National Microbiological Baseline Study in Broiler Chicken
December 2012 – December 2013

Results

Sampling

Of a total of 10,023 planned samples, 9,615 (or 95.9%) were collected and tested, including 7,961 abattoir and 1,654 retail samples indicating a high compliance with the sampling plan from participating establishments and provinces. Table 1 summarizes the number of planned, received and tested samples during the study. More than 91% of planned samples of each sample type were collected and tested with the exception of weep fluid samples with a proportion of 79.4%. Eight of 38 (or 21.1%) federally-registered poultry establishments did not produce bulk packs of whole carcasses during the study which resulted in a lower sampling rate for this sample type. The sampling plan of a single abattoir in Quebec was modified during the course of the study to reflect the reduction in slaughter activities in favour of a new poultry registered establishment from the Maritimes which started production operations in November 2012 and began sampling for the MBS in May 2013. The inclusion of the establishment in the sampling frame increased the total number of establishments to 38.

Many samples were declared unfit for testing at reception mostly due to their long transit time (3.0%) and/or surface temperature exceeding the acceptable temperature range (2.4%). Replacement samples were collected to ensure that at least 90% of planned samples were obtained and tested for each season for valid statistical analysis. Several part samples that were close to or less than 500 g sample weight did not yield sufficient rinse fluid to perform all microbiological tests during the first month of the study, in particular the qualitative test for Campylobacter. It appeared that part of the BPW added to these samples during the rinsing procedure at the laboratory was absorbed by the skin or muscle tissues, particularly by the SOBI thighs. Therefore, the minimum sample weight for part samples was increased to 700 g for the remainder of the study in order to yield sufficient rinse volume for all tests. In addition, we extended the study for one month to December 2013 to compensate for the reduced number of Campylobacter tests done during the winter season. Sample testing results for the month of December 2013 were combined with those for the winter period (December 2012 to February 2013) for analysis.

Slaughter

A total of 4,541 broiler chicken lots originating from various farms across Canada were sampled at slaughter in 38 federally-registered establishments and tested for Campylobacter and Salmonella. The unweighted prevalence or proportion of positive lots for Salmonella in Canada was 25.6% (CI: 24.3% – 26.9%) and ranged from a minimum of 17.4% (CI: 14.4% – 20.5%) in Midwest provinces to a maximum of 34.3% (CI: 31.4% – 37.2%) in Ontario (Table 2); the prevalence being significantly higher in Ontario than any other province or region except in Atlantic region. The geographical distribution of positive lots shows that the provinces from eastern Canada have higher prevalence than those from western Canada. No seasonal pattern was apparent for Salmonella in broiler chicken lots as little variation was observed during the year (Figure 1). The concentration of Salmonella in broiler chicken caeca was found to be highly variable from <0.3 to >11,000 MPN/g of caecal content (Tables 3a and 3b). Of 502 positive caeca samples tested, 64.9% had a concentration greater than 110 MPN/g.

The proportion of positive broiler chicken lots for Campylobacter in Canada was 24.1% (CI: 22.8% – 25.4%) and ranged from a minimum of 15.7% (CI: 13.4% – 18.0%) in Quebec to a maximum of 41.3% (CI: 37.7% – 44.9%) in British Columbia (Table 4); the prevalence being significantly higher in British Columbia than any other province or region. With the exception of the Maritimes, the distribution of positive lots for Campylobacter shows a spatial trend increasing gradually from Quebec towards the western provinces. The analysis of seasonal variation showed a small decrease in proportion of positive lots from winter to spring followed with a significant increase during summer and fall (Figure 2). Of 4,445 caeca samples tested, 3,370 (or 75.8%) samples were found negative or below the limit of detection of 50,000 CFU per g of caecal content. Out of 1,075 quantifiable caeca samples, at least 1,016 (or 96.9%) had a Campylobacter concentration between 106 to 109 CFU per g of caecal content (Table 5).

Processing

A total of 1,643 whole carcasses and 1,668 part samples from abattoirs were tested for the presence of Salmonella (Table 6). The prevalence of Salmonella on whole carcasses was 16.9% (CI: 15.1% - 18.7%) and significantly lower than the 29.6% (CI: 27.4% – 31.7%) prevalence observed in parts. When analyzed separately, the prevalence of Salmonella on SLBL breasts was 28.3% (CI: 25.6% – 31.0%) and not significantly different from the prevalence observed on SOBI thighs with 31.7% (CI: 28.0% – 35.4%).The seasonal variation of Salmonella on fresh abattoir chicken was examined for whole carcasses and parts separately as they have sufficient sample size for valid statistical comparison. There were no clear seasonal patterns for Salmonella, regardless of sample type. The variation of Salmonella prevalence on whole carcasses or parts was relatively small over the study period such that no significant difference in prevalence was observed among all seasons (Figure 1).

Of the 3,333 fresh abattoir chicken products tested, 781 were enumerated and only 11 (or 1.4%) exceeded 11  MPN/mL of rinse fluid (Table 7). The concentration of Salmonella on fresh abattoir chicken was below 3 MPN/mL for 94.9% of samples, of which 164 (or 21.0%) were below the limit of detection. The geometric mean concentration of Salmonella in rinse fluid of fresh abattoir chicken products ranged from 0.10 to 0.12 MPN/mL and there was no statistically significant difference between carcasses and parts or within part types (Tables 8a and 8b).

A total of 1,646 whole carcasses and 1,675 part samples from abattoirs were tested for the presence of Campylobacter using a quantitative and a qualitative method run in parallel. The prevalence of Campylobacter on fresh abattoir chicken was higher by the qualitative method, regardless of the sample type due to the enrichment broth allowing for cell resuscitation (Table 9). By combining results of both tests, the prevalence on whole carcasses was 27.4% (CI: 25.2% – 29.6%) and significantly lower than on parts with 39.0% (CI: 36.7% – 41.4%). The prevalence of Campylobacter on SLBL breasts was 39.0% (CI: 36.0% – 41.9%) and not significantly different from the prevalence observed on SOBI thighs with 39.2% (CI: 35.3% – 43.0%).The analysis of seasonal variation showed a decrease in proportion of positive whole carcass (significant) and part samples from winter to spring followed with a marked and significant increase during summer and fall for both sample types (Figure 2). For instance, the prevalence of Campylobacter on whole carcasses was 19.8% (CI: 15.9% - 23.7%) in spring and increased to 31.2% (CI: 26.7% - 35.7%) in summer.

Overall, 79.0% of fresh abattoir chicken samples were negative for Campylobacter (below the limit of detection, i.e., <1 CFU/mL) by the quantitative method of MLG 41.01 (Table 10). Out of 701 quantifiable samples, at least 346 (or 49.4%) had a concentration below 10 CFU per mL of rinse fluid. The geometric mean concentration of Campylobacter in rinse fluid of fresh abattoir chicken products ranged from 1.98 to 5.65 CFU/mL and was significantly higher on whole carcasses than on parts, but not when compared with SOBI thighs only (Tables 11a and 11b). The concentration of Campylobacter was also significantly higher on SOBI thighs than on SLBL breasts.

Of the 77 weep fluid samples collected from bulk packs containing multiple carcasses, 28 (or 36.4%) were positive for Salmonella and 15 (or 19.5%) for Campylobacter. The concentration of Salmonella in weep fluid was low ranging from below the limit of detection to 2.4 MPN/mL while the concentration of Campylobacter largely varied from undetected to a maximum of 660 CFU/mL.

A total of 1,643 whole carcasses and 1,591 part samples from abattoir were tested for the presence of generic E. coli (Tables 12a, 12b and 12c). The prevalence of generic E. coli on whole carcasses was 83.9% (CI: 82.2% – 85.7%) and 87.6% (CI: 86.0% - 89.2%) on parts. When parts were analyzed separately, the prevalence of generic E. coli on SLBL breasts was 83.4% (CI: 81.1% – 85.7%) and significantly lower than on SOBI thighs with 95.0% (CI: 92.9% – 96.5%). Out of 2,773 quantifiable samples, 2,294 (or 82.7%) had a concentration within the range of 11 to 1000 CFU per mL of rinse fluid. The geometric mean concentration of generic E. coli in rinse fluid of SOBI thighs was 96.1 CFU/mL and significantly higher than those observed on whole carcasses and SLBL breasts with 50.6 CFU/mL and 34.3 CFU/mL, respectively (Tables 13a and 13b).

Retail

A total of 404 whole carcasses and 1,239 part samples from the retail market were tested for the presence of Salmonella (Table 14). The prevalence of Salmonella on whole carcasses was 21.0% (CI: 17.1% – 25.0%), significantly lower than on parts with 31.6% (CI: 29.0% - 34.2%). When analyzed separately, the prevalence of Salmonella on SLBL breasts (P: 31.4%, CI: 28.3% – 34.6%) was not significantly different from that of SOBI thighs (32.1%, CI: 27.6% – 36.6%).The seasonal variation of Salmonella on fresh retail chicken was examined for all products grouped together as the sample size was not sufficient to analyze each sample type individually for valid statistical comparison. No seasonal pattern was apparent for Salmonella on fresh retail chicken products. The prevalence was higher during the spring (P: 32.4%, CI: 27.7% – 37.1%) and lower in summer (P: 25.3%, CI: 21.1% – 29.5%), but no significant difference was observed among all seasons (Figure 1).

The concentration of Salmonella on fresh retail chicken was low on most samples regardless of sample type, with 95.2% below 3 MPN/mL of rinse fluid (Table 15). The geometric mean concentrations of Salmonella in rinse fluid of fresh retail chicken products ranged from 0.07 to 0.09 MPN/mL and no significant difference was observed between carcasses and parts or within part types (Tables 8a and 8b).

A total of 404 whole carcasses and 1,247 part samples from the retail market were tested for the presence of Campylobacter using a quantitative and a qualitative method run in parallel. The prevalence of Campylobacter on fresh retail chicken was higher by the qualitative method, regardless of the sample type (Table 16). By combining results from both tests, the prevalence was 37.9% (CI: 33.1% – 42.6%) on whole carcasses, not significantly lower than on parts with 43.1% (CI: 40.3% – 45.8%). The prevalence of Campylobacter on SLBL breasts (P: 43.3%, CI: 39.9% – 46.6%) was found to be similar to that of SOBI thighs (P: 42.6%, CI: 37.8% – 47.4%).The analysis of seasonal variation shows a small decrease in proportion of positive retail products from winter to spring (P: 30.1%, CI: 25.5% - 34.7%) followed with a significant increase during summer (P: 50.3%, CI: 45.4% - 55.2%) and fall (Figure 2).

Overall, 78.9% of fresh retail chicken samples were negative for Campylobacter (below the limit of detection (<1 CFU/mL) by the quantitative method (Table 17). Out of 348 quantifiable samples, 231 or 66.4% had a concentration below 10 CFU per mL of rinse fluid. The geometric mean concentration of Campylobacter in rinse fluid of fresh retail chicken products ranged from 1.13 to 3.23 CFU/mL and was significantly higher on whole carcasses than on parts (Table 11), but not when compared with SOBI thighs only (Table 11). The concentration of Campylobacter was also significantly higher on SOBI thighs than on SLBL breasts.

Date modified: