Surveillance

• Abattoir
• Animal Clinical Isolates
• Feed
• Human clinical Isolates
• On-Farm
• Retail
  - Current status (May 2007)

Retail

Retail meat represents a logical sampling node for surveillance of antimicrobial resistance in the food chain as it is the endpoint of the food pathway (i.e. the point of consumer exposure prior to the kitchen). The objective of CIPARS Retail Surveillance is to investigate AMR among bacteria found in food at retail. This surveillance framework can be modified (e.g. food commodities, bacteria, regions) as necessary and function as a research platform to investigate specific questions regarding antimicrobial resistance in the agri-food sector.

The unit of concern is the bacterial isolate cultured from one of the commodities of interest and tested for susceptibility to a standard panel of antimicrobials. Commodities of interest are raw meat and poultry products commonly consumed by Canadians and mirror those commodities sampled in the CIPARS Abattoir and On-Farm Surveillance components including: chicken (skin-on chicken legs or wings), pork (shoulder chops) and beef (ground beef). For ground beef, a systematic selection of extra lean, lean, medium and regular ground beef is employed to reflect the potential heterogeneity of this product in terms of the combinations of fed beef and cull dairy as well as domestic vs. imported meat content. The meat cuts and types were also chosen based on high prevalence with regard to the targeted bacteria and low cost of purchase (Ravel, 2002).

The bacteria of interest in chicken are Campylobacter spp., Salmonella spp, Enterococcus spp., and generic E. coli. In pork and beef, only generic E. coli is cultured, given the low prevalence of Campylobacter spp. and Salmonella spp. in these commodities which was determined in the early phases of the retail program.

The target population is consumers of retail meat in Canada. The majority of retail sampling involves continuous, weekly sample submissions from randomly selected census divisions, weighted by population, in each of the participating provinces. In 2005, retail surveillance data were collected in Québec, Ontario, and Saskatchewan. Using Statistics Canada data, between 15 and 18 census divisions were selected per province by stratified random selection. The strata are formed by creating cumulative population quartiles from a list of divisions in a province sorted by population in ascending order and are summarized below:

In Québec and Ontario:

Strata One	10 divisions selected with two sampling days per division per year
Strata Two	four divisions selected, with five sampling days per division per year
Strata Three	two divisions selected with 10 sampling days per division per year
Strata Four	one division, 20 sampling days per year

In Saskatchewan:

Strata One	nine divisions selected with two sampling days per division per year
Strata Two	five divisions selected, with three sampling days per division per year
Strata Three	two divisions selected with five sampling days per division per year
Strata Four	one division, seven sampling days per year

Field workers in Québec and Ontario conduct one sampling day per week. In Saskatchewan, one sampling day is assigned every other week. Sampling is currently less frequent in Saskatchewan due to present funding constraints, limited laboratory capacity and to avoid store related over-sampling. Samples are collected on Monday or Tuesday for submission to the laboratory (LFZ, Saint-Hyacinthe, Québec) by Wednesday. Samples submitted from outside Québec are sent via 24-hour courier. In each province, two census divisions are sampled on each sampling day. In each census division, a slate of four stores is selected prior to the sampling day based on the type of store. Generally, three chain stores and one independent market or butcher shop are selected for sampling. An exception to this protocol is made in densely populated urban divisions, (e.g. Toronto and Montreal) where two chain stores and two independent markets or butcher shops are sampled to reflect the perceived shopping behaviour in such areas. From each store, one sample of each commodity of interest is collected, providing 11 meat samples per division per sampling day. At one store in each division, one beef sample is dropped in order to minimize over-sampling of this commodity.

If possible, specific store locations are to be sampled only once per sampling year. Using prevalence estimates, sampling protocols are optimized to yield 100 isolates per commodity per province per year (anticipated), plus 20% for lost or damaged samples.

Personal Digital Assistants (PDAs) are used to capture the following store and sample data:

• Type of store
• Number of cash registers – a surrogate measure of store volume
• Sell-by or packaging date
• Product Origin: Canada / USA / Other
• Federal Inspection stamp: Y / N
• “May Contain Previously Frozen Meat” label: Y / N
• Final Processing in store: Y / N
• Price/kg

Individual samples are packaged in sealed ‘zipper’ type bags and placed in 16 litre thermal coolers for transport. The ambient environmental temperature determined the number of ice packs placed in each cooler (e.g. one ice pack for temperatures below 20ºC and two ice packs for temperatures 20ºC or above). Temperature data recording instruments (Ertco Data Logger, West Patterson, NJ, USA) are used to monitor the temperature experience of samples in one or two coolers per sampling day.

Preliminary Results (2005)

The recovery rates and number of isolates submitted for antimicrobial susceptibility testing for each province where retail sampling was performed in 2005 are listed in the table below.

Year	Species	Province	E. coli		Salmonella		Campylobacter spp.		Enterococcus spp.
			Recovery rate	n	Recovery rate	n	Recovery rate	n	Recovery rate	n
2005	Beef	Ontario	81%	184
		Québec	56%	126
		Saskatchewan	79%	119
		Total	71%	429
	Pork	Ontario	59%	179
		Québec	26%	78
		Saskatchewan	30%	48
		Total	40%	305
	Chicken	Ontario	95%	145	9%	26	40%	120	99%	150
		Québec	95%	142	9%	26	34%	103	100%	150
		Saskatchewan	96%	81	14%	21	37%	52	98%	80
		Total	95%	368	10%	73	37%	275	99%	380

Note: Shaded areas represent microorganisms and commodities where no antimicrobial resistance results were presented.

Current status (May 2007)

In addition to the regular CIPARS Retail Surveillance activities in Québec, Ontario and Saskatchewan described above, brief retail sampling pilots have been conducted in the maritime provinces (New Brunswick, Nova Scotia, Prince Edward Island) in 2004 and 2007 and British Columbia in 2005 and 2006.  More regular retail surveillance sampling has been planned and is occurring in British Columbia for 2007. Beyond geographical expansion, CIPARS-Retail continues to be used as a platform to conduct other retail-related research projects involving other commodities such as milk fed veal and lamb.