Skip to main content
NCBI home page
The Canadian Veterinary Journal logoLink to The Canadian Veterinary Journal
. 2012 Aug;53(8):841–848.

Metrics for quantifying antimicrobial use in beef feedlots

Katharine M Benedict 1, Sheryl P Gow 1, Richard J Reid-Smith 1, Calvin W Booker 1, Paul S Morley 1,✉
PMCID: PMC3398520  PMID: 23372190

Abstract

Accurate antimicrobial drug use data are needed to enlighten discussions regarding the impact of antimicrobial drug use in agriculture. The primary objective of this study was to investigate the perceived accuracy and clarity of different methods for reporting antimicrobial drug use information collected regarding beef feedlots. Producers, veterinarians, industry representatives, public health officials, and other knowledgeable beef industry leaders were invited to complete a web-based survey. A total of 156 participants in 33 US states, 4 Canadian provinces, and 8 other countries completed the survey. No single metric was considered universally optimal for all use circumstances or for all audiences. To effectively communicate antimicrobial drug use data, evaluation of the target audience is critical to presenting the information. Metrics that are most accurate need to be carefully and repeatedly explained to the audience.

Introduction

Antimicrobial drugs (AMDs) are used in the feedlot for prevention of disease (prophylaxis/metaphylaxis), treatment of disease, and improvement of production efficiency (1–3). Antimicrobial drug use (AMU) in cattle and other animals is controversial (4–8), due to concerns about the potential impact on public health and the development of antimicrobial resistance (AMR).

Obtaining accurate data on AMU in food animals is critical in order to assess the development, dissemination, and persistence of AMR in food animals and the subsequent impact on human health (9–14). In order to report accurate usage and investigate potential associations with AMR, accurate and practically relevant measures to objectively quantify AMU (15–17) are required. A variety of metrics have been used to describe AMU in agriculture (Table 1). Each of these methods of reporting drug usage has advantages and disadvantages (18–21), and metrics that have been most commonly used in discussions and debate about AMU in agriculture have been widely criticized by scientists investigating the issue of antimicrobial resistance. The primary objective of this study was to investigate the perceived accuracy and clarity of metrics used for reporting antimicrobial drug use data that are collected from beef feedlots.

Table 1.

Definitions of antimicrobial drug use (AMU) metrics

AMU metric Definition
Sales value cost of the antimicrobial drug in standard currency
Drug mass in kilograms kg of active ingredient
Number of animals treated count of animals treated with antimicrobial drug
Treatment rate the percentage of animals receiving a given treatment in a given population
Animal defined daily dose (ADDD) number of days of treatment for an animal based on an assumed average maintenance dosage
ADDD per 1000 animals standardized exposure rate based upon the ADDD relative to a fixed number of animals; used to make standardized comparisons in drug exposure among populations or over time
Open in a new tab

Materials and methods

Sampling procedures

The study was conducted as a cross-sectional survey. The study population represented major stakeholders invested in the issues related to the usage of AMDs or the development of AMR in large cattle populations. The individuals targeted included owners and operators of beef production facilities, veterinarians, beef industry representatives, and public health officials familiar with AMU in the beef industry. Potential participants were contacted through e-mail LISTSERVS managed by relevant professional associations or agencies (see Acknowledgments) and by e-mail sent to a list of individuals compiled through recommendations of beef industry and public health leaders. Additionally, participants were encouraged to freely distribute the survey to other knowledgeable and interested colleagues.

For each association or agency, the association president or another administrative leader was contacted by e-mail to determine if their group was willing to participate. If so, an invitation was posted to the association’s LISTSERV. This same invitation was also sent to individuals specifically identified by stakeholders as being knowledgeable and interested in the topic. Direct access to the web-based survey instrument (SurveyMonkey.com Portland, Oregon, USA) was provided in the e-mail invitation as a hyperlink. This e-mail invitation included a second hyperlink which allowed the invited participant to specifically decline the opportunity to participate. A reminder e-mail was sent through each LISTSERV and to the list of individuals approximately 2 to 3 wk after the initial invitation. The web-based survey was available for completion for a 3-month period between June and August 2009. The survey collection instrument was set to only allow 1 response to be submitted per computer.

Study participation was voluntary and anonymous. The research protocol and survey instrument were reviewed and approved by the Colorado State University Institutional Review Board.

Survey instrument

The survey consisted of 22 questions within 3 general categories: 1) participants’ demographics and activities related to the beef industry, 2) opinions on the issues of AMU and AMR, and 3) perceptions about how information regarding AMU is best reported for beef cattle. The survey is available upon request from the corresponding author. Most of the questions required participants to select from a closed series of responses or Likert scale categories. For all questions, response options of “Unknown” and “No Preference” were available at the end of each set of categories. Additionally, open-ended responses were solicited on some questions to allow elaboration if desired by the participant. The questionnaire was pretested by 9 subject matter experts who matched the demographics of the intended study population.

Demographics

Individual participants were asked to report their number of years of active involvement with the beef industry, the primary state/province and nation of their professional activities, and their highest level of education (high school diploma/GED, degree/diploma from a technical school or community college, bachelors degree/BS/BA, advanced degree — specify). Participants were also asked to indicate the primary professional role in which they used AMU information (producer, production consultant, veterinarian, federal government representative, state government representative, university employee, nutritionist, feed salesperson, pharmaceutical industry representative, other — specify) and the top 3 sources from which they obtain information about AMDs (feed or drug companies, veterinarians, government extension officers, universities, farm magazines and newsletters, friends/relatives/neighbors, internet/world wide web, peer reviewed journals, beef specialists, other — specify).

Perceptions about AMR and AMU

Regarding AMR as a health issue, participants were asked whether their concerns or perception of risk had changed over the 10 y prior to the study (much greater, somewhat greater, no different, somewhat less, much less) at different organizational scales (locally/individual operations, regionally, nationally, globally). A parallel series of questions asked participants to provide their perceptions about changes for the true risk of health problems as a result of AMR. Questions regarding “perceived risk” were asked separately from questions regarding “true risk” in an attempt to differentiate changing awareness from changes in the actual probability of adverse events occurring related to AMR in bacteria. Participant perceptions regarding the importance of 5 uses of AMDs in feedlot cattle were also solicited to establish whether or not these uses are necessary to the management of feedlot cattle (feedlots need AMDs for specified uses, feedlots would be difficult to manage without specified uses, feedlots could be managed without specified uses, feedlots do not need AMDs for specified uses). The 5 uses of AMDs investigated were 1) prophylaxis/metaphylaxis at arrival, 2) prophylaxis/metaphylaxis after arrival, 3) use in feed or water for treatment of disease, 4) injectable drugs for treatment of disease, and 5) use in feed to prevent liver abscesses.

AMU Metrics

In order to investigate the appropriateness of different methods of quantifying AMU for various purposes, participants were asked to select the first and second most appropriate methods of quantifying AMU relative to hypothetical scenarios. Scenarios included a comparison of AMU for 2 AMDs in a large cattle population (feedlot), describing AMU data for investigation of AMR in a scientific paper, and reporting of AMU data to the general public. Participants were also asked to identify the least appropriate quantification method for reporting AMU to the general public. The scenarios were all structured around hypothetical situations which summarized the use of 2 AMDs according to label instructions for respiratory disease in an “average” population of feedlot steers (weighing approximately 250 kg) shortly after placement. The quantification methods investigated in this survey were number of treated animals, total mass of active drug, Animal Defined Daily Dose (ADDD), ADDD per 1000 animals, treatment rate, and sales value (Table 1). Definitions of each method were provided in each relevant section of the survey to ensure that participants were able to appropriately distinguish among the different metrics.

In the context of summarizing AMU for large cattle populations, participants were asked to specify the clarity and accuracy for 2 of the investigated metrics: number of animals treated and ADDD per 1000 animals (clarity categories: very clear, clear, somewhat clear, not clear, unknown; accuracy categories: very accurate, accurate, somewhat accurate, not accurate, unknown). In reference to an ongoing prospective surveillance program, participants were asked to select the best method for summarizing AMU for different organizational scales (local/individual operations, regional, national, global) and if a different definition or measurement was more appropriate for surveillance programs than the ones provided in this survey (unknown, no, yes — specify).

Regarding use of the ADDD method, participants were asked if data should be 1) calculated separately and reported separately for high and low dose exposures of the same drug, 2) calculated separately for high and low dose exposures, summed and reported as one summary number, or 3) calculated using a common dose regardless of exposure and reported together. With the definition available, an open-response question asked participants to interpret “400 ADDD of tetracycline” in their own words.

Data analysis

Overview

Survey responses were downloaded directly from the web-based collection instrument into a computer spreadsheet and summarized. Odds ratios with associated 95% CIs were calculated for contingency tables and the χ2 test was performed with statistical software (StataCorp. 2007. Stata Statistical Software: Release 10; StataCorp LP, College Station, Texas, USA). For the purposes of analysis, some response categories were collapsed to facilitate evaluation of simple associations.

Demographic classification for categorical analysis

The responses for the number of years of active involvement in the beef industry were dichotomized as being < or ≥ the median of the response distribution. Participant locale was categorized into North American (US and Canada), and non-North American. Professional role was categorized as veterinarian, university employee, producer, and other. The preferred sources of AMD information were categorized as peer reviewed journal, veterinarians, feed or drug companies, and others.

Categorical analyses of opinions about AMR and AMU metric classification

In order to facilitate analyses, Likert scale responses were dichotomized into categories for greater (much greater and somewhat greater) and not greater (no difference, somewhat less, much less, and unknown). Quantification metrics were grouped into 3 categories: 1) ADDD or ADDD per 1000 animals, 2) number of animals treated or treatment rate, and 3) sales value or total mass of active ingredient. Responses of unknown and no preference about appropriate metrics were excluded from analysis due to low response frequency for these categories. Responses to 2 separate questions regarding the clarity and accuracy of ADDD per 1000 animals and number of animals treated were dichotomized into clear or accurate (very clear/accurate, clear/accurate) and not clear or not accurate (somewhat clear/accurate, and not clear/accurate). A single evaluator (KMB) categorized the open-response question for defining “400 ADDD of tetracycline.” Responses which indicated participant understanding of the definition of ADDD were considered correct. Other responses were designated as incorrect if an obvious misunderstanding was described in open-response or as unknown if the participant volunteered their lack of understanding of this metric.

Results

Survey participants

Twenty associations and agencies were identified as having goals or interests that would be relevant to the issue of AMU in cattle and were contacted regarding participation in this study. Administrative leaders from 10 organizations agreed to post the invitation to their LISTSERVS and an additional 6 associations or agencies provided a list of individuals to contact directly with an invitation to participate. The survey was initiated by 250 individuals and 156 of these participants fully completed the survey. Only responses from completed surveys were summarized. An additional 98 individuals specifically declined to take the survey.

Respondents resided in 33 US states, 4 Canadian provinces, and 8 other countries (Belgium, Denmark, Germany, Ireland, Italy, Portugal, South Africa, and United Kingdom); 2 respondents did not specify their country of residence. Most respondents were from the US (81%; 124/154) and Canada (12%; 19/154). The median number of years of reported involvement in the beef industry was 20 (Q1 = 10, Q3 = 34). Veterinarians (51%; 79/156), university professionals (19%; 29/156), and beef producers (10%; 16/156) were the professional roles most commonly reported by the participants. Other participants reported their professional roles as pharmaceutical industry representatives (8%; 13/156), federal government representatives (5%; 8/156), feed sales representatives (1%; 2/156), state government representatives (0.6%; 1/156), production consultants (0.6%; 1/156), or other (4.5%; 7/156). As their highest earned degree, 90% (140/156) of participants held advanced degrees (e.g., MS, PhD, DVM), 6.4% (10/156) had baccalaureate degrees, 1.3% (2/156) had degrees from a technical school or community college, and 2.6% (4/156) had high school diplomas.

Sources of information

Seventy-two percent (112/156) of participants used peer-reviewed journals as one of the top 3 sources of information about AMDs, 60% (93/156) obtained information from veterinarians, 59% (92/156) gained their knowledge from feed or drug companies, and 37% (58/156) referenced universities. The world wide web (29.5%; 46/156) was used more often as 1 of the top 3 sources of AMD information than beef specialists (13%; 20/156), government extension officers (5%; 8/156), and farm magazines or newsletters (5%; 8/156).

Importance of antimicrobial drug resistance and use

The study attempted to differentiate perceptions about differences in awareness or perceived risk from differences in true risks related to AMR. Participants had greater concern about AMR as either a human health issue or an animal health issue than they did 10 years prior to the study (Figure 1). Participant concerns about AMR as a global health issue for humans were no different than concerns about AMR as a global health issue for animals (P = 0.17). When comparing responses regarding AMR as a global health issue for humans to other scales, no differences were found at the national level (P = 0.77). However, fewer participants had greater concern about AMR as a local (41%; 64/156; P = 0.0032) and regional (49%; 76/156; P = 0.027) health issue for humans compared with the global level. No differences were detected between responses about concern expressed regarding AMR in animals at the global scale compared with the local (P = 0.76), regional (P = 0.88), or national (P = 0.81) scales.

Figure 1.

Figure 1

Open in a new tab

Change in participants’ level of concern about antimicrobial resistance (AMR) as a global health issue in humans and animals during the previous decade (n = 156).

* Distributions of participant perception of greater or less true risk of health problems due to AMR in humans and in animals compared with their perception of true risk 10 years prior to the study were no different than the distribution of concern level for animals presented here (P > 0.05).

Despite the majority of participants having greater concern about AMR as a global human health issue, fewer participants believed that the true risk of global human health problems associated with AMR was greater than that 10 years prior to the study (41%; 64/156; P = 0.006). The percentage of participants with increased perceptions of true risk in animal health due to AMR was not statistically different than that for human health (45%; 70/156; P = 0.33). Compared with the global level, perceptions of true risk of human health problems due to AMR on the local (P = 0.20), regional (P = 0.17), and national (P = 0.64) scales were not statistically different. Likewise, the true risk of animal health problems associated with AMR were not statistically different on the local (P = 0.74), regional (P = 0.51), and national (P = 0.77) scales compared with the global level.

When asked about the importance of AMDs in feedlot production settings, participants responded similarly to 4 scenarios regarding AMDs used for prophylaxis/metaphylaxis at arrival, prophylaxis/metaphylaxis after arrival, use in feed or water for treatment of disease, and use in feed to prevent liver abscesses (Figure 2). In contrast, the majority of participants indicated that injectable AMDs were needed for treatment of diseases of feedlot cattle.

Figure 2.

Figure 2

Open in a new tab

Participants’ perceived need for 5 different uses of antimicrobial drugs (AMDs) in feedlot cattle (n = 156).

AMU metrics

When asked about measurements for reporting AMU, participants preferred different metrics for different types of use scenarios (Figure 3). When asked about the least appropriate metric (as opposed to the most appropriate metric), respondents indicated that sales value (44%; 69/156) and total mass of active drug (31%; 49/156) were the most inappropriate metrics for reporting AMU data to the general public. All other metrics for this question were each selected by less than 8% (12/156) of the participants.

Figure 3.

Figure 3

Open in a new tab

Participant selection of the top 2 antimicrobial drug use (AMU) metrics (cumulatively presented) most appropriate for 3 separate scenarios; 1) when comparing the amount of 2 hypothetical drugs in a large cattle population (stripes), 2) when describing AMU data relative to antimicrobial resistance (AMR) in a scientific paper (black), 3) when clearly reporting data regarding AMU to the general public (gray) (n = 156).

* Categories available for participant selection not displayed in this figure include Unknown and No Preference.

Summarizing large-scale surveillance

Quantifying AMU as the number of animals treated was most commonly judged by respondents as the easiest metric to understand when reporting AMU for large-scale surveillance in cattle populations (e.g., on a national or international scale). In contrast, quantifying AMU as ADDD per 1000 animals was deemed the most accurate metric for representing AMU for large-scale surveillance of AMU and AMR. The majority (88%; 137/156) of the respondents specified that number of animals treated was clearly understood, but only 36% (56/156) specified that it was accurate. Conversely, only 32% (50/156) of respondents specified that the ADDD per 1000 animals method was clearly understood, while 76% (119/156) specified that it was an accurate metric.

Approximately half of the participants selected ADDD per 1000 animals as the best method of summarizing AMU for prospective surveillance programs at the state/provincial (51%; 79/156), national (53%; 83/156), and global (50%; 78/156) organizational scales. All other metrics were selected by less than 15% (24/156) of participants for these purposes. It should be specifically noted that sales value was selected by less than 3% (5/156) of participants as the best method for summarizing AMU. For surveillance programs, most participants (74%; 116/156) were unaware of a more appropriate definition or measurement than the ones provided in this survey. Participants who indicated that a different definition or measurement was more appropriate than the ones provided in the questionnaire commented on refining specific definitions and stratifying metric summaries according to different confounders.

Despite this prevailing opinion that AMU can be accurately portrayed as ADDD or ADDD per 1000 animals, only 64% of participants (98/153) could provide an appropriate definition for what “400 ADDD of tetracycline” might represent if it were reported from a surveillance summary. Sixteen percent of participants (24/153) indicated that they did not know how to interpret this phrase, and the remaining 20% (31/153) gave incorrect interpretations in their free-text responses.

Utilizing the ADDD metric

Participants indicated that the use of the ADDD or ADDD per 1000 animals metrics should allow for differentiation of AMU-related to different prevention/treatment strategies. For example, most participants (74%; 116/156) recommended that separate use estimates should be reported between high and low dose exposures of the same drug (differentiation of low-dose, in-feed exposures to tetracycline compounds for prevention of liver abscesses versus high dose, intramuscular exposures to the same class of drugs for prevention or treatment of respiratory disease). Less than 10% of participants (14/156) held an opposing view that ADDDs should be summed or averaged in situations of disparate use strategies.

Response associations

Participants with ≥ 20 y of beef industry involvement were half as likely [odds ratio (OR) = 0.5, 95% confidence interval (95% CI): 0.2 to 0.9, P = 0.02] to have increased concern about AMR today as a global human health issue compared with 10 y prior to the study when compared with participants with < 20 y of beef industry involvement. These experienced participants were also more likely than the participants with < 20 y to believe that AMU in feed or water for treatment of disease was not needed rather than needed in the management of feedlots (OR = 11.3, 95% CI: 2.4 to 58.4, P = 0.01). When comparing 2 different classes of AMDs (OR = 2.6, 95% CI: 1.2 to 5.5, P = 0.01) or describing AMU data relative to AMR (OR = 2.6, 95% CI: 1.2 to 5.9, P = 0.02), participants with ≥ 20 y of beef industry involvement were more likely to select a method other than ADDD and ADDD per 1000 animals than participants with < 20 y of involvement.

Non-North American participants appeared to have different perspectives on the health risks of AMR, the necessity of AMDs in the management of feedlots, and appropriate AMU metrics compared with North American participants. Participants from non-North American countries were more likely to believe that the true risk of AMR-related health problems is greater today, in humans and animals, than it was 10 y ago (humans: OR = 4.0, 95% CI: 1.1 to 14.7, P = 0.03; animals: OR = 5.2, 95% CI: 1.4 to 19.0, P = 0.01). Non-North American participants were also more likely to indicate that AMDs are not needed in feedlots for prophylaxis/metaphylaxis at arrival (OR = 32.0, 95% CI: 4.0 to 264.3, P < 0.001) and in feed for prevention of liver abscesses (OR = 5.5, 95% CI: 1.0 to 28.8, P = 0.05). When asked about the best metric, non-North American participants were more likely than North American participants to select ADDD or ADDD per 1000 animals as appropriate for the scenarios of comparing 2 different AMDs (P = 0.01) and for reporting data regarding AMU to the public (P = 0.01).

Both professional role and highest degree earned were associated with differences in participants’ responses. Participants with professional roles in universities were more likely than other veterinarians to believe that the true risk of AMR as a health issue to humans is greater today than it was 10 y prior to the study for the regional (OR = 5.6, 95% CI: 2.3 to 13.9, P < 0.001), national (OR = 3.5, 95% CI: 1.4 to 8.3, P < 0.001), and global (OR = 3.2, 95% CI: 1.3 to 7.5, P = 0.01) scales. When describing AMU data relative to the occurrence of AMR, producers were more likely than other participants to select metrics considered most inappropriate by respondents (sales value and total mass of active drug) (P = 0.03). Similarly, participants with advanced degrees were more likely than participants with a bachelors degree (P = 0.05) or high school diploma (P = 0.05) to select ADDD or ADDD per 1000 animals as the most appropriate measure for the same scenario.

Discussion

Results of the present study suggest that several different reporting metrics are considered appropriate and useful for describing AMU in beef feedlots; no single method was considered to be optimal in all circumstances. Ideally, 1 metric would encompass both of the qualities investigated, accuracy, and clarity. Since such a metric does not exist, the use of 2 metrics simultaneously to capture both qualities would facilitate appropriate reporting of AMU. To effectively communicate information regarding AMU, it is important to carefully consider composition of the target audience in order to decide which metrics will most clearly be understood. Metrics that can be used to most accurately relay AMU information (e.g., ADDD) are not the easiest to understand and therefore may need to be carefully and repeatedly explained to the audience. In the past, reports of AMU that focused on sales value or mass of active ingredient metrics as estimates of AMU have not allowed for appropriate investigation of associations between exposures to AMDs and AMR. Theoretically, differences in the physical characteristics of AMDs, the dosages, numbers of animals treated, route of administration, duration of administration, and the reasons for use all modify the effect that exposure to AMDs has on AMR. Metrics relying on sales value or mass of active ingredient do not account for any of these differences. Incorporating such factors that impact how AMDs alter populations of bacteria (selection pressures) into AMU metrics is crucial to improve our understanding of the development, persistence, and dissemination of AMR. Quantification of AMU with metrics which do not account for factors that impact selection pressures distorts discussion regarding the impact of AMU and cannot be used to investigate AMR.

Data regarding AMU has been presented in a variety of formats depending on the purpose of reporting and the intended audiences. Some surveillance programs have quantified specific use of AMDs with a direct focus of investigating the immediate impact of exposures to AMDs on AMR. This work can be performed on a small scale, such as within a single facility or on a grander scale, such as on a national level (22–24). With data specifically and accurately gathered, researchers can evaluate associations or lack of associations between exposures to AMDs and AMR. In contrast, other reports have summarized various estimates of the quantity of AMDs used to illustrate discrepancies in use between humans and animals or between reasons for AMU (25–26). These latter reports commonly present AMU information in terms of mass of active ingredient or sales value of the AMDs. When summarizing AMU information in this manner, failure to differentiate compounds such as ionophores, which have no impact on AMR, and including their mass or sales value obviously distorts the exposure summaries relative to AMR (27–28).

At least in the short-term, AMU in humans and animals creates a selection pressure that contributes to a local increase in AMR. In theory, bacteria susceptible to the AMD are eliminated and resistant bacteria in the previously heterogeneous bacterial population persist (29). However, the probability of occurrence of this phenomenon in association with exposure to AMDs and the strength of this association is unknown. Additionally, little is understood about the duration of persistence within populations of animals and humans as well as about the likelihood of transmission of resistance between populations (30). To better elucidate the existence of human and animal health risks as well as the burden of such risks associated with exposure to AMDs, a proper quantification and reporting metric is needed (15,31).

Choosing an appropriate metric for reporting data regarding AMU is a deceptively complex matter. Challenges in the accuracy and clarity of reporting AMU vary by the AMD of concern and the organizational scale of reporting (e.g., comparisons between farms versus comparisons between regions or countries). Pharmaceutical companies provide AMDs in different combinations of ingredients that are administered by different routes and dosing schedules (32–33). Selection pressures against target bacteria are analogous between the formulations of the same AMD or between similarly structured AMDs (same class of drug) or both (33). However, even the same formulation of a drug (e.g., tetracycline) can apply selection pressures differently. For example, a “High Dose” situation to treat disease could apply a stronger pressure to a population of bacteria and eliminate bacteria with marginal susceptibility to the AMD. Yet, a “Low Dose” of the same AMD to improve production efficiency might lead to the quicker development of AMR since the marginally susceptible and the resistant bacteria would survive and commingle resistance traits (34–38). Further field-based research is needed to clearly indicate which circumstance might have greater consequence, which will require metrics that differentiate these different types of exposures.

The web-based format of this study was an easy and quick method to solicit the opinions from a variety of individuals affected by policy decisions regarding AMU in cattle. However, there are limitations that must be considered when interpreting these results. Conducting an extensive survey using probability-based sampling strategies was not possible since the total population (sampling frame) of experts in beef cattle AMU and AMR was unknown. With no prior knowledge of the sampling frame, a convenience sample was considered the best method to quickly and easily obtain a reasonably wide distribution of the questionnaire to individuals who were knowledgeable and interested in the topic. Therefore, the representativeness of the sampled individuals to the theoretical target population of experts in beef industry AMU and reporting could not be validated. However, the associations, agencies, and individuals targeted by invitation to the questionnaire were all recognized as important stakeholders.

Although this survey may not have included or represented all experts in the beef industry, stakeholders with advanced degrees were well-represented. Likely, holding an advanced degree would aid an individual in evaluating appropriate AMU metrics since the complexities of the related issues are not directly intuitive. However, the ADDD or ADDD per 1000 animals metrics were not clearly definable by all of the highly educated respondents to this survey. The lack of participants’ knowledge about the ADDD metrics may have contributed to nondifferential mis-classification when selecting useful metrics. These participants might have been drawn towards or away from selecting ADDD metrics if they were unable to distinguish them from other metrics or if they ignored metrics they did not understand, respectively. Additionally, stakeholders who were willing to participate may not have submitted a complete survey (thus, not included in this report) if they were not comfortable with their grasp on the intricacies of the ADDD metrics. If stakeholder groups, which were not well-represented in this study were included, different distributions of responses might be expected.

Response associations were tested to investigate similarities and differences between the different demographic groups to gain some insight into their perspectives. Experienced participants were found to have different perceptions on the presented issues than the participants with less number of years of involvement. This may reflect a reluctance to fully recognize or understand new approaches to the dynamic issues surrounding AMU. Differences noted between responses from North American and non-North American participants might be connected to industry differences in beef production or cultural differences in the relationships among producers, veterinarians, and regulatory agencies. Varying perspectives about AMU within different professional roles and levels of education further illustrate how crucial it is to evaluate the audience when selecting a metric for reporting. Investigating associations was not the primary focus of the study, so further consideration is needed.

This study investigated a finite number of quantification methods which represent categories of a large number of metrics that have been used. The ones used in the study were chosen to encompass the metrics most commonly used (sales value and total mass of active ingredient) and those which more fully account for selection pressure (ADDD and ADDD per 1000 animals). Since more than 1 metric may be viewed as appropriate in a specific situation, we solicited responses which allowed for the top 2 choices. The results have been presented here as a cumulative percentage of the top 2 choices because the interval perceived by the participant between their 2 choices can vary. Depending on the purpose of the research, how the data were collected, the organism being investigated for AMR, and the AMDs of interest, other metrics could be appropriate. However, in every case the concepts of clarity and accuracy of reporting should always be highly regarded.

Accuracy and clarity together are hard to come by in quantifying AMU. If we describe AMU to the absolute detail of what it represents, often we lose the simplicity of the representation. Participants indicated that although ADDD metrics lack clarity, they are quite accurate as opposed to the clearly understood metric, number of animals treated, which lacks accuracy. Participants of this study indicated the superior accuracy of the ADDD metric, yet not all participants were able to correctly define an ADDD metric. Therefore, though the details behind this metric may not be wholly understood, participants were still able to recognize that this metric better represents AMU than the other metrics. The absence of a complete understanding of ADDD does not invalidate audience ability to interpret reports.

In designing the questionnaire, AMU in the feedlot setting was the primary focus. Data regarding exposures to AMDs and AMR in North American feedlots have not been available in the past and current efforts to develop an appropriate surveillance system in Canada are underway. Therefore, our research group was specifically interested in the responses of stakeholders in the feedlot industry. Other animal production industries such as swine, poultry, or dairy operations may utilize AMDs differently, but types of use are generally similar. Other animal agribusinesses with scenarios which are not analogous to the feedlot scenarios presented would need further investigation since we found that the appropriate metric depends on the user audience of the information and the research question. Ideally, a common method or pairing of methods would be used in all animal agribusinesses as well as within the public health sector for more closely comparable estimates.

Acknowledgments

We thank the Academy of Veterinary Consultants, the American Association of Bovine Practitioners, the American College of Veterinary Clinical Pharmacology, the American College of Veterinary Preventive Medicine, the American College of Veterinary Internal Medicine, the American Public Health Association, the Association of State and Territorial Health Officials, the Association for Veterinary Epidemiology and Preventive Medicine, Canadian Cattlemen’s Association, the Centers for Disease Control and Prevention, the Council of State and Territorial Epidemiologists, the EPIVET LISTSERV, the United States Food and Drug Administration, Feedlot Health Management Services, the International Conference on the Use of Antimicrobials in Cattle Production, the National Association of State Public Health Veterinarians, the National Cattlemen’s Beef Association, the Public Health Agency of Canada, the Texas Cattle Feeders Association, and the United States Department of Agriculture. These associations and agencies were contacted to solicit their participation in the survey. CVJ

Footnotes

Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (hbroughton@cvma-acmv.org) for additional copies or permission to use this material elsewhere.

Financial support provided in part by the Public Health Agency of Canada.

References

  • 1.Apley M. Antimicrobial therapy of bovine respiratory disease. Vet Clin North Am Food Anim Pract. 1997;13:549–574. doi: 10.1016/s0749-0720(15)30313-3. [DOI] [PubMed] [Google Scholar]
  • 2.Apley M. Bovine respiratory disease: Pathogenesis, clinical signs, and treatment in lightweight calves: Stocker cattle management. Vet Clin North Am Food Anim Pract. 2006;22:399–411. doi: 10.1016/j.cvfa.2006.03.009. [DOI] [PubMed] [Google Scholar]
  • 3.Barton MD. Antibiotic use in animal feed and its impact on human health. Nutr Res Rev. 2000;13:279–299. doi: 10.1079/095442200108729106. [DOI] [PubMed] [Google Scholar]
  • 4.Barbosa TM, Levy SB. The impact of antibiotic use on resistance development and persistence. Drug Resist Update. 2000;3:303. doi: 10.1054/drup.2000.0167. [DOI] [PubMed] [Google Scholar]
  • 5.Casewell M, Friis C, Marco E, McMullin P, Phillips I. The European ban on growth-promoting antibiotics and emerging consequences for human and animal health. J Antimicrob Chemother. 2003;52:159–161. doi: 10.1093/jac/dkg313. [DOI] [PubMed] [Google Scholar]
  • 6.Phillips I, Casewell M, Cox T, et al. Does the use of antibiotics in food animals pose a risk to human health? A critical review of published data. J Antimicrob Chemother. 2004;53:28–52. doi: 10.1093/jac/dkg483. [DOI] [PubMed] [Google Scholar]
  • 7.van den Bogaard AE, Stobberingh EE. Epidemiology of resistance to antibiotics: Links between animals and humans. Int J Antimicrob Agents. 2000;14:327–335. doi: 10.1016/s0924-8579(00)00145-x. [DOI] [PubMed] [Google Scholar]
  • 8.Witte W. Medical consequences of antibiotic use in agriculture. Science. 1998;279(5353):996–997. doi: 10.1126/science.279.5353.996. [DOI] [PubMed] [Google Scholar]
  • 9.Caprioli A, Busani L, Martel JL, Helmuth R. Monitoring of antibiotic resistance in bacteria of animal origin: Epidemiological and micro-biological methodologies. Int J Antimicrob Agents. 2000;14:295–301. doi: 10.1016/s0924-8579(00)00140-0. [DOI] [PubMed] [Google Scholar]
  • 10.McEwen SA, Singer RS. Stakeholder position paper: The need for antimicrobial use data for risk assessment: Animal Antimicrobial Use Data Collection in the United States: Methodological Options. Prev Vet Med. 2006;73:169–176. doi: 10.1016/j.prevetmed.2005.09.017. [DOI] [PubMed] [Google Scholar]
  • 11.Phillips I. The 1997 Garrod Lecture. The subtleties of antibiotic resistance. J Antimicrob Chemother. 1998;42:5–12. doi: 10.1093/jac/42.1.5. [DOI] [PubMed] [Google Scholar]
  • 12.van den Bogaard AE, Stobberingh EE. Antibiotic usage in animals: Impact on bacterial resistance and public health. Drugs. 1999;58:589–607. doi: 10.2165/00003495-199958040-00002. [DOI] [PubMed] [Google Scholar]
  • 13.FAAIR Scientific Advisory Panel. Policy Recommendations. Clin Infect Dis. 2002;34:S76. doi: 10.1086/340243. [DOI] [PubMed] [Google Scholar]
  • 14.Livermore DM, Macgowan AP, Wale MCJ. Surveillance of antimicrobial resistance. Centralised surveys to validate routine data offer a practical approach. BMJ. 1998;317(7159):614–615. doi: 10.1136/bmj.317.7159.614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Carson CA, Reid-Smith R, Irwin RJ, Martin WS, McEwen SA. Antimicrobial use on 24 beef farms in Ontario. Can J Vet Res. 2008;72:109–118. [PMC free article] [PubMed] [Google Scholar]
  • 16.Filius PMG, Liem TBY, van der Linden PD, et al. An additional measure for quantifying antibiotic use in hospitals. J Antimicrob Chemother. 2005;55:805–808. doi: 10.1093/jac/dki093. [DOI] [PubMed] [Google Scholar]
  • 17.Singer RS, Reid-Smith RJ, Sischo WM. Stakeholder position paper: Epidemiological perspectives on antibiotic use in animals: Animal Antimicrobial Use Data Collection in the United States: Methodological Options. Prev Vet Med. 2006;73:153–161. doi: 10.1016/j.prevetmed.2005.09.019. [DOI] [PubMed] [Google Scholar]
  • 18.Chauvin C, Madec F, Guillemot D, Sanders P. The crucial question of standardisation when measuring drug consumption. Vet Res. 2001;32:533–543. doi: 10.1051/vetres:2001145. [DOI] [PubMed] [Google Scholar]
  • 19.Jensen VF, Jacobsen E, Bager F. Veterinary antimicrobial-usage statistics based on standardized measures of dosage. Prev Vet Med. 2004;64:201–215. doi: 10.1016/j.prevetmed.2004.04.001. [DOI] [PubMed] [Google Scholar]
  • 20.Merlo J, Wessling A, Melander A. Comparison of dose standard units for drug utilisation studies. Eur J Clin Pharmacol. 1996;50:27–30. doi: 10.1007/s002280050064. [DOI] [PubMed] [Google Scholar]
  • 21.Kritsotakis EI, Gikas A. Surveillance of antibiotic use in hospitals: Methods, trends and targets. Clin Microbiol Infect. 2006;12:701–704. doi: 10.1111/j.1469-0691.2006.01415.x. [DOI] [PubMed] [Google Scholar]
  • 22.Dunowska M, Morley PS, Traub-Dargatz JL, Hyatt DR, Dargatz DA. Impact of hospitalization and antimicrobial drug administration on antimicrobial susceptibility patterns of commensal Escherichia coli isolated from the feces of horses. J Am Vet Med Assoc. 2006;228:1909–1917. doi: 10.2460/javma.228.12.1909. [DOI] [PubMed] [Google Scholar]
  • 23.Bager F. DANMAP: Monitoring antimicrobial resistance in Denmark. Int J Antimicrob Agents. 2000;14:271–274. doi: 10.1016/s0924-8579(00)00135-7. [DOI] [PubMed] [Google Scholar]
  • 24.Bunner CA, Norby B, Bartlett PC, Erskine RJ, Downes FP, Kaneene JB. Prevalence and pattern of antimicrobial susceptibility in Escherichia coli isolated from pigs reared under antimicrobial-free and conventional production methods. J Am Vet Med Assoc. 2007;231:275–283. doi: 10.2460/javma.231.2.275. [DOI] [PubMed] [Google Scholar]
  • 25.Institute of medicine. Human health risks with the subtherapeutic use of penicillin or tetracyclines in animal feeds. Washington, DC: National Academy Press; 1989. pp. 65–7. [Google Scholar]
  • 26.Mellon M, Benbrook C, Benbrook KL. Hogging it: Estimates of antimicrobial abuse in livestock. Cambridge: Union of Concerned Scientists Publications; 2001. [Google Scholar]
  • 27.Callaway TR, Edrington TS, Rychlik JL, et al. Ionophores: Their use as ruminant growth promotants and impact on food safety. Curr Issues Intest Microbiol. 2003;4:43–51. [PubMed] [Google Scholar]
  • 28.Russell JB, Houliban AJ. Ionophore resistance of ruminal bacteria and its potential impact on human health. FEMS Microbiol Rev. 2003;27:65–74. doi: 10.1016/S0168-6445(03)00019-6. [DOI] [PubMed] [Google Scholar]
  • 29.Levy SB, Marshall B. Antibacterial resistance worldwide: Causes, challenges and responses. Nat Med Supplement. 2004;10:S122–S129. doi: 10.1038/nm1145. [DOI] [PubMed] [Google Scholar]
  • 30.Singer RS, Ward MP, Maldonado G. Can landscape ecology untangle the complexity of antibiotic resistance? Nat Rev Microbiol. 2007;4:943–952. doi: 10.1038/nrmicro1553. [DOI] [PubMed] [Google Scholar]
  • 31.Menendez Gonzalez S, Steiner A, Gassner B, Regula G. Antimicrobial use in Swiss dairy farms: Quantification and evaluation of data quality. Prev Vet Med. 2010;95:50–63. doi: 10.1016/j.prevetmed.2010.03.004. [DOI] [PubMed] [Google Scholar]
  • 32.McEwen SA, Fedorka-Cray PJ. Antimicrobial use and resistance in animals. Clin Infect Dis. 2002;34:S93. doi: 10.1086/340246. [DOI] [PubMed] [Google Scholar]
  • 33.Bywater RJ. Veterinary use of antimicrobials and emergence of resistance in zoonotic and sentinel bacteria in the EU. J Vet Med B Vet Public Health. 2004;51:361–363. doi: 10.1111/j.1439-0450.2004.00791.x. [DOI] [PubMed] [Google Scholar]
  • 34.Craig W. Does the dose matter? Clin Infect Dis. 2001;33(Suppl 3):S233–S237. doi: 10.1086/321854. [DOI] [PubMed] [Google Scholar]
  • 35.Funk JA, Lejeune JT, Wittum TE, Rajala-Schultz PJ. The effect of subtherapeutic chlortetracycline on antimicrobial resistance in the fecal flora of swine. Microb Drug Resist. 2006;12:210–218. doi: 10.1089/mdr.2006.12.210. [DOI] [PubMed] [Google Scholar]
  • 36.Ghosh S, LaPara TM. The effects of subtherapeutic antibiotic use in farm animals on the proliferation and persistence of antibiotic resistance among soil bacteria. ISME J. 2007;1:191–203. doi: 10.1038/ismej.2007.31. [DOI] [PubMed] [Google Scholar]
  • 37.Guillemot D, Carbon C, Balkau B, et al. Low dosage and long treatment duration of {beta}-lactam: Risk factors for carriage of penicillin-resistant Streptococcus pneumoniae. JAMA. 1998;279:365–370. doi: 10.1001/jama.279.5.365. [DOI] [PubMed] [Google Scholar]
  • 38.Kohanski MA, DePristo MA, Collins JJ. Sublethal antibiotic treatment leads to multidrug resistance via radical-induced mutagenesis. Mol Cell. 2010;37:311–320. doi: 10.1016/j.molcel.2010.01.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Canadian Veterinary Journal are provided here courtesy of Canadian Veterinary Medical Association

RESOURCES