There have not been recent reports on the prevalence of Campylobacter in retail broiler meat in the USA. Most of the studies include products with skin, and the samples are taken during processing where the carcasses are still intact and before portioning. The more recent publication summarizing the prevalence of Campylobacter spp. in processed carcasses comes from the nationwide microbiological baseline data collection program by the USDA FSIS. These data were collected from July 2007 through June 2008 and showed a prevalence of 40% Campylobacter positive in carcasses post-chill . Yet, most of the broiler meat sold in stores across the US is sold in tray packs and include boneless, skinless products.
Because Campylobacter spp. are at low numbers in retail broiler meat in the USA , concentration by centrifugation  and filtration have been performed to increase the number of Campylobacter cells before plating [8, 22]. Bolton broth was used in this study because this medium has been used most frequently for isolation of Campylobacter from poultry samples [23, 24], and it appears to be one of the best available alternatives to compromise between the inhibition of competitors and the growth of Campylobacter spp. . The data in Table 1 are similar to most recent reports on the prevalence of Campylobacter spp. in retail samples in the US [9, 10, 21]. This prevalence is similar to the data from Belgium , but lower than the reports from Ireland , England [28, 29], Canada , Japan  and Spain . The prevalence among different countries varies from as low as 25% in Switzerland to as high as 100% in the Czech Republic [31, 33, 34].
The low prevalence of Campylobacter spp. in tenderloins has been previously reported [9, 10]. The fluctuation in the prevalence of C. coli and C. jejuni by year has not been previously addressed. However, more surveillance data is necessary to understand the extent of this fluctuation, which may be comprised of an actual variability by year and/or an artifactual variability due to the methodology used for isolation. It has been shown that analyzing more than 25 g of sample increases the chances of recovering positive samples for Campylobacter spp. . Therefore, there is no optimal methodology to determine the true prevalence of these bacteria, and for all purposes the actual prevalence of Campylobacter spp. in retail broiler meat may be underestimated. An optimal methodology that could detect the true number of positive samples and/or the samples with the highest number of Campylobacter spp. would provide a more accurate prevalence for surveillance purposes of these pathogens in retail broiler meat.
There is substantial information suggesting that the predominant Campylobacter spp. present in commercial broiler products are C. jejuni and C.
coli, a trend that is especially clear in industrialized nations [27, 28, 36]. Because Campylobacter spp. are inert, very few biochemical tests are used for identification of species. These tests are mainly performed in qualified laboratories studying the taxonomy of these bacteria where several controls are evaluated in parallel to avoid false identification. Therefore, molecular techniques, mainly the polymerase chain reaction validated by sequencing and Southern blotting, provide simple, robust identification to the species level. In a recent summary of the current Campylobacter spp. worldwide prevalence, C. jejuni was the predominant Campylobacter spp. isolated from retail poultry with the exception of Thailand and South Africa, where the predominant species was C. coli. In some countries, C. coli represents less than 20% of all the Campylobacter isolates found in retail broiler meats [31, 37, 38]; yet, they are at a prevalence that exceeds 20% in live broiler chickens. This difference may be explained by the isolation procedure: direct plating is used to analyze fecal material from live animals, while enrichment is used to analyze retail broiler meat. Both Campylobacter spp. have been found in enriched retail samples , but it is not clear if enrichment procedures hinder one species versus the other, or favor the species that contain more vegetative cells at the beginning of the enrichment procedure.
Although other countries, such as Denmark, have shown a strong seasonal correlation in the prevalence of Campylobacter spp. in broiler flocks and in retail broiler meat , there were no seasonal variations detected in C. jejuni. Although statistical differences were seen for C. coli, a larger database is needed to confirm these results. There is no long-term data to assess the changes in the prevalence of Campylobacter spp. present in retail broiler meats. The results from 2005 clearly show that C. coli was the predominant species. These strains were tested with the same PCR assays as the rest of the data set; therefore, there is no bias in the methodology for identification. These data suggest that the product, the processing plant, the region, and even the season, may impact the prevalence of these pathogens in retail broiler meats.
A large diversity in the PFGE profiles of Campylobacter spp. has been reported in the literature [8, 28], with the greatest diversity found in Campylobacter isolates from broiler chickens . This diversity can be related to the larger database available for broiler chickens. This diversity may also be due to a true variability of types, meaning that Campylobacter strains found in chickens show more diversity than the Campylobacter strains isolated from other animal species. The diversity of Campylobacter strains by PFGE has also been demonstrated in clinical samples. For instance, throughout an infection involving 52 patients, one patient had two different Campylobacter species and four patients had different Campylobacter strains based on PFGE analysis. Although human infections with more than one Campylobacter strain are rare, changes in the PFGE profiles throughout an infection complicates the epidemiological studies of Campylobacter spp. . The collection and analysis of retail samples immediately before consumer exposure is the most appropriate sampling point for the collection of data that can be factored into risk analysis models. Therefore, a PFGE database of retail isolates that could be compared to PFGE patterns from human isolates may provide invaluable information to assess the actual risk of humans acquiring campylobacteriosis via consumption of retail meats.