Attitudes towards antimicrobial use and factors associated with antimicrobial use in western Canadian cow-calf herds

Abstract

One hundred cow-calf producers in western Canada were surveyed to determine their perceptions regarding antimicrobial use (AMU) and how these perceptions and other herd management factors were associated with AMU. Veterinarians were the most important source of AMU information. Half of the producers considered antimicrobial resistance (AMR) when choosing antimicrobials, while 24% considered the influence of AMU on AMR in human health. Younger producers < 30 y were most likely to consider AMR when choosing antimicrobials. Injectable products were used for disease prevention in 17% of herds; 5% used medically important antimicrobials in feed and 6% in water. Use of injectable antimicrobials of very high importance to human health was reported in 34% of herds. Producers with higher calf mortality were more likely to report AMU in feed or water. The use of Health Canada Category I antimicrobials was most common when calves were retained after weaning.

Introduction

The emerging global threat associated with antimicrobial resistance (AMR) and the importance of antimicrobial stewardship are priority areas for research and education for the Canadian beef industry (1). However, there is limited information on the extent to which individual cattle producers consider the threat posed by AMR when making treatment decisions for their cattle.

Although veterinary oversight of antimicrobial use (AMU) is increasing with changing Health Canada policy and regulations (2), there remains a need to understand producer perceptions of AMU and AMR and identify opportunities for enhancing on-farm stewardship. This is particularly important in cow-calf production in which there has been relatively limited routine involvement of veterinarians in herd health (3,4) compared to feedlot, swine, and poultry production systems. A 2010 survey of 310 cow-calf clients of western Canadian veterinary practices found that only 63% of producers consulted their veterinarian regarding treatment of a sick calf in the previous calving season. In addition, only 24% had a veterinarian examine or treat a sick calf (3).

Further context for examining veterinary oversight of AMU was provided by a 2012–2013 survey (4). Forty-two percent of Saskatchewan cow-calf producers responded that the use of veterinarians for individual animal services had decreased in the past 10 y. While decreasing individual animal consultation suggested reduced oversight of AMU for treatment of specific cases, the use of herd health services was reported to have concurrently increased by 32% (4). This could suggest potential opportunities for veterinary involvement in AMU protocols.

A key objective of the increase in veterinary oversight of AMU is to improve and support antimicrobial stewardship (2). Researchers have previously examined decisions by feedlot veterinarians related to AMU (5). However, because in Canadian beef production systems, antimicrobial treatments have traditionally been administered by the producer or animal health staff, on-farm decisions by producers will continue to influence AMU. Only 1 published North American study was identified that included questions on how cow-calf producers perceive AMR and factors influencing their treatment decisions (6).

In addition to examining producer attitudes to AMU, measurable indicators are needed for AMU practices to monitor the impact of future stewardship programs. While a number of metrics have been discussed (7), most rely on detailed records of AMU not available in many cow-calf herds. Of particular concern is the use of Health Canada Category I antimicrobials, defined as being of very high importance to human health (8). These drugs are preferred options for treatment of serious human infections, in which there are limited or no available alternatives.

The overall goal of this study was to examine producer perceptions regarding AMU in cow-calf herds, to identify producer and herd factors associated with these perceptions, and to determine how these perceptions and other herd characteristics influence AMU practices. The first objective was to describe producer perceptions about changes in antimicrobial effectiveness, the importance of AMU to profitability, and the impact of AMR in treatment decisions. Producer decisions in response to antimicrobial treatment failure were also reported. The second objective was to determine whether these producer perceptions and decisions in response to treatment failure varied based on producer age, herd size, and productivity.

The third objective of this study was to investigate factors associated with AMU practices of interest including: whether > 30% of either cows or calves received antimicrobials for treatment of disease, use of medically important antimicrobials (MIA) for disease prevention in any class of cattle, use of any antimicrobials in feed or water, and use of Health Canada Category I antimicrobials (8). Antimicrobials listed as Category I, II, or III by Health Canada are considered MIA (9). Medically important antimicrobials were defined recognizing that some antimicrobials are more important in the treatment of life-threatening bacterial diseases, and that resistance to these drugs could have more serious implications for human health (2,8). The final objective was to describe cow-calf producers’ most commonly used sources of information on antimicrobials.

Materials and methods

Antimicrobial use survey development and content

Researchers designed a paper-based survey to assess AMU and producer attitudes regarding AMU (10). An antimicrobial handbook listed commercial and generic drug names and included color photographs of product packaging for antimicrobials approved for use in cattle in Canada. This served as a reference for producers and acted as an aid to recall. The period of interest was July 1, 2013 to June 30, 2014. The first part of the survey assessed AMU in bulls, cows, unweaned calves, and weaned calves. For each class of animal, participants were asked whether and which antimicrobials were used for prevention or treatment of a list of common diseases. Finally, the percent of animals treated (< 5%, 6% to 30%, 31% to 70%, > 70%) was reported for each combination of product, reason for use, and class of animal.

The second section of the survey examined producer perceptions on AMU and decisions made in response to treatment failure. The first set of questions asked if the producer had observed a decrease in antimicrobial effectiveness for common conditions (lameness, respiratory, and gastrointestinal disease), or if they thought their profitability would decrease if they were unable to use antimicrobials. Producers were also asked if AMR development or the potential to impact AMR in human medicine influenced treatment decisions. The producers were presented with a statement and asked if: they agreed, were neutral, disagreed, or had no opinion. The second set of questions asked participants to rank potential actions if the first antimicrobial treatment for 3 common disease problems failed. Finally, producers were asked about their primary source of information regarding antimicrobials. Approval for the study was provided by the University of Saskatchewan Behavioural Research Ethics Board (#14-07) and a copy of the survey is available by e-mail request.

Herd recruitment and data collection

The participants were Canadian cow-calf producers recruited for a surveillance network targeting Alberta, Saskatchewan, and Manitoba. Both moderate-sized herds with 100 to 300 cow-calf pairs and large herds with > 300 cow-calf pairs were actively recruited for the study. Enrollment targets were based on the proportion of cow-calf herds located in each province as well as the geographic distribution of large and moderate-sized herds (11).

Cow-calf veterinarians in each region identified clients with basic production records that routinely pregnancy tested. Producers who agreed to participate were mailed a consent form and survey to collect baseline data from the 2013 calving season.

The second survey examining AMU from July 2013 to June 2014 was mailed mid-July 2014 to 104 producers who had returned the baseline survey. Reminders were sent in August 2014 and 100 AMU surveys were returned between July 2014 and March 2015. In February 2015, a third survey was distributed. The completed third survey obtained information on calving outcomes and herd management for 2014, and was returned between March 2015 and January 2016. Participants were provided with an honorarium.

Data entry and statistical analysis

Data from the AMU survey were linked with herd attribute data collected at the time of enrollment and management and production information.

All producer responses were recoded based on whether they agreed with statements regarding AMU and whether they ranked treatment options and sources of information as their first choice. The association between producer and herd characteristics (Table 1) and dichotomized responses to questions regarding AMU perceptions were examined using generalized linear models (GLM). The models used a Poisson distribution and log link function with a robust variance estimate. The resulting direct estimation of relative risk (RR) with 95% confidence interval (CI) (12,13) avoided overestimation of effect from odds ratios (OR) when the event of interest is common.

Table 1.

The unconditional associations between suspected risk factors and producer attitudes to AMU and effectiveness (n = 99 herds).

Variable	Relative risk	95% CI		P-valuea
		Lower	Upper
Decrease in the effectiveness of antimicrobials for treating respiratory disease, gastrointestinal disease, or lameness
• At least 1 decision-maker < 30 y	0.44	0.07	2.72	0.38
• Number of cows breeding season 2013 > 500	1.43	0.82	2.48	0.20
• Sale of purebred cattleb	1.50	0.86	2.62	0.15
• Calf mortality in 2013 > 5%	1.67	1.01	2.78	0.048d
• Calves retained after weaningb	1.50	0.90	2.51	0.12
Agreed with the statement “My profitability would decrease if I could not use injectable antimicrobials.”
• At least 1 decision-maker < 30 y	1.09	1.03	1.16	0.005
• Number of cows breeding season 2013 > 500	1.11	1.03	1.20	0.005
• Sale of purebred cattleb	0.89	0.72	1.10	0.28
• Calf mortality in 2013 > 5%	1.06	0.95	1.18	0.29
• Calves retained after weaningb	1.15	1.04	1.26	0.005
Agreed with the statement “My profitability would decrease if I could not use antimicrobials in feed or water.”
• At least 1 decision-maker < 30 y	0.48	0.08	2.99	0.44
• Number of cows breeding season 2013 > 500	1.61	0.92	2.85	0.10
• Sale of purebred cattleb	0.62	0.25	1.55	0.31
• Calf mortality in 2013 > 5%	0.95	0.49	1.83	0.87
• Calves retained after weaningb	1.58	0.91	2.74	0.11
Consider the potential for AMR development when choosing an antimicrobial for treating cattle on their farms
• At least 1 decision-maker < 30 y	2.27	1.80	2.86	0.001d
• Number of cows breeding season 2013 > 500	0.88	0.51	1.52	0.65
• Sale of purebred cattleb	1.38	0.88	2.14	0.16
• Calf mortality in 2013 > 5%	0.90	0.55	1.50	0.70
• Calves retained after weaningb	1.24	0.82	1.88	0.31
Consider the potential for AMR development in human medicine when choosing an antimicrobial for treating cattle on their farms
• At least 1 decision-maker < 30 y	3.10	1.56	6.18	0.001d
• Number of cows breeding season 2013 > 500	0.74	0.28	1.95	0.55
• Sale of purebred cattleb	1.85	0.90	3.81	0.09
• Calf mortality in 2013 > 5%	0.78	0.32	1.88	0.58
• Calves retained after weaningb	1.20	0.59	2.42	0.62
Treat with a different antimicrobial as first choice when treatment failed regardless of the disease under treatment, rather than consult a veterinarian or other sources of informationc
• At least 1 decision-maker < 30 y	2.98	1.51	5.91	0.002
• Number of cows breeding season 2013 > 500	0.72	0.27	1.87	0.50
• Sale of purebred cattleb	1.77	0.87	3.61	0.12
• Calf mortality in 2013 > 5%	2.34	1.23	4.52	0.01
• Calves retained after weaningb	0.66	0.30	1.44	0.30

^aP-value < 0.20 retained for consideration in the final multivariable models.

^bBased on information reported from 2014 to 2016.

^cn = 100 herds.

^dIndicates the only variable retained in the final model and final effect estimates for the outcome.

CI — confidence interval.

The associations between producer attitudes to AMU, herd management and productivity, and AMU practices of interest were also investigated using the previously described GLM. The outcomes of interest included: if > 30% of either cows or calves received treatment with antimicrobials for any disease, use of MIA for disease prevention in any class of cattle, AMU in feed or water, or use of any Health Canada Category I antimicrobials. All continuous risk factors (age and herd size) were examined and tested for violation of the linearity assumption. Producer attitudes to AMU, herd management, and productivity factors considered are listed in Table 1. Age was recoded to capture if a primary decision-maker was < 30 y. Herds were also categorized as to whether any purebred animals were sold, if calving started before March 2014, and if herd size was > 300 or > 500 cows.

Models were constructed by first screening all unconditional associations to identify potential risk factors with P < 0.20 and reported in tables. Multivariable models were constructed using manual backwards stepwise selection. All variables with P < 0.05 were retained in the final model and were considered statistically significant. Risk factors not significant in the full model were evaluated to see if removal changed other important effect estimates by > 20%. If so, the factor was retained as a confounder given that it was not an intervening variable. First-order interactions were retained if P < 0.05. Akaike information criteria (AIC) were used to select the best fit model where more than 1 variable represented the same value (e.g., continuous or categorical representation of herd size and age). Residuals were examined for outliers. All models were constructed using Stata 15 (StataCorp, College Station, Texas, USA). Only final models are presented in the text unless otherwise stated.

Results

Study population

One hundred herd owners returned the AMU survey; 49% from Alberta, 31% from Saskatchewan, 19% from Manitoba, and 1% from British Columbia. All 100 completed the initial enrollment survey describing the 2013 calving season; 90 of these producers also returned the third survey describing the 2014 calving season (89 with complete calving data). In 2013, > 300 cows calved in 34% (34/100) of herds and > 500 in 11% (11/100). In 2014, > 300 cows calved in 29% (26/89) of herds and > 500 in 9% (8/89). Of the 100 participants, 23% reported having sold at least some purebred cattle and 37% retained some calves after weaning.

In 2013, 25% (25/100) reported > 5% of calves died between birth and weaning. In 2013, the median month of weaning was November and ranged from September to February. In 2014, 17% (15/89) reported > 5% of calves died between birth and weaning. In 2014, half of all herds started calving before the end of March and the start of calving ranged from December to May. The median length of the 2014 calving season was 95 d [interquartile range (IQR): 73 to 114]. The average age of the person making decisions for the herd was 47 y (range: 23 to 79 y); 4 operations reported 2 primary decisionmakers. A primary decision-maker was: < 30 y in 6% of herds, < 40 y in 34%, and < 50 y in 55% of herds. Most participants contracted veterinarians to examine bulls before breeding season (88%, 87/99) and to pregnancy test cows (92%, 92/100).

Producer attitudes towards AMU and effectiveness

One producer did not complete the questions related to perceptions regarding AMU. Only 12% (12/99) of producers reported noticing decreased antimicrobial efficacy when treating respiratory disease during the previous year. Decreased antimicrobial efficacy was reported more frequently for treating gastrointestinal disease such as calf scours (17%, 17/99) or lameness (18%, 18/99). Thirty-four percent (34/99) of producers reported a decrease in the effectiveness of antimicrobials for treating ≥ 1 condition. After considering all risk factors, herd owners reporting > 5% calf mortality from birth to weaning in 2013 were most likely to report decreased antimicrobial effectiveness (Table 1).

Most producers (91%, 91/99) agreed with the statement “My profitability would decrease if I could not use injectable antimicrobials.” After accounting for all risk factors in the final model, primary decision-makers who were < 30 y [relative risk (RR): 1.12, 95% CI: 1.02 to 1.22, P = 0.014], had > 500 cows at breeding in 2013 (RR: 1.06, 95% CI: 1.01 to 1.11, P = 0.02), and had calves that were retained after weaning (RR: 1.13, 95% CI: 1.03 to 1.23, P = 0.006) were more likely to report their profitability would decrease if they could not use injectable antimicrobials.

Only 33% (33/99) of all producers reported that their profitability would decrease if they could not use antimicrobials in feed or water. There were no significant associations between the risk factors examined and concerns that profitability would decrease if antimicrobials could not be used in feed or water (Table 1). Twenty of the 34 (59%) producers using antimicrobials in feed or water reported their profitability would decrease if they could not use antimicrobials in feed or water.

Almost half of producers (47%, 47/99) reported considering the potential for AMR development when choosing an antimicrobial for treating their cattle. In the final model, herds in which a primary decision-maker was < 30 y were more than twice as likely to have AMR considered in choosing treatment options than herds in which the primary decision-maker was ≥ 30 y (Table 1).

The potential influence of AMU on AMR in human medicine was considered by 24% (24/99) of producers when choosing an antimicrobial for treating cattle. After considering all risk factors, decision-makers who were < 30 y were 3 times more likely to report considering AMR development in human medicine than were older participants (Table 1).

When producers were asked what they do when their first-choice antimicrobial failed treating respiratory disease, most used a different antimicrobial (52%, 52/100); however, consulting a veterinarian was the second most common option (44%, 44/100). The results were similar with failure of antimicrobial treatment for lameness; 48% (48/100) used a different antimicrobial and 30% (30/100) consulted a veterinarian. The results differed slightly with treatment failure for gastrointestinal disease for which 45% (45/100) consulted a veterinarian first and only 35% (35/100) treated with a different antimicrobial. After considering all significant risk factors in the final model, primary decision-makers who were < 30 y (RR: 2.41, 95% CI: 1.31 to 4.44, P = 0.005) and decision-makers in herds with calf mortality > 5% in 2013 (RR: 2.13, 95% CI: 1.10 to 4.11, P = 0.024) were most likely to consistently switch to a different antimicrobial as their first choice when treatment failed regardless of the disease, rather than consult a veterinarian or other information source.

Most producers reported that their veterinarian was their main source of information regarding antimicrobials (91%, 91/100). The most commonly reported second sources of information were: producer publications (24%), friends (21%), and feed or drug company representatives (16%).

Factors associated with AMU for disease treatment

The most common reasons for AMU across all classes of cattle were respiratory disease, lameness, diarrhea, and navel ill (Table 2). Only 2 herds did not report AMU. Injection was the most common administration route (97% of herds). Of 100 participants, 13% reported AMU in > 30% of cows or calves to treat a clinical disease. More than 30% of cows were treated in 4% of herds and > 30% of pre-weaned calves were treated with antimicrobials for ≥ 1 disease in 11% of herds. Oxytetracycline was most likely to be used to treat disease in > 30% of cows, and oxytetracycline and florfenicol were most likely to be used to treat disease in > 30% of unweaned calves.

Table 2.

Summary of the antimicrobials used, method of administration, and reasons for use in bulls, cows, and calves before and after weaning from July 2013 to June 2014 in 100 cow-calf herds.

Antimicrobial category and method of administration	Total number (%) of herds reporting AMU (n = 100)
	Use for any reason	Arthritis	Diarrhea	Disease prevention	Eye	Lameness	Mastitis	Navel ill	Other	Reproductive tract	Respiratory
Injectable cephalosporin (3rd gen)	18	2	10	0	1	2	4	4	2	2	4
Injectable fluoroquinolone	9	1	4	1	0	0	0	0	0	0	3
Injectable macrolide	44	3	1	6	1	14	2	2	6	5	31
Injectable penicillin	41	12	0	4	18	23	10	8	10	17	2
Injectable phenicol	81	14	21	0	1	7	3	47	5	0	71
Injectable sulfonamide	34	1	25	0	0	0	4	5	2	1	8
Injectable tetracycline	84	34	5	10	32	68	9	19	13	29	31
Intramammary	18	0	0	0	0	0	18	0	0	0	0
Intra-uterine bolus	7	0	0	0	0	0	0	0	0	7	0
Oral antimicrobial	68	1	54	26	0	3	0	1	6	0	4
Topical antimicrobial	12	0	0	0	11	0	0	0	2	0	0
Total herds reporting use of any type of antimicrobial in any animal classa	98	50	74	36	51	82	34	68	30	50	85

^aTotal does not reflect a sum of the columns as some herds used more than 1 product for a reason within a particular animal group or among different animal groups.

Starting calving earlier in the year and calves retained after weaning were unconditionally associated (P < 0.20) with treating > 30% of either cows or calves for ≥ 1 disease (Table 3). In the final multivariable model, the risk of treating > 30% of cows or calves with antimicrobials for disease was not significantly higher if calving started before March 2014 (RR: 2.70, 95% CI: 0.99 to 7.38, P = 0.052) or if calves were retained after weaning (RR: 2.39, 95% CI: 0.90 to 6.31, P = 0.08).

Table 3.

The unconditional associations between suspected risk factors and AMU practices or strategies (n = 100 herds).

Variable	Relative risk	95% CI		P-valuea
		Lower	Upper
> 30% of either cows or calves received treatment with antimicrobials
• Producer has noticed a decrease in antimicrobial efficacy	1.50	0.54	4.14	0.43
• Potential for AMR considered when making treatment decisions	0.69	0.24	1.98	0.49
• Potential for AMR in human medicine considered in making treatment decisions	0.57	0.13	2.40	0.44
• When antimicrobial treatment fails first choice is to consult a veterinarian	0.55	0.13	2.31	0.41
• At least 1 decision-maker < 30 y	1.31	0.20	8.51	0.78
• Number of cows calving 2013 > 500	1.47	0.37	5.83	0.58
• Sale of purebred cattleb	1.37	0.42	4.50	0.61
• Calving started before March 2014	2.30	0.81	6.57	0.12
• Calf mortality in 2013 > 5%	1.88	0.67	5.23	0.23
• Calves retained after weaningb	1.99	0.72	5.49	0.19
Antimicrobials used in the feed or water
• Producer has noticed a decrease in antimicrobial efficacy	0.95	0.54	1.70	0.87
• Potential for AMR considered when making treatment decisions	0.98	0.57	1.70	0.95
• Potential for AMR in human medicine considered in making treatment decisions	1.13	0.61	2.07	0.71
• When antimicrobial treatment fails first choice is to consult a veterinarian	0.97	0.51	1.87	0.94
• At least 1 decision-maker < 30 y	1.52	0.64	3.56	0.34
• Number of cows calving 2013 > 500	1.08	0.47	2.50	0.86
• Sale of purebred cattleb	0.98	0.47	2.01	0.95
• Calving started before March 2014	1.01	0.58	1.76	0.98
• Calf mortality in 2013 > 5%	1.64	0.95	2.81	0.07
• Calves retained after weaningb	1.51	0.88	2.60	0.13
Medically important antimicrobials used for disease prevention in cows or preweaned calves
• Producer has noticed a decrease in antimicrobial efficacy	0.78	0.26	2.36	0.66
• Potential for AMR considered when making treatment decisions	0.49	0.16	1.50	0.21
• Potential for AMR in human medicine considered in making treatment decisions	1.39	0.47	4.13	0.56
• When antimicrobial treatment fails first choice is to consult a veterinarian	0.86	0.26	2.86	0.81
• At least 1 decision-maker < 30 y	1.21	0.19	7.81	0.85
• Number of cows calving 2013 > 500	0.62	0.09	4.35	0.63
• Sale of purebred cattleb	0.56	0.13	2.33	0.42
• Calving started before March 2014	1.92	0.72	5.14	0.20
• Calf mortality in 2013 > 5%	0.45	0.06	3.25	0.43
• Calves retained after weaningb	1.70	0.65	4.49	0.28
Any Health Canada Category I drugs used in the herd
• Producer has noticed a decrease in antimicrobial efficacy	1.38	0.80	2.38	0.24
• Potential for AMR considered when making treatment decisions	1.40	0.81	2.44	0.23
• Potential for AMR in human medicine considered in making treatment decisions	0.96	0.50	1.84	0.91
• When antimicrobial treatment fails first choice is to consult a veterinarian	0.42	0.16	1.08	0.07
• At least 1 decision-maker < 30 y	2.09	1.10	3.97	0.02
• Number of cows calving 2013 > 500	0.78	0.28	2.15	0.64
• Sale of purebred cattleb	1.41	0.76	2.58	0.28
• Calving started before March 2014	1.28	0.74	2.21	0.38
• Calf mortality in 2013 > 5%	1.43	0.82	2.51	0.21
• Calves retained after weaningb	1.92	1.12	3.29	0.02
Number of reasons for use and production groups where Health Canada Category I drugs were used in the herd
• Producer has noticed a decrease in antimicrobial efficacy	1.70	0.77	3.76	0.19
• Potential for AMR considered when making treatment decisions	2.73	1.33	5.65	0.006c
• Potential for AMR in human medicine considered in making treatment decisions	1.17	0.49	2.83	0.71
• When antimicrobial treatment fails first choice is to consult a veterinarian	0.75	0.18	3.17	0.70
• At least 1 decision-maker < 30 y	2.49	1.07	5.79	0.04
• Number of cows calving 2013 > 500	0.72	0.19	2.74	0.64
• Sale of purebred cattleb	2.10	0.91	4.84	0.08
• Calving started before March 2014	1.12	0.51	2.46	0.77
• Calf mortality in 2013 > 5%	1.87	0.84	4.14	0.13
• Calves retained after weaningb	2.06	0.94	4.52	0.07

^aP-value < 0.20 retained for consideration in the final multivariable models.

^bBased on information reported from 2014 to 2016.

^cIndicates the only variable retained in the final model and final effect estimates for the outcome.

Use of oral antimicrobials

Most herd owners (68%) reported use of oral antimicrobials for treatment or prevention (Table 2). Most (52%) provided antimicrobial boluses to treat calves. However, just 34% provided antimicrobials (including ionophores) for any reason to some part of the herd in feed or water; 6% in water and 30% in feed at least once. Only 5% of herds reported MIA in feed.

Calf mortality in 2013 and retaining calves after weaning were unconditionally associated (P < 0.20) with AMU in feed or water (Table 3). In the final model, the herds most likely to use antimicrobials in feed or water were those with the highest calf mortality in 2013. For every 1% increase in calf mortality before weaning in 2013, the risk of AMU in feed or water increased by 1.08 times (95% CI: 1.00 to 1.18, P = 0.048).

Factors associated with AMU for disease prevention

Medically important antimicrobials were used for disease prevention in 21% of herds (any class) at least once (3% in feed, 1% in water, 17% injectable). As well, MIA were administered at least once for disease prevention in cows or preweaned calves in 14% of herds (1% in feed and 13% injectable). There were no factors associated with MIA use for disease prevention in cows and preweaned calves where P < 0.20 (Table 3).

Antimicrobial choices and importance in human health

Health Canada Category I (Category I) antimicrobials were used in 34% of herds at least once. Within Category I, injectable third generation cephalosporins (ceftiofur) were used at least once in 18% of herds and 9% reported using injectable fluoroquinolones (enrofloxacin). Injectable third generation cephalosporins were used by ≥ 1 herd to treat animals in each disease category but were not used for disease prevention (Table 2). An intramammary preparation of ceftiofur hydrochloride was also used in 1 herd. Injectable fluoroquinolones were primarily used to treat diarrhea and respiratory disease in unweaned calves. One herd reported using enrofloxacin for disease prevention in unweaned calves. No Category I products were administered orally.

A third Category I drug, Polymyxin B, was included in a commercial mixed intramammary preparation containing antimicrobials and hydrocortisone, which was used in 9% of herds to treat mastitis (Special Formula 17900-Forte Suspension; Zoetis Canada, Kirkland, Quebec). This commercial product was also used as a topical treatment for ocular disease in 9% of herds. Other antimicrobials included in this product were penicillin G procaine, dihydrostreptomycin, and novobiocin.

Consulting a veterinarian when the first-choice antimicrobial failed, having a primary decision-maker < 30 y, and retaining calves after weaning were unconditionally associated with use of Category I antimicrobials (Table 3). After considering all factors, only herd owners who retained calves after weaning (RR: 1.94, 95% CI: 1.13 to 3.30, P = 0.016) and who were < 30 y (RR: 2.14, 95% CI: 1.02 to 4.53, P = 0.045) were significantly more likely to use Category I products. Consulting a veterinarian when the first-choice antimicrobial failed was not significant after accounting for retaining calves and producer age, but the association did suggest a trend for decreasing use of Category I antimicrobials when herd owners first consulted a veterinarian (RR: 2.23, 95% CI: 0.93 to 5.39, P = 0.07).

The factors unconditionally associated (P < 0.20) with the number of reasons for use and production groups in which Category I antimicrobials were used included: reporting a decrease in antimicrobial effectiveness, considering the potential for AMR when making treatment decisions, having a primary decision-maker < 30 y, sale of purebred cattle, reporting calf mortality in 2013 > 5%, and retaining calves after weaning (Table 3). In the final model following multivariable analysis, the herd owners who were most likely to use Category I antimicrobials to treat more types of problems or in more classes of animals were those who reported considering the potential for AMR when making treatment decisions (RR: 2.73, 95% CI: 1.33 to 5.65, P = 0.006).

Discussion

This study provides baseline data on perceptions regarding antimicrobial effectiveness and factors affecting AMU practices from a sentinel cohort of cow-calf producers in western Canada. Almost all participants stated their veterinarian was their primary source of information on antimicrobials. However, when asked what they would do when their first choice antimicrobial treatment failed for respiratory disease in calves or lameness in cows, producers reported switching antimicrobials before calling their veterinarian. This could be because many commonly used antimicrobials did not require a veterinary prescription (14) or because some producers utilized treatment protocols from their veterinarian that addressed the potential for treatment failure.

Those most likely to switch antimicrobials before consulting a veterinarian were < 30 y and those who reported calf mortality > 5%. This differed for gastrointestinal disease in calves for which most producers contacted their veterinarian when treatment failed before switching products. Earlier surveys recommended more education and communication to increase veterinary involvement in AMU (3,4,6). This study suggests that producers already recognize veterinarians as an important resource. However, the producers in this study were recruited through veterinary clinics and likely represent producers with relatively strong veterinary-client-patient relationships. In addition, the herds recruited to this study were from western Canada, moderate to large in size, and most pregnancy tested. The results may not reflect smaller producers, those less likely to use herd management tools, and producers from eastern Canada. While many cow-calf producers have not historically sought veterinary advice for treatment decisions (3,4), changes in availability of antimicrobials based on new Health Canada regulatory frameworks are expected to increase consultation with veterinarians (2).

A third of participants reported observing decreased antimicrobial effectiveness for treating common conditions in calves or cows. However, < 1 in 5 producers reported a decrease in effectiveness for each individual condition of interest: respiratory disease, calf scours, and lameness. The extent to which these observations reflect AMR is unknown, but they do reflect consideration and concern by producers about AMU effectiveness. Almost all producers identified access to injectable antimicrobials as necessary for profitability. However, only a third of all producers and < 2/3 of producers who actually used feed or water products identified access to feed or water products as important.

While about half of producers reported considering AMR when choosing antimicrobials, only 1 in 4 reported that they considered the potential influence of AMU choices on AMR in human health. Participants who were < 30 y were most likely to consider the potential for AMR in their herd or in human health indicating awareness of the importance of AMR particularly in the younger producers. However, it is unlikely that AMR strongly influenced AMU decisions as most choices would be practically determined by products available on farm, for convenient purchase, or by veterinary prescription. Cost, ease of administration, previous experience, and withdrawal times are other examples of potentially important AMU determinants.

The only report identified with comparable data resulted from a 2007 study of beef producers in Tennessee (6). In that study, 16% of cow-calf producers reported that antimicrobials worked less effectively than in the past. However, 65% agreed with the statement that AMR bacteria in food are an important problem and 72% with the statement that prudent AMU in the beef industry is important for consumer confidence.

The final objective of this study was to investigate factors associated with AMU practices of interest. The use of MIA in a large proportion of the herd was relatively uncommon for either treatment or for prevention. Less than 1 in 8 producers reported AMU to treat ≥ 1 disease in > 30% of either cows or calves. The use of injectable products for disease prevention in any class of animal was reported in < 1 in 6 herds. Use of MIA in feed or water was even less common for treatment or prevention at ≤ 1 in 20 herds. The only significant association observed was between increasing calf mortality and the use of any antimicrobials in feed or water.

Use of injectable Category I antimicrobials was reported at least once in a third of herds. Third generation cephalosporins were used for treating a variety of diseases; however, fluoroquinolones were primarily limited to treatment of diarrhea and respiratory disease. While third generation cephalosporins were not used for disease prevention in any herds, fluoroquinolones were used for prevention in one. Polymyxin B was found in a combination product used either for mastitis or topically off label to treat pinkeye. While topical exposure is not considered an important risk for AMR, cattle can ingest small amounts of the drug through grooming behaviors or nursing. The polymyxin class also contains colistin or polymyxin E. The recent international emergence of colistin resistance in livestock is a concern within the medical community as colistin, in spite of its toxicity, has become a last resort drug when patients do not respond to other options (15).

The use of Category I antimicrobials is not new. In a 2002 study from western Canada, use of third generation cephalosporins was reported in 7.9% of calves and 1.5% of cows (16). At the time, off-label use of fluoroquinolones was reported in calves in 1.0% of herds. The reasons for use of specific products were not reported in the 2002 study. The reasons for use of Category I drugs in the current study differed slightly from that reported in Beef 2007–2008 report from the National Animal Health Monitoring System in the United States (17). Fluoroquinolones and cephalosporins were used most frequently for treating diarrhea and respiratory disease in calves in the United States (17). Cephalosporins were also a common treatment choice for respiratory disease in cows in that study.

Producers who retained calves after weaning and who were < 30 y were more likely to report use of Category I antimicrobials. There was an interesting trend towards decreasing use of Category I products by herd owners who contacted a veterinarian before switching drugs when faced with treatment failure. Although not significant, this finding suggests that plans to increase veterinary involvement in AMU decisions may support stewardship initiatives (2).

Those who considered the potential for AMR when making treatment decisions were also more likely to use Category I antimicrobials in more classes of animals or for more disease problems. This perhaps suggests an inherent contradiction. However, it is most likely that these producers were considering the potential for treatment failure associated with resistance to Category II and III antimicrobials in their herds and making decisions to optimize treatment responses or possibly that they did not understand the relationship between Health Canada Category of AMU and AMR.

The accuracy of data from this survey is limited in that the data were collected at the end of the survey period; however, most participating herd owners reported having health records and several other aids to recall were provided to participants (10). Another limitation was that there was no question regarding motivation for choosing specific antimicrobials, such as cost, convenience, effectiveness, route of administration, and withdrawal time that could provide important background for stewardship programs.

In summary, about half of western Canadian moderate to large cow-calf producers consider AMR in making treatment decisions, although the relative importance of other factors in making product choices requires further study. Veterinarians are important sources of advice on AMU. That role is expected to grow as regulations enhancing veterinary oversight of AMU come into effect and all MIA are available only by prescription. This study provides a baseline for cow-calf producers’ perceptions of antimicrobial effectiveness and attitudes influencing AMU as well as other factors associated with AMU practices. This work should be repeated to determine the impact of planned changes in AMU oversight by veterinarians and other stewardship initiatives. CVJ