*Article* **Knowledge, Attitude, and Practices (KAP) Survey among Veterinarians, and Risk Factors Relating to Antimicrobial Use and Treatment Failure in Dairy Herds of India**

**Deepthi Vijay <sup>1</sup> , Jasbir Singh Bedi 1,\*, Pankaj Dhaka <sup>1</sup> , Randhir Singh <sup>1</sup> , Jaswinder Singh <sup>2</sup> , Anil Kumar Arora <sup>3</sup> and Jatinder Paul Singh Gill <sup>1</sup>**


**Abstract:** The indiscriminate usage of antimicrobials in the animal health sector contributes immensely to antimicrobial resistance (AMR). The present study aims to assess the antimicrobial usage pattern and risk factors for AMR in animal husbandry sector of India. A cross-sectional survey about Knowledge, Attitude, and Practices (KAP) among veterinarians was carried out using a questionnaire comprising of 52 parameters associated with antibiotic use and the emergence of AMR in dairy herds. Respondents' KAP scores were estimated to rank their level of knowledge, attitude, and practice. Furthermore, risk factors associated with treatment failure were analyzed by univariable and multivariable analyses. Out of a total of 466 respondents, the majority had average knowledge (69.5%), neutral attitude (93.2%), and moderate practice (51.3%) scores toward judicious antibiotic usage. Veterinarians reported mastitis (88.0%), reproductive disorders (76.6%), and hemoprotozoan infections (49.6%) as the top three disease conditions that require antibiotic usage. Most of the veterinarians (90.6%) resorted to their "own experience" as the main criteria for antibiotic choice. The use of the highest priority critically important antimicrobials (HPCIA) listed by the World Health Organization (WHO) in animals, particularly quinolones (76.8%) and third-generation cephalosporins (47.8%), has been reported. On multivariable regression analysis of the risk factors, the lack of cooperation of the dairy farmers in the completion of a prescribed antibiotic course by the veterinarian and the demand for antibiotic use even in conditions not requiring antibiotic use were found to be significantly associated with the outcome variable "treatment failure" having respective odds of 1.8 (95%CI: 1.1–3.0) and 3.6 (95%CI: 2.3–5.8) (*p* < 0.05). The average KAP score of veterinarians, poor farm management practices, lack of awareness among farmers on prudent antibiotic use, and lack of antibiotic stewardship are the significant factors that need attention to combat the rising AMR in veterinary sector in India.

**Keywords:** antimicrobial resistance; antimicrobial usage; bovine; India; KAP survey; veterinarians

#### **1. Introduction**

Antimicrobial resistance is one of the greatest public health threats that has been projected to cause globally 10 million deaths and US\$100 trillion economic loss by 2050 [1]. In order to meet the food security of burgeoning human population, the economic scale production of food animals favor the high-density farming operations, which could double the antibiotic consumption by livestock in developing countries by 2030 [2,3]. The widespread application of antibiotics to food animal populations imposes strong selection pressure,

**Citation:** Vijay, D.; Bedi, J.S.; Dhaka, P.; Singh, R.; Singh, J.; Arora, A.K.; Gill, J.P.S. Knowledge, Attitude, and Practices (KAP) Survey among Veterinarians, and Risk Factors Relating to Antimicrobial Use and Treatment Failure in Dairy Herds of India. *Antibiotics* **2021**, *10*, 216. https://doi.org/10.3390/ antibiotics10020216

Academic Editor: Jeroen Dewulf

Received: 13 November 2020 Accepted: 15 January 2021 Published: 22 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

which contributes to the emergence, spread, and persistence of resistant pathogens to other animals, humans, and the environment [4]. The awareness on antibiotic resistance in human medicine has gained momentum; however, the role of animal husbandry practices in tackling antibiotic resistance are still being discussed with limited awareness among stakeholders, especially in developing countries [5].

India is bestowed with huge livestock wealth comprising of 193.5 million cattle and 109.9 million buffaloes [6]. The emerging intensive farming practices of the country has been posited as the hotspots of antibiotic resistance, and by 2030, the use of antibiotics in food animals has been projected to increase by 82% [2]. The threat of antibiotic resistance from the foods of animal origin has been discussed in many recent studies in India, highlighting the need for the judicious use of antibiotics in the animal health sector of the country [7–9]. Albeit, a "National Action Plan on Antimicrobial Resistance" has been enforced for optimizing antibiotic use in the country, strict enforcement still needs to be executed at the ground level [10–12].

While the reliability of data on the usage of antibiotics in the animal husbandry sector is questioned in general, some developing countries including India have a negligible amount of data [2,13]. In the midst of antimicrobial resistance crises with limited existing treatment options, mitigation strategies mainly revolve around awareness and proper stewardship for antibiotic usage among the key stakeholders. Thereby, understanding of knowledge, attitude, and practices (KAP) among the main stakeholders (e.g., veterinarians) with regard to antimicrobial use and resistance can help in the development of tailored intervention strategies to address poor practices, lack of knowledge, and negative attitude. Keeping in view the fact that there is no systematic KAP study along with prevailing antibiotic usage patterns and resistance in animal husbandry sector in Indian settings, the objectives of the present study were to assess Knowledge, Attitude, and Practices (KAP) among veterinarians in relevant to antimicrobial usage in animal husbandry sector through cross-sectional surveys, and identify the risk factors for the development of antimicrobial resistance (AMR) in India.

#### **2. Material and Methods**

#### *2.1. Study Design and Questionnaire Development*

The descriptive study was designed as a questionnaire-based cross-sectional analysis among the veterinarians of India during February 2020 to June 2020. A comprehensive review of the literature has been conducted to identify the factors influencing knowledge, attitude, and practices (KAP) on antimicrobial usage and resistance among veterinarians [14–16]. The questionnaire design was guided by the results from qualitative interviews and focus group discussions with veterinary academicians and farm animal practitioners of the Guru Angad Dev Veterinary and Animal Sciences University, Ludhiana, India. The questionnaire consisted of close-end questions, Likert scale statements, and open-ended questions exploring the existing knowledge, antimicrobial prescribing behaviors, perceptions on antimicrobial usage, and field practices associated with antimicrobial resistance. In addition, the veterinarian's recommendations were also requested for suggesting the interventions to combat antimicrobial resistance in the animal husbandry sector.

The questionnaire was divided into five sections: (1) Personal information; (2) Health services; (3) Knowledge, attitude, and practices toward antibiotic use; (4) Knowledge, attitude, and practices toward antimicrobial resistance; and (5) Miscellaneous section covering practices and recommendations for combating antimicrobial resistance.

The preliminary draft of the questionnaire having 58 questions was reviewed by five expert researchers to identify ambiguity and content validity. Later, the questionnaire was piloted among 20 veterinarians to assess its duration, clarity, and sequence. During the processing, six questions were omitted that were inappropriate, resulting in a total of 52 questions in the final questionnaire (Supplementary Material File S1).

#### *2.2. Sampling Procedure*

The source population of the present study comprised of registered veterinarians (Veterinary Council of India and/or State Veterinary Council) of India, and the study population included veterinarians who fulfilled the inclusion criteria of being farm animal practitioners. The sample size was calculated using the 'Raosoft calculator' (Raosoft: http://www.raosoft.com/samplesize.html?nosurvey). The sample size of 377 was estimated based on 50% response distribution, a 5% margin of error, and a 95% confidence interval. The expected response proportion of 50% was assumed based on the fact that both responses and response rates were completely unknown, since there are no previously published similar studies from India. Thereby, a total of 800 questionnaires were sent to the veterinarians selected through registered emails and/or personal contacts from professional societies and social media groups. The questionnaire was administrated by using the online interface of Google Forms (Google LLC, Mountain View, CA, USA) to the target population, and the survey remained open from May 2020 to June 2020.

#### *2.3. Ethical Statement*

The research was conducted in accordance with the Declaration of Helsinki and national standards. All the required ethical considerations have been taken into account. The nature of the study was completely voluntary, and informed consent was obtained from study participants. The details of the participants were anonymous, and data confidentiality was properly maintained.

#### *2.4. Statistical Analysis*

The completed questionnaires were manually checked for data quality before coding on Microsoft® Office Excel 2010. The study variables were summarized using proportions for qualitative variables and median and median absolute deviation for quantitative variables. The Likert-scale questions were condensed into two categories for analysis. A scoring system was generated by the subject experts of the University, in which the participants were given a score for knowledge, attitude, and practices based on the number of correct or appropriate responses. The overall score was determined based on the sum of correct answers to the eleven knowledge-based questions, four attitude-based questions, and thirteen practice-based questions. The respondent's level of knowledge/attitude/practices were categorized as "high/positive/good", "average/neutral/moderate", or "low/negative/poor" using the ≥75th percentile, <75th to 25th percentile, and <25th percentile of the individual scores, respectively. The Mann–Whitney U test/Kruskal–Wallis H test were used to determine the relationship between demographic characteristics of the veterinarians and their KAP scores. The correlation among the knowledge, attitude, and practice scores were assessed by the Spearman correlation. A *p*-value of ≤0.05 was interpreted as significant. The logistic regression analysis was performed to estimate predictors for the outcome variable, "frequent treatment failure". The outcome variable "frequent treatment failure" depicting the failure of response of the animal to the first line of antibiotic treatment by the veterinarians was ascertained from the questionnaire. Various risk factors associated with "frequent treatment failure" were used as predictors determined by univariate odds ratio. The multicollinearity was checked to rule out the relationship amongst the independent variables based on the Variable Inflation Factor value (VIF) calculated in an iterative manner. The associations between the selected variables for multivariable analysis had a VIF of less than 2. The interactions between the predictors were checked and were found to be non-significant. The model was constructed by considering all these explanatory variables using the backward stepwise approach using the Likelihood Ratio Test (LRT). The analyses were conducted using SPSS version 24.0 (SPSS Inc., IBM, Armonk, NY, USA).

#### **3. Results**

A total of 478 (59.7%) responses were received out a total of 800 questionnaires, of which 466 were with complete information. The questionnaires that contained incomplete (*n* = 7) and vague information (*n* = 5) were excluded from the study.

#### *3.1. Demographic Information*

The demographic profile of the participants belonged to twenty-five states of India, which were grouped into six geographical regions (Table 1). Out of 466 participants with a median age of 32 years, 70.0% were males and 30.0% were females. The highest number of respondents belonged to the 30–40 age group (37.5%). It was observed that 48.1% of veterinarians had post-graduate qualifications. Most of the veterinarians (62.9%) had less than 10 years of field experience. The majority of the respondents were working in veterinary hospitals (85.6%), while 14.4% were in veterinary polyclinics that had established laboratory facilities.


**Table 1.** Demographic information of respondents.

**Table 1.** *Cont.*


#### *3.2. Common Diseases Requiring Antibiotic Usage in Bovines*

Major disease conditions found in bovines requiring antibiotic usage are listed in Figure 1. The veterinarians reported mastitis (*n* = 410), reproductive disorders (*n* = 343), and hemoprotozoan infections (*n* = 231) as the top three disease conditions in bovines where antibiotics are widely used.

**Figure 1.** Major disease conditions requiring antibiotic use in bovines in India \* (\* Question: Top 03 disease conditions that require antibiotic use in bovines. Each veterinarian was asked to choose up to three disease conditions).

#### *3.3. Antibiotic Prescribing Decisions*

The decision over the choice of antibiotics in various diseases/conditions in bovines was influenced by different factors (Figure 2). The majority of the veterinarians (90.6%) depended on their own experience as the top criteria for choosing antibiotics followed by the availability (63.3%) and cost (59.0%) of the antibiotic. Recommendations from other veterinarians (31.8%) and pharmaceutical companies (7.1%) also influenced their decision regarding antibiotic use. Around 28% of the veterinarians took into account positive culture and sensitivity test results, whereas the withdrawal period of the drug influenced only 15% of veterinarians in prescribing the antibiotics. In addition, 24.9% veterinarians reported that the demand and expectation of farmers influences the prescription behaviors of antibiotics, even for conditions that do not require their use.

**Figure 2.** Factors determining the choice of antimicrobial use by veterinarians \* (\* Question: What are the top three factors in determining the choice of antibiotics use in your treatment? Choose among the following (Own experience/Availability of antibiotic(s)/Recommendations from other veterinarians/Cost of antibiotic/Positive culture and susceptibility tests/Drug withdrawal times/Recommendations from pharmaceutical company). Each veterinarian was asked to choose up to three top factors determining the choice of antimicrobials).

> The veterinarians reported the use of "highest priority critically important antimicrobials" (HPCIA) mentioned by World Health Organization (WHO) [17] in their choices for treatment, *viz.*, quinolones (76.8%; *n* = 358), third-generation cephalosporins (47.8%; *n* = 223), and fourth-generation cephalosporins (6.0%; *n* = 28) (Figure 3a). The quinolones (71.9%, *n* = 335) were the most commonly prescribed antibiotic for mastitis followed by third-generation cephalosporins (64.2%; *n* = 299) (Figure 3b). In case of metritis, thirdgeneration cephalosporins (55.6%; *n* = 259) followed by tetracycline (50.6%; *n* = 236) and quinolones (50.0%; *n* = 233) were the top three commonly used antibiotics (Figure 3c).

> However, 45.5% of the veterinarians were aware of the 'critically important list of antimicrobials' of the WHO [17], while 59.2% opined that restriction on the WHO suggested 'priority antibiotics for human-use only' is not possible in veterinary therapeutics. The antibiotics in the 'reserve group' as proposed by the WHO [18], particularly fourthgeneration cephalosporins, were used by 13.5% of the veterinarians in mastitis and by 6.9% of veterinarians in metritis. Moreover, 1.9% of the veterinarians reported the use of fifth-generation cephalosporins in mastitis. In addition, uses of alternate therapies such as herbal medicines were reported by 74.0% veterinarians, whereas 67.2% used probiotics, 43.8% used homeopathic medicine, and 2.4% used indigenous remedies for different disease conditions.

#### *3.4. Knowledge, Attitude, and Practice (KAP) Analysis*

The knowledge of the respondents on antimicrobial use and resistance was assessed by scoring eleven questions, with the score 1 given to correct answer while 0 was given to incorrect or not sure response (Table 2; Supplementary Material Table S1). The knowledge was scaled as high with a score ≥ 9, average with a score 6–9, and low with a score < 6. The median knowledge score of the respondents was 8.0 ± 1.0. Only 14.2% of the respondents had a high knowledge score, whereas most respondents (69.5%) had an average knowledge score. The majority of the respondents (73%) were regularly updating themselves on antimicrobial resistance, where the internet was the most common information source (Figure 4). A significantly higher knowledge score was observed among the veterinarians who regularly updated themselves compared with those who did not (U statistic: 4.6, *p*-value: 0.00).

**Figure 3.** Commonly used antibiotics in bovines ((**a**) overall use; (**b**) use in mastitis; (**c**) use in metritis)\* (\* Questions: (**a**) Top three frequently used antibiotics in the treatment of bovines; (**b**) Top three frequently used antibiotics for the treatment of mastitis in bovines; (**c**) Top three frequently used antibiotics for the treatment of metritis in bovines. Each veterinarian was asked to choose up to three most commonly used antibiotics).


**Table 2.** Knowledge, attitude, and practices (KAP) of veterinarians regarding antibiotic use and resistance.

\* Partially correct answers. ¶ Each question was scored with score of 1 for correct, 0.5 for partially correct, and 0 for incorrect or not sure responses.

**Figure 4.** Information sources referred by veterinarians on antibiotic use and resistance \* (\* Question: What are the major information sources that you refer regularly to for knowledge on antibiotic use and resistance? (The question was open-ended with the provision to answer more than one source of information)).

Attitude toward antibiotic use and associated resistance was assessed by four questions (Table 2, Supplementary Material Table S1) with score of 1 for correct, 0.5 for partially correct, and 0 for incorrect or not sure response. The attitude score of ≥2.5 was classified as positive, 0.5–2.5 was classified as neutral and < 0.5 was classified as negative. The majority of the respondents (93.3%) had attitude score in the neutral range, with an overall median of 1.5 ± 0.5.

The practice scores were assessed for thirteen questions (Table 2; Supplementary Material Table S1) with a score of 1 for correct, 0.5 for partially correct and 0 for incorrect practice. The practice scale with a score of ≥7.5 was classified as good, 4.5–7.5 was classified as moderate and <4.5 was classified as poor. The respondents had a median practice score of 6.0 ± 1.5. The majority of the respondents (51.3%) had a moderate practice score and 27.7% stated poor practice toward antimicrobial usage. In addition, 27.2% of veterinarians had attended training programs on antibiotic usage and resistance. The veterinarians who attended the training program had significantly higher practice scores (U statistic: 5.3, *p*-value: 0.00) and knowledge scores (U statistic: 3.8, *p*-value: 0.00).

#### *3.5. Association of KAP Scores with Demographic Characteristics*

The association of demographic characteristics and KAP scores were analyzed using the Mann–Whitney U test/Kruskal–Wallis H test (Table 3). A significant difference was observed among the age groups, with higher knowledge (*H* statistic: 10.9, df: 4, *p*-value: 0.03) score in the <30-year age group. Post hoc analysis revealed that the knowledge scores of veterinarians having age <30 differed significantly from the other age groups. The veterinarians with PhD degrees had significantly higher knowledge scores (*H* statistic: 37.8, df: 2, *p*-value: 0.00), and on post hoc analysis, the knowledge scores of all the groups having different educational qualifications differed significantly from each other. Moreover, a higher knowledge (*H* statistic: 19.1, df: 3, *p*-value: 0.00) score was observed among veterinarians having less than 10 years of experience, and post hoc analysis revealed that the knowledge score of veterinarians having less than 10 years of experience and veterinarians with 20–30 years of experience differed significantly from the knowledge score of veterinarians with 30–40 years of experience. The knowledge and attitude scores had no significant difference between the regions, while a higher practice score was observed amongst the veterinarians from the Western region (*H* statistic: 13.7, df: 5, *p*-value: 0.02), and post hoc analysis revealed that the practice score of veterinarians of the Western region differed significantly from that of respondents of the Northern and Eastern region. The

veterinarians working in veterinary polyclinics had higher knowledge scores than those working in veterinary hospitals (*U* statistic: 2.2, *p*-value: 0.03).


**Table 3.** Demographic characteristics and associated KAP scores.

\* Significant *p*-values are presented in bold characters. **¶** Kruskal–Wallis H test; **#** Mann–Whitney U test.

#### *3.6. Correlation between Knowledge, Attitude and Practice Scores*

The present study revealed weak linear correlations between knowledge–attitude (*r* = 0.23, *p* < 0.000), knowledge–practice (*r* = 0.20, *p* < 0.000), and attitude–practice (*r* = 0.18, *p* < 0.001) as per the criteria by Cohen (2013) (0–0.25 = weak correlation, 0.25–0.5 = fair correlation, 0.5–0.75 = good correlation, and >0.75 = excellent correlation) [19].

#### *3.7. Risk Factors Associated with Treatment Failure*

Most of the veterinarians (86.0%) admitted about ongoing antibiotic abuse in therapeutics, and 98.7% considered antimicrobial resistance as a serious public health issue. Frequent treatment failure has been reported by 21.7% of veterinarians, and therapeutic failure has been observed in mastitis treatment against HPCIA such as quinolones (13.5%), third-generation cephalosporins (11.4%), and high-priority antimicrobials such as synthetic penicillin (11.6%), penicillin (11.4%), and aminoglycosides (9.2%). For metritis treatment, veterinarians reported therapeutic failure against quinolones (2.4%), tetracyclines (2.1%), synthetic penicillins (1.9%), and third-generation cephalosporins (1.7%). The failure of

effective therapeutic response to antimicrobials other than antibiotics was reported by 66.1% of veterinarians for antiparasitic drugs and 9.4% for antifungal drugs.

The majority of the veterinarians (86.5%) attributed unauthorized practitioners (commonly called "quacks") followed by farmers and para-vets (43.6% each) as responsible for irrational use of antimicrobials in livestock (Figure 5). The practice of farmers directly acquiring antibiotics from a pharmacy without prescription was reported by 82.8% of the veterinarians, whereas 39.5% of the veterinarians reported non-cooperation of the farmers in the completion of the antibiotic course prescribed by them. However, 31.8% veterinarians organized awareness camps on antibiotic usage and resistance for farmers.

**Figure 5.** Personnel responsible for irrational use of antibiotics in field \*. (Quacks: unauthorized practitioners; Paraveterinarians: diploma holders in Veterinary Science, Class IV: helping staff in veterinary hospitals) (\* Question: Whom do you think as responsible for the irrational use of antibiotic in bovines at the field level (select all that apply)? (The question was having the provision to select more than one option)).

> Around 16.3% of veterinarians considered themselves responsible for the injudicious use of antimicrobials, and 39.1% of veterinarians used antibiotics for prophylaxis, especially to prevent outbreaks. The majority of the veterinarians (62.2%) rarely performed antibiotic susceptibility testing to complement their treatment, while 70.6% of veterinarians reported lack of laboratory facilities for performing antibiotic sensitivity testing in/near their hospital. Moreover, only 20.8% veterinarians were aware about the recommendations of the National Antimicrobial Resistance Plan of 2017, India [20].

#### *3.8. Univariable and Multivariable Analysis*

The univariable analysis for frequent treatment failure associated risk factors pertaining to veterinarian's and farmer's practices was carried out by calculating the odds ratio (Table 4). All the variables of univariable analysis were used for building logistic regression models using independent predictors of practices associated with veterinarians and farmers in respect to frequent treatment failure.

On multivariable logistic regression analysis with a backward stepwise approach using the Likelihood Ratio Test (LRT), the final model contained two variables as depicted in Table 5. With respect to the risk factors associated with veterinarians, "skipping doses of antibiotics" and "allowing farmer to inject subsequent doses of antibiotics after administering first dose of the treatment" were significantly found to be associated with frequent treatment failure, with respective odds ratios of 1.7 (95%CI: 1.1–2.6) and 1.8 (95%CI: 1.1–2.8) (*p*-value: <0.05) (Table 5a). The adjusted odds ratio of "illegitimate demands of farmers for antibiotic use" and "farmer's non-cooperation in completion of antibiotic course" were found to be significantly associated with "frequent treatment failure", with respective odds ratios of 3.6 (95%CI: 2.3–5.8) and 1.8 (95%CI: 1.1–3.0) (*p*-value: <0.05) (Table 5b). The Hosmer–Lemeshow test for goodness of fit was found to be non-significant for both the models of veterinarians and farmers (Table 5).


**Table 4.** Univariable analysis: (**a**) Veterinarians; (**b**) Farmers.

**Table 5.** Multivariable logistic regression analysis: (**a**) Veterinarians; (**b**) Farmers.


#### *3.9. Veterinarian's Recommendations*

The respondents were asked to provide a single best suggestion to combat antimicrobial resistance. The suggestions overlapped in many cases, and the duplicate suggestions were removed and are categorized into field level, policy level, and research level suggestions in Supplementary Material Table S2.

#### **4. Discussion**

In developing countries, possible factors for antibiotic resistance include increased and indiscriminate use of antibiotics in animal production, poor farm biosecurity, inadequate infection control practices in consort with lack of compliance with regulatory frameworks [21]. In Indian dairy herds, more than 70% of production losses have been incurred by mastitis, which remains the condition requiring the most antibiotic use [22]. Similarly, in the present study, veterinarians reported mastitis as the most common condition in bovines requiring antibiotic use followed by reproductive disorders and hemoprotozoan infections.

There are limited studies from India on antibiotic usage patterns for various conditions in animal husbandry [12]. The present study listed major disease conditions of bovines requiring antibiotic usage. Our study reports the use of HPCIA in animal therapeutics, with quinolones and third-generation cephalosporins as prime antibiotics used for mastitis and metritis. However, studies from western countries reported the use of non-HPCIA predominating in animal agriculture, while the use of critically important antimicrobials was limited to the treatment of diarrhea and respiratory diseases in bovines [23]. Similarly, in Australia, the major antibiotics in bovine therapeutics were tetracycline/doxycycline, penicillin, synthetic penicillin, and trimethoprim–sulfamethoxazole [16]. In addition, the alternate systems of medicine are prevalent both in the human and veterinary sector in India [24–26], and the veterinarians in the study also reported the widespread usage of herbal medicines and homeopathy in bovine therapeutics.

While choosing the antibiotics, previous experience of veterinarians remained the topmost criteria, which is in accordance with previous studies where veterinarian's prior experience of a drug was decisive for antibiotic selection [27]. Moreover, the cost of antibiotics had a moderate influence on antibiotic choice, as also reported by Australian veterinarians [16]. The lower use of antimicrobial culture and susceptibility testing in choosing antibiotics was in accordance with the study on New Zealand veterinarians [28]. The recommendations from the pharmaceutical company were a minor factor in the choice of antibiotics in contrary to the previous reports, where half of the veterinarians were influenced by the pharmaceutical companies [29].

In the present study, 69.5% of veterinarians had average knowledge score similar to earlier regional study from India, where 58.3% of veterinarians had a medium level of awareness on antibiotic resistance [30]. The majority of veterinarians had attitude in the neutral range and moderate practice scores, suggesting the need for more directed efforts on improving attitude and practices toward judicious antibiotic use. The highest knowledge and attitude scores were in the age group of <30 years and in veterinarians with <10 years of experience, which is in similar to earlier studies, where Dutch veterinarians with more years of experience were found to be less concerned about the possible contribution of veterinary antibiotic use to antimicrobial resistance [14]. The higher knowledge score among veterinarians working in veterinary polyclinics with established facilities is in accordance with reported higher social responsibility among veterinarians working in referral clinics [31]. The regional differences noted in the present study with a higher practice score for the Western region is in accordance with earlier studies where regional differences were observed [32], which might be due to the higher awareness of activities on animal husbandry practices, including farm biosecurity.

The highest consumption of antimicrobials in livestock has been reported in low- and middle-income countries where antibiotics are used for therapeutics, growth promotion, and prophylaxis [2]. In the present study, 39.0% of veterinarians reported the use of antibiotics for prophylaxis, mainly to prevent disease outbreaks, on contrary with developed nations where most veterinarians had abandoned the practice of using antibiotics for prophylaxis [33].

The reliability on diagnostic and antibiotic sensitivity testing is posited to be crucial for responsible antimicrobial use, while in the present study, 37.8% of veterinarians resorted to bacterial culture and susceptibility test results for choosing antibiotics. In addition, 70% veterinarians were not having access to well-equipped laboratory facilities for antibiotic susceptibility testing. This is in accordance with previous studies where in both veterinary [23,33] and human medicine [34], the use of antibiotic susceptibility testing for choosing antibiotics was less frequent. The lack of access to laboratory facilities for the majority of the veterinarians for confirming the root cause of treatment failure might have led to the assumption that treatment failure was due to antimicrobial resistance. Even though treatment failure may also arise due to other causes, such as the inadequate antimicrobial spectrum of the prescribed antibiotics due to the use of ineffective drugs or incorrect dosage or incorrect diagnosis, in the present study, more emphasis has been laid on antimicrobial resistance as leading causes of treatment failure, which might pose a limitation to the study.

The majority of veterinarians (87%) believed there is an ongoing antibiotic abuse in therapeutics in India, while a lower proportion of Australian livestock veterinarians opined the current usage of antibiotics as "significant" for antibiotic resistance [16]. Moreover, 98.7% veterinarians believed that antibiotic resistance was a serious public health issue, in similar line with the previous studies [35,36]. In addition, earlier studies also have reported a large number of untrained personnel (quacks) in veterinary practice in India, which might be due to unaffordable professional veterinary services for marginalized farmers [12,30,37].

The present study analyzed the possible risk factors of farmers and veterinarians for the development of treatment failure. The "illegitimate demands of farmers for antibiotic use" was significantly associated with treatment failure in accordance with the earlier studies, where around 33% veterinarians reported explicit demand of farmers for antibiotics [30]. On contrary, other study from Australia reported that the expectations of the client had a minimal influence on antibiotic prescription [16].

The majority of the veterinarians (82.8%) reported the purchase of antibiotics without prescription by farmers in accordance with earlier studies from India, where the lack of adequate knowledge among farmers and easy access to antibiotics without prescriptions were considered as possible drivers of this risk practice [38]. Around 31.8% of veterinarians have conducted training programs to improve knowledge of farmers on antibiotic usage. Earlier studies also reported that the majority of veterinarians believed in educating farmers on good management practices for reducing antimicrobial use [15,39].

In accordance with earlier studies where the Australian veterinarians have highlighted the need for cost-effective culture and susceptibility testing as well as rapid and affordable diagnostic tests for facilitating judicious antibiotic use [16], the present study has also put forward similar suggestions at the field level, regulatory level, and research level. The participating veterinarians of the present study have also emphasized the need for a datadriven interdisciplinary approach that is crucial for combating antimicrobial resistance. The present study could not have the exact proportional number of respondents from different regions of the country, which might pose a limitation. However, the study is the first of its kind to have a comprehensive approach on the existing antibiotic usage practices, KAP survey, and veterinarian's recommendations to address antimicrobial resistance.

#### **5. Conclusions**

To conclude, the facilitating changes in the attitude and practices of veterinarians can be augmented by the implementation of continuing veterinary education programs. The effective flow of information from veterinarians to farmers can create a paradigm shift in the perceptions of the farmers for judicious antibiotic use as well as less reliability on quacks. There is need to strengthen the laboratory surveillance networks, research and diagnostics, and judicious antimicrobial stewardship. More stringent guidelines on the use of HPCIA in the animal sector and the compliance with responsible antimicrobial prescription behaviors by veterinarians need to be implemented. A "One Health" framework facilitating behavioural change interventions in farmers and veterinarians by bringing all the stakeholders together and promoting prudent antimicrobial use and judicious antimicrobial stewardship is the need of the hour.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2079-63 82/10/2/216/s1, File S1: Questionnaire for veterinarians, Table S1: List of KAP parameters and their correct responses in questionnaire, Table S2: Recommendations at various levels from the study.

**Author Contributions:** Conceptualization, D.V., J.S.B, P.D., A.K.A. and J.P.S.G.; Methodology, D.V., J.S.B., P.D., J.S. and R.S.; Software, D.V., P.D. and J.S.B.; Formal Analysis, D.V., J.S.B., P.D. and J.S.; Investigation, D.V., P.D., J.S. and J.S.B.; Validation, R.S., A.K.A., J.S.B. and J.P.S.G.; Writing—Original Draft Preparation: D.V., J.S.B., and P.D.; Writing—Review and Editing: R.S., J.S., A.K.A., J.S.B. and J.P.S.G.; Supervision, A.K.A. and J.P.S.G.; and Project Administration, A.K.A. and J.P.S.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research was supported by grants from the Indian Council of Agricultural Research, Niche Area of Excellence project on 'Antibiotic Resistance: Animal Human Interface' (ICAR/Edn.10(8)/2016-EP&HS).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Advisory Committee of Guru Angad Dev Veterinary and Animal Sciences University, Ludhiana. The study was survey based and there was no involvement of any invasive procedures or experimental protocols on the study subjects.

**Informed Consent Statement:** The study was completely voluntary and proper consent of the participants were obtained before enrollment in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy and confidentiality agreements to the participants.

**Acknowledgments:** Authors are thankful to Guru Angad Dev Veterinary and Animal Sciences University, Ludhiana for providing necessary support for the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Antimicrobial Use in Animals in Timor-Leste Based on Veterinary Antimicrobial Imports between 2016 and 2019**

**Shawn Ting 1,\* , Abrao Pereira <sup>1</sup> , Amalia de Jesus Alves <sup>1</sup> , Salvador Fernandes <sup>2</sup> , Cristina da Costa Soares <sup>2</sup> , Felix Joanico Soares <sup>2</sup> , Onofre da Costa Henrique <sup>2</sup> , Steven Davis <sup>1</sup> , Jennifer Yan 1,3, Joshua R. Francis 1,3 , Tamsin S. Barnes <sup>4</sup> and Joanita Bendita da Costa Jong <sup>2</sup>**


**Abstract:** Monitoring veterinary antimicrobial use is part of the global strategy to tackle antimicrobial resistance. The purpose of this study was to quantify veterinary antimicrobials imported into Timor-Leste between 2016 and 2019 and describe the antimicrobial import profile of importers. Data were obtained from import applications received by the Ministry of Agriculture and Fisheries (MAF) of Timor-Leste. Import quantities were analysed by antimicrobial class, importance for human medicine, recommended route of administration and type of importer. An average of 57.4 kg (s.d. 31.0 kg) and 0.55 mg/kg (s.d. 0.27 mg/kg) animal biomass of antimicrobials was imported per year. Tetracyclines (35.5%), penicillins (23.7%), and macrolides (15.9%) were the commonly imported antimicrobial classes. Antimicrobials imported for parenteral administration were most common (60.1%). MAF was the largest importer (52.4%). Most of the critically important antimicrobials for human medicine were imported by poultry farms for oral administration and use for growth promotion could not be ruled out. In conclusion, the use of antimicrobials in animals in Timor-Leste is very low, in keeping with its predominantly subsistence agriculture system. Farmer education, development of treatment guidelines, and strengthening of the veterinary service is important for addressing the potential future misuse of antimicrobials especially in the commercial poultry industry.

**Keywords:** antimicrobial use; antimicrobial resistance (AMR); Timor-Leste; antibiotic; antimicrobial; veterinary; prudent use; critically important antimicrobials; growth promotion; poultry

### **1. Introduction**

The emergence of antimicrobial resistance is a major global health threat for the 21st century [1]. It is also a One Health challenge that requires coordinated action as transmission of resistant bacteria can occur between humans, animals, plants and the environment [2–4]. This emergence has been rapid and is linked to the overuse and misuse of antimicrobials in humans and animals [5,6]. Despite this, it is projected that the use of antimicrobials in humans and animals will continue to rise over the next decade [7,8]. In particular, the use of antimicrobials in food producing animals has received attention due to high levels of use globally for disease prevention and growth promotion [9,10]. While some developed countries have demonstrated a reduction in usage levels [11–15], usage in many developing countries have risen due to farm intensification and demand for animal-

**Citation:** Ting, S.; Pereira, A.; Alves, A.d.J.; Fernandes, S.; Soares, C.d.C.; Soares, F.J.; Henrique, O.d.C.; Davis, S.; Yan, J.; Francis, J.R.; et al. Antimicrobial Use in Animals in Timor-Leste Based on Veterinary Antimicrobial Imports between 2016 and 2019. *Antibiotics* **2021**, *10*, 426. https://doi.org/10.3390/ antibiotics10040426

Academic Editor: Clair L. Firth

Received: 16 March 2021 Accepted: 9 April 2021 Published: 12 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

based protein associated with rising incomes [16–18]. This puts low- and middle-income countries at a higher risk for emergence of resistance.

Antimicrobial resistance limits the effectiveness of antimicrobial therapy which has a greater impact in low and middle-income countries due to their weaker health systems, higher prevalence of infectious diseases and limited access to more expensive treatment alternatives [19,20]. To preserve the effectiveness of antimicrobials, a global strategy has been developed to tackle antimicrobial resistance [21]. This strategy is wide-ranging and multi-sectoral and includes initiatives to strengthen monitoring of antimicrobial use in animals [21].

Monitoring of antimicrobial use in animals at the national level enables a country to identify trends of use over time and assess the impact of policy measures to promote prudent use in animals [22]. When analysed in conjunction with data on antimicrobial resistance in animal and humans, it can also identify potential associations between antimicrobial use and resistance patterns [23,24]. To harmonize antimicrobial use data collection, the World Organisation for Animal Health (OIE) has published guidelines for monitoring the use of antimicrobials in food producing animals [25]. The guidelines acknowledge that antimicrobial use data can be obtained from different levels such as import, manufacturing, sales, dispensing records or from end-use sources [25]. While many higher income countries have been collecting data for many years [26,27], some low to middle income countries in Africa and Asia-Pacific are still facing challenges such as a lack of regulation, under-reporting and unreliable data when monitoring antimicrobial use in animals [10,28,29].

Timor-Leste is a lower-middle income country [30] located in the south-east portion of the Malay Archipelago with a population of 1.3 million [31]. Subsistence farming is the main livelihood for most of the rural population [32,33], with a high proportion of households owning livestock [34]. Chicken and pigs are the two most commonly reared species in the country [35]. Commercial animal farming is uncommon [36,37] but may increase with rising income levels [33]. Currently, there are two large commercial layer farms [38] and a growing number of commercial broiler farms. There are no major commercial livestock farms for other species. There is no local manufacture of antimicrobials, and all antimicrobials are imported into the country. All applications to import veterinary medicines into the country must be submitted to the Ministry of Agriculture and Fisheries (MAF) and there is no re-export of veterinary antimicrobials.

The aims of this study were to quantify veterinary antimicrobial imports into Timor-Leste between 2016 to 2019; and to describe these imports based on antimicrobial class, importance for human medicine, recommended route of administration and type of importer. The findings can help improve monitoring and control of veterinary antimicrobial use in Timor-Leste.

#### **2. Materials and Methods**

#### *2.1. Data Collection for Antimicrobial Imports*

All applications to import veterinary medicines into Timor-Leste submitted to MAF between January 2016 to December 2019 were screened to identify veterinary antimicrobials using OIE's list of antimicrobials of veterinary importance [39]. Data on the date of application, name of importer, brand name, quantity imported, name of active ingredient, concentration of active ingredient, route of administration and target species was extracted for each veterinary antimicrobial. Any missing details on the name of active ingredient, concentration of active ingredient and route of administration was obtained from the technical product sheets. Data collection was performed by two MAF staff who received training on recording antimicrobial import data from received import applications through three workshops and ongoing side-by-side mentorship [40]. The data was stored on an Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA). Data accuracy was checked independently by three researchers from Menzies School of Health Research between November and December 2020.

#### *2.2. Data Categorisation for Antimicrobial Imports*

Using the name of the active ingredient, each antimicrobial was classified into an antimicrobial class/subclass based on OIE guidelines [41]. The name of the active ingredient was also used to classify antimicrobials as a critically important antimicrobial (CIA), highly important antimicrobial (HIA) or an important antimicrobial (IA) using the World Health Organization (WHO) List of Critically Important Antimicrobials for Human Medicine [42]. The importer name was used to classify importers into 6 types to understand their individual import patterns: "MAF", "agriculture shops", "veterinary clinics", "layer farms", "broiler farms", and "education institutions". Layer and broiler farms were placed in separate categories because they may have different antimicrobial use patterns. In Timor-Leste, agriculture shops are enterprises where veterinary medicines can be procured without a prescription.

#### *2.3. Animal Biomass Calculation*

Data for biomass calculation (i.e., number of live animals, number of animals slaughtered and meat product quantity) were obtained from the Food and Agricultural Organization Global Statistical Database (FAOSTAT) [43,44]. Common animal species in Timor-Leste (buffalo, cattle, chicken, goats, horse, pigs, and sheep) [34] were included in the biomass calculation. Ducks, rabbits, dogs, and cats were excluded because data were not available. Total animal biomass was calculated for each year between 2016 and 2019 using an OIE method [25] except for bovine biomass because the proportion of animals in different age groups are not known. The data for total animal biomass calculation and estimates for annual biomass can be found in Supplementary Table S1.

#### *2.4. Data Analysis*

The weight of active ingredient in one unit of imported product per pharmaceutical form (e.g., bottle, bag, or tube) was estimated by multiplying the strength of the antimicrobial active ingredient by the volume or weight. All weights were expressed in kilograms (kg). Conversion factors based on OIE guidelines was used to mathematically convert international units (IU) into kilograms [29].

The weight of each active ingredient imported between 2016 to 2019 was calculated by multiplying the weight of active ingredient in one unit of product by the quantity imported. Adjustment for animal biomass was achieved by dividing the total weight of active ingredient by the total animal biomass. The result was expressed in milligram (mg) of active ingredient per kilogram (kg) of animal biomass.

Annual and total imports were calculated for each active ingredient, antimicrobial class, WHO class of importance in human medicine, route of administration and type of importer. Total annual imports of all antimicrobials by weight and weight adjusted for biomass were summarized as mean ± s.d. Spearman's rank correlation coefficient (rs) was used to test the hypothesis of a monotonic (increasing or decreasing) trend in imports by total weight, total weight adjusted for biomass, individual active ingredient and type of importer. Data analysis was performed using Stata 15 software (StataCorp, College Station, TX, USA)

#### *2.5. Ethical Approval*

The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by Human Research Ethics Committee of the Northern Territory (NT) Department of Health and Menzies School of Health Research (2020-3841) and Institute Nacional de Saude in Timor-Leste (MS-INS/DE/IX/2020/1411).

#### **3. Results**

#### *3.1. Import Quantities and Trends*

Between 2016 to 2019, a total of 229.8 kg of active ingredients of veterinary antimicrobials were imported into Timor-Leste (mean: 57.4 ± 31.0 kg per year). Import quantities

were lower in 2017 and 2018 compared to 2016 and 2019 (see Table 1). After adjusting for animal biomass, the average amount of imported antimicrobials was 0.55 ± 0.27 mg/kg biomass per year. There was no evidence of a significant monotonic trend in antimicrobial imports based on total weight (rs: −0.40, *p* value: 0.60) or weight adjusted by biomass (rs: −0.40, *p* value: 0.60) (see Figure 1).

**Figure 1.** Antimicrobial import weight (mg) adjusted by animal biomass (kg) into Timor-Leste between 2016 and 2019.

A total of 21 antimicrobial active ingredients belonging to 8 classes of antimicrobials were imported during the study period. The import quantities of different antimicrobials between 2016 and 2019 can be found in Table 1. The active ingredients imported in the largest quantities were oxytetracyline (81.7 kg; 35.5%), amoxicillin (34.8 kg; 15.2%), tylosin (25.2 kg; 11.0%) and dihydrostreptomycin (25.8 kg; 11.2%). The classes of antimicrobials imported in the largest quantities were tetracyclines (81.7 kg; 35.5%), penicillins (54.4 kg; 23.7%), macrolides (36.5 kg; 15.9%) and aminoglycosides (25.8 kg; 11.3%). There was some evidence of monotonic increase in imports of neomycin (rs: 0.95, *p* value: 0.05) but quantities imported each year were extremely small. There was also some evidence of a monotonic decrease in imports of tylosin (rs: −0.95, *p* value: 0.05) driven by a relatively large import in 2016 and sulfamonomethoxine (rs: −0.95, *p* value: 0.05) although quantities imported each year were extremely small. There was no strong evidence of a monotonic trend in the import of any of the other individual antimicrobials (see Table 1). Based on WHO classification, most of the imported veterinary antimicrobials were CIAs (117.9 kg; 51.3%) followed by HIAs (111.8 kg; 48.7%).

*Antibiotics* **2021**, *10*, 426



#### *3.2. Import Pattern by Recommended Route of Administration*

Recommended routes of administration for imported veterinary antimicrobials during the study period were parenteral (138.0 kg; 60.1%), oral (91.5 kg; 39.8%), and topical (0.3 kg; 0.1%). The majority of tetracyclines (81.3 kg; 99.6%), aminoglycosides (25.8 kg; 99.9%), sulphonamides (13.4 kg; 96.2%), and cephalosporins (0.01 kg; 100%) were for parenteral administration, while the majority of penicillins (37.6 kg; 69.2%), macrolides (36.3 kg; 99.5%), polypeptides (10.9 kg; 100%), and fluoroquinolones (6.0 kg; 100%) were for oral administration. The quantities of different antimicrobial classes for parenteral, oral and topical administration are shown in Figure 2. The weight of antimicrobial classes recommended for administration through different routes for each year over the study period can be found in Supplementary Table S2.

#### *3.3. Import Pattern by Importer Type*

Between 2016 and 2019, the biggest importers of antimicrobials were MAF (120.4 kg; 52.4%), followed by layers farms (81.1 kg; 35.3%) and agriculture shops (15.9 kg; 6.9%) (See Figure 3). There was very strong evidence of a monotonic increase in antimicrobial imports by MAF (rs: 1.0, *p* value: <0.001) and evidence of a monotonic increase in antimicrobial imports by broiler farms (rs: 0.95, *p* value: 0.05) but no evidence of a monotonic trend in antimicrobial import patterns for other types of importers (see Figure 4A,B). The pattern of imports by layer farms was unique as imports were high in 2016 (58.6 kg) and 2019 (22.5 kg) but negligible between those years (see Figure 4A). Educational institutions imported a relatively small amount (0.6 kg) of antimicrobials once in 2016. Colistin, neomycin and enrofloxacin were only imported by layer or broiler farms. Cephalosporins were only imported by veterinary clinics. The weights of individual antimicrobials and antimicrobial classes imported by different type of importers for each year during the study period can be found in Supplementary Table S3.

**Figure 3.** Total active ingredients imported between 2016 and 2019, by type of importer.

**Figure 4.** (**A**,**B**): Trend of antimicrobials imported by different importers between 2016 and 2019. Bigger importers are represented in (**B**) and smaller importers in (**B**), thus *y*-axes differ between diagrams.

The biggest importers of CIAs were layer farms (81.1 kg), MAF (25.6 kg) and broiler farms (9.0 kg). Layer and broiler farms imported CIAs almost exclusively; while CIAs accounted for less than a quarter of imports by MAF, agriculture shops and veterinary clinics (see Figure 5). Almost all antimicrobial imports by layer and broiler farms were for oral administration; while almost all imports by MAF and agriculture shops were for parenteral administration (see Figure 6).

**Figure 5.** Profile of antimicrobial imports of different importer types by WHO classification of importance to human medicine. Important antimicrobials for human medicine represented less than 0.02 kg (0.01%) of total imports and were therefore not included in the diagram.

**Figure 6.** Profile of antimicrobial imports of different importer types, by recommended route of administration. Antimicrobials for administration via the topical route represented less than 0.3 kg (0.1%) of total imports and were therefore not included in the diagram.

#### **4. Discussion**

#### *4.1. Strengths of the Study*

This is the first study to describe veterinary antimicrobial imports into Timor-Leste. It showed a very low level of antimicrobial use in animals. Future studies of a similar nature will enable analysis of long-term trends and identification of changes in import patterns arising from interventions. Import data is a reasonable proxy for actual antimicrobial use for Timor-Leste since there is no local manufacture of veterinary antimicrobials and no re-export of antimicrobials. The data collection method was implemented consistently as it was performed by trained personnel using a written protocol. The accuracy of data was checked rigorously by authors to minimise data entry errors, and calculations were done with methods aligned with international guidelines. The training during data collection strengthened the capacity of MAF personnel to record antimicrobial import data and facilitated the timely reporting of results to the OIE, which is often a challenge in developing countries.

#### *4.2. Quantity of Antimicrobial Import*

The quantity of antimicrobials imported for use in animals in Timor-Leste after adjusting for biomass (0.55 mg/kg biomass) is very low compared to the global average

of 144.39 mg/kg and regional average (Asia, Far East, and Oceania) of 237.72 mg/kg in 2016 [29]. The use of veterinary antimicrobials in Timor-Leste was even lower than countries such as New Zealand, Norway, and Iceland which are known to have some of the lowest use levels in the world [45,46]. The low level of use is likely due to the subsistence agriculture system in Timor-Leste [33,47] where there is poor access to veterinary services and medicines. The low level of use is also consistent with another study in Timor-Leste which showed that only 1% of backyard chicken farmers used commercial medicines in their animals [48]. It would be interesting to compare the results from Timor-Leste to other countries with a similar agriculture background but similar studies from such countries could not be found [10]. Although antimicrobial use levels are currently low, use may increase in the future with farming intensification, as seen in other developing countries [49,50]. In this study, there is already evidence of increasing use in the broiler industry, with import levels rising by 119% between 2018 and 2019.

#### *4.3. Trend of Antimicrobial Import*

Trends in antimicrobial imports over the study period can be explained by looking individually at each importer. For MAF, the rise in antimicrobial imports during the study period represented increased procurement following annual feedback that government employed animal health professionals (e.g., veterinary and livestock technicians) faced shortages for field use [51].

For layer farms, it is likely that the import quantities were inconsistent between years because this group included only two large commercial layer farms that import antimicrobials in bulk quantities for use over a few years. For broiler farms, antimicrobial imports occurred only after 2018 following the import of day-old chicks from Indonesia after the lifting of avian-influenza related import restrictions [52]. The easing of restrictions was followed by a government effort to promote the growth of the broiler industry. The use of antimicrobials may also reflect the lack of resources to implement farm biosecurity and vaccination programmes on these farms [53,54]. Use of antimicrobials on broiler farms could be expected to rise in the future, mimicking the trends seen in neighbouring Indonesia where there was a rise of antimicrobial use due to industry growth, lack of alternative disease control options and a relatively low cost of antibiotics [55]. Therefore, farmer education programmes to improve knowledge on good animal husbandry practices and biosecurity could be useful [56]. The availability of quality vaccines would provide further options for disease prevention and control [57].

For agriculture shops, the reason for a decrease in imports during the study period was unclear but could be partially attributed to non-adherence to the MAF import application process resulting in data not being captured. For veterinary clinics, the low quantities imported reflect the small size of the industry—there were only four veterinary practices operating in Timor-Leste during the study period. The closure of one veterinary clinic in 2018 coincided with a drop in antimicrobial imports by veterinary clinics that year. For education institutions, there was only a once off import of antimicrobials by an agriculture school in 2016. There were no direct imports of antimicrobials by other types of commercial livestock farms apart from poultry, but animals on these farms could still receive antimicrobials imported by MAF or agriculture shops.

#### *4.4. Antimicrobial Class and Importance for Human Medicine*

The common antimicrobial classes in Timor-Leste (tetracycline, penicillin, and macrolide) are consistent with global and regional (Asia, Far East, and Oceania) usage patterns [29]. The most imported antimicrobials in Timor-Leste (oxytetracycline, amoxicillin, tylosin, and dihydrostreptomycin) were consistent with antimicrobials used in poultry and pig production in developing countries in Asia and Africa [18,50,58,59]. Oxytetracycline is popular because of its broad-spectrum action, low cost, and availability in long-acting formulations [60,61] and it is likely that similar reasons underpin its popularity in Timor-Leste. Amoxicillin and tylosin were imported almost exclusively in oral formulation by

commercial poultry farms, and the popularity of these antimicrobials in small scale poultry farms were also reported in other studies in other countries [50,62]. Dihydrostreptomycin was commonly imported in formulations with benzylpenicillin by MAF due to the combination's broad-spectrum action across a wide range of livestock species. It was positive that colistin, which is an antibiotic of last resort for human medicine that is commonly used in developing countries [63,64] contributed to less than 5% of imports to Timor-Leste with the majority imported in 2016. However, the broiler industry has been importing colistin albeit in small quantities in recent years and this should be closely monitored.

There has been a strong push towards reducing the use of medically important antimicrobials in livestock globally [65]. The almost exclusive imports of CIAs by commercial poultry farms could be attributed to the lack of awareness on antimicrobial resistance and its impact on public health, which has been observed in studies elsewhere [66,67]. On the other hand, the low proportion of CIA imports by MAF (21.2%) and veterinary clinics (21.7%) puts the professional veterinary service in positive light in terms of preserving critically important antimicrobials for use in human health. Of important concern is the import of fluroquinolones, polymyxins, and 3rd and 4th generation cephalosporins which are highest priority critically important antimicrobials for human medicine. Although the combined quantity of these classes contributed to less than 8% of total imports, future import and distribution of these antimicrobials should be closely monitored because of the potential risk of the development and transmission of antimicrobial resistance from livestock to humans [68,69]. To address the high proportion of CIA usage in the commercial poultry sector, a jointly developed antimicrobial treatment guideline between government and industry preferencing the use of non-CIA antibiotics may be effective [70].

#### *4.5. Route of Administration*

In this study, antimicrobials recommended for oral administration (39.8%) were less common than reported in some countries [12,14,71]. Antimicrobials imported by MAF and agriculture shops were mostly for parenteral administration. This is likely to be because they were mainly for use in species such as pigs and cattle that are reared extensively on small-holder livestock farms. On the other hand, commercial poultry farms probably imported mainly antimicrobials for oral administration because they are convenient for mass administration in poultry reared in semi-intensive or intensive production environments. The use of orally administered antimicrobials should be monitored in Timor-Leste as it has been demonstrated elsewhere that this route is more prone to misuse from inappropriate dosing and promotes the development of antimicrobial resistance [72].

#### *4.6. Use of Antimicrobials for Growth Promotion*

The import of antimicrobials intended for oral administration raises the concern of use of antimicrobials for growth promotion. Ideally antimicrobials should not be used for growth promotion without a public health risk assessment and any use should be phased out especially for critically important antimicrobials [21,65]. According to MAF, antimicrobials are not known to be used for growth promotion in the country. However, oral bacitracin and tylosin that were imported by poultry farms have been used for growth promotion worldwide [28,73]. In addition, the technical fact sheet of some antimicrobials indicated that the products could be administered for growth promotion. Therefore, it is possible that commercial farmers are administering antimicrobials at low doses, as recommended for growth promotion, without being aware. This has also been reported in another study [49]. The possible use of antimicrobials for growth promotion in Timor-Leste should be further investigated.

#### *4.7. Use of Antimicrobials in Aquaculture*

Although antimicrobial use is common practice in aquaculture systems worldwide and regionally [74–76], there were no aquaculture importers in this study and no products were indicated for use in aquatic animals. The absence of antimicrobial use in this sector is likely due to the relatively small and underdeveloped aquaculture sector [77].

#### *4.8. Limitations*

Timor-Leste is not immune to the challenges and limitations of monitoring antimicrobial use. The study only included import data after 2016 because of the five-year holding limit of hardcopy applications in the MAF office and the lack of digital record keeping. Thus, only data from 2016 to 2020 were available. The study excluded data from 2020 because the calendar year of 2020 had not yet ended at the point of data collection. This short study period limited the power of the study to detect trends in import quantities. However, digital record keeping was initiated as part of the study which will enable future studies to cover a longer time period.

It is likely that the study provided an under-estimation of the total amount of antimicrobials used in Timor-Leste due to the non-submission of import applications by some importers as reported in another study [71]. The possible reasons for non-submission include an importer's desire to avoid waiting times for approval, a weak regulatory framework and the lack of enforcement. Antimicrobials intended for human use could have also been administered to animals although elsewhere this is usually limited to companion animals [78]. The authors predict that the underestimation would not result in more than a doubling in the total amount of imported antimicrobials during the study period. Even if this happened, Timor-Leste would still demonstrate one of the lowest use rates compared to other countries that have reported usage data.

There may be a small degree of inaccuracy for the animal biomass estimation because the data obtained from FAOSTAT was based on extrapolations. This source of data was used because annual census data was unavailable during the study period. The OIE method for biomass calculation involved the use of European conversion coefficients and breeding cycles that may be different to Timor-Leste. However, these default parameters were used as no suitable alternative for a Timor-Leste context was found. The exclusion of minor species such as ducks, rabbits, dogs, and cats from the biomass calculation is likely to have only a marginal impact on the result since the population is relatively small [25].

It was not possible to quantify the antimicrobials that were administered to different animal species based on the import data due to the multi-species indication for many of the antimicrobials. However, a rough estimation of the division of antimicrobial use between livestock and companion animals could be estimated by assuming that antimicrobials imported by veterinary clinics were administered exclusively to companion animals, and antimicrobials imported by all other importers were administered exclusively to livestock.

#### *4.9. Future Directions*

Although the use of antimicrobials in animals is Timor-Leste is very low, there is potential for future misuse and overuse with farming intensification. Future studies investigating the knowledge, attitudes and practices of animal health professionals and farmers on antimicrobial use would be useful for identify strategies for promoting prudent use of antimicrobials in animals as identified in other studies [79–81]. Even in the absence of such studies, early action can be informed by studies conducted in other developing countries [55,58,82]. In addition to farmer education, which was mentioned previously, improving farmers access to animal health professionals [64], and training of animal health professionals to engage with farmers on prudent antimicrobial use has been shown to be effective elsewhere [82]. The strengthening of laboratory capacity in bacterial culture and antimicrobial susceptibility testing will also facilitate better decision making on antimicrobial use [65].

To improve the quality of data collected, MAF is engaging with importers such as agriculture shops to understand their reservations on submitting import applications and exploring legislative tools to improve compliance on import application submission. Future monitoring could focus on collecting data more proximal to the site of usage such as at end-user level to elucidate species and production type usage patterns [83,84].

#### **5. Conclusions**

This baseline study demonstrated very low levels of antimicrobial use in animals in Timor-Leste consistent with its subsistence agriculture system. Antimicrobial classes imported in the largest quantities were tetracyclines, penicillins, and macrolides. This is very similar to usage patterns in other countries globally and regionally. Import of CIAs for administration via the oral route was high in the poultry industry, and antimicrobial use for growth promotion could not be ruled out. Antimicrobial use in the poultry industry is expected to rise due to industry growth and the limited alternative disease control strategies. Education of farmers, development of antimicrobial treatment guidelines and improving access to veterinary services can help to ensure good antimicrobial stewardship in the animal health sector. Through this study, in-country capacity to monitor antimicrobial imports according to OIE reporting requirements was developed.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/antibiotics10040426/s1, Table S1: Animal biomass for buffalo, cattle, chicken, goats, horses, pigs and sheep in Timor-Leste between 2016 and 2019. Table S2: Import weight of various veterinary antimicrobial classes by route of administration between 2016 and 2019. Table S3: Total weight of active ingredients of different classes of veterinary antimicrobials imported into Timor-Leste, by type of importer between 2016 and 2019.

**Author Contributions:** Conceptualization, S.T., J.Y., J.R.F. and J.B.d.C.J.; Methodology, S.T., A.P., A.d.J.A. and S.D.; Software, S.T.; Investigation and data curation, S.T., A.P., A.d.J.A., S.D., C.d.C.S., S.F., F.J.S., O.d.C.H. and J.B.d.C.J.; Formal analysis, S.T., A.P., A.d.J.A., S.D. and T.S.B.; Writing original draft preparation, S.T., A.P., A.d.J.A. and S.D.; Writing—review and editing, S.T., A.P., A.d.J.A., S.D., C.d.C.S., S.F., F.J.S., O.d.C.H., J.Y., J.R.F., T.S.B. and J.B.d.C.J.; Visualization, S.T., A.P., A.d.J.A., S.D. and T.S.B.; Supervision, S.T.; J.Y. and J.R.F.; Project administration, S.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by the Fleming Fund Country Grant for Timor-Leste (FF/17/233). The Fleming Fund is a UK aid investment programme to tackle antimicrobial resistance in low- and middle-income countries around the world and is managed by the UK Department of Health and Social Care.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved Human Research Ethics Committee of the Northern Territory (NT) Department of Health and Menzies School of Health Research (2020-3841) and Institute Nacional de Saude in Timor-Leste (MS-INS/DE/IX/2020/1411).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available in the Supplementary Materials.

**Acknowledgments:** We acknowledge the staff from the Ministry of Agriculture and Fisheries of Timor-Leste for their support in this study. We thank Winnie Chen (Menzies School of Health Research) and Agnes Agunos (Public Health Agency of Canada) for critically reviewing the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Antibiotics* Editorial Office E-mail: antibiotics@mdpi.com www.mdpi.com/journal/antibiotics

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com ISBN 978-3-0365-4081-8