraying gun in combination with the ATP-bioluminescence for determining the microbial load was investigated. In each case 3.8-cm(2) pig-rind was rinsed with buffered peptone water using a modified sprinkle-bottle. The sample-suspensions were gathered in a one-way syringe and sent through a filter cascade consisting of a coarse filter in order to remove bigger particles and a sensitive filter to attach the micro-organisms as well as somatic cells. This procedure was followed by an extraction of the ATP from the sensitive filter with CellSolver FBC-solution. The destructively gained samples were determined on their TVC by the conventional method. Results of the carcass-surface spray test combined with bioluminescence compared to the conventional destructive method followed by plate-count showed a correlation r=0.93 with a detection limit down to 10(4) cfu/cm(2).     

Contamination of carcasses with Salmonella during poultry slaughter

Successively slaughtered poultry flocks were sampled for Salmonella to study the relationship between gastrointestinal colonization of the birds and contamination of the carcasses after slaughter. Samples from 56 broiler flocks and 16 spent layer and breeder flocks were collected in six slaughterhouses. Salmonella isolates were serotyped and further characterized by pulsed-field gel electrophoresis (PFGE). Although only 7 (13%) broiler flocks were colonized with Salmonella at slaughter, carcasses of 31 (55%) broiler flocks were contaminated after slaughter. Concerning the layer and breeder flocks, 11 (69%) flocks were colonized in the gastrointestinal tract, but after slaughter, carcasses of all flocks were contaminated. The Salmonella status determined at the farm did not always correlate to the status at slaughter. On the other hand, the slaughter of Salmonella-colonized flocks did not always result in the contamination of the carcasses with the same PFGE types isolated from the gastrointestinal tract. When only uncolonized flocks were slaughtered, the carcasses of flocks were on some occasions still contaminated with Salmonella. This indicates possible cross-contamination from the slaughter equipment or transport crates. These observations show that it is difficult to reach the benefits of logistic slaughter in commercial poultry slaughterhouses.     

Resistance patterns of Campylobacter spp. strains isolated from poultry carcasses in a big Swiss poultry slaughterhouse

The aim of this study was to determine resistance patterns of strains of Campylobacter spp. isolated from poultry carcasses in one of the two big Swiss poultry slaughterhouses. A variety of antibiotics with clinical relevance in human and/or in veterinary medicine was tested. In addition, the results of the disc diffusion method, E-test and microdilution broth methods were compared. Of the 195 Campylobacter jejuni strains isolated from 195 poultry carcasses from 21 flocks, 134 strains were susceptible in vitro to all tested antibiotics. Sixty-one strains (31.3%, from eight flocks) showed resistance. Forty-one strains were resistant to a single antibiotic-34 to streptomycin, 6 to ampicillin and 1 to ciprofloxacin. Eighteen strains (from two flocks) showed combined resistance to erythromycin and streptomycin, two strains to ciprofloxacin and streptomycin. None of the isolates was resistant to tetracycline. The data of this first study in Switzerland show a favourable resistance situation for C. jejuni strains against erythromycin, tetracycline and ciprofloxacin. The disc diffusion method was found to be a reliable and easy tool for monitoring the prevalence of resistant C. jejuni strains. For surveillance of changes in the susceptibility concentration levels to antimicrobial agents, however, a MIC method should be used. Further investigations along the whole poultry production chain (farm, slaughterhouse and retail levels) are now necessary in order to confirm the resistance situation.     

Microbiological quality of retail poultry carcasses in Spain.

A total of 40 eviscerated and refrigerated chicken carcasses were collected from five retail outlets (three supermarkets and two poulterers' shops) in León (Spain). The level of microorganisms on chicken carcasses was assessed using the excised breast-skin technique. Mean counts (log10 CFU/g) of psychrotrophs, pseudomonads, fluorescent pseudomonads, enterococci, Micrococcaceae, Staphylococcus aureus, and yeasts and molds were 4.84, 4.11, 3.32, 2.72, 3.80, 3.67, and 2.99, respectively. A significant correlation coefficient was found between pseudomonads and fluorescent pseudomonad counts (r = 0.827; P < 0.001) and between Micrococcaceae and S. aureus counts (r = 0.915; P < 0.001). Levels of psychrotrophs, pseudomonads, fluorescent pseudomonads, and yeasts and molds were significantly (P < 0.05) higher in supermarkets than in poulterers' shops, possibly due to the longer period of time the carcasses spent in the supermarkets (between 1 and 2 days, as opposed to only 4 to 16 h in the case of poulterers' shops). Carcasses from poulterers' shops showed higher (P < 0.05) counts of enterococci. Micrococcaceae, and S. aureus, which suggests higher storage temperatures in these outlets. Only S. aureus counts (especially those from poulterers' shops) exceeded the established values in the microbiological criteria for poultry meat consulted.