MONITORING OF ANTIMICROBIAL RESISTANCE IN FOOD ANIMAL PRODUCTION
Location: Bacterial Epidemiology and Antimicrobial Resistance
Title: Antibiotics and the R’s – Mechanisms of resistance
Submitted to: American Association of Swine Veterinarians Annual Meeting
Publication Type: Proceedings
Publication Acceptance Date: March 5, 2011
Publication Date: March 5, 2011
Citation: Cray, P.J. 2011. Antibiotics and the R’s – Mechanisms of resistance. American Association of Swine Veterinarians Annual Meeting. March 5-8,2011. Phoenix, Arizona.
Illness attributable to foods and food stuff in the United States (U.S.) does not appear to be diminishing.1 Collectively, Scharff reports the total cost of food borne illness in the U.S. at $152 billion a year. 2 While antimicrobial treatment of uncomplicated gastroenteritis in not typically indicated, when it is warranted3, the presence/development of resistance can result in treatment failure.
Global concern regarding the development of antimicrobial resistant bacteria has not waned since the introduction of penicillin. In fact, concern over the emergence of new resistance patterns, including the development of multidrug resistance (MDR), continues to rise.4 Surveillance/monitoring programs can allow for detection of emerging/increasing resistance over time. However, the specific design and harmonization of antimicrobial surveillance programs is difficult. Currently, the World Health Organization has convened an Advisory Group on Integrated Surveillance of Antimicrobial Resistance (WHO-AGISAR) to minimize the public health impact of antimicrobial resistance associated with the use of antimicrobials in food animals.5 In particular, the Advisory Group assists WHO on matters relating to integrated (global) surveillance of antimicrobial resistance and containment of food-related antimicrobial resistance. One major goal of AGISAR is to develop harmonized schemes for monitoring antimicrobial resistance in zoonotic and enteric bacteria; appropriate sampling is included.
Since 1996 the U.S. has conducted surveillance on food borne and commensal antimicrobial resistance bacteria through the National Antimicrobial Resistance Monitoring System - Enteric Bacteria (NARMS) to prospectively monitor changes in antimicrobial susceptibilities of zoonotic pathogens from human and animal clinical specimens, from healthy farm animals, and from raw product of food-producing animals at slaughter.6,7 Non-typhoid Salmonella was selected as the sentinel organism. The goals and objectives of the monitoring program are to 1) monitor trends in antimicrobial resistance among food borne bacteria from humans, retail meats, and animals, 2) disseminate timely information on antimicrobial resistance to promote interventions that reduce resistance among food borne bacteria, 3) conduct research to better understand the emergence, persistence, and spread of antimicrobial resistance, and 4) assist the FDA in making decisions related to the approval of safe and effective antimicrobial drugs for animals.8
Information from NARMS is released through the publication of annual reports covering the current status of resistance in humans, retail meats, and food animals.8-12 The program underwent a review by the FDA Science Board in 2007 that, while lauding the program’s accomplishments, suggested sampling strategies and in particular animal sampling should be reevaluated and modified. As reviewed by others sampling and provision of contextual information are two major sources of bias which must be addressed within surveillance systems.6-18 For NARMS as a whole, a cost-effective, unbiased and unambiguous process for collecting this information has not been established and a review and analysis of existing sampling strategies is needed within the animal arm of NARMS.
Currently, isolates are collected as part of the USDA FSIS HACCP risk-based inspection program.19 Isolates received under this program are biased toward plants with high microbial risk which exceed the guidelines for Salmonella testing. As a result, there tends to be an over representation of isolates originating from these plants and this does not necessarily provide national representation of the resistance problem in food animals. This bias calls for a re-evaluation and re-design of the sampling strategy.
Materials and methods
Experimental Design on Farm Sampling – Contribution of food borne and commensal isolates originating from animals is best obtained by sample of an individual animal while it resides in the production environment on farm. This is also the only collection point in which drug use and development/persistence of resistance can be evaluated. Since Salmonella is the sentinel organism for NARMS, all calculations are based on Salmonella prevalence. The animal arm of NARMS has been collecting data since 1996 and in order to preserve continuity of information, the current sampling and testing program will continue. However, simultaneous to our continued testing we will begin on-farm sampling in swine, continue participation in NAHMS (when studies occur), and continue discussion with FSIS to explore other sampling and data analysis options (Figure 1).
Initially, a longitudinal study will be conducted in pork (finishing swine). This first study will begin in collaboration with Dr. Wondwossen Gebreyes and the National Pork Board. In the design of this study, sample size calculations were performed using Neyman-Pearson hypothesis testing, where the null hypothesis is no difference between the proportion of Salmonella recovered between samplings at a significance level a=0.05 and power=0.80. The FDA NARMS (Pat McDermott, personal communication) is requesting a minimum of 300 isolates per study population. Given that, approximately 4,200 floor pen samples must be collected to yield a minimum of 300 isolates.
Using recent projections from USDA NASS20, operations with more than 5,000 head accounted for 61.1% of the total hog and pig inventory (projected at approximately 68 million pigs) in 2008 with the ten largest states (Iowa, North Carolina, Minnesota, Illinois, Indiana, Nebraska, Missouri, Oklahoma, Ohio, and Kansas) accounting for over 85% of this inventory. Therefore, we propose to start initial sampling in the top three swine production states, Iowa (20+ million pigs), North Carolina (10+ million pigs) and Minnesota (7+ million pigs) where total inventory would equal 37 + million pigs. Capturing 95% of that inventory would equal 35 million pigs (37 million * 0.95) which is representative of 51.5% of total production inventory (35 million/68 million, total inventory). Based on the collection of 30 pen floor samples per barn, a total of 150 barns could be solicited for collection of samples. Dividing this among the 3 states equals 50 barns per state. Based on National Animal Health Monitoring System Swine 2006 projections of 7.2% Salmonella prevalence on-farm21, the expected total number of Salmonella isolates recovered would equal approximately 324 which exceed the NARMS requirement.
Seasonal differences in prevalence among bacteria are known and after the first year of collection we will attempt to sample these same barns seasonally. Further considerations will be given to increasing the numbers of samples collected, number of states sampled, inclusion of small and medium producers, as well as expansion into the other commodities using similar sampling calculations as funding and additional collaborators are added.
Susceptibility testing and mulit-drug resistance - Antimicrobial resistance determination is conducted using custom made plates containing up to 16 antimicrobics in a semi-automated minimal inhibitory concentration format system (SensititerTM, Trek Diagnostic) according to manufacturer’s directions. Antimicrobials tested are the same across all three arms of NARMS. Isolates will be classified as susceptible, intermediate, or resistant based on Clinical Laboratory Standards Institute established breakpoints, where available.22,23 Multi-drug resistance (MDR) is defined as resistance to greater than or equal 2 antimicrobial classes.
Results and discussion
On-farm sampling – Implementation of a new sampling design is expected within the second quarter of 2011. Results will be incorporated into the NARMS reports and listed as ‘on-farm’. It is expected that this information will significantly impact our understanding of resistance throughout the food production continuum