Antimicrobial prescribing practice in UK equine veterinary practice
Reasons for performing study: Antimicrobial use is known to select for the emergence of resistant strains of bacteria; therefore prudent use in both human and veterinary medicine is essential to preserve their efficacy.
Objectives: To characterise antimicrobial prescribing patterns in UK equine practice and evaluate factors associated with prescribing.
Methods: A postal questionnaire including 4 clinical scenarios was sent to 740 veterinary surgeons that treat horses. Data were collected on the clinician, their practice and sources of information regarding antimicrobials and their use. Multivariable logistic regression analysis was used to ascertain factors associated with 5 outcomes; prescribing off-licence drugs; prescribing at incorrect doses; prescribing 3rd or 4th generation cephalosporins; prescribing fluoroquinolones and prescribing potentiated sulphonamides.
Results: Questionnaires were completed by 38% of veterinary surgeons who were sent questionnaires. Less than 1% of practices had antimicrobial use guidelines. Trimethoprim-sulphonamides were most commonly prescribed in each clinical scenario. Eleven percent of prescriptions were for antimicrobial drugs not licensed for use in horses in the UK. Five percent of prescriptions for licensed antimicrobials were under the recommended dose rate and 56% over the recommended dose rate. Fluoroquinolones and 3rd and 4th generation cephalosporins accounted for 1 and 3% of prescriptions, respectively. Veterinary surgeons working at referral practices were more likely to prescribe 3rd and 4th generation cephalosporins and fluoroquinolones and off-licence antimicrobials whereas those working in first-opinion practices were more likely to prescribe potentiated sulphonamides. Sources of information regarding antimicrobials also had an effect on prescribing.
Conclusions and potential relevance: Prescription of antimicrobials at inappropriate doses was common even when provided with the weight of the animal. It is uncommon for UK equine veterinary practices to have guidelines for antimicrobial use. The introduction of such guidelines could lead to more prudent use of antimicrobials.
Antimicrobial products are used in the treatment of many infectious conditions in horses and their use is important in maintaining health and welfare. However, increasingly, bacteria are developing resistance to antimicrobial products and this has led to problems in the treatment of some infections in animals . In addition, due to the high levels of contact between people and companion animals, including horses, this may have an impact on the treatment of human infections through the transfer of antimicrobial-resistant organisms from animals to people or via the transfer of resistance genes between bacteria . It is well established that antimicrobial use is one of the main drivers for selection of antimicrobial resistance [1,3–5], therefore appropriate use of antimicrobials in both human and veterinary medicine is essential to preserve their efficacy. Heightened concern over the emergence of antimicrobial resistance in animals has resulted in the use of antimicrobials in veterinary medicine coming under increased scrutiny. It has been suggested that certain classes of antimicrobial (cephalosporins and fluoroquinolones) should not be available for veterinary use in the UK in order to preserve their efficacy in human medicine, which could carry consequences for animal health and welfare .
Antimicrobials are widely used in equine practice in the UK to treat a variety of infectious conditions and can only be prescribed by a veterinary surgeon . A limited number of antimicrobial classes are licensed for use in horses including penicillins, trimethoprim-sulphonamides, aminoglycosides, cephalosporins and oxytetracycline . Recent publications from Canada conclude that inappropriate use of antimicrobials is common in the horse [9,10]; however, objective data on how antimicrobials are used in UK equine veterinary practice currently do not exist. The sales of antimicrobials used in animals in the UK is published annually by the Veterinary Medicines Directorate . Yet, from these sales figures, it is not possible to calculate the number of horses treated or the types of regimens likely to have been used. For some particular groups of antimicrobials, such as the fluoroquinolones or cephalosporins, a large proportion of the drugs may have been administered to horses . Without detailed data on indication-based usage of antimicrobials in veterinary practice, it is impossible to investigate potential associations between the prevalence of antimicrobial resistance in companion animals and patterns of antimicrobial use.
Judicious use of antimicrobials is advocated by UK veterinary bodies [13,14], although guidelines vary and it is unclear if these are widely followed in practice. The main objectives of this study were to determine the current prescribing practices of antimicrobials in equine medicine and evaluate factors associated with inappropriate prescribing using 4 scenarios.
Materials and methods
Study population and design
A cross-sectional, questionnaire-based survey of UK equine veterinary surgeons was undertaken in order to gather data on antimicrobial use by equine clinicians. An electronic copy of the ‘RCVS Directory of Veterinary Practices’ was obtained from the Royal College of Veterinary Surgeons (RCVS) and practices indicating that they treated horses comprised our study population (n = 1278), from which a random sample was selected (n = 740). The final sample size was calculated using Win Episcope 2.0 to estimate the usage levels of a particular antimicrobial between 5 and 50% with 5% precision and 95% confidence intervals (CIs). For each randomly selected veterinary practice, one veterinary surgeon was randomly selected and the address for mailing obtained from the RCVS ‘Register of Members’.
A postal questionnaire sent to individual equine veterinary surgeons (one clinician per practice) was used to collect data. The questionnaire contained 2 sections. The first section collected data about the equine practice at which the clinician worked, the clinician (Table 1) and antimicrobial prescribing practices of the clinician (sources of information regarding antimicrobials and their use, antimicrobial prescribing habits, restrictions on classes of antimicrobials available for use and the importance of client expectations, financial constraints, clinical signs, cytology, microbiological culture and ease of administration when prescribing antimicrobials). The second section consisted of 4 clinical scenario questions: contaminated limb wound; coughing pyrexic yearlings; coughing adult horse and pastern pyoderma (Item S1). Respondents were asked if they would prescribe antimicrobial drugs to the animals described in the scenarios and, if so, were asked to complete a table detailing the trade name of the antimicrobial, dosage prescribed, route of administration, frequency of administration (per day) and duration of the course. The scenario questions were not designed to have a correct or incorrect answer, but as a method to collect objective data on antimicrobial use and to determine any incorrect dosing or use of off-licence preparations.
Table 1. Independent variables (potential risk factors) included in the univariable and multivariable analysis
|Type of veterinary practice (No. respondents = 251)||First-opinion||244||97.2|
|Out of hours||50||20.0|
|Total number of veterinary surgeons employed by the veterinary practice (n = 251)||1–5||116||46.2|
|more than 15||17||6.8|
|Practice has an antimicrobial policy (n = 249)||2||0.8|
|Age group of the veterinary surgeon (n = 250)||20–30||90||36.0|
|Gender of the veterinary surgeon (n = 250)||Female||115||46.0|
|Position held within the veterinary practice (n = 250)||Partner||101||40.4|
|Year graduated with veterinary since degree (n = 249)|| || |
|Veterinary surgeon holds post graduate qualifications (n = 250)||47||18.8|
|Main sources of information regarding antimicrobials and their use (n = 251)||Practice policy||40||15.9|
|Veterinary science degree course notes||130||51.8|
|Veterinary Medicines Directorate||47||18.7|
|NOAH Veterinary Compendium||207||82.5|
|Continuing Professional Development courses||194||77.3|
Questionnaires were piloted on 50 equine clinicians randomly selected as described above. The pilot study was successful and no changes were made to the questionnaire. Following the pilot study, questionnaires, a cover letter explaining the study and a postage paid return envelope were sent to 740 equine veterinary surgeons during September 2009. A reminder postcard was sent to veterinary surgeons that had not responded 2 weeks after the questionnaire had been sent and a further questionnaire was mailed to nonrespondents 2 weeks after the reminder postcard had been sent in order to maximise the response rate.
Data were entered into a Microsoft Access databasea. The accuracy of data entry was checked by selecting and examining a random sample of 30% of the questionnaires. A descriptive analysis of the distributions of antimicrobial use for each scenario was performed. In the UK, medicines for use in specific animal species are licensed by the Veterinary Medicines Directorate (VMD) . The correct dose rates, prescribing regimens and whether a drug is licensed for use in particular animal species, for all antimicrobial drugs used in veterinary medicine in the UK, are provided by the VMD in the form of a Summary of Product Characteristics document (SPC), which is necessary for the antimicrobial drug to be granted a licence. Through comparison with the recommendations published by the VMD, it was determined whether the use of antimicrobials reported by respondents was licensed for use in horses and whether the dose prescribed was in keeping with the guidelines, route or was an under- or overdose (doses within 2% of the guidelines were considered correct). Analyses were carried out to determine associations between 5 dependant variables; whether off-licence antimicrobials were prescribed; whether antimicrobials were prescribed at incorrect doses; use of 3rd and 4th generation cephalosporins; use of fluoroquinolones and use of potentiated sulphonamides. Independent variables included the type of veterinary practice at which respondents worked, personal details about themselves and information on factors that influenced their antimicrobial prescribing practices (Table 1). Associations between independent and dependent variables were analysed using univariable and multivariable logistic regression models. Variables with a P value<0.2 were considered for inclusion in a multivariable logistic regression model which was built using a step-wise, backward elimination procedure. Variables remained in the model if they significantly improved the fit (P<0.05). The goodness-of-fit of the model was assessed using the Hosmer–Lemeshow test statistic. All analyses were carried out using the statistical software package Stata 8b.
A total of 281 of 740 (38%) questionnaires were returned. However, 30 questionnaires were incomplete and were excluded from the analysis leaving 251 completed questionnaires and a usable response rate of 34%.
The majority of respondents worked for first-opinion practices (n = 244; 97.2%). Twelve percent of practices (n = 30) were second-opinion referral centres and 7.6% (n = 19) were classed as veterinary hospitals (some veterinary practices were classed in more than one category, as they were both first-opinion and referral practices). A large proportion of practices were mixed practices (practices that treated all veterinary species) (n = 212; 84.5%) and a smaller proportion specialised in equine practice (n = 39; 15.5%). A large proportion of veterinary surgeons (n = 243; 97.2%) reported that they could prescribe antimicrobial drugs at their own discretion; however, 15.6% (n = 39) of respondents reported that they were restricted in the classes of antimicrobial available for use. Only 2 veterinary surgeons (0.8%) reported that their practices had written antimicrobial use guidelines (both were first-opinion, mixed practices). Nearly 19% (n = 47) of respondents reported that they had at least one post graduate qualification. The National Office of Animal Health (NOAH) Compendium of Data Sheets for Animal Medicines , was the most popular source of information concerning antimicrobials and their use for 82.5% (n = 207) of respondents. Respondents indicated that Continuing Professional Development (CPD) courses (n = 194; 77.3%), pharmaceutical companies(n = 170; 67.7%) and veterinary science degree course notes (n = 130; 51.8%) were also important sources of information. The VMD , which is an Executive Agency of the Department for Environment, Food and Rural Affairs (DEFRA) responsible for the authorisation of all veterinary medicines and monitoring their safety and efficacy following authorisation, was considered an important source of information about antimicrobials by 18.7% (n = 47) of respondents. Scientific journals and textbooks were considered important by 11.6% (n = 29) of respondents.
Respondents were asked whether they routinely weigh or use a weight tape to determine the weight of horses for which they prescribe antimicrobials. Sixty-one percent (n = 152) reported that they rarely or never weigh, or use a weight tape to determine the weight of horses for which they prescribe antimicrobials. Approximately 15% (n = 38) of respondents always weigh and 23.7% (n = 59) sometimes weigh or use a weight tape to determine the weight of horses. When asked to rank the importance of a number of factors when deciding to treat an animal with antimicrobials, clinical signs were ranked as important by 93.2% respondents, ease of antimicrobial administration by 64.5%, bacterial culture by 39.0%, financial constraints by 35.1%, client expectation by 27.5% and cytology by 15.9% of respondents.
Clinical case scenario analysis
The majority of respondents prescribed antimicrobial therapy for the contaminated limb wound scenario (241/249, 96.8%), 61.2% (153/250) prescribed antimicrobials for the coughing pyrexic yearlings, 53.2% (133/250) for coughing adult horse and 80.0% (200/250) for pastern pyoderma. Of these 31.1, 12.4, 9.8 and 23.0% prescribed 2 antimicrobial drugs for therapy for scenarios 1–4, respectively. The antimicrobial drugs that veterinary surgeons reported they would have used in each clinical scenario, grouped by antimicrobial class, are shown in Table 2.
Table 2. Antimicrobial drugs that veterinary surgeons reported they would have used in each of the 4 clinical scenario cases, grouped by antimicrobial class. The number of nonlicensed preparations of the antimicrobial is also shown Δ
|Amoxicillin and clavulanic acid||0||0||0||0||0||0||0||0||0||4||1.63||4|
|Penicillins and aminoglycosides combined||Procaine benzylpenicillin and dihydrostreptomycin||39||12.30||0||11||6.40||0||12||8.28||0||19||7.76||0|
|Procaine benzylpenicillin and neomycin||20||6.31||0||6||3.49||0||5||3.45||0||9||3.67||0|
|Penethamate hydriodide, dihydrostreptomycin and framycetin||0||0||0||0||0||0||0||0||0||1||0.41||1|
|Penicillins, aminoglygocides and aminocoumarins combined||Novobiocin, procaine benzylpenicillin, dihydrostreptomycin and neomycin||0||0||0||0||0||0||0||0||0||3||1.22||3|
|3rd generation cephalosporins||Ceftiofur||7||2.21||0||6||3.49||0||1||0.69||0||3||1.22||0|
|4th generation cephalosporins||Cefquinome||5||1.58||0||4||2.33||0||1||0.69||0||1||0.41||0|
|Trimethoprim and sulphadiazine||151||47.63||0||116||67.44||1||113||77.93||0||114||46.53||0|
|Trimethoprim abd sulphadoxine||7||2.21||0||3||1.74||0||1||0.69||0||1||0.41||0|
|Fusidic acid||Fusidic acid||0||0||0||0||0||0||0||0||0||34||13.88||34|
When compared with guidelines on the correct dosage regimen for each antimicrobial issued by the VMD, across all clinical case scenarios 5.8% of prescriptions (n = 65/1128) were under the recommended dose and 56.9% (n = 642/1128) were over the recommended dose. Table 3 shows the distribution of over- and underdosing for each antimicrobial.
Table 3. Percentage of each antimicrobial drug licensed in the UK for use in horses that veterinary surgeons reported they would have used that was prescribed at the correct dose, under- or overdosed. This table includes the dosage given to all 4 yearlings mentioned in scenario 2, coughing pyrexic yearlings. Additionally, it is not always possible to calculate the correct dose, where one type of drug has been given by 2 different routes, the dosage has been calculated for each, given if the antimicrobial was applied topically or full information of the antimicrobial prescribed was not stated in the answer to the scenario, hence the differences in totals to Table 2
|Penicillins and aminoglycosides combined||Procaine benzylpenicillin and dihydrostreptomycin||68||16.15||7||10.77||35||5.45||110|
|Procaine benzylpenicillin and neomycin||45||10.69||2||3.08||6||0.93||53|
|3rd generation cephalosporins||Ceftiofur||21||4.99||0||0||1||0.16||22|
|4th generation cephalosporins||Cefquinome||14||3.33||0||0||0||0||14|
|Trimethoprim and sulphadiazine||155||36.82||53||81.54||535||83.33||743|
|Trimethoprim and sulphadoxine||16||3.80||2||3.08||2||0.31||20|
Nine percent (n = 30/317) of prescriptions for the contaminated limb wound scenario where for antimicrobial drugs that are not currently licensed for use in horses in the UK. Approximately 3.0% (n = 5/172) of prescriptions were not licensed in the coughing pyrexic yearlings scenario. In the coughing horse scenario, 3.4% (n = 5/145) of antimicrobials prescribed were not licensed. Finally, 24.9% (n = 61/245) of antimicrobials prescribed in the pastern pyoderma scenario were not licensed.
Factors that were found to be significantly associated with the use of each antimicrobial class, as determined by univariable logistic regression analysis, are provided in the supplementary information. Multivariable analysis (Table 4) showed that respondents working at referral practices were more likely to prescribe 3rd and 4th generation cephalosporins (OR = 3.2, 95% CI = 1.2–8.3) and fluoroquinolones (OR = 6.2, 95% CI = 1.3–29.3) than those not working in these types of practices. Additionally, respondents working in first-opinion practices were more likely to prescribe potentiated sulphonamides (trimethoprim-sulphonamide class, OR = 13.8, 95% CI = 2.4–79.1). Clinicians who used the VMD as their main source of information regarding antimicrobials and their use were less likely to prescribe trimethoprim-sulphonamides to horses in any of the 4 scenarios (OR = 0.21, 95% CI = 0.1–0.5). Whilst, respondents who used practice policy as a main source of information regarding antimicrobials and their use were 8 times (95% CI = 1.1–63.7) more likely to prescribe the wrong dose of antimicrobial compared with respondents who did not use practice policy as their main source of information.
Table 4. Adjusted odds ratios (OR) with 95% confidence intervals (CIs) for prescribing antimicrobials at the wrong dose, off-licence or the prescription of 3rd and 4th generation cephalosporins, fluoroquinolones and trimethoprim/sulphonamides in any of the 4 scenario questions
|Type of veterinary practice (No. respondents = 251)||First-opinion||244||97.2|| || || || || || || || ||13.8 (2.4–79.6)||0.003|
|Referral||30||12.0||2.62 (1.2–5.9)||0.02|| || ||3.2 (1.2–8.3)||0.02||6.2 (1.3–29.3)||0.02|| || |
|Main sources of information regarding antimicrobials and their use (n = 251)||Practice policy||40||15.9|| || ||8.5 (1.1–63.7)||0.04|| || || || || || |
|Veterinary medicines directorate||47||18.7|| || || || || || || || ||0.2 (0.1–0.5)||0.001|
|Continuing professional development courses||194||77.3||2.5 (1.2–5.4)||0.02|| || || || || || || || |
|Other||66||26.3||2.1 (1.1–3.8)||0.02|| || || || || || || || |
Veterinary surgeons working at referral practices, those who reported using CPD courses or other information sources such as colleague's knowledge and textbooks as a source of information about antimicrobials, were all more likely to prescribe off-licence antimicrobials.
Tables S1–S5 contain crude and adjusted odds ratios with 95% confidence intervals for risk factors associated with prescribing antimicrobials in any of the 4 scenario questions.
To date, there have been limited studies in the UK that have investigated how antimicrobials are used in veterinary practice. Major concerns have been the use of antimicrobials in situations in which bacterial infection is unlikely, inappropriate drug choice, the administration of antimicrobial drugs at incorrect dosage rates and for inappropriate durations . The lack of data concerning antimicrobial prescribing practices has meant it has not been possible to determine whether these concerns are unfounded or not. This study attempted to collect antimicrobial prescribing data from UK equine clinicians through the use of a scenario-based questionnaire. The scenario questions were not designed to have a correct or an incorrect answer, rather to collect objective data on the antimicrobials that clinicians would prescribe in 4 clinical scenarios. There are limitations associated with self-reported behaviour studies. For example, it is possible that responses of the veterinary surgeons to the clinical scenarios in our study are different to what they would practice in reality and respondents may be more likely to answer with what they think is the correct response. However, despite these drawbacks, this type of clinical scenario questionnaire design has previously been used successfully in studies investigating the antimicrobial prescribing practices of medical practitioners .
In this study, surprisingly, less than 1% of respondents reported that their practice had written antimicrobial use guidelines. A similar finding was reported by Prescott et al.  who reported that only 3 of 21 veterinary hospitals surveyed in Canada and the USA had antimicrobial use policies. The introduction of antimicrobial use guidelines to a small animal veterinary teaching hospital in Canada was shown to have a positive effect on the prescribing practices of clinicians, including a general reduction in the quantity of antimicrobials used and an increase in the relative use of first-line compared with third-line drugs . Similar effects have been documented in human hospitals and, in some cases, a reduction in the prevalence of antimicrobial resistant bacteria has been noted [18–20]. The introduction of antimicrobial use guidelines would appear to lead to more prudent use of antimicrobials and is a measure that should be implemented by veterinary practices across the UK.
Clinicians in this study most commonly used the NOAH compendium and CPD courses as sources of information about antimicrobials and their use. The NOAH compendium  contains datasheets for all preparations licensed for veterinary use in the UK, based upon the SPC of the drug. Therefore, the compendium is a good source of information regarding all aspects of antimicrobial drugs including dose rates, prescribing regimens and whether a drug is licensed for use in particular animal species. However, optimal antimicrobial selection is as important as prescribing the selected antimicrobial at the correct dose and for the correct duration of therapy . A compendium will not provide guidance to prevent the potential overuse of available broad-spectrum antimicrobials. It is likely that the most up-to-date information regarding appropriate drug selection will be provided by the scientific literature , which only 11.6% of respondents reported to use as a main source of information regarding antimicrobials. Continuing Professional Development courses were also identified as an important source of information about antimicrobials (77.3% of respondents). Continuing Professional Development courses tend to address very specific topics and as a result may not give a broad overview of antimicrobial use. Furthermore, respondents who reported CPD courses to be their main source of information regarding antimicrobials, were more likely to prescribe drugs not licensed for use in horses. Many antimicrobial drug preparations are not licensed for use in horses due to the large cost involved with obtaining a licence for use in a specific species. However, there are circumstances where off-licence use of antimicrobials is permitted as long as the drugs are prescribed according to the cascade , which provides a legal mechanism allowing veterinary surgeons to use their clinical judgement to prescribe a medicine where no authorised medicine exists.
Results of multivariable analysis revealed that veterinary surgeons working at referral practices were more likely to prescribe 3rd and 4th generation cephalosporins and fluoroquinolones in the clinical case scenarios. It is possible that clinicians working at referral centres are more likely to encounter animals with infections caused by multi-drug resistant bacteria due to prior treatment failures. They therefore may be more likely to use antimicrobials that would not normally be considered first-line options. Despite clinicians working at referral practices being significantly more likely to prescribe these classes of antimicrobials, only 3.2% (n = 28/879) and 1.0% (n = 9/879) of all prescriptions written for the clinical case scenarios were for 3rd and 4th generation cephalosporins and fluoroquinolones, respectively. This may suggest that the use of these drugs is not common practice in UK equine medicine in general but may be more common in equine referral practices (3rd and 4th generation cephalosporins and fluoroquinolones were used by 23.3 and 10.0% of respondents, respectively who worked at referral practices compared to 9.0 and 1.8% of respondents, respectively who worked in other practice types). Fluoroquinolones are not licensed for use in horses in the UK ; however, more recently, the fluoroquinolone enrofloxacin (Baytril) has gained popularity in UK equine practice because it can be administered orally.
Clinicians working in first-opinion practice were significantly more likely to prescribe trimethoprim-sulphonamide antimicrobials. Combinations of trimethoprim and sulphonamides are commonly used in antimicrobial treatment of horses because of their broad-spectrum of activity, the convenience of oral administration, lack of apparent toxic effects and low cost ; therefore, it is not surprising that they are associated with first-opinion practices. Furthermore, clinicians who used the VMD as their main source of information regarding antimicrobials and their use were less likely to prescribe trimethoprim-sulphonamides to horses in any of the 4 scenarios. High levels of antimicrobial resistance to trimethoprim-sulphonamides have been reported in bacteria isolated from horses [23–25] and this may be a cause for concern over the popularity of trimethoprim-sulphonamides in UK equine practice. Studies from the UK have demonstrated high levels of trimethoprim resistance in E. coli in faecal samples from horses both in an equine hospital  and also from horses in the community [25,27]. In a similar study performed in a human hospital, the use of trimethoprim-sulphonamides was associated with a decreased susceptibility of a range of bacteria to other antimicrobials .
The prescription of antimicrobials at doses under and over the correct dose, as stated by the VMD  occurred in each clinical scenario despite the weight of the horses being stated in each clinical scenario. Sixty-one percent of veterinary surgeons surveyed reported that they rarely either weigh or use weight tapes, to determine the weight of horses when they prescribe antimicrobials. It is often not possible to weigh a horse using weigh scales in first-opinion practice as most consultations will occur outside a clinic where weight scales are unlikely to be available. However, it is possible to estimate the weight of a horse using a weight tape . The same problem exists in human medicine; dosing regimens for antimicrobial drugs do not routinely take into consideration the bodyweight of the person for whom the antimicrobial is being prescribed despite recommended dose rates being available . The decreasing susceptibility of many common pathogens to antimicrobial agents in both human and veterinary medicine means that it is important to avoid practices such as substandard dosing that may select for the evolution of antimicrobial resistance .
In conclusion, with an increasing incidence of antimicrobial resistance and a decline in antimicrobial drug development programmes , the options for antimicrobial therapy are becoming narrower for both medical and veterinary clinicians. It is vital that antimicrobials are used appropriately to maintain their effectiveness. This study has provided objective data on antimicrobial prescribing practices in UK equine practice, which is necessary to assess whether the use of antimicrobials follows existing guidelines. There are examples of human and veterinary hospitals that have implemented successful antimicrobial stewardship programmes to improve antimicrobial prescribing practices ranging from the production of practice antimicrobial use guidelines to having to ask for permission when prescribing certain antimicrobials [15,20]. Less than 1% of veterinary practices in this study reported that they had an antimicrobial use policy. It has been suggested that any institution where antimicrobials are prescribed, both for the treatment of people or animals, should be encouraged to produce an antimicrobial use policy and ensure that it is reviewed and updated regularly .
Conflicts of interest
The corresponding author has funding from a pharmaceutical company for another project on antimicrobial resistance.
Source of funding
This work was funded by the UK Department for Environment, Food and Rural Affairs (DEFRA).
The authors would like to thank all the veterinary clinicians who took part in this study.
a Microsoft; Redmond, Washington, USA.
b Stata Corporation, College Station, Texas, USA.