Food Protection Trends - November 2009 - 781

INTRODUCTION In Alberta, Canada, campylobacteriosis is the most common bacterial enteric illness, with 36.1 cases per 100,000 people reported in 2005 (25, 27). Campylobacter jejuni (C. jejuni), the most frequently isolated species in human disease, is responsible for approximately 85% of all human Campylobacter infections (21). While consumption of contaminated poultry meat is generally considered the primary source of infection for people (14), other routes of transmission may exist. Similarity between human and domestic livestock Campylobacter isolates has been reported based on molecular typing studies. (6, 12, 18, 22). In studies in Alberta feedlot cattle near the end of the feeding period, fecal prevalences for Campylobacter spp. and for C. jejuni have been estimated to be up to 87% and 61%, respectively (2, 11, 16). Other species of Campylobacter of potential public health importance, including C. coli, C. fetus, C. hyointestinalis, and C. lanianae, have also been isolated from cattle feces in Alberta (16, 17). However, research into the prevalence of Campylobacter spp. in retail ground beef in Alberta has been limited. In Edmonton, Alberta, a city in northern Alberta which was not part of the sampling area for our study, a recent retail ground beef survey reported no positive samples from the 100 packages tested (4). The prevalence of Campylobacter spp. in retail ground beef has ranged from 0–20% worldwide on the basis of culture and biochemical or molecular identification of species; however, commonly less than 5% of samples tested have identified campylobacters (4, 7, 28, 30). The goals of this project were to assess the prevalence of Campylobacter spp. (in particular C. jejuni) and to investigate risk factors potentially associated with the presence of Campylobacter spp. in retail ground beef. This paper reports the results of a culture survey of retail ground beef (n = 1,200) and PCR of a subset of these (n = 142) from 60 retail grocers of four major chains in three cities in southern Alberta. MATERIALS AND METHODS Sample size calculation For a survey using simple random sampling, 179 packages of ground beef would have been necessary to measure a 3% expected prevalence of C. jejuni (29) with 2.5% precision and 95% confidence (Epi-Info, version 3.01, CDC, USA, 2003). After application of an inflation factor formula (9) to account for clustering of the expected frequency of Campylobacter within retail stores, the survey required 1,200 packages from 60 stores (assuming an intraclass correlation coefficient (ICC) of 0.3, an unadjusted sample size of 179, and collection of 20 packages per store). An ICC describing clustering of C. jejuni within source was not available from previous publications; the choice of 0.3 was slightly more conservative than previously published ICCs for non-enteric cattle conditions (19). Sampling protocol The goal of sampling was to identify grocery chains likely to supply the largest sales volume of ground beef to consumers. Four chains with the highest numbers of retail stores from three cities in southern Alberta were identified, and a sampling frame of individual stores was compiled from telephone book white and yellow pages (chain name and pharmacy headings) and internet searches (chain name). Stratified random sampling (by city and by chain within city) ensured that meat samples were taken from all chains in all cities. Fifteen stores were sampled from chain 1, 22 from chain 2, 16 from chain 3 and seven from chain 4. Forty-six stores were sampled in city 1, six stores in city 2 and eight stores in city 3. Five packages per store per collection were randomly sampled from the 60 stores, using a hand-held randomization program (Handy Randy, Stevens Creek Software, Cupertino, CA, USA), for a total of 1,200 retail packages of regular or lean ground beef. Three hundred packages were purchased during each of four collection periods: two winter (Nov. 21–23, 2004, and Jan. 9–11, 2005) and two summer (May 30–31, June 1, 2005 and July 18–20, 2005). After purchase, each package of ground beef was placed into a pre-labeled Ziploc bag (SC Johnson, Racine, WI, USA) and then packed into a cooler (The Coleman Company Inc., 5286B, Wichita, KS, USA) with six ice packs (Ice-Pak/HotPak, Montreal, QC, Canada). A Hobo H08 Pro temperature monitor (Onset Computer Corporation, Pocasset, MA, USA) was included in one cooler from each of the 12 meat shipments. Each cooler was sealed and shipped to the Vaccine and Infectious Disease Organization (VIDO, Saskatoon, SK, Canada) by bus (Greyhound Transport Canada Corporation) overnight. Ground beef packages were processed within approximately 24 hours of collection. Transport temperature ranges were evaluated from two hours after closure to two hours before the cooler was opened. Employees knowledgeable about in-store meat practices were identified by phone inquiry or observed directly working with meat, and were asked questions regarding their meat department practices. Information on the cutting and packaging of raw poultry, the type of meat used to produce the ground beef (coarse tubes, market trim or both) and whether the ground beef contained meat that had previously been frozen were collected. Experimental inoculation of retail ground beef as sensitivity analysis A pure culture of C. jejuni (NCTC 11168) that had been previously suspended in 25% glycerol/50% Brain Heart Infusion broth and frozen to -70°C was used as the source strain for this experiment. The culture was thawed on ice and plated on a Mueller-Hinton agar plate. The plate was then incubated microaerobically (85% N2, 10% CO2, 5% O2) at 42°C for 48 hours and checked to ensure the culture was pure by use of the Gram stain. The culture was then suspended in 0.85% NaCl (normal saline) to an absorbance of 0.5 at 600nm (Ultrospec® 3000, Pharmacia Biotech) to form a 109 colony forming units (CFU)/ml solution. To create the final 1 × 104, 1 × 103, 1 × 102, or 1 × 101 CFU/g dilutions, C. jejuni stock solution was further diluted with normal saline to a total volume of 1 ml, which was added NOVEMBER 2009 | FOOD PROTECTION TRENDS

Food Protection Trends - November 2009

Table of Contents for the Digital Edition of Food Protection Trends - November 2009