Section 3: 2008 Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) annual report – Research collaboration

Section Three – Public Health Agency of Canada Research Collaborations

Box 1. Antimicrobial-resistant bacteria in companion animals in Ontario.

Several CIPARS-affiliated research projects have been undertaken to investigate the existence of antimicrobial-resistant bacteria in dogs and cats in Ontario. The findings of 3 recent studies are described here briefly.

Occurrence of antimicrobial-resistant bacteria in healthy dogs and cats presented to private veterinary hospitals in Southern Ontario: a preliminary study.

Murphy C,^{Footnote 1} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Prescott JF,^{Footnote 3} Bonnett BN,^{Footnote 1} Poppe C,^{Footnote 2} Boerlin P,^{Footnote 3} Weese JS,^{Footnote 3} Janecko N,^{Footnote 1} McEwen SA^{Footnote 1}

The prevalence and patterns of antimicrobial susceptibility of fecal bacteria were determined for healthy dogs (n = 188) and cats (n = 39) from private veterinary hospitals in Southern Ontario. The animals had no recent exposure to antimicrobials. The study was carried out in the summer of 2002. Escherichia coli was recovered from all dogs and cats. On the other hand, no Salmonella, extended-spectrum β-lactamase-producing E. coli, methicillin-resistant Staphylococcus aureus, or methicillin-resistant Staphylococcus pseudintermedius were recovered.

The prevalence of antimicrobial resistance in E. coli was as follows: ampicillin—dogs, 13% and cats, 4%; cephalothin—dogs, 13% and cats, < 1%; streptomycin—dogs, 17% and cats, 2%; and tetracycline—dogs, 11% and cats, 2%. Eleven percent of dogs and 15% of cats had E. coli isolates that were resistant to at least 2 antimicrobials. Cephamycinase (blaCMY-2)-producing E. coli were cultured from the feces of 2 dogs. The prevalence of resistance in commensal E. coli from this group of animals was lower than that previously reported for companion animals: however, a small percentage of dogs may be a reservoir for blaCMY-2 E. coli.

Accepted for publication: Canadian Veterinary Journal.

Corresponding author: Colleen Murphy (cmurph02@uoguelph.ca)

Pet-related management factors associated with the presence of Salmonella in the feces of dogs in Ontario.

Leonard EK,^{Footnote 1} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Janecko N,^{Footnote 1} Pearl D,^{Footnote 1} Finley R,^{Footnote 3} Peregrine A,^{Footnote 4} Weese JS^{Footnote 4}

Between October 2005 and May 2006, 138 dogs from 84 households in Ontario were enrolled in a cross-sectional study. The goal of the study was to identify pet-related management factors associated with the presence of Salmonella in feces of dogs from volunteer households. Twenty-three percent (32/138) of dogs had at least 1 fecal sample with positive results for Salmonella, and 25% (21/84) of the households had at least 1 dog shedding Salmonella. Twelve serovars of Salmonella were identified. The most common were S. Typhimurium (33%), S. Kentucky (15%), S. Brandenburg (15%), and S. Heidelberg (13%).

Important risk factors associated with Salmonella shedding included having contact with livestock, receiving a probiotic in the month prior to sample collection, consuming a commercial or homemade raw food diet, consuming raw meat and eggs, and having more than 1 dog in the household. Antimicrobial susceptibility testing of the Salmonella isolates has been completed and epidemiological analyses are in progress.

Corresponding author: Erin Leonard (eleonard@uoguelph.ca)

Pet-related management factors associated with the presence of Campylobacter, Salmonella, and Giardia in the feces of pet dogs visiting veterinary clinics in Ontario.

Leonard EK,^{Footnote 1} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Janecko N,^{Footnote 1} Pearl D,^{Footnote 1} Finley R,^{Footnote 3} Peregrine A,^{Footnote 4} Weese JS^{Footnote 4}

From July 2008 until May 2009, 240 dogs from 7 veterinary clinics in the Region of Waterloo, Ontario were enrolled in a cross-sectional study. The purpose of this study was to identify pet-related management factors that may be associated with the presence of Campylobacter spp., Salmonella, and Giardia spp. in the feces of dogs visiting veterinary clinics. Twenty-two percent (52/240) of the dogs had at least 1 fecal sample positive for Campylobacter. Among Campylobacter -positive dogs, 89% were positive for C. upsaliensis, 14% were positive for C. jejuni, and 1 dog had both C. upsaliensis and C. jejuni. Six percent (14/240) of the dogs had at least 1 sample positive for Giardia spp., and 2% (4/240) had at least 1 sample positive for Salmonella.

Significant risk factors for a dog testing positive for any species of Campylobacter included being less than 1 year of age, participating in a group activity (e.g. obedience or agility training), and having homemade cooked food as their diet or added to their diet. Treatment with antimicrobials in the month prior to sample collection was negatively associated with Campylobacter shedding. Important risk factors for a dog testing positive for Giardia included being less than 1 year of age, living in a rural small town, having a previous enteric illness (infection with Giardia, Salmonella, Campylobacter, or Clostridium difficile ), and drinking well water. Antimicrobial susceptibility testing of the Campylobacter- and Salmonella -positive samples has been completed, and antimicrobial resistance patterns will be compared with those of generic Escherichia coli recovered from the same dogs.

Corresponding author: Erin Leonard (eleonard@uoguelph.ca)

Box 2. Prevalence of selected veterinary and zoonotic pathogens isolated from environmental samples collected from veterinary clinics in Southern Ontario.

Murphy CP,^{Footnote 1>} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Boerlin P,^{Footnote 3} Weese JS,^{Footnote 3} Prescott JF,^{Footnote 3} Janecko N,^{Footnote 1} Hassard L,^{Footnote 4} McEwen SA^{Footnote 1}

The importance of hospital-based infection control in veterinary medicine is increasingly recognized, whereas the role of the clinic environment in hospital-acquired infections is largely unknown. The purpose of this study was to evaluate environmental contamination with Escherichia coli and other selected veterinary and zoonotic pathogens in community veterinary hospitals in Southern Ontario. Over the study period (May through August, 2005), environmental samples were collected from 101 companion animal hospitals. The proportion of hospitals with positive environmental swabs was as follows: E. coli, 92%; Clostridium difficile, 58%; methicillin-resistant Staphylococcus aureus (MRSA), 9%; bla CMY-2 E. coli, 9%; methicillin-resistant Staphylococcus pseudintermedius, 7%; and Salmonella, 2%. Vancomycin-resistant Enterococcus spp., canine parvovirus, and feline calicivirus were not isolated. The prevalence of antimicrobial resistance in the E. coli isolates was low. All Salmonella isolates were susceptible to all antimicrobials evaluated. Susceptibility testing was not performed on the other bacterial isolates.

This study demonstrated that there is an environmental reservoir of pathogens in veterinary hospitals. Important potential veterinary and human pathogens were recovered including Canadian epidemic strains MRSA-2, MRSA-5, and C. difficile ribotype 027. Additional studies are required to characterize risk factors associated with hospital-acquired infections in companion animals, including the role of the environment.

Accepted for publication: Canadian Veterinary Journal.

Corresponding author: Colleen Murphy (cmurph02@uoguelph.ca)

Box 3. Antimicrobial use and resistance on sheep farms in Ontario.

Sheep are considered a minor food animal commodity in Canada, and few antimicrobials are approved for use in sheep and lambs in Canada. Consequently, it was hypothesized that much antimicrobial use would be extra-label drug use (ELDU), which is the use of a drug in any manner inconsistent with label instructions. This antimicrobial use practice in livestock may have public health implications. An antimicrobial use and resistance project was therefore initiated to prospectively gather antimicrobial use and resistance data from sheep farms in Ontario. Each component of this project is presented independently below.

Antimicrobial use on sheep farms in Ontario, Canada.

Moon CS,^{Footnote 1} Berke O,^{Footnote 1} Avery BP,^{Footnote 1},^{Footnote 2} McEwen SA,^{Footnote 1} Reid-Smith RJ,^{Footnote 1}, Scott L,^{>Footnote 1} Janecko N,^{Footnote 1} Menzies P^{Footnote 1}

Producers from 49 lamb-producing sheep farms in Ontario maintained antimicrobial treatment records for a 12-month study period between 2006 and 2008. Farm-level data (i.e. animal management practices and inventories of the number of lambs, ewes, and rams on the farm) were collected via a questionnaire administered to producers at the beginning and end of the study period. Antimicrobial exposure rates (AERs) and rates of extra-label drug use (ELDU; indication, dose or sheep class inconsistent with label instructions) were calculated by use of treatment records and sheep inventories. Treatment-level and farm-level variables were investigated for associations with rates of antimicrobial use by means of Poisson rate regression analysis fit with a generalized estimating equation to control for clustering at the farm level.

Overall, the mean AER for lambs and adult sheep was approximately 66 sheep-days treated per 1,000 sheep-days at risk. Chlortetracycline, an in-feed antimicrobial approved for use in lambs to prevent losses from enterotoxaemia, had the highest mean AER in both lambs (32.7 sheep-days treated per 1,000 sheep-days at risk) and adult sheep (10.6 sheep-days treated per 1,000 sheep-days at risk). Other antimicrobials with high AERs included long-acting oxytetracycline (not licensed for use in sheep) and short- and long-acting penicillin (both products licensed for use in sheep). Among sheep treated with a licensed antimicrobial, on average, the approved product was used in an extra-label manner in 811.6 sheep-days per 1,000 sheep-days treated. The mean rate of using an antimicrobial not licensed for any use in sheep was 191.2 sheep-days per 1,000 sheep-days treated with any antimicrobial. In summary, approximately 20% of use involved a non-licensed product and approximately 80% of licensed antimicrobial use involved some form of ELDU.

Commonly reported diseases such as respiratory illnesses, wounds/infections, or non-specific disease states (e.g. depressed, off feed, or febrile) were significantly ( P≤ 0.05) associated with a lower AER in both lambs and adult sheep. Treatment of non-specific disease, mastitis/udder conditions, and ewes post-lambing were significantly associated with lower rates of non-licensed antimicrobial use in all sheep. Less commonly reported disease states (e.g. abortion or gastrointestinal problem) were significantly associated with higher rates of non-licensed use. These results suggest that the need to treat less common diseases is driving ELDU in sheep in Ontario, presumably because the less common the disease, the less likely it is to be included as a labeled use for any antimicrobial.

Direct comparison of ELDU rates could not be made because of limited documentation in other species. However, the results presented here will be useful in determining whether public health concerns about antimicrobial use in Ontario sheep are warranted and in the development of drug use and licensure strategies for the Canadian sheep industry.

Corresponding author: Catherine Moon (cmoon5@uwo.ca)

Prevalence of antimicrobial resistance among Escherichia coli, Salmonella, and Campylobacter isolated from Ontario sheep.

Scott L,^{Footnote 1} Moon CS,^{Footnote 1} Menzies P,^{Footnote 1} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Berke O,^{Footnote 1} Avery BP,^{Footnote 1},^{Footnote 2} Janecko N,^{Footnote 1>} McEwen SA^{Footnote 1}

Inventories and treatment records for 49 Ontario sheep flocks, including 1 sheep feedlot, were maintained for a 12-month study period between 2006 and 2008. At the initial and final visits, pooled fecal samples were collected from 5 animals from each of 2 groups: weaned lambs and adult ewes. The samples were processed for culture of generic Escherichia coli, Salmonella, and Campylobacter, and all bacterial isolates were subjected to antimicrobial susceptibility testing. Preliminary analysis revealed the prevalence and type of resistance detected at the 2 collection times were similar. Therefore, only the results from the final farm visits are presented here.

A total of 137 pooled fecal samples were collected from 48 flocks. Fecal samples were not collected from 1 farm at the final visit because of flock health problems. All pooled samples had positive culture results for E. coli, and 3 isolates per sample were selected for antimicrobial susceptibility testing (n = 411 isolates). Fourteen percent (56/411) of E. coli isolates were resistant to at least 1 antimicrobial. Resistance to tetracycline was detected in 13% of isolates tested, resistance to streptomycin in 3%, and resistance to sulfisoxazole in 3%. One percent or less of isolates were resistant to each of ampicillin, kanamycin, trimethoprim-sulfamethoxazole, and chloramphenicol. Multidrug resistance among E. coli isolates was low (5%), and no resistance was detected to antimicrobials classified as Category I (Very High Importance in Human Medicine). Only 2 of the pooled fecal samples yielded Salmonella: 1 S. Enteritidis isolate and 1 Salmonella IIIb 61:k1,5,7 isolate. Neither Salmonella isolate was resistant to the antimicrobials tested. The prevalence of Campylobacter spp. was 62% (85/137). Of 85 isolates (1 isolate per positive sample), 86% were C. jejuni, 11% were C. coli, 1% were C. lari, and 2% were other Campylobacter species. Of 82 Campylobacter isolates tested for antimicrobial susceptibility, 53% were resistant to 1 or more antimicrobials. Resistance to tetracycline was detected in 41% of isolates tested, to nalidixic acid in 4%, and to ciprofloxacin in 2%. One percent of isolates were resistant to each of azithromycin, clindamycin, erythromycin, and telithromycin. Little multidrug resistance (4%) was detected among Campylobacter isolates. Further analyses will examine associations between antimicrobial use and resistance in the E. coli and Campylobacter isolates collected from Ontario sheep flocks.

Corresponding author: Lisa Scott (lscott@uoguelph.ca)

Box 4. Prevalence of antimicrobial-resistant bacteria in retail meat from a Northern Ontario First Nations community.

Varughese M,^{Footnote 1},^{Footnote 2} Reid-Smith RJ,^{Footnote 1},^{Footnote 2} Janecko N,^{Footnote 1} Edge V,^{Footnote 3} McEwen SA^{Footnote 1}

Antimicrobial resistance is a critical issue in global healthcare and the transmission of resistant bacteria through the food supply is a growing concern. Although there are documented food- and waterborne outbreaks in First Nations communities, rates of sporadic illness and detection of resistance in food- and waterborne bacteria (i.e. Salmonella, Campylobacter, and Escherichia coli) in humans and via the food supply have not been studied specifically in First Nations communities. As such, a pilot retail meat surveillance project, following the methods established by the CIPARS, was initiated in September 2007 in a remote Northern Ontario First Nations community.

The community was only accessible by plane, but road access was possible for 6 to 8 weeks during the winter. Samples of meat were purchased from the local grocery store, packaged, and shipped by the field worker. Samples were received within 24 hours after the date they were sent from the community and were processed for culturing of E. coli and Salmonella at the Canadian Research Institute for Food Safety, University of Guelph. A portion of each chicken sample was sent to the Laboratory Services Division, University of Guelph for Campylobacter isolation. Salmonella and E. coli isolates were sent to the Laboratory for Foodborne Zoonoses (LFZ) in Guelph, Ontario for antimicrobial susceptibility testing (broth microdilution method) and serotyping/phage typing for Salmonella.Campylobacter isolates were sent to the LFZ in Saint-Hyacinthe, Québec for susceptibility testing (broth microdilution method). Eighty frozen chicken, pork, and beef samples were collected between 2007 and 2008.

Corresponding author: Betsy Varughese (Marie_Varughese@phac-aspc.gc.ca)

Box 5. Antimicrobial-resistant bacteria isolated from wild small mammals in Ontario.

The prevalence of enteric bacteria and antimicrobial resistance has, in general, been well studied in humans and livestock. However, little work has focused on the presence of antimicrobial-resistant bacteria in free-living, wild animals. To determine whether wildlife play a role in the maintenance and dissemination of these bacteria, CIPARS has entered into several research collaborations with the University of Guelph. The results of 1 study are presented below. Other projects investigating antimicrobial resistance in wildlife are currently underway. Together, these studies will provide essential information that will improve our understanding of the role of wildlife in the spread of antimicrobial resistance among bacteria in the environment and of the potential public health risk posed by wildlife. They will also enable us to improve and refine existing surveillance and control programs.

Antimicrobial resistance in Escherichia coli isolates from swine and wild small mammals in proximity to swine farms and in natural environments in Ontario

Kozak GK,^{Footnote 1} Boerlin P,^{Footnote 1} Janecko N,^{Footnote 2} Reid-Smith R,^{Footnote 2},^{Footnote 3} Jardine C^{Footnote 1}

This study was conducted to evaluate the effect of habitat (farm or natural area) on the presence of antimicrobial resistance in generic Escherichia coli isolates obtained from wild small mammals (i.e. mice, voles, and shrews). Additionally, we compared the types and distribution of antimicrobial resistance in E. coli isolated from pigs living on the same farms from which wild small mammals were collected.

Wild small mammals were trapped between June and November, 2007. In total, 42 E. coli isolates were recovered from 22 wild small mammals trapped on farms, and 37 isolates were recovered from 20 wild small mammals trapped in natural areas. Fecal samples from swine were collected between 2005 and 2008, with additional sampling in 2007 to correspond with the wild mammal trapping. All E. coli isolates from wild small mammals and 25 isolates from pooled fecal samples collected from each of 5 swine farms were tested for susceptibility to 15 antimicrobials (Table).

Small mammals caught on farms were 5 times as likely to carry tetracycline-resistant E. coli as were those living in natural areas. Resistance to tetracycline was the most commonly detected resistance in isolates recovered from swine (83% of isolates). Our findings suggest that wild small mammals living on farms are more likely to carry E. coli than are those from natural areas believed to be less impacted by humans and agricultural activities. We hypothesize that proximity to food-animal agriculture increases the likelihood of antimicrobial resistance in E. coli isolated from wild animals, possibly through exposure to resistant E. coli from livestock, to their resistance genes, or to antimicrobials through contact with animal feed.

Published in Applied and Environmental Microbiology 2009; 75:559-566.

Corresponding author: Patrick Boerlin (pboerlin@uoguelph.ca)

Box 6. Methicillin-resistant Staphylococcus aureus in retail meat: 2008-2009.

Weese JS,^{Footnote 1} Rousseau J,^{Footnote 2} Avery BP,^{Footnote 2} Reid-Smith R^{Footnote 1},^{Footnote 2}

Methicillin-resistant Staphylococcus aureus (MRSA) is a critically important human pathogen. Over the past 10 to15 years, there has been a dramatic increase in community-associated MRSA infections internationally, and the role of animals and food has been questioned. In Europe, a particular strain of MRSA, ST398, has emerged in food animals in previously low MRSA-prevalence countries and is now accounting for a large and increasing percentage of human infections. Direct or indirect contact with food animals is a risk factor for MRSA infection and concerns have been expressed about the potential role of meat as a vehicle for MRSA transmission. Given these concerns, prospective surveillance of retail meat was performed.

Retail meat samples were purchased via CIPARS sampling and tested for MRSA contamination. The first study identified MRSA contamination in 31/402 (8%) samples of pork chops, ground pork, and pork shoulders from British Columbia, Saskatchewan, Ontario, and Québec. A strain most commonly associated with horses and horse personnel, CMRSA-5, accounted for 39% of all isolates, whereas 32% were the food-animal-associated strain ST398 and 29% were strain CMRSA-2, a common human epidemic clone. A study was then conducted to detect and quantify MRSA in beef and pork in British Columbia, Saskatchewan, and Ontario. Isolates of MRSA were recovered from 8/127 (6%) ground pork samples, 14/89 (16%) pork chops, and 11/198 (6%) ground beef samples. Fifty-nine percent of positive pork samples were only positive on enrichment culture, with detected levels in quantifiable samples ranging from 20 to 3,590 colony-forming units (CFU)/g. Similarly, 45% of beef samples were positive only on enrichment culture. Therefore, most samples presumably contained very low quantities of MRSA and even samples that were quantifiable tended to have low levels of contamination. Of the quantifiable samples, levels ranged from 20 to 240 CFU/g. All isolates were classified as CMRSA-2. The predominance of this human MRSA clone raises questions about the origin of contamination of meat, particularly considering that ST398, the strain most commonly associated with food animals, was not detected. Retail chicken was also evaluated, with MRSA present in only 1/250 (0.4%) of samples, on both direct and enrichment culture. Only 1 colony was present on direct culture, indicating very low-level (approximately 20 CFU/g) contamination.

Found relatively commonly in retail meat in Canada, MRSA has been reported in some other regions. Strains found in meat are of concern because of their role in human disease, although currently ST398 infections are rare in people in Canada. The relevance of MRSA contamination is unclear. While it is plausible that food could act as a vehicle for MRSA transmission, no objective evidence is yet available. The source of contamination is also unclear, particularly given the 6% prevalence in retail beef yet the inability to find MRSA in feedlot cattle in Canada, based on results of a feedlot study in Alberta where MRSA was not isolated from any of the almost 500 cattle. Other sources of contamination such as slaughterhouse and processing environments, as well as from people in slaughterhouses to retail stores, are also possible.

Corresponding author: J. Scott Weese (jsweese@uoguelph.ca)

Box 7. Clostridium difficile in retail meat.

Weese JS,^{Footnote 1} Rousseau J,^{Footnote 1} Avery BP,^{Footnote 2} Reid-Smith R^{Footnote 1},^{Footnote 2}

Clostridium difficile infection is an important cause of enteric disease in people. Once primarily a hospital-associated pathogen, it appears to be emerging as an important cause of community-associated disease. Further, the epidemiology of C. difficile infection (CDI) is changing, with increased morbidity, mortality, and relapse rates. Much of this has been attributed to the emergence of ribotype 027/NAP1. There is some indication that another strain, ribotype 078/toxinotype V, may be over-represented in community-associated CDI in people. Because these 2 strains have been the most common strains identified in food animals and preliminary studies of food, food has been hypothesized to be a source of infection.

After initial studies indicating the presence of C. difficile in retail meat in Canada, additional studies were undertaken to better estimate the prevalence, strain distribution, and regional distribution in the country. Clostridium difficile was isolated from 7/393 (2%) retail pork samples from British Columbia, Saskatchewan, Ontario, and Québec. The most common strain was ribotype 027. Whereas the infectious dose of C. difficile for humans is not known and is probably variable, the level of meat contamination may be an important factor. Accordingly, a study was conducted to detect and quantify C. difficile in retail pork and beef from British Columbia, Saskatchewan, Ontario, and Québec. Clostridium difficile was isolated from 14/115 (12%) ground beef and 14/115 (12%) ground pork samples. For ground beef, 10 of 14 positive samples were positive on enrichment culture, with samples that were quantifiable only having 120 to 240 spores/g. For ground pork, 10 of 14 samples were positive on enrichment culture only, and 20 to 60 spores/g were identified in quantifiable samples. Ribotype 078 predominated in both beef and pork, and ribotype 027 was also identified. Clostridium difficile was also isolated from 26/208 (13%) retail chicken meat samples from Ontario. All isolates from chicken were ribotype 078 and were only positive with enrichment culture, suggesting that C. difficile was present at very low (< 20 colony forming units/g) levels.

Clostridium difficile is present in a variety of retail meat products across Canada. In general, the levels are low. The relevance of this is unclear. Low-level exposure to C. difficile in meat, water, and vegetables and from environmental sources may be common, and meat may not be a serious concern. It is also possible that food is only a relevant source of infection for people already at high risk, such as those being treated with antimicrobials, people with concurrent disease, and immunosuppressed individuals. However, the presence in retail meat of C. difficile strains that are important in community-associated infections and the ability of C. difficile spores to survive cooking indicate that further study of the relevance of this is needed.

Corresponding author: J. Scott Weese (jsweese@uoguelph.ca)

Box 8. Characterization of antimicrobial resistance in Escherichia coli, enterococci, and Salmonella recovered from retail meat in Alberta.

Aslam M,^{Footnote 1} Bohaychuk V,^{Footnote 2} Checkley S,^{Footnote 2} Diarra MS,^{Footnote 3} Avery B,^{Footnote 4} Reid-Smith RJ^{Footnote 4}

The objective of this study was to characterize antimicrobial resistance in Escherichia coli, Enterococcus, and Salmonella isolated from retail meat samples in Alberta. The sampling plan used by CIPARS was followed and involved continuous weekly sampling from retail stores in randomly selected census divisions, weighted by population. A total of 564 samples comprising chicken (n = 206), beef (134), pork (133), and turkey (91) meats were collected. Generic E. coli, enterococci, and Salmonella were isolated, and isolate identities were confirmed by means of standard culture, biochemical, and polymerase chain reaction methods.

A total of 849 E. coli isolates and 1,079 Enterococcus isolates comprising 2 isolates from each of the 4 meat types were analyzed for antimicrobial resistance. Three isolates of Salmonella were selected from each positive sample for a total of 333 isolates. Antimicrobial susceptibility to 15 antimicrobials for E. coli and Salmonella and 17 antimicrobials for enterococci was determined by use of an automated system. The results were interpreted according to the Clinical Laboratory Standard Institute guidelines.

Antimicrobial resistance was more common in E. coli isolates recovered from chicken and turkey samples than in isolates from beef and pork samples. Thirty-six percent and 23% of¿ E. coli isolates from chicken were resistant to amoxicillin-clavulanic acid and ceftiofur, respectively. Both of these antimicrobials are classified as Category I agents (Very High Importance in Human Medicine). Resistance to more than 2 antimicrobials was also common among these chicken E. coli isolates.

Enterococcus faecalis was the most common (> 90%) enterococcal species identified, followed by E. faecium (4%). High percentages of enterococci isolated from chicken samples were resistant to erythromycin (47%), lincomycin (94%), and tylosin (27%). All of these antimicrobials belong to Category II of the Veterinary Drugs Directorate's ranking of antimicrobials (High Importance in Human Medicine). A comparatively small number of enterococci from beef, pork, and turkey meats were resistant to these antimicrobials. All enterococci were susceptible to vancomycin.

Salmonella was recovered from chicken, turkey, and pork samples; no Salmonella was recovered from beef samples. The most common Salmonella serotypes identified were Hadar (27% of isolates), Heidelberg (23%), and Kentucky (16%). In Salmonella isolated from chicken and turkey samples, resistance was common to the following antimicrobials: tetracycline (51% chicken; 45% turkey), streptomycin (31% chicken; 30% turkey), amoxicillin-clavulanic acid (22% chicken; 27% turkey), ampicillin (22% chicken; 27% turkey), ceftiofur (22% chicken; 27% turkey), and cefoxitin (22% chicken; 27% turkey). Intermediate susceptibility to ceftriaxoneFootnote 35 (19% chicken; 27% turkey) was also identified. The Salmonella antimicrobial susceptibility results presented here are preliminary because susceptibility results for 99 isolates were pending at the time of publication of this report.

In summary, these preliminary data suggest that resistant E. coli, enterococci, and Salmonella are more prevalent in retail chicken meat (40%) and turkey (27%) than in pork (2%) and beef (0%).

Corresponding author: Mueen Aslam (Aslamm@agr.gc.ca)