# **Antimicrobial Prescribing and Stewardship 1st Volume**

Edited by Diane Ashiru-Oredope Printed Edition of the Special Issue Published in *Antibiotics*

www.mdpi.com/journal/antibiotics

#### **Antimicrobial Prescribing and Stewardship, 1st Volume**

#### **Antimicrobial Prescribing and Stewardship, 1st Volume**

Editor

**Diane Ashiru-Oredope**

MDPI Basel Beijing Wuhan Barcelona Belgrade Manchester Tokyo Cluj Tianjin

*Editor* Diane Ashiru-Oredope HCAI and AMR Division UK Health Security London United Kingdom

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Antibiotics* (ISSN 2079-6382) (available at: www.mdpi.com/journal/antibiotics/special issues/prescribing antibiotic).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.

**ISBN 978-3-0365-7255-0 (Hbk) ISBN 978-3-0365-7254-3 (PDF)**

© 2023 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

#### **Contents**


Reprinted from: *Antibiotics* **2021**, *10*, 136, doi:10.3390/antibiotics10020136 . . . . . . . . . . . . . . **105**

#### **Leah F. Jones, Heidi Williamson, Petronella Downing, Donna M. Lecky, Diana Harcourt and Cliodna McNulty**


#### **Decraene and Rajesh Rajendran et al.**

Implementation of a Delayed Prescribing Model to Reduce Antibiotic Prescribing for Suspected Upper Respiratory Tract Infections in a Hospital Outpatient Department, Ghana Reprinted from: *Antibiotics* **2020**, *9*, 773, doi:10.3390/antibiotics9110773 . . . . . . . . . . . . . . . **235**


#### **About the Editor**

#### **Diane Ashiru-Oredope**

Professor Diane Ashiru-Oredope PhD FFRPS, FRPharmS is the Lead Pharmacist for healthcare-associated infections (HCAI) and antimicrobial resistance (AMR) at the UK Health Security Agency, and she is the chair of the English Surveillance Programme for Antimicrobial Utilisation and Resistance (ESPAUR). She is the Honorary Chair and Professor of Pharmaceutical Public Health at the University of Nottingham.

An antimicrobial pharmacist by background, Diane has led several projects that have shaped national and international policy in tackling antimicrobial resistance, including creating the global Antibiotic Guardian campaign in 2014. From 2016 to March 2022, she was an advisor and Global AMR leader for the Commonwealth Pharmacists Association. Diane remains active in research areas, with more than 70 peer-reviewed publications, and she is one of the editors for the BMC Public Health journal. She recently led a UK-wide evidence review on Pharmaceutical Public Health, commissioned by the four UK Chief Pharmaceutical Officers. In 2022, she was credentialed as a consultant pharmacist through the Royal Pharmaceutical Society's national credentialing and assessment processes. She has been recognised nationally, including through awards, nominations to deliver two TEDx talks, and fellowships with the Royal Pharmaceutical Society (FRPharmS and FFRPS).

#### **Preface to "Antimicrobial Prescribing and Stewardship, 1st Volume"**

Antimicrobial stewardship as coherent set of actions which promote using antimicrobials in ways that ensure sustainable access to effective therapy for all who need them"(Dyar, O.J.; 2017) is critical (alongside, e.g., infection prevention and control strategies) for tackling antimicrobial resistance/drug-resistant infections.

This Antimicrobial Prescribing and Stewardship Issue (Volume 1) consist of manuscripts, including original research, review articles, case series, and opinion papers for topics related to antimicrobial (antibiotic, antifungal) stewardship, including:

Disease-based/organism-based antimicrobial stewardship;

Diagnostic stewardship;

Influence of antimicrobial utilisation changes in antimicrobial resistance;

Impact of antimicrobial stewardship on quality performance measures and patient outcomes;

Novel antimicrobial stewardship education and training approaches or interventions aimed at the public and/or healthcare workers;

Behavioural change approaches to antimicrobial stewardship;

Collaborative practice agreements in antimicrobial stewardship;

Antimicrobial stewardship in special populations (e.g., paediatrics, geriatrics, emergency medicine, hematology/oncology);

Tackling AMR through antimicrobial stewardship in low- and middle-income countries;

Antimicrobial stewardship for animal health;

Antimicrobial stewardship in alternative settings (e.g., community practice, long-term care, and resource-limited small and rural hospitals);

Antimicrobial use and stewardship in the context of the COVID-19 pandemic;

Global collaborations to tackle AMR through antimicrobial stewardship.

This edition also aimed to include articles that recognize the theme for World Antimicrobial Awareness Week 2020, "United to preserve antimicrobials", with papers which consider the impact of global collaborations and health partnerships being invited to contribute.

World Antimicrobial Awareness Week (WAAW), led globally by the WHO, aims to increase awareness of global antimicrobial resistance (AMR) and to encourage best practices among the public, health workers, and policy makers to avoid the further emergence and spread of drug-resistant infections.

We also built on the success of this edition and there is a second volume on "Antimicrobial Prescribing and Stewardship, 2nd Volume".

> **Diane Ashiru-Oredope** *Editor*

*Article*

#### **Feasibility Study of the World Health Organization Health Care Facility-Based Antimicrobial Stewardship Toolkit for Low- and Middle-Income Countries**

#### **Gina Maki 1,\*, Ingrid Smith <sup>2</sup> , Sarah Paulin <sup>2</sup> , Linda Kaljee <sup>3</sup> , Watipaso Kasambara <sup>4</sup> , Jessie Mlotha <sup>4</sup> , Pem Chuki <sup>5</sup> , Priscilla Rupali <sup>6</sup> , Dipendra R. Singh <sup>7</sup> , Deepak C. Bajracharya <sup>8</sup> , Lisa Barrow <sup>9</sup> , Eliaser Johnson <sup>9</sup> , Tyler Prentiss <sup>3</sup> and Marcus Zervos 1,10**


Received: 20 July 2020; Accepted: 28 August 2020; Published: 29 August 2020

**Abstract:** Antimicrobial stewardship (AMS) has emerged as a systematic approach to optimize antimicrobial use and reduce antimicrobial resistance. To support the implementation of AMS programs, the World Health Organization developed a draft toolkit for health care facility AMS programs in low- and middle-income countries. A feasibility study was conducted in Bhutan, the Federated States of Micronesia, Malawi, and Nepal to obtain local input on toolkit content and implementation of AMS programs. This descriptive qualitative study included semi-structured interviews with national- and facility-level stakeholders. Respondents identified AMS as a priority and perceived the draft toolkit as a much-needed document to further AMS program implementation. Facilitators for implementing AMS included strong national and facility leadership and clinical staff engagement. Barriers included lack of human and financial resources, inadequate regulations for prescription antibiotic sales, and insufficient AMS training. Action items for AMS implementation included improved laboratory surveillance, establishment of a stepwise approach for implementation, and mechanisms for reporting and feedback. Recommendations to improve the AMS toolkit's content included additional guidance on defining the responsibilities of the committees and how to prioritize AMS programming based on local context. The AMS toolkit was perceived to be an important asset as countries and health care facilities move forward to implement AMS programs.

**Keywords:** antimicrobial resistance; antimicrobial stewardship; low- and middle-income countries; barriers and enablers

#### **1. Introduction**

The misuse of antimicrobials is one of the main drivers for the development of antimicrobial resistance (AMR) [1,2]. Antimicrobial stewardship (AMS) programs have been shown to be effective in reducing unneeded antimicrobial use and slowing AMR in high-income countries; however, there are limited data on the feasibility of AMS programs in low- and middle-income countries (LMIC) [3–6]. As a response, the World Health Organization (WHO) has developed a practical toolkit for health care facility-based AMS programs in LMIC (hereafter referred to as the "AMS toolkit") [7].

In 2015, at the 68th World Health Assembly, AMR was recognized as a threat to public health. A global action plan, including an objective to optimize the use of antimicrobials, was endorsed during the Assembly. AMS programs aim not only to optimize antimicrobial use, but also to improve patient outcomes, decrease rates of AMR, and reduce health care costs [8–11]. With few new antimicrobials being produced and the decreased effectiveness of existing antimicrobials, AMS programs are an essential component of a One Health approach to address AMR [12]. Many LMIC have inadequate AMS policies and treatment guidelines at both the national and health care facility levels, resulting in a disproportionate impact of AMR in these countries [1,2,11,13–15]. Intraregional, interdisciplinary collaborations and partnerships are needed at the national and facility level to adapt, implement, and disseminate AMS programs that are locally salient [16–18].

The AMS toolkit is divided into six sections: (1) Structural Core Elements for AMS Implementation at the National Level; (2) Structural Core Elements for AMS Implementation at the Facility Level; (3) Planning AMS programs; (4) Performing AMS interventions; (5) Assessing AMS programs; (6) Education and Training. As part of the AMS toolkit's development, a feasibility study was undertaken in Bhutan, the Federated States of Micronesia (FSM), Malawi, and Nepal. The study objectives were to (1) assess local knowledge and perceptions regarding AMR, AMS, and key concepts of the AMS toolkit; (2) identify barriers and facilitators for the implementation of AMS programs and policies; (3) identify recommendations to revise the AMS toolkit draft to ensure it meets the needs of a broad range of LMIC settings; (4) to provide recommendations to the countries on initiating and strengthening AMS programs. The study countries were selected based on geographic regions where AMR is a significant issue and where, to date, there are limited resources and AMS programs. These countries also represent different regions and varying healthcare systems (Table 1).


**Table 1.** Information on Health Care Systems, AMR Stewardship, and Pharmaceutical Sales by Country.

**Table 1.** *Cont*.


<sup>1</sup> World Health Organization Western Pacific Region. 2017. Federated States of Micronesia: WHO Country Cooperation Strategy, 2018-2022. <sup>2</sup> Ministry of Health and Population, Republic of Malawi. The Health Care System. Available at: https://www.health.gov.mw/index.php/2016-01-06-19-58-23/national-aids.

#### **2. Results**

The following results are organized by national and health care facility core elements and health care facility-based interventions as described within the AMS toolkit. In addition, we have included data on actionable items and recommendations for the draft toolkit. A summary of demographic information is found in Table 2.


#### **Table 2.** Clinical Staff Demographics.

AMR—antimicrobial resistance; IPC—infection prevention and control.

Key findings from this study are shown in Table 3.



AMR—antimicrobial resistance; AMS—antimicrobial stewardship; IPC—infection prevention and control.

#### *2.1. National Core Elements*

The national core elements include (1) National Plan and Strategies; (2) Regulations and Guidelines; (3) Education and Training; (4) Supporting Technologies and Data.

#### 2.1.1. National Plan and Strategies

All study countries identified AMR as a growing threat and have drafted an AMR National Action Plan (NAP); however, the countries were at various levels in terms of NAP implementation. Relatively few respondents outside of the national government were aware of the NAP content.

**Remark 1.** *"I have heard of the National Plan, but it is not being implemented in our medical college* . . . *." (Hospital Administrator, Nepal)*

Study participants aware of the NAP reported dedicated high-level national leadership in support of the NAP. In addition, some countries have obtained outside funding to support portions of the NAP. However, across the four countries, numerous barriers were identified that affected the implementation of NAP policies and programs. These barriers included the need for additional financial support and technical assistance, limited laboratory capacity, including infrastructure and expertise, and lack of technical expertise on AMR-specific issues.

**Remark 2.** *"* . . . *again, it involves a lot of resources. You need to see the sensitivity and (do) surveillance* . . . *to review sensitivities* . . . *A lot of budget is involved in that. So, funding is another part (in terms of) feasibility or not. And of course, the human side and human resources are also important...." (National Level, Nepal)*

#### 2.1.2. Regulations and Guidelines

Facilitators for AMR regulations and guidelines included the existence of drafted standard antibiotic prescribing guidelines, which have been developed in Bhutan, Malawi, and Nepal. In FSM, there is an antibiotic prescribing guideline that has been reviewed with national stakeholders and external technical expertise. FSM, Malawi, and Nepal are in the process of developing and implementing policies for prescription-only antibiotic sales. In Bhutan, prescription-only regulations are implemented and enforced. The key barriers identified included inadequate monitoring and evaluation at the national and regional levels for infection prevention and control (IPC) and AMS including adherence to guidelines, inadequate laboratory facilities to provide empiric diagnostic data and support, implementation of the AWaRe (Access, Watch, Reserve) classification of antibiotics, challenges with the supply chain of medications in hospital pharmacies, and lack of prescribers (physicians) in remote areas.

**Remark 3.** *"* . . . *.I'm not claiming that just writing in law will be su*ffi*cient to really restrict prescriptions for selling* . . . *we have so many pharmacies which have been already selling without prescriptions* . . . *we need to really go and then take action against those who are selling the antibiotics* . . . *at the same time in the public sector we are promoting some antibiotics to be used by [non-physicians]* . . . *because neonatal mortality [is] very high* . . . *" (National Level, Nepal)*

**Remark 4.** *"* . . . *If we don't have the drugs [in the hospital], the doctor will prescribe and then the patient is obligated to buy in a pharmacy* . . . *." (National Level, Malawi)*

#### 2.1.3. Awareness, Training, and Education

At the national level, there was strong support to enhance education and training in AMR/AMS for physicians, nurses, pharmacists, and laboratory staff. At the broader community level, AMS awareness, training, and education included public antibiotic information campaigns. Barriers included lack of dedicated financial support and limited training and technical expertise. Respondents also reported the need for more community advocacy and awareness of AMR.

**Remark 5.** *"* . . . *and we are also thinking to develop some sort of dramatizing* . . . *learning materials and publishing to the media (for the community). So that type of materials we're planning to do. That requires a little bit budget so we are constrained with budget so we're planning anyway* . . . *and also planning to do some training for health professionals* . . . *primary, community levels–upper-level health professionals also need to have training* . . . *." (National Level, Nepal)*

#### 2.1.4. Supporting Technologies and Data

Bhutan, Malawi, and Nepal are recipients of Fleming Fund grants to expand microbiology laboratory capacity and strengthen surveillance systems [16]. Respondents perceived this support as a starting point to significantly enhance NAP implementation. However, respondents also recognized the need for more sustained funding and training in diagnostic testing, laboratory surveillance, and information technologies, and the struggle with inadequate local expertise in monitoring antimicrobial use and consumption.

**Remark 6.** *"So, we are so glad that at least one of our key priorities for the NAP–AMR surveillance—we have a country grant to support improving lab capacity and surveillance. So, that's a very good plus for us* . . . *." (National Level, Malawi)*

#### *2.2. Health Care Facility Core Elements*

Health care facility core elements include (1) Leadership Commitment; (2) Accountability and Responsibilities; (3) Education and Training; (4) Monitoring and Surveillance; (5) Reporting and Feedback.

#### 2.2.1. Leadership Commitment

A key component of facility-based AMS is dedicated support from facility leadership to encourage the development of AMS and IPC committees and implement programs. Facility leaders stated that the recommended AMS activities in the toolkit are feasible and necessary to decrease AMR. Barriers to leadership commitment included inadequate dedicated human and financial resources for AMS programs, and inadequate internal communication between administrators, management, and staff.

**Remark 7.** *"A successful stewardship program can only work when there is good collaboration between top management and middle-level management. Because middle-level management, they are the ones that are in touch with the sta*ff *on the ground there." (Hospital Administrator, Malawi)*

**Remark 8.** *"Stressing the importance of leadership commitment is really important. Because nothing really happens without leadership commitment." (National Level, FSM)*

#### 2.2.2. Accountability and Responsibilities

Respondents described active IPC committees, quality improvement teams, and drug and therapeutic committees. These committees are engaged in routine activities and provide important feedback and training to clinical personnel. Building on these committees, leaders and clinical staff were enthusiastic about future implementation of AMS committees and antibiotic prescribing guidelines.

**Remark 9.** *"* . . . *we don't have an AMS team but the infection control committee they have nurse sta*ff*. They do regular monitoring of cultures of di*ff*erent areas of the theatre* . . . *they take a routine culture on a regular basis* . . . *." (Hospital Administrator, Nepal)*

Barriers included physicians' reluctance to change their prescribing practices, even with the available evidence-based guidelines, and provider-heavy workloads, which decreased interest in devoting time to an AMS committee.

**Remark 10.** *"We have an issue of shortages of sta*ff . . . *we also have issues of people wearing too many hats, so that is part of, I guess, it's kind of like we're not sure who's going to take care of it, and wonder who which program is." (Clinical Sta*ff*, FSM)*

#### 2.2.3. Education and Training

There was a high level of enthusiasm for expanding AMS education and training within health care facilities. Currently, some study facilities have an established education infrastructure (e.g., continuing medical education) that can support additional AMR/AMS training. However, despite enthusiasm, time and resources were barriers. Many facilities lack space and available expertise to support trainings. At some sites, even IPC trainings are limited due to inadequate resources. In addition, some respondents emphasized the need for a 'hands-on' approach to support sustained knowledge.

**Remark 11.** *"Major barriers as of right now is lack of knowledge mainly. We have high sta*ff *turnover, so in the last three years there have been no new trainings. Almost 75% of the sta*ff *may not be aware of the IP (infection prevention) practices* . . . *." (Clinical Sta*ff*, Malawi)*

**Remark 12.** *"So hands-on is a very, very important part of it. As compared to just reading and looking at the modules. But you have to apply, the application of that knowledge needs to be implemented as well. With the hands-on skills I think it will stick and will stay there longer, and be more useful to the people. In my opinion, I think two or more models of education is probably the best suitable for us in this setting."*

#### 2.2.4. Monitoring and Surveillance

Among microbiologists and laboratory staff, there was a strong desire for capacity building and international support for up-to-date equipment and supplies and development of AMR surveillance systems. However, in most health care facility sites, there was inadequate capacity to conduct point prevalence surveys and routine surveillance of susceptibility patterns, including health care facility-specific antibiogram data.

**Remark 13.** *"For the lab, our major challenges are both human and material resources* . . . *we have a very big challenge procuring laboratory microbiology supplies. Either maybe because of the budget or our major supplier—the central medical stores—they don't have them in stock* . . . . . . *we have the knowledge, but resources are not there* . . . *." (Clinical Sta*ff*, Malawi)*

#### 2.2.5. Reporting and Feedback

In health care facilities with developed IPC committees, there are existing reporting and feedback structures in place that can be expanded to include AMS. However, in most facilities, there are inadequate structures for reporting these data to facility management and clinical staff. Study participants noted there was little communication between laboratory and clinical personnel, decreasing opportunities for information exchange about AMR patterns and impact of the use of specific antibiotics.

**Remark 14.** *"* . . . *We are very much lacking in reporting and feedback. We can do something but this is one area we really have to really have to think and discuss with regards to core elements* . . . *there are so many challenges, which we have to sit together and discuss and see how to move things forward." (Clinical Sta*ff*, Bhutan)*

#### *2.3. Action Items at the National and Health Care Facility Level*

Multiple action items were identified at both the national and facility level to move forward with implementation of the NAP and the WHO toolkit and development and implementation of AMS committees and other supporting programs and policies. Identification of sustainable funding and technical expertise in human, animal, and environmental health was essential across each of the core elements. Action items at the national and facility level are found in Boxes 1 and 2.

#### **Box 1.** National-level action items to support AMS and toolkit implementation.


AMR—antimicrobial resistance; AMS—antimicrobial stewardship.

**Box 2.** Health care facility-level action items to support AMS and toolkit implementation.


AMR—antimicrobial resistance; AMS—antimicrobial stewardship; CME—continuing medical education; DTC—drug and therapeutic committees; IPC—infection prevention and control; QIT—quality improvement teams.

#### *2.4. Health Care Facility-Based AMS Interventions*

The WHO toolkit provides a detailed overview of evidence-based health care facility AMS interventions including (1) persuasive, educational, and feedback; (2) restrictive; (3) structural interventions. The toolkit also includes information on planning, implementing, and assessing AMS programs. Identified enablers to support AMS interventions included strong leadership support at the health care facility administration level, overall strong interest in education and training in IPC, AMR, and AMS among clinical staff, and perceptions among staff that AMS interventions decrease unnecessary antibiotic use. Barriers included inadequate local infectious disease, AMR and AMS expertise, and limited financial and human resources to implement interventions and conduct program monitoring and evaluation.

**Remark 15.** *"When it comes to interventions I think they are very much appropriate because many of these problems do exist in our* . . . *day-to-day practices. But we are not really assessing them* . . . *to the fullest extent about the interventions." (Clinical Sta*ff*, Bhutan)*

#### *2.5. Summary of Recommendations for the Draft WHO Toolkit*

Study participants suggested recommendations on the improvement of content, organization, and presentation of materials, which were incorporated into the final version of the WHO AMS toolkit (Table 4). Participants also noted that the review of the AMS toolkit needs to be an iterative process as implementation of the toolkit and AMS programs and policies progress in each country.




**Table 4.** *Cont*.

**Remark 16.** *"I think it's a good start. It has more stu*ff*, areas that need to be stringent, if we kind of have an impact on this issue. And I think it's pretty comprehensive in a sense* . . . *But I think it should be an organic process, as we move along and identify issues, we address them and continue to make improvements." (Clinical Sta*ff*, FSM)*

#### **3. Discussion**

The toolkit was universally well-received by policy makers, facility management, and clinical staff levels throughout the four study countries. Data were obtained from a diverse multinational and interdisciplinary group of stakeholders. Identification of possible enablers and barriers for toolkit implementation at the national and facility level supported revisions to ensure that the toolkit meets the needs of a broad range of LMIC settings. Each study country presented different contextual factors to consider regarding AMS implementation and use of the AMS toolkit. Varying factors included different health priorities at both the national and facility level, current status of nationalized universal health care plans, variances in public health funding, availability and use of antibiotics, and development and enforcement of prescription-only regulations. These factors must be considered on a country-by-country basis for stakeholder engagement and evaluating pathways to toolkit implementation.

Key facilitators and enablers included strong leadership commitment at the national, local, and facility levels, increases in funding mechanisms to support development of surveillance systems within countries, and increased awareness of AMR. Key barriers to AMS implementation included limited human and financial resources, inadequate supporting technologies (e.g., monitoring and surveillance), and communication challenges between facility administration and staff, and between staff members. These barriers can be mitigated using a clear step-by-step approach, as indicated in the WHO AMS toolkit, tailored to specific country and facility contexts and needs. In addition, a multidisciplinary training and education approach can potentially strengthen AMS commitment and communication within health care facilities.

Prior studies have described the core elements of AMS programs in LMIC settings, including the need to build laboratory capacity, enhance IPC, and establish surveillance systems of both infections and antibiotic use [15,19–25]. Feasibility study data support these needs, as well as other essential elements of AMS. Overall, respondents stated that leadership support at the national and senior facility management levels was needed for successful implementation. In addition, stepwise implementation strategies were universally considered to be useful. At the national level, respondents supported the urgent need for the development and implementation of antibiotic treatment guidelines. AMS must be identified as a national priority and included in facility key performance indicators, requiring dedicated support, accountability, and assigned roles and responsibilities. Within health care facilities, AMS must include written strategies, implementation of a formal multidisciplinary structure including laboratory surveillance and IPC, and identification of dedicated staff with clearly

defined roles. These strategies must include support and expertise on infection management, access to timely laboratory/imaging/information technology services and available trained and experienced professionals in AMR and infectious disease.

Respondents reported the need for increased and consistent education and training. A first step towards strengthening educational initiatives includes understanding current clinical staff competencies and building tailored projects and programs that emphasize and develop knowledge and skills. Education must be ongoing, hands on, and practical, and include incentives and a broad range of resources (e.g., face-to-face and web-based).

Respondents felt the toolkit provided important information on appropriate antibiotic use and consumption and the means to utilize less expensive and technologically based approaches appropriate to LMIC contexts. AMS committees were considered at the heart of effective facility-based stewardship. Therefore, these committees must be formed and members must receive regular training in antimicrobial prescribing practices and stewardship. The committee should be responsible for reviewing and auditing courses of therapy for specified antimicrobial agents and clinical conditions. There is also the need for an established and effective communication strategy between the AMS committee, leadership, and clinical staff.

In terms of health care facility-based AMS, evidence-based antibiotic treatment guidelines were identified as a key component of AMS. Where possible, guidelines should be based on local antibiotic resistance patterns and availability and cost-effectiveness of agents. Day-to-day guidelines should be kept simple and include empiric antibiotic selection, definitive antibiotic selection, organism and disease states, intravenous to oral conversion, renal dosing, and duration of therapy. In conjunction with those guidelines, AMS programs are needed to reduce the overuse and overprescribing of antibiotics, the use of broad-spectrum antibiotics and dose combinations, and delayed prescribing.

Health care facilities in LMIC need capacity building both in terms of human and technological resources to develop a formulary and auditing process, antibiotic prescribing documentation policies and procedures, and regulations regarding drug restrictions including use of the WHO AWaRe (Access, Watch, Reserve) classifications. Health care facility monitoring and surveillance capabilities need to be developed to support AMS initiatives including measures to monitor quality/quantity of antimicrobial use at the unit and facility-wide level, compliance with specific interventions, and identification of antibiotic susceptibility rates for locally significant pathogens.

#### *Limitations*

The feasibility study was conducted in only four countries with a sample size of 12 national leaders, 21 facility administrators, and 65 clinical staff. Only one country was selected from Africa and there were no LMIC from Latin America, the Caribbean, the Middle East, or Europe included. Despite the small sample size, purposeful sampling was undertaken to ensure that different regions of the study countries were included with a diverse group of national- and facility-based stakeholders. These data provide a general overview of barriers and facilitators for implementation of AMS programs and the AMS toolkit. As implementation of the toolkit moves forward, additional data from other countries will continue to contribute to future versions. In addition, the feasibility study was focused on health facilities which provide inpatient care. Respondents discussed the need for AMS within community health facilities and education for patients. Future research and development of community-based AMS training and interventions are needed to address the high consumption of antibiotics outside of inpatient facilities.

#### **4. Materials and Methods**

#### *4.1. Overview*

The feasibility study was conducted from February 2019 to May 2019. The project was a partnership inclusive of a multinational and interdisciplinary team with expertise in AMR and AMS, infectious

diseases, IPC, public health, nursing sciences, pharmacy, and social sciences. The study countries were selected by both WHO staff and the HFHS feasibility study team, based on geographic regions where AMR is a significant issue and where, to date, there are limited resources and AMS programs. These countries represent diverse contexts and challenges associated with the implementation of health care facility-based AMS programs. In addition, WHO and/or HFHS had worked with AMS leaders in the four selected countries, which facilitated rapid implementation of the feasibility study.

The study population included national- and local-level policymakers, facility administrators, and clinical staff. Study health care facilities were identified by in-country investigators and coordinators, and represented various facilities (e.g., public, private, and non-profit) and diverse geographic regions within each country. All facilities included inpatient care and ranged in size from 36 to 850 beds. Participating clinical staff included physicians, nurses, pharmacists, microbiologists, and laboratory technicians.

The study used a qualitative design based on key domains for program feasibility studies [26,27]. These included (1) acceptability of the toolkit; (2) demand and anticipated use; (3) practicality of the toolkit for use in LMIC; (4) integration of the toolkit within existing infrastructures; (5) adaptability of the toolkit within local contexts; (6) implementation and dissemination enablers and barriers to toolkit sustainability and scale-up within LMIC. The qualitative approach provided opportunity to engage with multiple partners from the selected sites throughout the development, implementation, and dissemination of the study. Through this engagement, conversations about AMR during meetings, workshops, and interviews provided visibility to local, national, and international issues related to AMR, the role of stewardship in the contexts of LMIC, and the potential for adaptation of the WHO toolkit to support AMR stewardship at the policy and programmatic levels in multiple settings.

In each country, the project was undertaken after an initial meeting with in-country study investigators and local governmental, nongovernmental, and health care facility stakeholders. After completion of the study, dissemination stakeholder workshops were convened within each country. These workshops provided opportunity for local input on (1) the interpretation of the feasibility study data; (2) ways to address facilitators and barriers to implementation of the AMS toolkit and AMS programs and policies; (3) identification of actionable items to promote implementation. This input from each country is reflected in the final reports and subsequently, in this paper.

#### *4.2. Sample Size and Recruitment*

Overall, 12 policy makers were recruited and interviewed, and 15 health care facilities were selected between the four countries. With the facilities, a total of 21 administrators, 20 physicians, 21 nurses, 11 pharmacists, and 13 laboratory personnel were interviewed (Table 5). When the study started, the research team estimated the sample size with the stipulation that it could be smaller or larger depending on data saturation. In each site, the team felt confident that the data collected reached saturation and no significant additional information was being recorded to justify additional interviews.

National-level respondents were identified by in-country principal investigators and coordinators as well as recommendations from the WHO Country Offices and included individuals involved in the development of AMS NAP and other experts in AMR and AMS. At the facility level, administrators or managers were invited to participate in the study. Clinical staff selection criteria included individuals engaged in current or past IPC programs and/or those engaged in existing AMS committees. A range of staff were interviewed, including ward physicians and nurses, laboratory staff, and pharmacists. In each country, the international and local partners worked closely together to approach potential respondents to explain the purpose of the study and request their participation. At the policy and hospital administration level, all of those approached made themselves available for the interviews. At the clinical level, the study team requested interviews with representatives from nursing, medicine, pharmacy, and microbiology/laboratory staff (Table 5). Potential participants were provided with an abbreviated version of the toolkit with sections specific to their position (e.g., policy maker, administrator, and staff) and a list of key topics to be covered in the interview [28].


**Table 5.** Study sites, health care facility types, and sample sizes for policy makers, facility administrators, and clinical staff.

#### *4.3. Research Instruments*

Three interview guides and demographic forms were developed specific to the population groups (policymakers, health care facility administrators, and clinical staff). Draft interview guides and demographic forms were provided to in-country investigators to review and revise to ensure they reflected local contexts. Interview guide items and probes focused on the research objectives, the toolkit core elements for health care facility-based AMS in LMIC, and the 6 program feasibility key domains. Demographic forms included items on current institutional affiliation (e.g., city/district, type of institution, and number of beds), respondents' education and employment (e.g., current position and years in current position), and engagement in AMR, AMS, and IPC programs at the health care facility or national levels (see Table S1).

#### *4.4. Data Collection and Management*

Data collection was led by team members from the Henry Ford Health System in partnership with local staff. In Bhutan, additional data collection support was provided by investigators/infectious disease specialists from Christian Medical College, Vellore, India. Interviews were conducted in English, with interpretation to local language verbally as required. Interviews were audiotaped and transcribed. Transcribed data were entered into Ethnograph, version 6, a qualitative data management software. A data coding dictionary was developed based on the interview guides and emergent themes from team members' field experiences and the transcribed text.

#### *4.5. Data Analysis*

After initial coding was completed, groups of code words were organized under common topics including the AMS core elements, the key domains for the feasibility study, and emergent themes. Searches were conducted within common themes, study country, and population (policymakers, administrators, and staff). Search documents were saved and reviewed to identify key findings and

recommendations within and across countries in terms of barriers and facilitators for health care facility-based AMS policies and programs, action items to strengthen AMS policies and programs including implementation of the AMS toolkit, and specific recommendations for the AMS toolkit's content and organization. Illustrative text within the transcripts were identified to support the summary conclusions. Country-specific draft reports were sent to in-country investigators and other stakeholders for their review and input.

#### *4.6. Ethical Approval*

The project was approved by ethical review boards in each country, the Henry Ford Health System Institutional Review Board (Detroit, MI, USA), and the WHO Ethical Review Committee (Geneva) (Approval February 2019, #ppp3131/004479)]. All participants provided written informed consent.

#### **5. Conclusions**

There was clear consensus that optimal implementation and use of the toolkit requires recognition of country-specific contexts. These include diagnostic challenges, laboratory capacity, and high burdens of infectious diseases. Health care workers responsible for prescribing antibiotics have a broad range of education, training, and experience. Development of antibiotic prescribing guidelines may often be limited by inadequate local data on disease burden and susceptibility patterns. Many LMIC have poorly regulated prescription-only policies or limited access to essential antibiotics

Despite these challenges, the consensus among study respondents was that the toolkit will be an important asset as countries and health care facilities move forward to combat AMR and implement AMS programs. More information will be needed to address implementation strategies and many barriers need to be addressed to increase likelihood of successful implementation within the study countries and other LMIC. The road ahead must include commitment at the national and facility levels to prioritize AMR and develop sustainable national and local initiatives. The WHO toolkit provides a comprehensive review of core elements of AMS, strategies for intervention adaptation and implementation, and expansion of training and educational platforms. With growing global concerns regarding AMR, the WHO toolkit can provide practical guidance and support to LMIC worldwide [1].

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2079-6382/9/9/556/s1, Table S1. Interview guides.

**Author Contributions:** Conceptualization, G.M., I.S., S.P., L.K., T.P., M.Z.; methodology, I.S., S.P., L.K.; formal analysis, G.M., I.S., S.P., L.K., P.C., T.P.; investigation, G.M., L.K., W.K., J.M., P.C., P.R., D.R.S., D.C.B., L.B., E.J., T.P., M.Z.; writing—original draft preparation, G.M., L.K., T.P., M.Z.; writing—review and editing, G.M., L.K., T.P., M.Z.; project administration, I.S., S.P., M.Z.; funding acquisition, M.Z. All authors have read and agreed to the published version of the manuscript. The authors S.P. and I.S. are staff members of the World Health Organization. The author alone is responsible for the views expressed in this publication and they do not necessarily represent the views, decisions or policies of the World Health Organization.

**Funding:** This research was funded by the World Health Organization, grant number D30521.

**Acknowledgments:** We would like to thank the following national experts and WHO country office staff involved in the feasibility studies: Sonam Yangchen and Pema Yangzom (Bhutan); Kelias Msyamboza (Malawi); Eunyoung Ko (Federated States of Micronesia); and Rajan Rayamajhi and Reuben Samuel (Nepal). We would also like to thank the many individuals who participated in the study in the four study countries.

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

#### **References**


© 2020 World Health Organization; Licensee MDPI, Basel, Switzerland. This is an open access article distributed under the terms of the Creative Commons Attribution IGO License (CC BY) license (http://creativecommons.org/licenses/by/3.0/igo/legalcode), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In any reproduction of this article there should not be any suggestion that WHO or this article endorse any specific organisation or products. The use of the WHO logo is not permitted.

*Article*

#### **Implementation of the WHO Approved "Tailoring Antimicrobial Resistance Programs (TAP)" Reduces Patients' Request for Antibiotics**

**Nasser M. Kaplan <sup>1</sup> , Yousef S. Khader <sup>2</sup> , Mahmoud A. Alfaqih 3,\*, Rami Saadeh <sup>2</sup> and Lora Al Sawalha <sup>4</sup>**


Received: 7 June 2020; Accepted: 4 August 2020; Published: 12 August 2020

**Abstract:** The misuse of antibiotics is a worldwide public health concern. Behavioral Intervention programs that aim to reduce patients' own request for antibiotics during their visit to primary care clinics is an attractive strategy to combat this problem. We tested the effectiveness of a behavioral modification method known as the Tailoring Antimicrobial resistance Programs (TAP) in reducing the request for antibiotics by patients visiting primary care clinics for mild upper respiratory tract infections (URTIs). A stratified cluster randomized design with two groups pre-post, comparing intervention with the control, was conducted in six health centers. TAP was implemented for eight weeks. Request for antibiotics was assessed before (period 1) and after introducing TAP (period 2). The percentage of patients or their escorts who requested antibiotics in period 1 was 59.7% in the control group and 60.2% in the intervention group. The percentage of patients who requested antibiotics did not significantly change between period 1 and 2 in the control group, who continued to receive the standard of care. The above percentage significantly decreased in the intervention group from 60.2% to 38.5% (*p* < 0.05). We conclude that behavioral change programs including TAP are a viable alternative strategy to address antibiotic misuse in Jordan.

**Keywords:** antibiotics; microbial resistance; upper respiratory tract infections

#### **1. Introduction**

Antimicrobial resistance is a threat to the public health sector worldwide [1]. Many factors contribute to this problem; however, the misuse and/or overuse of antibiotics are established as the major driving forces [2,3]. Indeed, it was estimated that up to 50% of all antimicrobials globally prescribed to patients are not even necessary [4,5]. It is interesting to note that most of the unnecessary antibiotic prescription takes place in the primary care setting [6], with the biggest percentage of unnecessary antibiotics being prescribed for patients with upper respiratory tract infections (URTIs) [6].

Jordan is a developing country in the Middle East and North Africa (MENA) region. The misuse of antibiotics by consumers, including the use of antibiotics without prescriptions, was widely documented in Jordan and in the region [7–12]. Other forms of antibiotic misuse in the MENA region include the use of antibiotics for improper indications, including to fight viral infections [13–16].

Despite the magnitude of the antibiotic misuse problem, most of the countries in the MENA region have no laws and/or legislations that prohibit dispensing antibiotics without a proper prescription [8,17]. Moreover, countries that have relevant legislation in place do not have proper surveillance systems and/or do not adequately enforce relevant laws [18]. Interestingly, knowledge about antibiotic misuse and antimicrobial resistance by itself, without being coupled with behavioral change interventions, does not seem to be an efficient strategy to enforce better antibiotic stewardship [19–21]. This observation might be explained by the complexity of the factors that affect antibiotic misuse which appears to be influenced by a plethora of cultural and social factors [19,21]. For example, several reports demonstrated that the specialty of the health care provider, patient education and other patient socio-economic factors guide the antibiotic prescription patterns of physicians [22–25].

Tailoring Antimicrobial Resistance Programs (TAP) is a behavioral change methodology developed by the World Health Organization (WHO) Eastern Mediterranean Regional Office (EMRO) to modify the behaviors that drive antimicrobial resistance (AMR). TAP methodology not only aims to identify barriers against proper behavior but also identifies the incentives that drive such a behavior. TAP proposes guidelines for (a) the design of proper behavioral change strategies, (b) implementation of such strategies and (c) evaluation of the results of any behavioral intervention. The Ministry of Health in Jordan joined the WHO TAP in November 2018 to pilot a behavioral change intervention that aims to reduce the prescription of antibiotics for viral URTIs in a primary healthcare setting. This study presents and discusses the findings of the TAP intervention, specifically its effect in reducing the percentage of patients that request antibiotics. Additionally, the study investigated the association of several socioeconomic factors with changes in antibiotic request by the patients following TAP intervention.

#### **2. Results**

#### *2.1. The Characteristics of the Study Subjects*

A total of 855 subjects (506 in the control group and 349 in the intervention group) participated in the study in period 1 before the implementation of the intervention. In period 2, following the intervention, a total of 1025 subjects (576 in the control group and 449 in the intervention group) were enrolled in the study (Figure 1).

A stratified cluster randomized trial with two groups pre-post design, comparing intervention with the control (standard care), was used in the study. The study was performed in six health centers in Amman. The centers were randomized into two groups (three centers each). In period 1, there was a pre-assessment of antibiotic request. In period 2, following application of the intervention or maintenance of standard treatment care, there was a re-evaluation of antibiotic request among enrolled patients.

The socio-demographic characteristics of the subjects of the control and intervention groups in periods 1 and 2 are shown in Table 1. In period 1, 17.8% of the subjects in the control group and 7.2% of the subjects in the intervention group were children (*p* < 0.001). In period 2, almost one quarter of the subjects in both groups were children (*p* = 0.320). In period 1, subjects of the intervention group were significantly younger. Moreover, a significantly higher number of the above subjects did not hold a university degree. More than half of the subjects (55.9%) in the control group and 33% of the subjects in the intervention group were new patients. In period 2, a significantly lower number of subjects in the intervention group received college/university education than subjects in the intervention group (*p* < 0.001). The characteristics of the subjects significantly differed between period 1 and period 2 in both control and intervention groups.

*Antibiotics* **2020**, *9*, x FOR PEER REVIEW 3 of 13

**Figure 1.** A flow chart that explains the design of the study. **Figure 1.** A flow chart that explains the design of the study.

with the control (standard care), was used in the study. The study was performed in six health centers in Amman. The centers were randomized into two groups (three centers each). In period 1, there was **Table 1.** The socio-demographic characteristics of patients in the control and intervention groups during period 1 and period 2.

A stratified cluster randomized trial with two groups pre-post design, comparing intervention


#### *2.2. E*ff*ect of TAP Intervention on Antibiotics Request among Study Subjects* Figure 2). The relative percent of reduction in the percentage of subjects who requested antibiotics between the two periods in the intervention group was 36% (absolute difference of 21.7%).

The percentage of patients or their escorts who requested antibiotics in period 1, before the implementation of the intervention, was 59.7% in the control group and 60.2% in the intervention group (*p* = 0.886) (Figure 2). While the percentage of requesting antibiotics did not change significantly between period 1 and 2 in the control group (*p* = 0.393), this percentage decreased significantly in the intervention group from 60.2% to 38.5% (*p* < 0.05) (expressed as no request in Figure 2). The relative percent of reduction in the percentage of subjects who requested antibiotics between the two periods in the intervention group was 36% (absolute difference of 21.7%). *2.3. Pattern of Antibiotics Request*  In the intervention group, the percentage of patients or their escorts who requested antibiotics for themselves decreased from 51.6% before the intervention to 23.4% following the intervention, with a relative percent of reduction of 45.7% (Figure 2). In the control group, the above described percentage did not change significantly between period 1 and period 2 (*p* = 0.359). About 45.3% and 46.9% of participants requested antibiotics for themselves in period 1 and period 2, respectively.

*Antibiotics* **2020**, *9*, x FOR PEER REVIEW 5 of 13

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0%

(**A**)

#### **Intervention group**

.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0%

(**B**)

**Figure 2.** The pattern of antibiotics requests in the control or intervention groups. A horizontal bar graph displaying the pattern of antibiotics request in period 1 (**blue**) and period 2 (**red**) in (**A**) Control or (**B**) Intervention groups. Each bar represents the percentage of individuals that either did not request the antibiotics, requested the antibiotics from themselves, requested the antibiotics for an adult patient attending the center, requested the antibiotics for an adult patient not attending the center, requested the antibiotics for a child patient attending the center or requested the antibiotics for a child patient not attending the center.

#### *2.3. Pattern of Antibiotics Request*

In the intervention group, the percentage of patients or their escorts who requested antibiotics for themselves decreased from 51.6% before the intervention to 23.4% following the intervention, with a relative percent of reduction of 45.7% (Figure 2). In the control group, the above described percentage did not change significantly between period 1 and period 2 (*p* = 0.359). About 45.3% and 46.9% of participants requested antibiotics for themselves in period 1 and period 2, respectively.

#### *2.4. Reasons for Requesting Antibiotics*

Overall, among those who requested antibiotics in both the control and intervention groups (n = 1014), the most common reasons were sore throat (36.7%) followed by cough (27.5%). The healthrelated complaints that motivated subjects of this study to request antibiotics differed between the control and intervention groups (Table 2). In the control group, the most common reasons for requesting antibiotics were sore throat, followed by flu, and then cough (Table 2). In the intervention group, the main complaints were sore throat, followed by pain upon swallowing, fever, and then cough (Table 2).



#### *2.5. Type of Antibiotics Requested*

Of those who requested antibiotics, 961 (94.8%) requested a specific type of antibiotics (data not shown). It is interesting to note that more than three quarters (78.1%) requested amoxicillin/clavulanic acid (Amoclan) (data not shown).

#### *2.6. Factors That Influence Antibiotics Request*

Table 3 shows a multivariate analysis of the effect of multiple factors on the decision of the study subjects to request antibiotics from the prescriber (i.e., physician). In the control group, the variables that were significantly associated with requesting antibiotics by the patients were age of the patient, type of patient (regular vs. first appointment), level of education of the patient and the specialty of the health care provider. Antibiotics were more likely to be requested by, or for, adult patients compared to patients who were children (Odds Ratio (OR) = 1.7) (Table 3). Regular patients were more likely to request antibiotics compared to patients visiting that specific physician for the first time (OR = 2.5) (Table 3). Patients with no formal education, primary education, secondary education or with professional training were more likely to request antibiotics than patients with college/university education. Patients who were visiting a family doctor were less likely to request antibiotics than patients visiting general practitioners (OR =0.5). On the other hand, in the intervention group, our analysis showed that antibiotic request by subjects was significantly lower following the implementation of

the intervention (OR = 0.4). The only other variable in the intervention group to significantly affect antibiotic request was primary or secondary education compared to having no formal education. Our results showed that having primary or secondary education significantly increased the odds of requesting antibiotics.


**Table 3.** Multivariate analysis of factors associated with antibiotics demand in the control and intervention groups.

#### **3. Discussion**

The misuse of antibiotics in primary care is a major contributor to antibiotic resistance [5]. URTIs are common presentations seen in the general practice [26]. URTI without complications is most often caused by a virus [27]. Antibiotics have no efficacy in the treatment of viral infections, but are nevertheless often prescribed for their treatment [28,29].

In this study, using a stratified cluster randomized trial with two groups pre-post design, we evaluated the effect of the TAP intervention in reducing the percentage of patients that request antibiotics. The above design allowed for a comparison of the intervention group receiving TAP with a control group in which standard care was maintained. In addition to collecting information on the percentage of patients who requested antibiotics, the research team collected data of several factors previously reported to affect antibiotic vigilance such as gender, age, level of education of the patient and the specialty of the health care provider. The study design also differentiated between patients visiting the physician for the first time and returning patients. Recruitment to the study was restricted to patients complaining from URTIs.

In this pilot study conducted on patients visiting six health centers in the capital city of Jordan, Amman, we were the first group to demonstrate the efficacy of the TAP program in reducing the percentage of patients that request antibiotics from their health care provider in a primary care setting. Notably, our findings also indicated that the TAP program achieved its goal independent of all the other variables that might influence antibiotic requests by the patients.

In the absence of any intervention (in our case the TAP program) our findings indicated that more than half of the patients diagnosed with mild URTIs would request antibiotics for the treatment of their illness. This result is analogous with other reports which suggested that patient pressure or "perceived pressure" is a major driver of the lack of antibiotic vigilance [30,31]. The above figure, showing that

most patients request antibiotics from their primary health care provider as a result of illnesses that do not normally require antibiotics, may reflect the lack of public awareness programs on the harmful effects of the unnecessary use of antibiotics. Although this observation is alarming from a public health standpoint, the fact that the TAP method successfully reduced the number of patients that request antibiotics shows that positive behavioral change could be achieved in the patient population and invites the application of the TAP method on a larger scale.

An interesting finding of this investigation was the difference observed in the percentage of patients that request antibiotics based on the specialty of the health care provider. For example, in the control group, it was observed that patients were more likely to request antibiotics from general practitioners vs. family medicine specialists. In Jordan, general practitioners start their appointment following one year of vocational training only (internship), without enrollment in any residency program. On the other hand, to become a family medicine specialist in Jordan, candidates must finish their vocational training, enroll in a structured residency program and pass a national board exam. The exact reason behind the above disparity in antibiotic request between patients seen by different specialists is unknown but could be related to family medicine specialists building better communication and assertive skills during their residency programs [32]. If the above explanation turns out to be partially responsible for this disparity, a solution for this problem would be to offer general practitioners Continuing Medical Education (CME) courses in communication skills and antibiotic stewardship. These courses would help mend the gap created by a longer study path to become a family medicine specialist.

This investigation has a few limitations. First, this study was conducted in Amman, the capital city of Jordan. Although the findings of this study are very promising, the adoption of the TAP program as a method to achieve better antibiotic vigilance requires testing the program across different geographic regions. For example, the level of education, a variable shown in this study to affect antibiotic stewardship, might be different in Amman from other geographic regions in the country. Second, the research team failed to collect information on the volume of patients examined by physicians on a single day in the clinic. This variable was shown in several reports to affect consultation time with the patient, and was significantly associated with an excess, often unnecessary, antibiotic prescription [33,34]. Indeed, it would be interesting to evaluate if implementation of the TAP actually increased the consultation time with the patients and how that affected the overall revenue of the medical practice/clinic. Despite these limitations, this study is the first in Jordan and should be informative to public health policy makers and health care workers interested in antibiotic stewardship with regards to the size of the antibiotic misuse problem and the feasibility of reducing this problem with a simple behavioral approach.

#### **4. Materials and Methods**

#### *4.1. Study Design, Site Selection and Randomization*

A stratified cluster randomized trial with two groups pre-post design, comparing intervention with the control (standard care), was conducted in the period between August and November of 2019. The standardized behavior change intervention was implemented for eight weeks in the intervention group. The demand for antibiotics was assessed among patients with mild URTIs attending the intervention and control centers before and after introducing the behavior change intervention. Written consent was requested from all patients before the interviews. No identifiers were collected. Approval from the Jordan Ministry of Health Ethical Review Board was obtained prior to conducting the study. All interviews were conducted in a closed room to ensure privacy and confidentiality.

Six health centers in Amman, Jordan were selected using a stratified cluster randomized sampling strategy. The centers were classified into three strata—small, medium and large—based on the number of physicians and monthly patient visits obtained from statistics of the year 2018. Out of each stratum, one center was randomized to the intervention group and another center to the control group, resulting in three centers in each group.

#### *4.2. Patient Recruitment*

General practitioners, family medicine specialists, pediatricians and internal medicine specialists were trained to interview patients attending the clinic for mild URTIs or to obtain medications for relatives with mild URTIs. The practitioners used a semi-structured questionnaire before and after the intervention to assess the demand for antibiotics. The questionnaire was pilot tested on 30 patients and revised accordingly. All consecutive patients of all ages diagnosed with URTIs who visited the selected health centers during the working hours for the duration of the study period were included. Only patients visiting general practitioners, family medicine specialists, pediatricians, and internal medicine specialists were included. Patients diagnosed with infections other than URTIs were excluded.

#### *4.3. Intervention*

A strategic behavior change intervention package was designed and implemented in the three intervention centers. As part of the intervention, physicians were trained, by a WHO expert in the area of antimicrobial resistance and a consultant in communication, to adopt a more proper behavior relevant to antibiotic prescription and to communicate with patients who insist on receiving antibiotics for viral URTIs. A 1-day training workshop was held in the premises of the Ministry of Health (MoH). During the training, physicians were trained on the current national guidelines for prescribing antibiotics to patients with URTIs and were trained on the best approaches to manage discussions with difficult patients. The physicians in the intervention centers received a copy of the clinical guidelines for the diagnosis and treatment of URTIs and were instructed to adopt the guidelines in their practices. A commitment was obtained from prescribers to become advocates for the proper use of antibiotics for URTIs and to join the intervention by signing a commitment board.

Posters were placed in the waiting areas to advise patients not to request antibiotics from their doctors and to always consult a doctor before antibiotics' administration, and leaflets about the proper use of antibiotics were distributed to patients. During the routine patient consultation in the intervention centers, physicians requested each patient to answer a quiz about the indication and proper use of antibiotics. Then, the physician held a short discussion (2–3 min) about the answers to the questions, encouraged patients to reduce their requests for antibiotics, and provided patients with information about the antibiotics and the consequences of improper prescription. Peer-to-peer weekly coffee sessions were held and moderated by the MoH staff to strengthen the bonds between colleagues and managers as a single entity that reduced the unnecessary use of antibiotics. Moreover, the strategies used to implement the behavior change, including the roles of the prescribers and patients, are shown in Table 4.


*Antibiotics* **2020**, *9*, 507

The MoH staff coordinating the project regularly visited the clinics during the intervention, observed the physicians' practices, and filled out the monitoring forms. The monitoring form included information on the physicians' adherence to the study protocol and the number of patients treated.

#### *4.4. Sample Size*

A minimum sample size needed to assess the effect of the intervention on the change in the percentage of patients who request antibiotics for URTIs in a pretest-posttest nonequivalent control group design was calculated using G\*Power. Assuming that the percentage of patients who request antibiotics for URTIs in the selected health centers is 50%, the sample size needed to detect a change of 12% in this percentage following the intervention (at a level of significance of 0.05 and a power of 80%) is 370 patients in the intervention group (370 at pretest and 370 at posttest) and 370 patients in the control group (370 at pretest and 370 at posttest). This is the minimum sample size with enough power to determine the impact of the intervention, taking into consideration that the analysis will be stratified by demographic and clinical characteristics.

#### *4.5. Statistical Analysis*

Data were analyzed using IBM SPSS, version 20 (IBM Corp., Armonk, NY, USA). Data were described using means, standard deviations, and percentages. Chi-square test was used to compare the percentage of patients who requested antibiotics for URTIs between intervention and control groups and between the two periods within each group. The same test was used to compare demographic and other categorical variables between intervention and control groups and between the two periods within each group. Binary logistic regression was used to test for the change in request for antibiotics over time, after adjusting for patients' characteristics. The interaction term between period (pretest (period 1)/ posttest (period 2)) and group (intervention/control) was tested. A *p*-value of less than 0.05 was considered statistically significant.

#### **5. Conclusions**

In conclusion, in this pilot study evaluating the TAP program as a measure to achieve proper antibiotic stewardship in Jordan, we provide evidence on its efficacy, simplicity and feasibility. Given the small scale of this investigation, we recommend testing the program on a larger scale and across multiple health sectors in the country. We anticipate that the interventional program described in this investigation might be adopted as a public health method to address the misuse of antibiotics in Jordan.

**Author Contributions:** Conceptualization, N.M.K.; Formal analysis, Y.S.K. and R.S.; Funding acquisition, L.A.S.; Methodology, Y.S.K.; Project administration, N.M.K., Y.S.K., M.A.A. and L.A.S.; Validation, R.S.; Writing—original draft, M.A.A.; Writing—review and editing, M.A.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the World Health Organization (WHO) office in Amman.

**Acknowledgments:** The authors would like to thank officials at the Ministry of Health in Jordan for facilitating completion of the study. BMJ 1998, 317, 609–610.

**Conflicts of Interest:** The authors would like to report no conflict of interest.

#### **Abbreviations**



#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

*Perspective*

#### **The Need for Ongoing Antimicrobial Stewardship during the COVID-19 Pandemic and Actionable Recommendations**

#### **Wei Ping Khor <sup>1</sup> , Omotayo Olaoye <sup>1</sup> , Nikki D'Arcy <sup>1</sup> , Eva M. Krockow <sup>2</sup> , Rasha Abdelsalam Elshenawy <sup>3</sup> , Victoria Rutter <sup>1</sup> and Diane Ashiru-Oredope 1,\***


Received: 14 November 2020; Accepted: 9 December 2020; Published: 14 December 2020

**Abstract:** The coronavirus disease (COVID-19) pandemic, which has significant impact on global health care delivery, occurs amid the ongoing global health crisis of antimicrobial resistance. Early data demonstrated that bacterial and fungal co-infection with COVID-19 remain low and indiscriminate use of antimicrobials during the pandemic may worsen antimicrobial resistance It is, therefore, essential to maintain the ongoing effort of antimicrobial stewardship activities in all sectors globally.

**Keywords:** antimicrobial stewardship; COVID-19; pharmacy

#### **1. Introduction**

Coronavirus disease (COVID-19), caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been exerting a significant impact on global health care delivery across both primary and secondary care, since it was first reported in December 2019 [1]. As of 20 September 2020, 30 million people globally have tested positive for COVID-19, of which 3.1% have died [2]. It is critical that normal acute infection management is maintained, and potential COVID-19 complications are anticipated. For the majority of patients, COVID-19 will run an uncomplicated course, hospital admission will not be required, and secondary infection will be uncommon [3].

The current SARS-CoV-2 pandemic occurs amid the already ongoing global health crisis of antimicrobial resistance (AMR). Infections caused by antimicrobial-resistant pathogens are estimated to cause 700,000 deaths each year globally and may complicate the care of COVID-19 patients, potentially leading to increased mortality [4], and result in significant economic burden. Resistant infections have previously been highlighted (pre-pandemic era) as causing economic burden and estimated to cost more than 100 trillion US dollars by 2050 if left unaddressed [4]; these are likely to be even more significant in the current pandemic and the post-pandemic era. At time of writing, it is reported that approximately 5% of COVID-19 patients require admission to intensive care units (ICU) and those with significant co-morbidity may require ventilatory support [5]. ICU admission and mechanical ventilation significantly increase the risk of patients acquiring secondary viral, bacterial and fungal infections [6,7].

One risk during a pandemic is that all resources may be diverted to treating patients infected with the pandemic agent, and other key health care priorities may be overlooked or deprioritised. However, antimicrobial stewardship (AMS) programmes remain essential and are likely more important at a time when needs for healthcare resources may exceed capacity [8]. Inappropriate access to and use of antimicrobials during the current SARS-CoV-2 pandemic may worsen AMR globally. This paper aims to contribute to highlighting the need for ongoing action to tackle the global AMR crisis during the COVID-19 pandemic and the need to uphold and continue the principal of AMS programmes among pharmacy teams. First, it focuses on reviewing the key challenges of optimising infection management and minimising AMR during the pandemic. For this, a literature search was performed by extracting published articles from PubMed with the search terms of "antibiotic stewardship", "antimicrobial stewardship", "antimicrobial resistance" and "COVID-19". Subsequently, the article makes actionable recommendations for clinical practice in the context of COVID-19.

#### **2. Key Challenges of Optimising Infection Management and Minimising AMR**

#### *2.1. Continued Occurrence of Common Infections*

Amid the current SARS-CoV-2 pandemic, common infections including e.g., seasonal influenza, bacterial infections, tropical infections or malaria will continue to be present [9]. There is no evidence that these common infections should be managed differently during the pandemic [9]. Local, and/or national primary and secondary care infection management guidelines and AMS principles should continue to be followed. Inappropriate use of antibiotics to treat viral infections and indiscriminate use of broad-spectrum antibiotics may reduce availability and lead to resistance and/or increased *Clostridiodes di*ffi*cile* infections. Chronic infections such as human immunodeficiency virus (HIV) and tuberculosis remain global health issues and may be heavily impacted by the ongoing pandemic, in which diagnosis and treatment may be delayed, inappropriate or interrupted [10].

As governments across the world are closing down cities and restricting movements to flatten the curve of the pandemic, a large proportion of healthcare resources as well as staff are being diverted to halt the spread of COVID-19 [11]. In addition, there is emerging evidence that healthcare staff that lead on AMS have been asked to prioritise COVID-19 response and management, leading to reduced AMS activities.

#### *2.2. Empiric Use of Antimicrobials in Patients with Suspected or Proven COVID-19*

Testing for SARS-CoV-2 is currently not widely available globally, and the reverse transcription polymerase chain reaction (RT-PCR) test that is currently the gold-standard diagnostic test has a high false negative rate [12]. In the absence of diagnostic confirmation of SARS-CoV-2 infection, it is important that clinical features are carefully assessed to determine the likely source of infection. However, clinical features of SARS-CoV-2 infection are non-specific and can be indistinguishable from bacterial or influenza pneumonia. Some published initial recommendations were to consider empirical broad-spectrum antibiotics and neuraminidase inhibitors when patients presenting with COVID-19 symptoms are admitted to intensive care units [13]. However, it is important to note that the use of broad-spectrum antibiotics can lead to *Clostridiodes di*ffi*cile* infection and a rise in AMR.

WHO guidance on the clinical management of COVID-19 suggests that antibiotics should not be prescribed for the prevention or treatment of mild COVID-19; while for suspected or confirmed moderate COVID-19 cases, antibiotic therapy should only be offered if there is clinical suspicion of bacterial infection [14]. However, for patients who have suspected or confirmed severe COVID-19, early empirical antimicrobials can be administered to treat all likely pathogens based on clinical judgement, patient host factors and local epidemiology [14].

Other international guidance, for example, The National Institute for Health and Care Excellence (NICE, London, UK) guidelines from the UK, suggest that antibiotics for the treatment or prevention of pneumonia in community settings should not be offered if SARS-CoV-2 is likely to be the cause or if the symptoms are mild [15]. Similarly, in Africa, the Uganda Ministry of Health guidelines do not advocate the use of antibiotics if the patient only suffers from mild COVID-19 symptoms [16].

In spite of this guidance, the empirical use of antibiotics in hospital settings is likely to increase globally because of the ongoing pandemic. In published case studies from China, it has been shown that 100% of severe and moderate cases were treated empirically with antimicrobials such as moxifloxacin and/or cephalosporin [17]. Indeed, this appears to have been standard practice in many hospitals in China [18].

In one hospital in Wuhan, 95% of admitted patients with suspected SARS-CoV-2 infection received antibiotics [19]. Of 191 patients included in the study, 181 patients received antibiotics, but this was shown to have no effect on survival (*p* = 0.15) while 41 patients received antiviral agents, which also had no effect on survival (*p* = 0.87) [19]. Similarly, in a review paper assessing 9 studies conducted in China and the United States including 806 SARS-CoV-2 positive patients, while some patients were found to develop bacterial or fungal co-infection, the comparatively low proportion (8%) did not justify the reported antimicrobial prescribing rates, which included 72% of patients receiving empirical broad-spectrum antimicrobial therapy [20]. A study conducted in the United Kingdom by Hughes et al. showed that the number of bacterial coinfections is low, occurring in 3.2% (27/836) of cases [21]. No evidence of fungal co-infection during early COVID-19 hospital presentation (0–5 days post-admission) was observed [21]. These findings demonstrate that there are currently limited data to support widespread usage of antimicrobial therapy on COVID-19 patients, and there is a need to develop global and regional antimicrobial policies and strengthen AMS interventions to prevent inappropriate use of antimicrobial therapy during the pandemic.

#### *2.3. Falsified and Substandard Antimicrobial Medicines*

Another challenge during the pandemic response in many countries is combating falsified and substandard medicines and pharmaceutical supplies. Falsified medicines are medicines which have no or little active ingredients and have not undergone any quality control evaluation, while substandard medicines are authorised medical products that fail to meet either their quality standards or specifications or both [22]. Past studies have shown that falsified or substandard antimicrobials are highly likely to promote the emergence and spread of AMR [23]. Recent studies have also shown that the resistance of *Escherichia coli* and *Mycobacterium smegmatis* to rifampin occurred as a result of exposure to substandard medicines; this presents a potential threat to tuberculosis treatment [24,25]. Over the years, it has been recognised that addressing the problem of substandard or falsified medicines will require the united action of all relevant stakeholders including government bodies, policy makers, regulatory and law enforcement agencies, public health professionals, patients and the general public. Pharmacists play a pivotal role in combating falsified and substandard medicines by working on strengthening supply chain procurement processes to ensure uninterrupted access to safe and effective medicines during the pandemic [26–29].

#### *2.4. Stock Management and Supply Chain of Antimicrobials*

Global antimicrobial supply chains are likely to be affected by the pandemic, and the cost may be increased due to travel restrictions or cancellations, therefore risking patients' lives and potentially contributing to drug resistance [11]. This is particularly challenging in countries that are highly dependent on imported medicines and pharmaceuticals [30]. Hence, there is a need to apply new innovative supply chain management strategies and diverse supply chains to ensure and protect the supply of essential medicines during the COVID-19 pandemic and beyond [31,32].

#### *2.5. Healthcare Associated Infections*

Whilst data are currently scant and there is no evidence to suggest that patients with COVID-19 are more likely to be infected by multidrug resistant bacteria and fungi, there is increased possibility that healthcare-associated infections will occur in COVID-19 patients with prolonged hospitalisation [33]. A large proportion of patients admitted to hospital due to severe symptoms required mechanical ventilation (75%), in a Seattle study [34]. Patients may be at increased risk of developing hospital-acquired

pneumonia (HAP) and ventilator-associated pneumonia (VAP), which are often associated with drug-resistant bacterial strains. In the first documented outbreak of SARS-CoV-2 in Wuhan, China, VAP occurred in 31% of patients requiring mechanical ventilation and was associated with increased mortality [19]. There is evidence to suggest that a large proportion of deaths during the 1918 Influenza pandemic were due to secondary bacterial infection [35]. Secondary fungal infections must also be considered; putative invasive pulmonary aspergillosis was found in almost one third of critically ill COVID-19 patients at a Parisian hospital [6].

#### **3. Recommendations for Adaptations of Clinical Practice in the Context of COVID-19**

#### *3.1. Consider Existing AMS Principles*

Adherence to the local, national and international guideline recommendations is vital to prevent over- and inappropriate prescribing of antimicrobials during the pandemic. To support ease of access to antimicrobial prescribing and COVID-19 management guidelines, the Commonwealth Pharmacists Association (London, UK) developed a new repository of resources on COVID-19 prevention and management via the Commonwealth Partnerships for Antimicrobial Stewardship (CwPAMS) app, which is a smartphone app that consists of national antimicrobial prescribing guidelines from Ghana, Tanzania, Uganda and Zambia as well as international guidelines from the World Health Organization and International Pharmaceutical Federation (FIP) (The Hague, The Netherlands) [36]. A repository of useful resources on COVID-19 may be found in the Supplementary Materials File S1.

Despite limited evidence for the effectiveness of AMS interventions in low- and middle-income countries, AMS alongside infection prevention and control (IPC) remain the cornerstone to tackle AMR [37,38]. Appropriate use of antimicrobials may also reduce the economic burden and ensure availability of antimicrobials given the economic crises during the pandemic [38]. National and/or local levels should continue to develop action plans and policies to promote and perform AMS programmes [38,39]. In recent development, the Commonwealth Pharmacists Association published a CwPAMS toolkit, which outlines strategies and projects that a healthcare facility can implement as part of an AMS workplan. This may serve as a guidance especially for resource-limited countries to initiate AMS programmes [40].

Patient education on the appropriate use of antimicrobials is important as there is no evidence that antibiotics can be used for the treatment of viruses, and research/clinical trials on the use of certain antimicrobials in the management of COVID-19 is still ongoing. On 23 April 2020, the Africa Centres for Disease Control released a statement on medications to treat COVID-19. In the statement, the Africa CDC made the following recommendation:

"Physicians should not prescribe, and individuals should not take, chloroquine or hydroxychloroquine to prevent or treat COVID-19 except under clinical trial or monitored emergency use of unregistered and investigational interventions (MEURI) as these drugs can cause neurologic, ophthalmic, cardiac, and other forms of toxicity and Physicians should not prescribe, and individuals should not take, Lopinavir/Ritonavir, Remdesivir or other medications to prevent or treat COVID-19 except under clinical trial or MEURI" [41].

The WHO's interim guidance on the clinical management of COVID-19 released on 27 May 2020 also makes the same recommendation stating that:

"Chloroquine and hydroxychloroquine (+/− azithromycin); antivirals including but not limited to Lopinavir/ritonavir, remdesivir, umifenovir, favipiravir; Immunomodulators, including but not limited to tocilizumab and Interferon-β-1a, and plasma therapy should not be administered as treatment or prophylaxis for COVID-19, outside of the context of clinical trials" [14].

Key AMS components that can be promoted and practiced amid the current ongoing COVID-19 pandemic are described below. Actionable recommendations are sub-divided into different sectors of care.

#### 3.1.1. Hospital Care

To improve infection management and reduce AMR in hospital patients, the following principles are likely to be vital [14,39]:


#### 3.1.2. Community Care/Primary Care

Primary care providers have the responsibility to support and guide patients through managing COVID-19 symptoms, explaining [42]:


#### *3.2. Harness the AMS Role of Pharmacists and Their Teams*

Pharmacy teams in the community especially, also play a strategic role in ensuring the rational use of medicines and, as a result, are critically placed to help address AMR [43]. Pharmacists, along with other healthcare professionals, are crucial to ensuring that knowledge and evidence are effectively gathered and provided to members of the public. This ensures the judicious use of medications and prevention of stockpiling of medicines, especially the precious commodities of anti-infective drugs. Individual countries should devise and strengthen prescription-monitoring schemes to monitor the safety and efficacy of any off-label drugs being used for COVID-19 management [44].

Community pharmacists' dual roles as healthcare providers and retailers has also become very apparent during the pandemic [45], emphasizing their indispensable place as providers of safe and

effective medicines and medicines information. Pharmacists and pharmacy associations have a vital role in engaging the public and providing education in communities to ensure timely delivery of scientifically proven and reliable information on COVID-19 prevention and management. In line with this, the CPA alongside other pharmacy bodies such as FIP (The Hague, The Netherlands) and the Royal Pharmaceutical Society (London, UK) organised webinars aimed at ensuring that pharmacists across the Commonwealth, worldwide and UK respectively were equipped with knowledge and the right resources to support the COVID-19 response during the peak of the pandemic. The webinars were designed strategically to prevent common occurrences of misinformation and rumours during the pandemic, which can lead to misuse of medicines and a negative impact on public health [46]. This has been highlighted in the cases of chloroquine, a drug which has a long history of being used for malaria treatment, and hydroxychloroquine, which is commonly used for autoimmune disease treatment. These drugs have come into the limelight through the media as potential COVID-19 treatments. There is currently no randomised controlled trial that suggests their efficacy in preventing or treating COVID-19, and indiscriminate promotion and widespread use of chloroquine and hydroxychloroquine have led to drug shortages, self-treatment, fatal overdoses and potential drug resistance [44]. This is of considerable health concern especially in countries, which are endemic with malaria or have poor assess to reliable and accurate health information.

Community pharmacists are well placed to promote AMS and often have the right knowledge, adequate opportunity and inherent dedication required. Despite this, there is limited information on AMS interventions at a community level with the community pharmacist role being less established and harnessed [47]. Similarly, there is a paucity of data on how community pharmacists have applied the principles of AMS to combat AMR in low- and middle-income countries during the COVID-19 pandemic. Community pharmacists can promote AMS in the context of COVID-19 through [47]:


Pharmacy teams across all sectors are at the forefront of contributing to the COVID-19 crisis emergency preparedness. A recent study on global contributions by pharmacists during the COVID-19 pandemic in nine countries discussed how pharmacists worked at the frontline of the pandemic to provide care spanning across a broad range of areas including community pharmacies, hospitals, clinics, public health and care homes among other vital areas [50]. It is, therefore, important for pharmacy teams to be equipped with emergency preparedness skills as well as knowledge on prevention measures for COVID-19 whilst carrying out their duties. This includes [51]:


#### *3.3. Address Issues of Falsified and Substandard Antimicrobial Medicines*

As professionals charged with the final custody of medicines, pharmacists have a vital role to play in ensuring that the quality and efficacy of medicines is maintained in these settings, especially with impending challenges in drug supply and access to quality medicines as a result of the pandemic. Substandard and falsified medicines pose significant safety, quality and efficacy risks to patients [52]. In the context of falsified medicines, community pharmacists as frontline healthcare workers have specific duties including:


#### *3.4. Manage Access to E*ff*ective Antimicrobials*

Preparedness to ensure hospitals do not run out of antibiotics and other critical drugs is key, and guidance for which medicines stocks to increase should be provided. The WHO has recently published a COVID-19 Essential Supplies Forecasting Tool, which provides guidance on essential drugs including antimicrobials and consumables required to treat severe or critically ill patients [53]. Individual countries will need to conduct active surveillance and establish early warning mechanisms to receive alerts whenever a drug shortage is anticipated by evaluating a drug utilisation review especially for antibiotics. This is particularly important for antibiotics, which are commonly used for community-associated bacterial to lower respiratory tract infections, as shortages are expected to increase urgent care consultations and potentially increase hospital admissions. There has been increasing demand and evidence for the incorporation of digital technology as a tool to monitor stock levels, which provides feedback mechanisms that would ensure the continuity of medication supply.

#### *3.5. Ensure E*ff*ective Infection Prevention and Control (IPC) Practices*

IPC and appropriate use of personal protective equipment (PPE) have been well recognised as ways to control AMR [54]. During a pandemic, these measures are critical and should be expanded to contain the spread of the infections (both the spread of SARS-CoV-2 as well as other hospital-acquired infections [54]. Dedicated, trained IPC teams should be in place where possible. In countries where IPC is limited or non-existent, minimum requirements for IPC must be implemented as soon as possible, both at the national and facility level.

The WHO has issued five strategies to prevent or limit the transmission of SARS-CoV-2 in health care settings. It specifies:


Hand hygiene and respiratory measures are essential, and all health care workers should be aware of and apply the WHO's 5 Moments for Hand Hygiene approach [54]. Appropriate selection of hand rub is equally important to ensure optimal antimicrobial efficacy [54]. In accordance with the WHO guidance on local production of hand rub formulation, the Commonwealth Pharmacists Association responded to the COVID-19 pandemic by launching a training video to support pharmacy teams in the production of WHO-formula alcohol-based hand sanitisers to further support IPC in hospitals and prevent the spread of infections, including COVID-19 [55,56].

Whilst access to PPE may be limited, IPC teams should, at the earliest opportunity, assess and quantify demand for masks and hand sanitisers. Local manufacturers of PPE should be identified and engaged to ensure supply where possible. Teams should be trained in appropriate use of gloves and masks—how to use, remove and dispose of them. Interventions to minimise the need for PPE include the use of telemedicine to identify COVID-19 cases, physical barriers such as windows at points of patient contact, restriction in access to patients by visitors and non-essential healthcare workers [57]. PPE should only be used where appropriate to minimise shortages. Environmental cleaning and disinfection procedures should be adhered to.

The use of IPC measures must be supported by administrative controls, including the availability of resources, appropriate infrastructure to allow the segregation of infected patients from the uninfected and distancing of healthcare workers and patients, the development of clear IPC policies and the access to laboratory testing.

#### *3.6. Advocate for AMS at the Governmental Level (State or Federal)*

The harm of AMR has a widespread effect on not only human health but also other critical priorities including the achievement of universal health coverage and sustainable development goals (SDGs). For example, combatting AMR is important for achieving SDG 3 of "good health and well-being", because the availability of effective antimicrobials is essential for restoring health where an infection is present. Effective antimicrobials also support the prevention of maternal, neonatal and childhood deaths as well as epidemics of communicable diseases such as tuberculosis, HIV and gonorrhea [58]. Other related SDG goals which AMR can impact include SDG 2 "zero hunger", SDG 8 "decent work and economic growth", SDG 6 "clean water and sanitation", SDG 12 "responsible consumption and production" and SDG 17 "partnerships for the goals" [58]. Considering that AMS is the key action to combat AMR, advocacy at the government level is therefore, an important determinant of its success. Identifying gaps is an important initial step for advocacy; for this, assessing the current level of AMS activities using the WHO checklist of essential national/regional and facility core elements for AMS programmes is recommended [59].

National and international advocacy as well as advocacy through civil societies is particularly important in resource-limited areas [60]. Through the effort of governmental and international collaborations to share established strategies, policies and skills; resource-limited countries may benefit

from the experience of countries with existing AMS programmes. This can provide a framework to kickstart and expand AMS programmes without unnecessary delays [60].

#### **4. Conclusions**

The COVID-19 pandemic is a significant and new public health threat, putting tremendous pressure on all healthcare professionals. However, the ongoing global crisis of AMR must not be neglected. We highlight key challenges of infection management including continued occurrence of common infections, empiric use of antibiotics to treat COVID-19 patients, problematic access to effective antimicrobials and hospital-acquired infections. Given these challenges, we advocate that urgent actions are required to continue AMS practices during the pandemic. Specifically, we highlight the need for the reliance on existing principles of AMS across the hospital sector, primary care and community pharmacy. Other recommendations include ensuring access to effective antimicrobials as well as upholding the principles of IPC. Finally, advocacy for AMS must continue at all levels during the current pandemic and in the post-pandemic era.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2079-6382/9/12/904/s1. Supplementary Materials File S1: A repository of useful resources on COVID-19.

**Author Contributions:** Conceptualisation, D.A.-O., V.R. and N.D.; funding acquisition, D.A.-O. and V.R.; methodology, D.A.-O., N.D., O.O. and W.P.K.; data Curation, O.O., W.P.K. and D.A.-O.; project administration, W.P.K.; supervision, D.A.-O. and N.D.; writing of the original draft, W.P.K. and O.O.; writing of review and editing, W.P.K., O.O., E.M.K., R.A.E., N.D. and D.A.-O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was funded as part of the CwPAMS supported by the Tropical Health and Education Trust (THET) and the CPA using official development assistance (ODA) funding, through the Department of Health and Social Care's Fleming Fund. The Fleming Fund is a £265 million UK aid investment to tackle antimicrobial resistance by supporting low- and middle-income countries to generate, use and share data on AMR. The programme is managed by the UK Department of Health and Social Care. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the Fleming Fund, the Department of Health and Social Care, THET or CPA.

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

#### **References**


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

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

#### *Article* **Assessing the Impact of COVID-19 on Antimicrobial Stewardship Activities/Programs in the United Kingdom**

**Diane Ashiru-Oredope 1,\* , Frances Kerr <sup>2</sup> , Stephen Hughes <sup>1</sup> , Jonathan Urch <sup>1</sup> , Marisa Lanzman <sup>1</sup> , Ting Yau <sup>1</sup> , Alison Cockburn <sup>2</sup> , Rakhee Patel <sup>1</sup> , Adel Sheikh <sup>1</sup> , Cairine Gormley <sup>3</sup> , Aneeka Chavda <sup>1</sup> , Tejal Vaghela <sup>1</sup> , Ceri Phillips <sup>4</sup> , Nicholas Reid <sup>4</sup> and Aaron Brady <sup>3</sup>**


**Abstract:** Since first identified in late 2019, the acute respiratory syndrome coronavirus (SARS-CoV2) and the resulting coronavirus disease (COVID-19) pandemic has overwhelmed healthcare systems worldwide, often diverting key resources in a bid to meet unprecedented challenges. To measure its impact on national antimicrobial stewardship (AMS) activities, a questionnaire was designed and disseminated to antimicrobialstewardship leads in the United Kingdom (UK). Most respondents reported a reduction in AMS activity with 64% (61/95) reporting that COVID-19 had a negative impact on routine AMS activities. Activities reported to have been negatively affected by the pandemic include audit, quality improvement initiatives, education, AMS meetings, and multidisciplinary working including ward rounds. However, positive outcomes were also identified, with technology being increasingly used as a tool to facilitate stewardship, e.g., virtual meetings and ward rounds and increased the acceptance of using procalcitonin tests to distinguish between viral and bacterial infections. The COVID-19 pandemic has had a significant impact on the AMS activities undertaken across the UK. The long-term impact of the reduced AMS activities on incidence of AMR are not yet known. The legacy of innovation, use of technology, and increased collaboration from the pandemic could strengthen AMS in the post-pandemic era and presents opportunities for further development of AMS.

**Keywords:** COVID-19; antimicrobial stewardship (AMS); antimicrobial resistance (AMR); coronavirus; SARS-CoV-2

**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/).

#### **1. Introduction**

The novel coronavirus, SARS-CoV-2, has dominated all aspects of healthcare since it was first identified at the end of 2019 [1,2]. The coronavirus disease (COVID-19) has overwhelmed healthcare systems in those countries affected and diverted resources away from established services, as clinical teams look to manage this pandemic [3]. The antimicrobial stewardship (AMS) services, established to optimize anti-infectives and minimize the spread and impact of antimicrobial resistance (AMR), have been severely impacted by COVID-19 [4]. Whilst we battle against this pandemic, it is essential that we do not lose sight of the long-term AMR priorities.

**Citation:** Ashiru-Oredope, D.; Kerr, F.; Hughes, S.; Urch, J.; Lanzman, M.; Yau, T.; Cockburn, A.; Patel, R.; Sheikh, A.; Gormley, C.; et al. Assessing the Impact of COVID-19 on Antimicrobial Stewardship Activities/Programs in the United Kingdom. *Antibiotics* **2021**, *10*, 110. https://doi.org/10.3390/antibiotics 10020110

Academic Editor: Albert Figueras Received: 1 December 2020 Accepted: 16 January 2021 Published: 23 January 2021

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

41

The long-term impact of COVID-19 on AMR has been much debated in the recent literature [5,6]. The highlighted importance of infectious disease and microbiology teams in managing this emerging pandemic, the increased awareness of and use of personal protective equipment and greater focus on hand hygiene are all expected to support existing AMR strategies. Limiting patient contact and social distancing may lead to reductions in healthcare-associated transmission of disease. These benefits are likely offset by prioritized allocation of isolation rooms to COVID-19 patients over those with multi-drug resistant organisms and the reallocation of resource to fight this pandemic. Many infectious disease and microbiology teams have been repurposed to manage complex COVID-19 patients and thus established AMS services have suffered. High antibacterial prescribing in patients presenting with COVID-19 is expected to propagate AMR and presents an immediate challenge for AMR [7]. Reports of low prevalence of confirmed bacterial and fungal coinfections with COVID-19 are emerging yet high rates of empiric antibacterial prescribing are evident [8,9]. Challenges differentiating COVID-19 presentations with classical bacterial pneumonia, the established concerns with bacterial co-infection with other viral infections (e.g., influenza), and often reduced diagnostic resources all contribute to difficulties when differentiating COVID-19 from potential concurrent bacterial infection [10–15]. Understandably, in the absence of robust evidence and clear guidance, antibacterials are often added as a precaution. This is complicated further by early conflicting evidence purporting the potential antiviral role of azithromycin, subsequently leading to increased use of macrolides for non-bacterial indications [16–19].

The infection pharmacist has been central to the delivery of care on the frontline and supporting the traditional AMS role. With the increased pressure on the health system during the pandemic, infection pharmacists have been called upon as key members of the healthcare team to support and alleviate the burden on over-stretched emergency departments, intensive care units, and to support medical staff with the management of high acuity patients. In addition, AMS roles have developed in response to local needs and resource availability. Availability of new technologies and reduced patient contact have also transformed traditional services and provide unique challenges and opportunities for antimicrobial teams [20]. The expected impact of COVID-19 on existing AMS services and on antimicrobial prescribing; thus, AMR remains unknown [6].

The challenge for pharmacists to balance the demands of daily clinical duties with those of maintaining an oversight of the rapidly emerging evidence base is great. Frequent reviews of the literature, drafting local guidelines, managing the effects of fragile medication supply chains, and introducing novel anti-infective therapies within trial or compassionate use settings as well as effectively communicating these changes have become an essential role for infection pharmacists. The Pharmacy Infection Network (PIN) of the United Kingdom Clinical Pharmacy Association (UKCPAPIN) during the first wave of the pandemic in the UK sought to support pharmacists, providing peer support, and creating the opportunities for shared learning to help reduce the burden for individual pharmacists. To better understand what was being done, what the barriers were, and the potential impact of COVID-19 on existing AMS services the UKCPAPIN developed a survey for distribution to all UKCPAPIN members within the United Kingdom. The survey was purposed to explore the intended and unintended changes of AMS services, to quantify (where possible) these changes at a national level, to guide future interventions by the UKCPA to better support colleagues and advocate for relevant actions based on recommendations from the survey results.

This manuscript provides an overview of this survey, conducted in June 2020, describing the challenges and opportunities that exist in the AMS teams across the UK and Ireland and identifies how the UKCPA can better support antimicrobial pharmacists in their goals to optimize patient care in these unprecedented times.

#### **2. Results and Discussion**

#### *2.1. Demographics of Respondents*

Overall, there were responses on behalf of 95 of 169 acute trusts or health boards (56%) in the UK: 79/143 acute trusts in England, 5/14 health boards (Scotland), 7/7 health boards (Wales), and 4/5 health and social care trusts (Northern Ireland) (Table 1). This is the widest survey to date that authors can locate on the effects of the COVID-19 pandemic on AMS activities, covering almost a hundred healthcare providers (56%) across the four nations of the UK. Majority of the responding organizations were hospital trusts consisting of district/general hospitals (41%) followed by teaching hospitals (26%) (Table 1).


**Table 1.** Country distribution of responses (*n* = 95).

The approximate number of hospitalized COVID-19 cases as estimated by the respondents in the organizations (up until 31 May 2020) ranged from 0 to >2000; the majority reported having more than 500 hospitalized cases of COVID-19 at the time of the survey and four organizations reported having more than 200 hospitalized cases.

The majority of the respondents were lead antimicrobial or infection pharmacists (90%; 85/95), members of the infection/AMS pharmacy team (7%; 7/95) or microbiologist (1%; 1/95) who would have good insight into the AMS challenges and changes within their organizations. Two (2%) of the respondents were clinical pharmacists. There were no AMS nurse respondents.

#### *2.2. Impact of COVID-19 on Antimicrobial Stewardship (AMS) Activities/Initiatives*

When asked how much of an impact COVID-19 had had on their routine AMS activities (i.e., "In your opinion, how much impact would you say COVID-19 has had on your routine AMS activities?), 65% (61/95) felt that COVID-19 had a negative impact on routine AMS activities, with 31% (29/95) stating it had a very negative impact and 34% (32/95) describing some negative impact. While no one felt it had a very positive effect, 7% (7/95) did feel that the overall effect of COVID-19 was positive, whereas 25% (25/95) respondents thought that overall there were both positive and negative effects on AMS with the COVID-19 pandemic. Only 2 (2%) participants felt that COVID-19 had no impact on AMS activity within their hospital and one respondent stated they were unsure/unable to assess. *Antibiotics* **2021**, *10*, x FOR PEER REVIEW 4 of 13 routine AMS activities?), 65% (61/95) felt that COVID-19 had a negative impact on routine AMS activities, with 31% (29/95) stating it had a very negative impact and 34% (32/95) describing some negative impact. While no one felt it had a very positive effect, 7% (7/95) did feel that the overall effect of COVID-19 was positive, whereas 25% (25/95) respondents thought that overall there were both positive and negative effects on AMS with the

> Most of the activities listed in Figure 1 were considered to have been negatively affected by the pandemic. The greatest impact was on audit, quality improvement initiatives, education, training, AMS meetings, and multidisciplinary workings including ward rounds. Qualitative data collected through open questions also supported this, with respondents highlighting core AMS work such as reviewing and writing non-COVID-19 guidelines as being the most affected. Respondents were concerned about increased antibiotic use, including increased use of broad-spectrum antibiotics, delayed IV to oral switches (IVOST), and prolonged antibiotic durations. However, they were not able to accurately quantify increases due to the impact on routine AMS surveillance activities. In addition, there were concerns of inappropriate prescribing of antimicrobials in patients with COVID-19 infection. Although these concerns cannot be accurately quantified at present due to the UK-wide decrease in audit activities undertaken by antimicrobial pharmacists, the suspicion of increased 'just in case' prescribing of antimicrobials is supported by PHE Fingertips data. Analysis of this national surveillance database indicates a substantial increase in antibiotic prescribing (DDD/1000 admissions) for the current COVID-19 period in comparison to all previous quarters going back to 2017 [21]. Notably, this trend was also seen across all NHS Acute Trusts in England. COVID-19 pandemic. Only 2 (2%) participants felt that COVID-19 had no impact on AMS activity within their hospital and one respondent stated they were unsure/unable to assess. Most of the activities listed in Figure 1 were considered to have been negatively affected by the pandemic. The greatest impact was on audit, quality improvement initiatives, education, training, AMS meetings, and multidisciplinary workings including ward rounds. Qualitative data collected through open questions also supported this, with respondents highlighting core AMS work such as reviewing and writing non-COVID-19 guidelines as being the most affected. Respondents were concerned about increased antibiotic use, including increased use of broad-spectrum antibiotics, delayed IV to oral switches (IVOST), and prolonged antibiotic durations. However, they were not able to accurately quantify increases due to the impact on routine AMS surveillance activities. In addition, there were concerns of inappropriate prescribing of antimicrobials in patients with COVID-19 infection. Although these concerns cannot be accurately quantified at present due to the UK-wide decrease in audit activities undertaken by antimicrobial pharmacists, the suspicion of increased 'just in case' prescribing of antimicrobials is supported by PHE Fingertips data. Analysis of this national surveillance database indicates a substantial increase in antibiotic prescribing (DDD/1000 admissions) for the current COVID-19 period in comparison to all previous quarters going back to 2017 [21]. Notably, this trend was also seen across all NHS Acute Trusts in England.

**Figure 1.** Impact of coronavirus 2019 (COVID-19) on antimicrobial stewardship (AMS) activities (*n*  = 95 survey respondents). **Figure 1.** Impact of coronavirus 2019 (COVID-19) on antimicrobial stewardship (AMS) activities (*n* = 95 survey respondents).

Furthermore, PHE Fingertips data also reported a reduction in the WHO-classified 'Access' group of antibiotics which are typically narrow-spectrum and indicated as firstline treatment agents. Conversely, an increase in both the WHO-classified 'Watch' and Furthermore, PHE Fingertips data also reported a reduction in the WHO-classified 'Access' group of antibiotics which are typically narrow-spectrum and indicated as first-line treatment agents. Conversely, an increase in both the WHO-classified 'Watch' and Reserve' groups of antibiotics (typically more broad-spectrum and/or last resort antibiotics) were recorded nationally [21].

This suggests that nationwide use of antibiotics is not only increasing in overall volume but, more concerningly, in the number and volume of broad-spectrum agents prescribed. It is beyond the scope of this paper, but this trend has obvious implications for antimicrobial resistance in the months and years to

Open questions within the survey indicated that respondents were concerned that cases of *Clostridioides difficile* infection (CDI) were rising in some hospitals. It is however difficult to attribute increasing CDI rates with reduced AMS activities as there are multiple confounding factors involved. National surveillance of CDI also shows that cases were already rising pre-COVID-19 pandemic [21]. Moreover, when inquiring into what causes the increased concern for pharmacists, we found that physical limitations on conducting ward rounds, the inability to conduct regular antimicrobial audits, and the inability to see patients in person to confirm patient medication histories were most commonly cited. Stock shortages were also identified as time consuming and difficult to manage due to overwhelmed supply chains for antibiotics, antivirals and in some cases personal protective equipment (PPE). Some stock shortages for some antimicrobials such as levofloxacin appear to have commenced worldwide before the pandemic [22]. Due to the lack of routine AMS activity, it was felt that the full picture was not yet available to fully quantify the impact of COVID-19 on AMS and AMR.

Positive outcomes were also identified, with technology being increasingly used as a tool to facilitate stewardship, e.g., virtual meetings and ward rounds. The COVID-19 pandemic was also seen to break down barriers, resulting in increased collaboration. Other outcomes which respondents considered as positive were the increased introduction of novel biomarkers (e.g., procalcitonin) for differentiating viral and bacterial infections and better use of technology including virtual platforms and remote working. In addition, the use of hospital electronic prescribing systems facilitated AMS activities by antimicrobial pharmacists; allowing them to target their activities, for example identification of patients receiving excessive durations of antibiotics. There has also been a positive increase in multidisciplinary working where pharmacist contributions have been welcomed in an ever-changing evidence-based environment and pharmacists feeling valued for their contribution. Increased awareness of antimicrobial guidelines and improvements seen in infection prevention control have also been highlighted as likely to have a positive impact on AMS and resistance in the longer term. Innovation has also been key with some adapting services such as outpatient clinics and outpatient parenteral antibiotic therapy (OPAT) and changing current inpatient processes such as COVID-19 patients receiving a senior review more quickly. A virtual hospital model has been suggested as helpful to tackle the COVID-19 pandemic [23].

The majority of the respondents (73%) did not consider that there were non-COVID-19 related confounding factors that might have impacted AMS activities since the declaration of the pandemic in the UK (March 2020). For those that highlighted that there were confounding factors, these included staffing challenges within the infection team (i.e., lack of a stewardship lead microbiologist, antimicrobial pharmacists either not being in post or pharmacist AMS leads being redeployed, or needing to focus on clinical trials), drug shortages, increased post infection reviews for MRSA bacteremia and *Clostridioides difficile* cases. Positive confounding factors were also highlighted for example suspension of local meetings and national quality improvement schemes which allowed more time to review patients or target patients on high risk antibiotics.

Recently, Lynch et al. suggested that "AMS has become a casualty of the COVID-19 pandemic" [24]. In this survey we highlighted that while routine AMS activities were indeed a casualty of the COVID-19 pandemic, there were some opportunities presented and some positive outcomes. A recent review by Monnet and Harbarth reviewed the various determinants that may result in either an increase or, inversely, a decrease in AMR. They found that these determinants to be balanced [25]. However, the true impact of the

COVID-19 pandemic on AMR will not become clear for months, possibly years, when full surveillance data on antimicrobial use and resistance become available. In addition, the changes in AMR will vary depending on the settings, e.g., hospital types/units (ICU vs. other units) and facilities available in these settings; the reduction in usual hospital activities (such as routine surgery), availability of electronic prescribing and stock management systems; community vs. hospital settings, the number of COVID-19 cases as well as AMS activities that continue to be implemented through the pandemic.

#### *2.3. COVID-19 Specific Changes to the Management of Pneumonia*

Figure 2 illustrates the identified changes in AMS activity in the management of patients with community acquired pneumonia (CAP) during the COVID-19 surge in April 2020 against a baseline of 31 January 2020. It highlights for example that the pandemic led to increased use of procalcitonin in the management of respiratory tract infection both within and outside of the ICU, guiding antibiotic de-escalation and initiation. 53% (50/95) of respondents had updated guidelines on CAP before the release of the COVID-19 rapid guideline: managing suspected or confirmed pneumonia in adults in hospital by the National Institute for Health and Care Excellence (NICE) on 1 May 2020 [26]. There was also decreased AMS monitoring through audits such as the Start Smart then Focus (SSTF) studies, and the use of the CURB65 scoring system decreased slightly. The NICE guideline for CAP highlighted that CURB65 tool for CAP had not been validated for people with COVID-19. NICE guidance for the management of pneumonia in adults in hospital specified that there is insufficient evidence to recommend routine procalcitonin testing to guide decisions about antibiotics and encouraged centers already using procalcitonin tests to participate in research and data collection [27]. However, many organizations incorporated adherence to the CURB 65 scoring and advocated use of procalcitonin within their guidelines for management of COVD-19 patients.

**Figure 2.** Changes to AMS initiatives as a result of the COVID-19 surge (*n* = 95). Key for the *Y*-axis: **Procalcitonin-Start Abx:** procalcitonin use to inform starting antibiotics. **Procali\_NON-ICU-de-escalation:** Procalcitonin use in non-intensive care unit (ICU) settings to inform de-escalation and stopping antimicrobial stewardship activity. **Procalcitonin\_ICUde-escalation:** Procalcitonin use in ICU settings only to inform de-escalation and stopping antimicrobial stewardship activity. **Regular SSTF audits:** Regular (weekly or monthly) audit of review of antimicrobial prescriptions (Start Smart then Focus principles). **CURB65:** CURB 65 is specified in the guideline for assessing severity of Community Acquired Pneumonia **OxyStats4CAP**: Oxygen Saturations is specified in the guideline for assessing severity of Community Acquired Pneumonia. **NEWS score for CAP:** NEWS2 score is specified in the guideline for assessing severity of Community Acquired Pneumonia. **Other measures, CAP:** Other measures specified in the guideline for assessing severity of Community Acquired Pneumonia. **Radiological imaging/appearance (X-ray/CT/MRI):** Radiological imaging/appearance (X-ray/CT/MRI) to facilitate antibiotic review (de-escalating or stopping antibiotic) **Amended guideline pre NICE:** Amended antimicrobial prescribing guidance for COVID19 (pre NICE Guidance publications). 46

#### *2.4. Participation in COVID-19 Clinical Trials\**

At the time of the survey, almost all responding organizations (*n* = 95) were participating in the Randomised Evaluation of COVID-19 Therapy (RECOVERY) clinical trial (98%) with 75% and 58% participating in the Easy Access to Medicine Scheme (EAMS)– Remdesivir and Randomised, Embedded, Multi-factorial, Adaptive Platform Trial for Community-Acquired Pneumonia (REMAP-CAP) clinical trials respectively. Other trials and schemes taking place within responding Trusts included the since-discontinued expanded access program (EAP) for remdesivir (9%), Accelerating COVID-19 Research & Development (ACCORD-2) (8%), Safety and Efficacy of Tocilizumab in Patients With Severe COVID-19 Pneumonia (COVACTA) (6%), Platform Randomised trial of INterventions against COVID-19 In older peoPLE (PRINCIPLE) (4%), Adaptive COVID-19 Treatment Trial (ACTT) (4%), and Azithromycin versus usual care In Ambulatory COVID-19 (ATOMIC2) (2%). Two respondents reported that none of these trials were taking place in their organization. More than half of respondents also stated that their organization were part of the Early Access to Medicines Scheme (EAMS) for remdesivir.

As of 30 June 2020, there were 1142 clinical trials recruiting patients for COVID-19 management in hospitals or ICU settings globally with 62 of these registered for patients in the UK [28,29]. As perhaps expected, all organizations except two participated in the RECOVERY trial (RECOVERY; ISRCTN50189673), which was one of the two clinical trials globally that received the greatest media and scientific attention at the time. The other trial was the WHO "Solidarity" trial (ISRCTN83971151), which did not include sites in the UK. The lead role that many AMS teams had in management of these clinical trials may have contributed to the impact noted on routine AMS activities. Lack of resources for AMS because of re-allocation to COVID-19 planning and management, such as multiple trials, has also been highlighted by others [30]

#### *2.5. Update of Local Guidelines and Implementation of National Guidelines*

A third of responding organizations (UK-wide) had updated their local guidelines based on the NICE national guidelines for CAP and hospital-acquired pneumonia (HAP) published in April 2020; with just over 40% stating they were already aligned with the published guidelines whilst 12% of organizations stated they did not plan to update their guidelines based on national guidelines (Table 2).


**Table 2.** Organizations updating guidance in line with NICE recommendations (n = 95 respondents).

A high proportion of organizations reviewed or updated their CAP, HAP, or healthcare associated infections (HCAI) guidelines as part of COVID-19 planning or during the COVID-19 surge. Three quarters of organizations (77%) also developed dedicated COVID-19 infection management guidelines (Figure 3). Other activities which had been affected during the COVID-19 first wave are highlighted in Table 3.

affected during the COVID-19 first wave are highlighted in Table 3

NICE criteria on when to stop antibiotics been implemented/promoted (*n* = 95)

**Figure 3.** Percentage of organizations that reviewed their guidelines (*n* = 95 respondents).


36.8 27.4 24.2 5.3 6.3

A high proportion of organizations reviewed or updated their CAP, HAP, or healthcare associated infections (HCAI) guidelines as part of COVID-19 planning or during the COVID-19 surge. Three quarters of organizations (77%) also developed dedicated COVID-19 infection management guidelines (Figure 3). Other activities which had been


#### *2.6. Communication Methods within Secondary Care Settings (n = 95) 2.6. Communication Methods within Secondary Care Settings (n = 95)*

Digital methods were the most common methods of communication within organizations during the COVID-19 first surge (Table 4). A variety of methods were employed to keep staff up to date with current best practice in an ever changing evidence base including the local intranet, an antibiotic app, and an increase in virtual meetings and tele-Digital methods were the most common methods of communication within organizations during the COVID-19 first surge (Table 4). A variety of methods were employed to keep staff up to date with current best practice in an ever changing evidence base including the local intranet, an antibiotic app, and an increase in virtual meetings and teleconferences.



#### *2.7. Staff Changes during COVID-19 Epidemic*

COVID-19 has had a considerable impact on the roles and responsibilities of antimicrobial pharmacists (Table 5). More than half (57%) of antimicrobial pharmacists were seconded to other clinical roles within the pharmacy team and wider hospital with many having to undertake more than one role (Table 5). The main roles pharmacists were seconded to were ICU and general medicine. A small proportion of pharmacists also were seconded to roles outside pharmacy.


**Table 5.** New responsibilities for infection management pharmacy teams.

Antimicrobial pharmacists and antimicrobial pharmacy teams also undertook additional responsibilities as demonstrated in Table 6, with the highest number reporting additional responsibility for managing drug shortages, for both antimicrobial and nonantimicrobial medication. Managing supply of medication to patients with COVID-19 and providing PPE advice was also an additional role undertaken by many pharmacists during the initial COVID-19 pandemic.It is also evident that antimicrobial pharmacists had considerable involvement in the provision of infection prevention and control advice which may well have been part of the multidisciplinary ward round activities. The extension of antimicrobial pharmacists' roles beyond traditional duties/activities has also been highlighted by Goff et al. (2020) [31]. In addition, a recent review proposed recommendation for harnessing the AMS role of pharmacists and their teams in the context of COVID-19 and importance of continuing to advocate for AMS [32].

**Table 6.** Additional activities undertaken by AMS (*n* = 95) pharmacy teams.


More than half of respondents (56%) (Table 7) had to undertake additional training on their own time with only 37% being able to complete additional training and learning needs around COVID-19 within work. There was no training available for COVID-19 during the first surge as it was new to all. More than three months after the start of the pandemic, no hospital was able to provide formalized mandated training on COVID-19; this is likely to change as understanding of COVID-19 progresses and when a vaccine becomes available, which will require large scale training before administration. Learning on the job, reading papers being published from China and joining various webinars were the typical opportunities available during the surge. Our survey results showed that 92% of the respondents undertook this learning in their own time or as on the job training. This is further emphasized by the respondents in the increase in multidisciplinary team (MDT) working as the teams were learning as a team.

**Table 7.** Opportunity for additional learning undertaken during the COVID-19 Pandemic (*n* = 95).


#### **3. Materials and Methods**

A quantitative survey-based approach was adopted using a 20-item questionnaire developed from the literature on AMS in the context of COVID-19 and consensus from infection/antimicrobial pharmacists (Supplementary Materials 1). Demographic data on the organization of each respondent included: which UK country, type of hospital (teaching, district/general, larger organization with multiple hospital sites, specialist), number of COVID-19 cases up until 31 May 2020, and the role of respondents. The survey was reviewed and refined by discussion with a working group comprised of members from the UKCPA Pharmacy Infection Network (UKCPAPIN), Association of Scottish Antimicrobial Pharmacists, All Wales Antimicrobial Pharmacists Group, and Northern Ireland Regional Antimicrobial Pharmacists Network. The survey was then hosted on Google Forms, a web-based survey platform, then pilot tested with five individuals across the UK. Following this initial testing, the final survey was disseminated by UKCPAPIN, Association of Scottish Antimicrobial Pharmacists, All Wales Antimicrobial Pharmacists Group, and Northern Ireland Regional Antimicrobial Pharmacists Network. The survey was also promoted via UKCPAPIN social media channels, and antimicrobial pharmacists/local network WhatsApp groups.

#### *3.1. Respondent Eligibility*

Pharmacy infection professionals (pharmacists, pharmacy technicians, and dispensers) across all UK secondary care and acute institutions/hospitals were the intended audience for the completion of the survey. Participation was voluntary, with the questionnaire being open for responses over a 2-week period (4 to 10 June2020).

#### *3.2. Data Management*

All data were held securely and in line with the General Data Protection Regulation 2016/679 (17). Study approval was also obtained from the UKCPAPIN

#### *3.3. Data Analysis*

Descriptive statistics on the frequency distributions and percentages were used to analyze the responses. Data were analyzed using Microsoft® Excel (2010). The survey tool is provided as Supplementary Information 1.

#### **4. Conclusions**

The findings of our survey provides, for the first time, quantitative and qualitative data on the impact of the COVID-19 pandemic on AMS activities undertaken across the UK.

Key stewardship activities that were negatively impacted include AMS ward rounds, MDT AMS meetings, quality improvement, audits, and education/training. The long-term impact of COVID-19 and the full extent of reduced AMS activities, as well as the impact of this on AMR, is unlikely to become clear for months and possibly years. We will know more when surveillance data on antimicrobial use and resistance become available, which will likely vary depending on setting and incidence of COVID-19 within each health system. Monitoring the impact of any harm caused by reduced AMS activities such as *C. difficile*, increased multidrug resistant organisms, and increased hospital admission or length of stay and mortality further reinforced the need to preserve this vital activity in future pandemic or COVID-19 surges. An additional survey to compare the impact of AMS activity during the first wave and subsequent waves or overall would add to the evidence.

Positive impacts identified within participating organizations highlighted through the survey (linked to measures to control the pandemic) included the increased acceptance of using procalcitonin to discriminate between viral and bacterial pneumonia-reducing inappropriate antimicrobial use in viral pneumonia patients in the post pandemic era. Technology was embraced to bring some of the historic AMS activities into the digital age and should be further harnessed and promoted. Using virtual platforms for education and training, multidisciplinary team meetings, AMS meetings, AMS rounds, and virtual clinics could also continue to be encouraged for AMS activities.

While the impact of the COVID-19 pandemic on AMS activity has been quantified, the psychological impact of additional roles, secondment to other specialties, and additional responsibilities that antimicrobial pharmacists have undertaken has yet to be evaluated and may form the basis of further studies. It is important that those who lead on AMS continue to have protected time to focus on AMS during current or future pandemics.

As expected wide scale participation in clinical trials for treatments of COVID-19 have been observed across the whole of the UK. The large number of participants has contributed to and will continue to progress the understanding of treatment options for COVID-19 for the benefit of future patients.

The legacy of innovation, use of technology, and increased collaboration/links with non-infection specialists, which the pandemic made necessary, could in fact strengthen AMS in the post-pandemic era and presents further opportunities for development of the antimicrobial stewardship roles. In addition, the networking and support network that has been developed will continue to support pharmacists in this role in future.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2079-638 2/10/2/110/s1. Survey Form UKCPAPIN COVID-19 Survey.

**Author Contributions:** Conceptualization, D.A.-O.; data curation, D.A.-O.; formal analysis, D.A.-O. and F.K.; methodology, D.A.-O., J.U., M.L., T.Y., A.C. (Aneeka Chavda), R.P., A.S., C.G., A.C. (Alison Cockburn), T.V., C.P., N.R., and A.B.; project administration, D.A.-O.; writing–original draft, D.A.-O., F.K., and S.H.; writing–review and editing, D.A.-O., F.K., J.U., M.L., T.Y., A.C. (Aneeka Chavda), R.P., A.S., C.G., A.C. (Alison Cockburn), T.V., C.P., and A.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project received no external funding.

**Institutional Review Board Statement:** The survey was conducted as part of service improvement of the Pharmacy Infection Network; no ethical approval was required

**Informed Consent Statement:** The following statement was included on the survey form and needed to be completed before completion of the survey "The survey is completely voluntary and part of UKCPA Pharmacy Infection Network support provision. You have the right to refuse to answer questions or withdraw at any time. By proceeding to the next page: I consent to UKCPAPIN collecting and using the information about me that I voluntarily provide for the purposes of the survey I have read, understand and agree to the information provided above".

**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 organisation privacy.

**Acknowledgments:** We are grateful to all infection colleagues across the NHS in England, Scotland, Wales and Northern Ireland who replied to the survey. The views expressed are not those of the NHS or their respective organizations. We also acknowledge Elizabeth Beech who made important contributions to the design of the survey.

**Conflicts of Interest:** The authors declare no conflict of interest. The organizations the authors of this manuscript have their substantive infection lead roles in are listed: D.A.-O. (Public Health England); F.K. (NHS Lanarkshire, Scotland); S.H. (Chelsea and Westminster Hospital NHS Trust); J.U.(Great Western Hospitals NHS Foundation Trust); M.L. (Royal Free London NHS Foundation Trust); T.Y. (St George's University Hospitals NHS Foundation Trust); A.C. (Alison Cockburn) (NHS Lothian, Scotland); A.S. (Portsmouth Hospitals University NHS Trust); R.P. (Dartford And Gravesham NHS Trust); C.G. (Western Trust, Northern Ireland); A.C. (Aneeka Chavda) (Imperial College Healthcare NHS Trust); T.V. (West Hertfordshire Hospitals NHS Trust); C.P. (Aneurin Bevan University Health Board, Wales); N.R. (Public Health Wales); A.B. (Queens University, Belfast).

#### **References**


#### *Article*

#### **A One Health Approach to Strengthening Antimicrobial Stewardship in Wakiso District, Uganda**

**David Musoke 1,\* , Freddy Eric Kitutu <sup>2</sup> , Lawrence Mugisha <sup>3</sup> , Saba Amir <sup>4</sup> , Claire Brandish <sup>5</sup> , Deborah Ikhile <sup>6</sup> , Henry Kajumbula <sup>7</sup> , Ismail Musoke Kizito <sup>8</sup> , Grace Biyinzika Lubega <sup>1</sup> , Filimin Niyongabo <sup>1</sup> , Bee Yean Ng <sup>5</sup> , Jean O'Driscoll <sup>5</sup> , Kate Russell-Hobbs <sup>5</sup> , Jody Winter <sup>9</sup> and Linda Gibson <sup>6</sup>**


Received: 29 August 2020; Accepted: 19 October 2020; Published: 31 October 2020

**Abstract:** Antimicrobial stewardship (AMS), as one of the global strategies to promote responsible use of antimicrobials to prevent antimicrobial resistance (AMR), remains poor in many low-and middle-income countries (LMICs). We implemented a project aimed at strengthening AMS in Wakiso district, Uganda using a One Health approach. A total of 86 health practitioners (HPs), including animal health workers, and 227 community health workers (CHWs) participated in training workshops, and over 300 pupils from primary schools were sensitized on AMR, AMS, and infection prevention and control (IPC). We further established two multidisciplinary online communities of practice (CoPs) for health professionals and students, with a current membership of 321 and 162, respectively. In addition, a Medicine and Therapeutics Committee (MTC) was set up at Entebbe Regional Referral Hospital. The project evaluation, conducted three months after training, revealed that the majority of the HPs (92.2%) and CHWs (90.3%) reported enhanced practices, including improved hand washing (57.3% and 81.0%, respectively). In addition, 51.5% of the HPs reported a reduction in the quantity of unnecessary antibiotics given per patient. This project demonstrates that AMS interventions using a One Health approach can promote understanding of the prudent use of antimicrobials and improve practices at health facilities and in communities.

**Keywords:** antimicrobial resistance; antimicrobial stewardship; community health workers; health practitioners; infection prevention and control; multidisciplinary; one health; Uganda; UK

#### **1. Introduction**

Antimicrobial resistance (AMR) poses a global public health concern that relates to humans, animals, and the environment. Several factors contribute to the escalation of AMR, including inappropriate prescription, misuse and overuse of, and lack of effective stewardship of antimicrobials [1]. In 2015, the World Health Assembly endorsed a global action plan to tackle the worldwide problem of AMR [1]. This plan has at its core the use of a One Health multi-sectoral approach, and calls for collaboration and co-ordination globally and locally. In response, Uganda developed and released its 5-year AMR National Action Plan (NAP) in 2018, which sets out a framework of actions to address the undertakings across the country [2]. The NAP acknowledges limited awareness of AMR and data pertaining to antimicrobial use, rising rates of AMR in the country, and the comprehensive steps that need to be taken to contain and control this threat to global health [2–4].

One Health refers to a collaborative, co-ordinated, and multidisciplinary approach to ensure the health and wellbeing of humans, animals, and the environment across different spatial levels [5]. A One Health approach is necessary as AMR is an ecological challenge that is affected by the interrelations between humans, animals and the environment [5]. The implementation of interventions and actions of multiple actors towards the optimization of antimicrobial use is known as antimicrobial stewardship (AMS) [6]. Despite increasing evidence for the need for a multidisciplinary approach to tackle AMR, the use of One Health in addressing AMS challenges has been minimal. A recent systematic review showed that there is a dearth in the practice and implementation of AMS programs across Africa [7]. Whereas there is increasing evidence on challenges affecting AMS in Uganda, there is little literature on antimicrobial use in animals and its relationship to human health and the environment [8–10]. This therefore calls for more interventions to use a multidisciplinary approach to improve AMS across the country as stipulated in the NAP [2].

With support from the Commonwealth Partnership for Antimicrobial Stewardship (CwPAMS) scheme [11], an initiative of Commonwealth Pharmacists Association (CPA) and Tropical Health and Education Trust (THET) under the Fleming Fund of the UK Department of Health and Social Care (DHSC), our health partnership aimed to strengthen AMS in Wakiso district, Uganda. The focus of the project was on capacity building, multidisciplinary stakeholder engagement, and knowledge exchange using a One Health approach. The project drew on a multidisciplinary partnership and expertise from: the Schools of Social Sciences, Animal, Rural and Environmental, and Science and Technology at Nottingham Trent University (NTU); Buckinghamshire Healthcare NHS Trust (BHT); Colleges of Health Sciences, and Veterinary Medicine, Animal Resources and Biosecurity at Makerere University (Mak); and Entebbe Regional Referral Hospital (ERRH) in Uganda.

It is acknowledged that antimicrobials are used widely in both humans and animals, and are commonly present in the environment [12], hence the need for a broader One Health approach in addressing AMR. Although most AMS interventions have been health facility based, a large proportion of antimicrobials are used in the community, both as part of outpatient care [13] and integrated community case management (iCCM) for the treatment of childhood illnesses. In iCCM, community health workers (CHWs) are involved in the diagnosis of malaria, diarrhea, and pneumonia among children under five years of age. CHWs are the frontline health cadre at the community level in many low- and middle-income countries (LMICs), including Uganda, and so, they have a key role in ensuring proper use of antimicrobials in their communities. Therefore, this project was designed with interventions at both health facility and community levels to ensure wide reach and impact. It was also planned for knowledge to cascade from healthcare professionals into the wider communities through the CHWs. Specifically, the project aimed to: strengthen AMR awareness and upskill human and animal health practitioners (HPs) in AMS and infection prevention and control (IPC); utilize a training of trainers approach with the HPs and CHWs to improve community-wide awareness of AMR; establish communities of practices (CoPs) for sustainable engagement and resource sharing to support AMS; and facilitate knowledge exchange and sharing of best practice between Uganda and

UK. In this paper, we describe the main activities and achievements of the project, including results from the evaluation of the HPs and CHWs who were involved in the training workshops.

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

#### *2.1. Project Site and Setting*

The 15-month project, as part of the CwPAMS scheme, was implemented in Wakiso district, central Uganda. Wakiso district has a total surface area of 2807.75 square kilometers, and a population of 2,007,700 people at an estimated growth rate of 4.1% [14]. ERRH, located in Entebbe municipality, Wakiso district, is the health facility where the main project interventions were implemented. ERRH has a 200-bed capacity, serving approximately 300 to 400 out-patients per day. Services offered at the hospital include but are not limited to: dental, pharmacy, peadiatrics, radiology, laboratory, maternity, maternal and child health, general surgery, internal medicine, and orthopedics. The hospital is led by a medical director, and approximately 10% of its staff are prescribers, including medical officers, dental surgeons, and clinical officers. ERRH serves a population of over 300,000, including the community in Entebbe municipality and neighboring areas, some being islands on Lake Victoria [14]. The community component of the project was implemented in Busiro South Health Sub District (HSD) in Wakiso district, which is comprised of three town councils (Kajjansi, Kasanje, and Katabi) and one sub county (Bussi). The HSD has a population of approximately 243,420 people. With a high number of households in Busiro South and the wider Wakiso district engaged in poultry and livestock farming, antimicrobials are used extensively [15,16]. The animal health workers involved in the project worked in Entebbe municipality either with the local government or as private practitioners. These health workers carry out diagnosis and treatment of animals mainly in the community. The CHWs in Busiro South HSD involved in the project not only treat childhood illnesses of diarrhoea, pneumonia, and malaria but also participate in educating the community on key public health issues, including AMR.

#### *2.2. Project Team and Reciprocal Visits*

This project was delivered as part of a 10-year international partnership between NTU, UK, and School of Public Health at Mak, Uganda. The partnership co-opted a multidisciplinary team for delivery of the project. This was essential due to the nature of the multifactorial challenges of AMR in humans, animals, and the environment. From the UK, specialist antimicrobial pharmacists and a medical microbiologist from BHT, a microbiology lecturer from School of Science and Technology, and an animal specialist from the School of Animal, Rural and Environmental Studies (ARES) at NTU took part in the interventions. In Uganda, project partners included: public health specialists, pharmacists including a clinical pharmacist, a veterinary doctor, and a microbiologist from Mak with support from health professionals from the Ministry of Health (MOH), ERRH, Wakiso district local government, and Entebbe Municipal Council. As part of the project, reciprocal visits between members from the UK and Uganda for planning, scoping, implementation, knowledge exchange, and sharing of best practices were held.

#### *2.3. Project Planning and Stakeholder Engagement*

The multidisciplinary project team conducted several meetings both physically and virtually before, during, and after implementation. The virtual meetings were held monthly using Skype and were attended by partners from both the UK and Uganda. The meetings facilitated project planning, implementation, monitoring, and evaluation as well as keeping track of the achievement of set goals. Other day-to-day communication to support implementation and timely completion of project tasks was achieved using a WhatsApp group. A google drive account was also set up for the team to access project documentation, such as previous meeting minutes, photos, and training materials. These communication avenues were invaluable for tracking progress of ongoing project activities and enhancing team work. Before and during project implementation, the project team held several meetings and engagements with various stakeholders in Uganda to ensure ownership, buy-in, and

participation in planned activities. These stakeholders included government ministries (MOH and Ministry of Agriculture, Animal Industry and Fisheries—MAAIF), governmental parastatals (such as the National Drug Authority), professional associations (the Pharmaceutical Society of Uganda), training institutions (Mak), local governments (Wakiso district), health facilities (ERRH and lower level health facilities, such as health centre IIs, IIIs and IVs), local leaders (such as local council chairpersons), and the general community. The project team specifically engaged the MOH Technical Working Committee (TWC) on AMS, optimal access, and use, which is mandated to provide technical oversight of all AMS activities in the country. This engagement involved collaborative planning as well as regularly providing project updates in the TWC meetings and getting feedback that informed ongoing activities.

#### *2.4. Enhancing Capacity of Health Practitioners, Community Health Workers, and School Pupils*

The project held training workshops for HPs from both human and animal health to create awareness on AMR/AMS/IPC using a One Health approach. The workshops targeted selected HPs from government health facilities, including ERRH, as well as animal health workers in Wakiso district. The selection of HPs involved in the workshops was done in consultation with contacts at ERRH, Entebbe Municipal Council, and Wakiso District Health Office. Using the 'training of trainers' model, selected trained HPs were involved in training workshops for CHWs in AMR/AMS/IPC also using a One Health approach. The CHWs were from Kajjansi town council in Wakiso district where earlier NTU–Mak partnership interventions had been implemented [17]. All CHWs in the town council available at the time were involved in the workshops. In addition to the training of HPs and CHWs, the project also sensitised pupils in two primary schools in Wakiso district (St. Theresa and Kawotto Saviours Primary Schools) on AMR/AMS/IPC. St. Theresa's is a government school in Entebbe municipality while Kawotto Saviours is a private school in Kajjansi town council, both in Wakiso district. These schools were selected in consultation with key stakeholders in Wakiso district.

#### *2.5. Establishment of One Health Communities of Practice and University Student Engagement*

The project set up two online CoPs involving individuals from human health, animal health, and the environment. The first CoP was for health professionals, including HPs, researchers, policy makers, and academics, while the second targeted undergraduate students of Mak from various disciplines, including environmental health, veterinary medicine, pharmacy, biomedical sciences, social sciences, and computer sciences. In addition to the CoPs, multidisciplinary students at Mak and microbiology students at NTU were involved in various activities to promote AMR/AMS/IPC, including seminars, webinars, and competitions. The competitions, which were to design AMS messages using innovative approaches, were held to commemorate WAAW 2019. The winners of the competition were recognized as a form of motivation to continue being antibiotic guardians in their respective settings.

#### *2.6. Establishment of a Medicine and Therapeutics Committee*

Working closely with the hospital pharmacist, the project supported the establishment of a MTC at ERRH. This included appointing a multi-disciplinary team, in accordance with the MOH MTC manual [18]. The aim of the MTC is to support the safe and effective use of medicines through evaluation of usage, and to develop guidelines and protocols for medicine prescription and administration, as well as other health commodity related activities at the hospital. The MTC at ERRH has 12 members and three sub-committees reflecting the main functions: supply chain; pharmacovigilance, and AMS. During the course of the project, the MTC held three meetings to elect members, establish roles and responsibilities, develop work plans including a procurement plan for the 2020/21 financial year, and define standard IPC practices at the hospital. The MTC was specifically involved in: selection of medicines to be used; monitoring and ensuring rational use of medicines as per the standard treatment guidelines; development of draft treatment protocols; as well as developing Standard Operating Procedures (SOPs) during the management of COVID-19 patients.

#### *2.7. Project Evaluation*

The final project evaluation involved assessment of practices of the beneficiary HPs and CHWs, which was carried out three months after their respective training workshop. Specifically, the evaluation was aimed at establishing any changes in practices of the HPs and CHWs related to AMR/AMS/IPC following the workshops conducted as part of the project. The practices assessed among the HPs included: increased use of the Uganda Clinical Guidelines (UCG) when prescribing antimicrobials; increased diagnosis based on laboratory results; improved handwashing; and more patient guidance and counselling when they do not require antimicrobials. For the CHWs, practices, such as improved hand washing with soap as well as increased community sensitization on avoiding self-medication, consulting animal health professionals whenever animals were ill, and avoiding the use of human prescribed medicines in animals including poultry, were assessed. The evaluation among the HPs used a self-administered questionnaire while for the CHWs, a researcher-administered questionnaire was used. In addition, key informant interviews using telephone were held with selected key individuals from the various stakeholders involved in the project. The interviews included HPs and CHWs who participated in the workshops, particularly those with leadership roles, CoP members, Wakiso district health office staff, and facilitators of the training. In this paper, we present a summary of the key findings from the evaluation. The diagrammatic summary of the project implementation is shown in Figure 1.

**Figure 1.** Summary of the project implementation.

#### **3. Results**

#### *3.1. Reciprocal Visits*

The project exchange visits lasted between one and two weeks depending on the timing of activities as well as the availability of the team members. The timing of these visits was crucial for the different teams to appreciate the level of, and approaches to AMS in the host country. The first UK team visit to Uganda in April 2019 was organised as a scoping visit. This was necessary to appreciate AMS activity in the country, including challenges, knowledge gaps, areas of sub-optimal antibiotic use, development needs at ERRH, and animal health AMS concerns to inform the project activities, including training of HPs and capacity support to ERRH. During this visit, several meetings were held with the ERRH administration, Wakiso district health and veterinary staff, MOH Department of Pharmaceuticals and Natural Medicines, and College of Veterinary Medicine, Animal Resources and Biosecurity at Mak. The second exchange visit was that of the ERRH pharmacist to the UK in June 2019. The pharmacist was involved in AMS discussions involving multi-disciplinary teams from BHT and NTU (City, Brackenhurst, and Clifton campuses). While at the Brackenhurst campus, which houses the School of Animal, Rural and Environment Sciences at NTU, the pharmacist learnt about the various animal projects there, including research on AMS. The pharmacist also spent some time in the microbiology department at the Clifton campus where he made a presentation to share experiences of AMS at ERRH and generally in Uganda among faculty. At BHT, the pharmacist spent time seeing AMS in practice on clinical ward rounds and in the pathology laboratory, and visited the medical microbiology laboratory where he gave a presentation to scientists. The pharmacist met with the medical director to discuss a potential memorandum of understanding with ERRH and how future work may be undertaken to support other collaborative activities. In the UK, the pharmacist was also involved in further project planning and work shadowing to learn more about AMS practices. The third exchange visit was of the UK team in September 2019 that involved a series of activities including: facilitation of three two-day AMS trainings among human and animal HPs; AMS seminars at Mak; and visits to primary schools to sensitize them on AMR and to launch an awareness competition. The final visit was by a Ugandan clinical pharmacist and lecturer at Mak in September 2020. This visit focused on AMS / AMR knowledge exchange, potential expansion of partnership work, discussions on final project activities including evaluation and dissemination, as well as exploring avenues for future collaboration between Uganda and the UK. In total, six UK members, including antimicrobial pharmacists, microbiologists, and an animal health specialist from NTU and BHT, travelled to Uganda over the two visits, while two Uganda pharmacists visited the UK as part of the project.

#### *3.2. Training of Health Practitioners*

In September 2019, the project team held training workshops for a multidisciplinary group of 86 health professionals from human and animal health on AMR/AMS/IPC. Among the trained HPs, 56 were from ERRH, 20 were from lower level health facilities in Wakiso district (Nsaggu HC II, Nakawuka HC III, Kajjansi HC IV, Zzinga HC II, Kitala HC II, Bussi HC II, and Kasanje HC III), and 10 were animal health professionals working within Entebbe municipality. Whereas all the human health HPs were selected from government health facilities, some of the animal health HPs were in private practice. The workshops for the HPs from both animal and human health were greatly appreciated by the participants as they shared experiences among themselves, and realised the similarities and close linkage between AMS among their professions. All HPs trained were involved in either prescription, administration, issuance of antimicrobials, or related clinical work. They included pharmacists, medical doctors, laboratory technicians, clinical officers, nurses, and veterinary practitioners. The two-day training workshops, held in Entebbe, were facilitated jointly by the UK and Uganda members of the project team. This ensured the local context of material and applicability was delivered during the training. The UK trainers included three antimicrobial pharmacists, two microbiologists, and an animal health specialist, while the Uganda team included three pharmacists, a veterinary doctor, and two

environmental health scientists. The workshop sessions included: introduction to AMR/AMS/IPC; the World Health Organization (WHO) AMR competency framework for human health; prudent antibiotic use, actions, and barriers in human and animal health; use of UCG incuding how to use the Microguide app that hosts these guidelines provided by CwPAMS; the One Health approach; hand hygiene; and sharing of gained knowledge with others using the capability, opportunity, and motivation behaviour (COM-B) change model [19]. The workshop sessions were interactive and engaged participants in order to facilitate adult learning. Specifically, sessions included a pre- and post-assessment, group discussions, e-Bug AMR balloon experiment [20], hand washing demonstration using the Glow Germ Gel kit™ [21], case studies from both human and animal health, interactive games, sharing of past experiences, and a certificate awarding ceremony. From the pre- and post-training assessment, there was an improvement in the knowledge levels of the HPs after the trainings. Out of the 80 HPs who participated in the post assessment, 64 (80%) correctly defined AMR compared to 63.9% (53/83) who had done so during the pre-assessment. In the post-training evaluation, 73.8% (59/80) of the HPs stated that inadequate hand hygiene is one of the contributing factors of AMR compared to 21.7% (18/83) who did so during the pre-training assessment.

#### *3.3. Training of Community Health Workers*

A total of 227 CHWs from Kajjansi town council took part in training workshops conducted by a team of three HPs who were among those trained in the project as part of the 'training of trainers' model. The one-day workshop for CHWs, in groups of approximately 30, was conducted using interactive and engaging sessions similar to those used with HPs. Given that CHWs are involved in treatment of only three diseases (malaria, diarrhoea, and pneumonia) for infants less than five years under iCCM, and health promotion on key public health issues, their training was largely focused on creating awareness on AMR/AMS/IPC in their communities. The workshops included sessions on: introduction to antimicrobials and AMR; AMS at the community level; prevention of AMR in animals and humans; water, sanitation, and hygiene; food hygiene and safety; and IPC in communities. Similar to the workshops of HPs, a pre- and post-training assessment was undertaken for the CHWs. There was a notable improvement in knowledge of the CHWs in the post-training assessment in comparison with the pre-training survey. Out of the 212 CHWs who participated in the post-training assessment, 97.1% (206/212) reported that microorganisms can fail to respond to antimicrobials compared to 49.3% (111/225) in the pre-training assessment. In addition, 96.2% (204/212) of the CHWs reported in the post-training assessment that antimicrobials dumped in the environment can lead to AMR compared to 33.3% (75/225) in the pre-training assessment. Furthermore, 97.6% (204/212) of the CHWs agreed in the post-training assessment that inappropriate use of antimicrobials in livestock can lead to AMR compared to 36.4% (82/225) in the pre-training assessment. In the post-training assessment, all the CHWs (100%) felt that they were knowledgeable enough to educate their communities on preventing AMR through improved AMS and IPC following the training.

#### *3.4. Increasing Awareness on AMS to Primary School Pupils and University Students*

The project introduced and created awareness on AMR/AMS among over 300 pupils at St. Theresa and Kawotto Saviours primary schools in Wakiso district. These sessions were attended by pupils mainly in the upper classes of the schools (primary 5 to 7) as well as their teachers. These schools were twinned with two schools in the UK (Wingrave Church of England and Longwick Church of England primary schools). Pupils in both the Ugandan and UK schools also participated in a competition to develop AMR/AMS/IPC messages in commemoration of World Antibiotic Awareness Week (WAAW) 2019, and the winners received various awards. Winners in Uganda and the UK were selected for best poster / song / performance. Award ceremonies were held at the schools in Uganda and the UK to share the winning material among pupils, teachers, and parents to reinforce the messages and illustrate the importance of the global issue of AMR. Winners (four from Uganda and four from the UK) received an award of 25 GBP, a young antibiotic guardian t-shirt, and a certificate of appreciation

for their participation. As part of the twinning of the Uganda and UK schools, ideas were shared on future collaboration, which will be explored in future. An initial activity carried out as part of the twinning was pupils from the Uganda schools writing pen pal letters to their UK counterparts and vice versa, which was well received by both groups and the administration of the schools. At Mak, the project organised three seminars and workshops on AMR / AMS /IPC, and interactions with NTU students. The seminars were attended by over 120 students from the UK and Uganda from different programs, including environmental health, pharmacy, microbiology, and environment. The project also launched an AMR awareness competition among Makerere University Environmental Health Students' Association (MUEHSA) and NTU students for them to design appropriate messages for community sensitization. During the World Antibiotic Awareness Week (WAAW) in November 2019, the project held an award giving ceremony in which the four winners of the competition from Mak and NTU each received a cash prize of 25 GBP as well as a certificate of appreciation.

#### *3.5. Establishment of One Health AMS Communities of Practice*

Through stakeholder engagement at initiation of the project, the need for two online CoPs, one for health professionals and the other for students (as opposed to the earlier planned one), was identified and established. The aim of the CoPs was to provide a platform for sharing resources, opportunities, and materials on AMR, AMS, and IPC targeting both human and animal health, as well as enhance sustainability as these would continue after the project duration. The *Antimicrobial Stewardship, Optimum Access and Use in Uganda* CoP for health professionals is hosted by the MOH TWC on AMS, optimum access, and use. The *Students for Antimicrobial Stewardship* CoP is a Facebook group that was formed after the realisation of the need for students to work together, in a multidisciplinary setting, at an early stage in their careers to tackle AMR. This platform is managed by five AMR champions from the following schools at Mak: veterinary medicine; biosecurity; biotechnical, and laboratory science; health sciences; and public health. Currently, the students' CoP has 162 members while the one for health professionals has 321 members, with membership of both groups steadily increasing. The health professionals' CoP has sent out over 50 emails with resources, opportunities, and other materials concerning AMR/AMS from Uganda, the UK, and globally. One example of how this has changed engagement with this work is that some members of the health professionals CoP have submitted abstracts to conferences, which they learnt about from the online platform. The students CoP has sent over 20 messages on opportunities for students to participate in. These opportunities have included attending webinars and conferences, such as the 4th National Conference on AMR held in Kampala, Uganda in 2019. In addition, students on their CoP have appreciated the importance of working in multi-disciplinary teams to tackle AMR, which they are likely to utilize during their future professional work.

#### *3.6. Evaluation of the Trained HPs and CHWs*

From the project evaluation, there was a positive change in practices among the HPs and CHWs following the training. Out of the 77 HPs who participated in the evaluation, 68 (88.3%) stated that they had adopted new practices from the project training. Out of the 68 HPs who adopted new practices, 39 (57.3%) reported improved handwashing, over half 36 (52.9%) reported an increase in use of the UCG when prescribing antimicrobials, and 35 (51.5%) reported a reduction in the quantity of unnecessary antibiotics given per patient. Among the 77 HPs, 48 (62.3%) reported having faced challenges when attempting to become an antimicrobial steward in their setting. These challenges included stock out of drugs, 29 (60.4%); lack of personal protective equipment (PPE), including gloves and masks, 19 (39.6%); and insufficient laboratory capacity, 17 (27.1%) (Table 1).


**Table 1.** Evaluation results of health practitioners.

\* Multiple response question; \*\* Other practices included advising against self-medication, encouraging use of organic feeds for poultry and livestock, health education on adherence to drugs, waste segregation, and training other HPs who did not attend the training; \*\*\* Other challenges included poor attitude of fellow HPs, patients demanding for antibiotics, self-medication by patients, and lack of flexibility among some prescribers.

From the qualitative evaluation of the project, the HPs reported that they were using the UCG more during the prescribing of antimicrobials, and they had also reduced the quantities of antibiotics given per patient when appropriate to do so. Improved prescription practices among the HPs also led to improved availability of antimicrobials at the various health facilities as demonstrated in the quotation below.

*"At the facility nowadays, I only give out amoxicillin where necessary as per the guidelines. I do not just give out antibiotics anymore. For this reason, I am now able to save amoxicillin tablets for patients who really need them, and I also give out the medication in the right dose. This has helped me reduce on the number of times I go to look for medicines from other facilities due to reduced stock-outs at my facility."* Health worker, Bussi health centre II

Among the 226 CHWs who participated in the evaluation, 204 (90.3%) reported improved practices attributable to the training. The majority, 183 (81%), of the CHWs reported increased handwashing with soap, 175 (77.4%) had encouraged community members to improve personal hygiene and general sanitation, 151 (66.8%) had encouraged community members to take the full dose of their prescribed medication, while 130 (57.2%) had encouraged farmers to always consult veterinary professionals whenever their animals were ill. Following the training, 69 CHWs (30.5%) reported having reached between 50 to 100 community members and health educated them on AMR, while 27 (12%) reached over 100 community members (Table 2).


**Table 2.** Evaluation results of community health workers.


**Table 2.** *Cont*.

\* Multiple response question. \*\* Others included advising against sharing of drugs, proper drug storage, advising against eating of dead animals, and burying dead animals. \*\*\* Other challenges included poor attitude of community members, poverty, and disrespect for CHWs.

Qualitative evaluation among the CHWs confirmed improved practices amongst them regarding the prevention of AMR in the community, particularly regarding animal husbandry, such as observing antibiotic withdrawal periods among animals before consuming their products, not eating deceased animals that were recently on treatment, reduction of self-prescription for animals, and reduction in use of human-prescribed antimicrobials among animals as mentioned in the quotations below.

*"Before the training, I was among the people who used to slaughter sick chicken which were under treatment or those that had died while receiving medication. During the training, I learnt that we should never slaughter sick animals undergoing treatment. I now also know that to slaughter a chicken recently on treatment for consumption, you must wait for 7 or more days so that the medication is no longer present in its body hence not consuming small doses of the drug from the chicken which can contribute to AMR."* Female CHW, Kajjansi Town Council

*"For us, we thought that medicine that has been prescribed for humans could be used the way we wanted to treat animals especially antiretroviral drugs which we used to give pigs, and amoxicillin capsules to chicken. However, at the training I learnt that it is very dangerous to give human medicine to animals, and we should always call the veterinary doctor to treat our animals and prescribe the medicine for them, other than treating the animals ourselves."* Male CHW, Kajjansi Town Council

#### **4. Discussion**

Improper use of antimicrobials including non-compliance with guidelines is a contributing factor to AMR [4], hence the need for interventions targeting improvement of access to and appropriate use of antimicrobials. There is limited data on AMS programs in Africa, including Uganda [7,22], hence our project provides an important contribution to the existing body of knowledge on how to address the growing burden of AMR. The need for a One Health approach to promote AMS is well established internationally [5], and is translated into the local Ugandan context through the NAP [2]. The One Health approach used in the project was made possible through the multidisciplinary nature of our team with expertise in human and animal health, and environmental health. The strength of this team was fostered through reciprocal relationships and engagement with a variety of in-country stakeholders, including policy makers, such as the MOH and MAAIF. The project was delivered through multiple parallel interventions, including the training of HPs and CHWs, the creation of AMS awareness among university students and school pupils, as well as the establishment of CoPs and a hospital MTC. Our project is one of the few that have delivered multiple interventions at both health facility and community levels using a One Health approach in Uganda so provides a good contribution to the NAP in the fight against AMR.

In line with WHO recommendations, our training approach focused on both human and animal HPs [9]. In addition, the training of HPs on AMR/AMS/IPC took a One Health approach where an emphasis was given to the use of antimicrobials in humans and animals, and their link to the environment. This was necessary to truly embed the principles of One Health into the training component of our project. Indeed, with a multidisciplinary team from Uganda and the UK facilitating the training, the goal of ensuring One Health was achieved. The project was able to translate AMS principles across various primary health care levels in Uganda as the training involved HPs from ERRH and lower level health facilities. Whereas these lower level health facilities in Uganda (health centre IIs, IIIs and IVs) are often ignored for AMS and other interventions, they provide health care to a good portion of the population, particularly in rural areas that have limited access to hospitals [23]. The use of UK trainers in addition to those from Uganda was of benefit for both countries. These trainers brought their unique expertise around infection prevention and the UK's principles of AMR and AMS practice. On the other hand, the UK trainers were exposed to the issues of AMS and AMR in a developing country context, and community-based approaches of tackling global health challenges, particularly AMR. The project also facilitated bi-directional learning between the two countries, which can inform other health partnerships. The use of participatory training techniques, such as demonstrating handwashing using the Glow Germ Gel Kits, facilitated learning and knowledge retention as opposed to traditional didactic methods.

In addition to the training of HPs, our project trained CHWs on AMR/AMS/IPC using the 'training of trainers' model. AMS programs are largely hospital based and have predominantly focused on HPs, such as laboratory scientists, nurses, clinicians, pharmacists, and microbiologists [24]. Our study went beyond this and involved CHWs as they are key to health service delivery at the community level in Uganda as well as other developing countries. The CHWs trained in our project have a primary responsibility of health education and promotion in their communities while some had an extra role on iCCM. From the evaluation results, it was evident that training of the CHWs enhanced their capacity to contribute to AMR through sensitization of the general population on the appropriate use of antibiotics, particularly in humans, but also in animals among other related issues. Miscommunication at the community level has been identified as a major challenge to addressing AMR in Africa [25], hence CHWs are a critical part of the health workforce to contribute to the NAP in the local setting. Given the training of CHWs had a component of IPC, they were subsequently invaluable in the promotion of sanitation and hygiene during the emergency public health measures undertaken in response to the COVID-19 pandemic in the communities. The use of the 'training of trainers' model, where the trained HPs later trained CHWs, is an indication of the health partnership's commitment towards building local capabilities in primary health care delivery. In addition, using trained practitioners to train another health cadre enhances health promotion and ensures cultural appropriateness [26].

Awareness of basic hygiene principles and infection control are at the core of AMR education [5]. The AMS activities we implemented in schools will help to promote intergenerational awareness on AMR, as well as facilitate pupils becoming antimicrobial guardians in the future. Such pupils are also likely to promote proper AMS and IPC practices, such as handwashing with soap, at critical times among peers and family members. Our project also involved undergraduate students from various disciplines concerning human and animal health from Uganda and the UK in AMR awareness-raising activities, including seminars, webinars, and competitions. Although these students may have earlier been exposed to AMR as part of their studies, it cannot be guaranteed that they had adequate knowledge and skills on AMS, which would be reflected in their practice. At the undergraduate level, many prescribers and students of other professions may not be confident in their preparedness to deal with AMS. For instance, a survey among fourth-year medical students in the United States revealed that only one-third of those surveyed considered themselves as being adequately prepared in basic antimicrobial use [27]. Similarly, a study among paramedical students in Ethiopia revealed that less than half of the students surveyed had adequate knowledge of AMR [28]. These gaps in knowledge among university students reflect the need for AMS awareness initiatives to be incorporated into educational training curricula. Involving pupils and students in AMS / IPC interventions is therefore important as they are the future generation, and it also contributes to ensuring the sustainability of project activities.

Our project facilitated the establishment of an MTC at ERRH to support appropriate prescribing of antimicrobials and related activities. Having an MTC at a hospital is one of the recommended steps in setting up an AMS program [29], hence it is an important contribution to the NAP. During the COVID-19 pandemic, ERRH was the first facility to handle cases with disease in the country, and the rapid and comprehensive support from the MTC to guide the hospital response was timely. The current existence and operation of the MTC at the hospital is a key achievement of our project given that it continues to support day-to-day operations at the facility, hence contributing to the sustainability of our project interventions. Establishment of the CoPs on AMS was also instrumental in ensuring sustainability as they continue to enhance knowledge exchange among human and animal health professionals and students. The CoPs continue to be integral to building awareness and sharing best practices on AMR/AMS/IPC. Given the COPs are online, the UK project team are able to stay engaged and be involved when not in Uganda. Indeed, online CoPs have been demonstrated to have a wider reach and keep participants engaged in comparison with physical ones, particularly in this digital age [30,31]. Having set-up the health professionals' CoP as part of the MOH TWC on AMS, optimum access and use will also contribute towards its sustainability beyond the project duration.

The pre- and post-training assessment and project evaluation revealed an improvement in HPs' and CHWs' knowledge of AMR and their practices to promote AMS and IPC. Specifically, the evaluation results showed that the training of HPs substantially improved the organizational culture for the majority, with 88.3% adopting new practices around AMR/AMS/IPC in line with the national requirement as prescribed in the UCG. Studies in different parts of Africa, such as South Africa [32], have also recorded change in organizational culture following the implementation of an AMS program. This change in practice among HPs involved in our study resulted in improved availability of antimicrobials as prescriptions and use were optimized. However, the evaluation revealed that health systems challenges, such as stock out of medicine, inadequate human resources, and lack of PPE for hospital staff, could impede AMS promotion if not equally tackled. In the case of CHWs, the evaluation results demonstrated how training them has a wider impact on improving positive health outcomes across communities, with many CHWs having sensitized between 50 and 100 community members on AMR/AMS within three months following the training. In addition, training the CHWs was instrumental in promoting the One Health approach, which was evident through encouraging livestock farmers to consult with a veterinary specialist regarding the health of their animals in the community. Animal husbandry is an important practice in the project community [15] and the East African region in general [25] so it is critical it is considered while implementing AMS interventions. This practice is reported to bring with it a high burden of what Ampaire et al. [25] referred to as "community-acquired infections". There is a high rate of antibiotics misuse and poor engagement with veterinary professionals among livestock farmers at the community level in Africa [33]. Therefore, the ability of CHWs to support improved practices regarding management of animal conditions should be integrated in future AMS activities to contribute to the fight against AMR in Uganda and beyond.

One of the strengths of our project is that it was implemented as part of a 10-year established health partnership between NTU and Mak with existing structures, intellectual capital, as well as local and global resources and networks that will contribute to strengthening the primary health care system in Uganda. In addition, our project embraced the One Health approach and targeted AMS interventions at both the health facility and community levels as well as including primary schools and university students, which is worth mentioning. A limitation of our project was its limited scope given it involved one hospital, a few lower level health facilities, two primary schools, and selected university students. The involvement of animal health workers in project activities was also low compared to those from human health, which can be improved in the future. Nevertheless, being a small pilot project, the achievements and lessons learnt will be instrumental in informing our future partnership activities to strengthen AMS in Uganda and further contribute to the NAP.

#### **5. Conclusions**

Adoption of a One Health approach in our project facilitated multidisciplinary efforts, including training human and animal HPs, to increase awareness and contribute towards improving AMS at health facilities and in the community. Reciprocal visits and establishment of CoPs fostered bi-directional learning and knowledge transfer on AMS between the UK and Uganda. The achievements of this project can inform the design of large-scale AMS interventions in support of implementation of the Uganda AMR National Action Plan.

**Author Contributions:** D.M. and L.G. are the Uganda and UK health partnership leads respectively and initiated the project idea. F.E.K., L.M., C.B., H.K., I.M.K., B.Y.N., J.O., K.R.-H. and J.W. contributed to conceptualization of the project. All authors including S.A., D.I., G.B.L. and F.N. participated in project implementation and writing the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was funded as part of the Commonwealth Partnerships on Antimicrobial Stewardship (CwPAMS) supported by Tropical Health and Education Trust (THET) and Commonwealth Pharmacists Association (CPA) using Official Development Assistance (ODA) funding, through the Department of Health and Social Care's Fleming Fund. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the Fleming Fund, the Department of Health and Social Care, THET or CPA.

**Acknowledgments:** We thank all stakeholders who supported the implementation of the project including from the: Ministry of Health; Ministry of Agriculture, Animal Industry and Fisheries; Entebbe Regional Referral Hospital; Makerere University College of Health Sciences; Makerere University College of Veterinary Medicine, Animal Resources and Biosecurity; Entebbe Municipal Council; Wakiso District Health Office; Kajjansi Health Centre (HC) IV; Nakawuka HC III; Kasanje HC III; Nsaggu HC II; Zzinga HC II; Kitala HC II; and Bussi HC II.

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

#### **References**


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

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

#### *Article* **Designing a National Veterinary Prescribing Champion Programme for Welsh Veterinary Practices: The Arwain Vet Cymru Project**

**Gwen M. Rees 1,2,\*, Alison Bard <sup>2</sup> and Kristen K. Reyher <sup>2</sup>**


**Abstract:** Antimicrobial use in agriculture has been identified as an area of focus for reducing overall antimicrobial use and improving stewardship. In this paper, we outline the design of a complex antimicrobial stewardship (AMS) intervention aimed at developing a national Veterinary Prescribing Champion programme for Welsh farm animal veterinary practices. We describe the process by which participants were encouraged to design and deliver bespoke individualised AMS activities at practice level by forging participant "champion" identities and communities of practice through participatory and educational online activities. We describe the key phases identified as important when designing this complex intervention, namely (i) involving key collaborators in government and industry to stimulate project engagement; (ii) grounding the design in the literature, the results of stakeholder engagement, expert panel input, and veterinary clinician feedback to promote contextual relevance and appropriateness; and (iii) taking a theoretical approach to implementing intervention design to foster critical psychological needs for participant motivation and scheme involvement. With recruitment of over 80% of all farm animal practices in Wales to the programme, we also describe demographic data of the participating Welsh Veterinary Prescribing Champions in order to inform recruitment and design of future AMS programmes.

**Keywords:** antimicrobial stewardship; veterinary; complex intervention

#### **1. Introduction**

Antimicrobial resistance (AMR) is a global One Health challenge of great significance [1]. The World Health Organisation describes AMR as a global health and development threat requiring urgent multisectoral action [2]. While the development and transmission of AMR is complex and not yet fully understood, antimicrobial use is known to be a major driver of resistance and there is broad consensus that antimicrobial stewardship (AMS) is a key component in addressing the issue [2–4]. Indeed, "the critical role of antimicrobial stewardship in tackling the problem of AMR is reflected in its inclusion as a key action in the UK five-year antibiotic resistance strategy" [5]. Extensive AMS programmes are commonly seen in human healthcare settings [4,6,7] and, although they form a part of many national and global AMR action plans [2,3,5], their implementation in veterinary practice remains sporadic and small scale [8].

Antimicrobial use in agriculture has been identified as an area of focus for reducing overall antimicrobial use and improving stewardship [3]. In the UK context, recent efforts have led to a decrease in overall antimicrobial use in food-producing animals of 45% since 2015 [9]. These reductions have been broadly industry-led, with industry bodies recognising a consumer demand for responsible antimicrobial use and an increasing political focus on the issue [10–12]. Responsible prescribing is defined by the UK AMR

**Citation:** Rees, G.M.; Bard, A.; Reyher, K.K. Designing a National Veterinary Prescribing Champion Programme for Welsh Veterinary Practices: The Arwain Vet Cymru Project. *Antibiotics* **2021**, *10*, 253. https://doi.org/10.3390/ antibiotics10030253

Academic Editors: Diane Ashiru-Oredope and Jeroen Dewulf

Received: 27 November 2020 Accepted: 25 February 2021 Published: 3 March 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/).

5-Year National Action Plan as "The use of antimicrobials in the optimal way, for the right pathogen, at the right dose, for the right duration, for the treatment or prevention of infectious disease." [12].

In Wales, agriculture, animal health, and animal welfare are devolved policy areas over which the Welsh government has legislative powers [13]. AMR has been a policy focus in recent years, with the establishment of an Animals and the Environment AMR Delivery Group leading to the publication of the Welsh government's five-year AMR Implementation Plan [11]. This plan includes the key focus areas of improving standards of antimicrobial selection and prescribing, as well as improving standards of antimicrobial supply. AMS has been recognised as a vital component of national AMR strategies, although there is work to be done to improve implementation [5]. The agricultural industry represents a proportionally greater percentage of the national economy in Wales than it does for the UK as a whole, and the majority of Welsh farming is based on beef, sheep, and dairy production [14]. As such, the health and welfare—and related antimicrobial prescribing of cattle and sheep, could be argued as being of relatively greater significance in Wales than the rest of the UK.

AMS programmes are complex interventions consisting of several interacting and inter-relational components, which present challenges to those designing, implementing, and analysing such programmes [15]. Successful design requires evaluation of the available evidence, engagement with theory and a good theoretical understanding of how an intervention may cause change [16]. A recent systematic review found the use of theory, engagement of end users, identifying barriers, and selecting appropriate intervention components to be key elements of the successful design of interventions for changing healthcare professionals' behaviour [17]. Additionally, involving stakeholders, understanding the intervention context and considering implementation in a "real world" setting have also been seen as essential principles for consideration [18].

The purpose of this paper is to outline the development and implementation of a national AMS scheme for farm animal veterinary practices through the establishment of a network of Veterinary Prescribing Champions (VPCs) as part of the wider Arwain Vet Cymru (AVC) programme in Wales. Arwain Vet Cymru is a collaborative initiative, which aims to train and support a national network of Veterinary Prescribing Champions across Wales to improve antibiotic prescribing in cattle and sheep. The project is participatory in approach, aiming to empower veterinary surgeons to develop and implement bespoke stewardship interventions, as well as share experiences and ideas. Both development and implementation of this scheme were informed by the self-determination theory (SDT), a broad theory of human motivation covering elements of interpersonal dynamics, goals and motives, individual differences, psychological needs, and psychological well-being [19]. SDT explicitly recognises that some behaviours are not intrinsically appealing and that the salient question when considering behaviour change is how to motivate individuals to value, self-regulate and (without external pressure) carry out and maintain, such behaviours. As such, SDT is particularly pertinent to the context of AMS, as it considers not just how and whether AMS behaviours are likely to be enacted, but the mechanisms by which these behaviours can become self-directed and, thus, maintained over time elements critical to a national AMS scheme.

#### *Aim and Objectives*

As the Veterinary Prescribing Champion Network is a novel intervention—with similar programmes now being considered for England and Scotland—this paper aims to inform future national stewardship programmes about its design, methodology, and enactment, providing a much-needed evidence base for future complex interventions in the veterinary sphere. Specific objectives are to examine:

• The process through which this national AMS scheme was appropriately contextualised, involving the integration of complementary knowledge pathways in the development of intervention goals;


#### **2. Methods**

The study obtained ethical approval from the University of Bristol Health Sciences Research Ethics Committee, Reference 99522.

#### *2.1. Study Setting*

Antimicrobial use in Wales is regulated by the Veterinary Medicines Directorate, an executive agency of the Department for the Environment, Food, and Rural Affairs. All antimicrobials used in food-producing animals are prescription-only medicines, which can only be prescribed by veterinary surgeons to animals "under their care" [20]. Farmers in Wales are in the relatively privileged position of being able to store antimicrobials on farm for use at a later date [21]. There is a requirement to maintain purchase and use records, although there is evidence that these records may not always be accurate [22]. Veterinary practices in Wales that provide farm animal services are members of one of two Veterinary Delivery Partnerships, established to allow the delivery of government tuberculosis testing across Wales. Practices are otherwise separate and private business entities, with farmers able to choose their veterinary practice freely. There are approximately 50 separate veterinary practices providing farm animal services in Wales, although some of these are located along the border in England.

#### *2.2. Theoretical Basis*

The AVC intervention aimed to facilitate a professional environment that would inspire VPCs to engage with and endorse the network and their new AMS behaviours. Given the disparate nature of current antimicrobial prescribing and stewardship in veterinary practice, it was recognised that each participant's context and behavioural opportunities would likely be different, and a one-size-fits-all approach to AMS was unlikely to be effective. This intervention was therefore founded on the selection, adoption, and implementation of AMS behaviour changes by AVC participants themselves, through participant involvement in the scheme cultivating a prescribing "champion" mindset to cement their intention to design and implement an AMS intervention within their own professional environments. In considering this target behaviour change through the lens of the widely used COM-B behaviour system (capability, opportunity and motivation) and the associated Behaviour Change Wheel [23], it was clear that achieving this goal necessitated a focus on delivering an intervention design that engaged core motivational drivers of individual AVC participants with regards to their engagement with AMS knowledge, principles, and activities. To this end, an evidence-based theoretical perspective was sought to inform the AVC process and activities with respect for—and targeted attention towards—fundamental VPC motivational needs. Few frameworks on motivation have spurred as much research as SDT, with a recent conceptual and empirical meta-analysis supporting key premises within the theory [24].

SDT identifies distinct types of motivation that are key to understanding how—and whether—behaviour becomes internalised by individuals and stimulates personal growth and change. The most fundamental distinction is between intrinsic motivation, which refers to carrying out a behaviour because it is inherently enjoyable or interesting, and extrinsic motivation, which refers to carrying out a behaviour because it leads to a separable outcome or instrumental value [25]. For example, a veterinary surgeon who spends her spare time reading a paper on responsible prescribing practices, purely because she is curious about the topic, does so because she is intrinsically motivated, whilst her colleague who carries out the same behaviour only because it has been mandated by their boss is extrinsically motivated.

Extrinsic motivation can be further classified by its underpinning reasons or goals, forming a continuum from internalised and agentic extrinsically motivated states to those that are more motivationally impoverished and externalised [25]. Those extrinsic behaviours that are more internalised (i.e., in line with an individual's closely held beliefs or values) are likely to be associated with better quality of engagement, more positive self-perception and greater persistence than those behaviours that are more externalised (i.e., those carried out due to external punishments and rewards or a focus on approval from others via, for example, pride, shame or guilt) [24,25]. As such, for the veterinary surgeon reading about responsible prescribing because his boss requires it, if he also views the activity as valuable in developing his professional knowledge and identity, he will be more effectively engaged than if he acts purely to avoid guilt or a reprimand.

Where the premise of this intervention was to encourage individuals to carry out behaviours that might not have been intrinsically motivated (otherwise, no intervention would arguably have been necessary), we believed promoting conditions that allowed VPCs to feel more in control—and to express internalised motivation in their AVC engagement and chosen AMS behaviours—to be critical to intervention success, given associated benefits in learning, engagement, creativity, and personal commitment [25]. SDT identifies three universal psychological conditions that—across cultures—are critical to promoting internalised forms of extrinsic motivation in individual behaviour: the needs to feel competence (perceived self-efficacy), autonomy (a sense of choice, being the origin of one's own behaviour), and relatedness (feeling understood and cared for by others) [26,27]. Significant consideration was therefore paid to fostering these conditions in all aspects of programme delivery to promote VPC self-direction in AVC activities and resulting AMS behaviour change goals (Table 1).

#### *2.3. Engagement through Key Collaborators*

The AVC project—which also includes quantitative antimicrobial use data collection and animal health planning schemes alongside the intervention—represents a collaboration between Bristol Vet School, Welsh government's Office of the Chief Veterinary Officer, the industry-controlled farmer cooperative Welsh Lamb and Beef Producers and the South Wales and North Wales Veterinary Delivery Partners, Iechyd da and Milfeddygon Gogledd Cymru, respectively. Development of a network of Veterinary Prescribing Champions is one part of the wider AVC project, which is aimed at addressing antimicrobial resistance in Welsh agriculture. This includes work to develop technology for improving the accuracy of medicine use recording by farmers, led by Welsh Lamb and Beef Producers, and benchmarking veterinary practice antimicrobial use, led by Iechyd Da. By engaging these key collaborators, the AVC intervention was supported by leading academic, governmental and industry representatives able to engage with potential participants and encourage active involvement in the programme. Each collaborator contributed to engagement in the following ways: The Office of the Chief Veterinary Officer was able to use established communication channels to encourage participation in an official capacity, with the Chief Veterinary Officer for Wales endorsing the programme and encouraging veterinary surgeons to take part. Welsh Lamb and Beef Producers are responsible for farm quality assurance schemes in Wales and, therefore, are a familiar industry body that Welsh farm animal veterinary surgeons understand to represent farmers' interests, which helped improve engagement. The Veterinary Delivery Partners were able to contribute to active recruitment by disseminating details of the project to their veterinary practice members through formal networks. The project lead (GR) had also worked as a farm animal veterinary surgeon in Wales and, therefore, was able to combine these formal recruitment pathways with informal networks to further promote engagement.

Participant demographic data were collected through an online questionnaire at the time of initial recruitment and registration.

**Table 1.** Operational conditions of self-determination theory posited by Silva, Marques, and Teixera [27] for consideration in intervention design and their adaptation to guiding principles for appropriate enactment within Arwain Vet Cymru.


Where SDT = self-determination theory, AMS = antimicrobial stewardship, AVC = Arwain Vet Cymru, VPC = veterinary prescribing champion, GR = Gwen Rees.

Support for relatedness

Feedback

Skills training

Empathy

Attunement

Dedication of resources

Dependability

prescribing champion, GR = Gwen Rees.

*Antibiotics* **2021**, *10*, x 6 of 18

Ensure VPCs have the opportunity to access relevant and non-judgmental feedback on their practice interventions throughout design and implementation processes, both individually (through accessibility of contact with G.R. as project lead) and in-group meetings where this is facilitated peer-to-peer within the network (i.e., workshops and discussion groups).

Commitment to providing education, training, guidance and support in key areas of AMS as identified through knowledge pathways in Phase One of intervention design, to ensure VPCs feel adequately equipped to identify and set their own AMS behaviour change goal(s).

Ensuring group meetings (discussion groups, workshops) offer opportunities for VPCs to explore and reflect on their colleagues' perspectives at both peer-to-peer and group levels. Facilitate alternate perspective taking on any contentious issues if they arise within the group.

prescribing and AMS challenges, in addition to genuine appreciation for VPC engagement.

Careful attention to, gathering knowledge about and responding to VPC perspectives both (i) by those coordinating the AVC scheme and (ii) facilitated peer-to-peer within the AVC network, to ensure VPCs needs to feel validated, accepted, affirmed, and significant within AVC are met [28], and to generate a felt sense of union with other VPCs in this process [29].

Emphasising where and how AVC coordinators and wider project collaborators (industry, government) are investing time and energy into the scheme, in addition to creating project opportunities (workshops, discussion groups) where VPCs are connected by volunteering their time and energy to drive the momentum of AVC.

Ensuring VPCs feel that support is available to them via AVC in case of need on their AMS behaviour change journey, through guidance on how they can seek the input and advice of the project lead (GR) throughout.

Champions is one part of the wider AVC project, which is aimed at addressing antimicrobial resistance in Welsh agriculture. This includes work to develop technology for improv-

Affection Those coordinating the AVC scheme taking care to convey a sense of care and concern for participants

#### programme. *2.4. Designing a National Stewardship Programme 2.4. Designing a National Stewardship Programme*  The development of the AVC intervention model occurred in two phases (Figure 2).

*2.3. Engagement through Key Collaborators*  The AVC project—which also includes quantitative antimicrobial use data collection and animal health planning schemes alongside the intervention—represents a collaboration between Bristol Vet School, Welsh government's Office of the Chief Veterinary Officer, the industry-controlled farmer cooperative Welsh Lamb and Beef Producers and the South Wales and North Wales Veterinary Delivery Partners, Iechyd da and Milfeddygon Gogledd Cymru, respectively. Development of a network of Veterinary Prescribing The development of the AVC intervention model occurred in two phases (Figure 2). Firstly, identifying critical elements of the intervention—through the integration of the four knowledge pathways representing subject experts, relevant stakeholders, practicing veterinary surgeons, and the current evidence base for effective interventions targeting prescribing practice—enabled the design of a context-specific and appropriate intervention. Secondly, grounding the delivery of this intervention within the SDT theoretical framework—by identifying operational SDT conditions relevant and applicable to AVC participation—allowed for an understanding how the intervention was proposed to engage VPCs' internalised motivation. Firstly, identifying critical elements of the intervention—through the integration of the four knowledge pathways representing subject experts, relevant stakeholders, practicing veterinary surgeons, and the current evidence base for effective interventions targeting prescribing practice—enabled the design of a context-specific and appropriate intervention. Secondly, grounding the delivery of this intervention within the SDT theoretical framework—by identifying operational SDT conditions relevant and applicable to AVC participation—allowed for an understanding how the intervention was proposed to engage VPCs' internalised motivation.

**Figure 2.** Design map of the Arwain Vet Cymru program, identifying the key phases of the design and implementation process.

#### 2.4.1. Phase One: Contextual Knowledge

Four knowledge pathways were explored to appropriately contextualise the aims of the AMS program for VPCs in Wales.

Relevant stakeholders: key stakeholders were identified in the areas of veterinary professional regulation, specialist veterinary membership organisations, farming body representatives, government policy departments, and human public health. Stakeholders included the British Veterinary Association's Welsh Branch Council, Welsh government's AMR in Animals and the Environment Delivery Group, the Sheep Veterinary Society, the British Cattle Veterinary Association, Public Health Wales, and the National Farmers Union, among others. These stakeholders were contacted and invited to input into the design of the new national stewardship programme. Stakeholders involved in ongoing animal health projects in Wales were contacted in order to coordinate efforts and avoid duplication.

Practicing veterinary surgeons: practising farm animal veterinary surgeons in Wales were informally surveyed by the Veterinary Delivery Partners in order to identify key issues they felt important to be included in the design of the programme. This took the form of utilising existing communication networks between farm animal veterinary surgeons, including email and WhatsApp communications, to invite suggestions for stewardship intervention strategies and feedback on current policy.

Expert input: a broad range of expertise was available through the University of Bristol's "AMR Force" multidisciplinary research group, consisting of clinical veterinary practitioners, epidemiologists, veterinary academics, and social scientists. By drawing upon this expertise, intervention design was informed by the current research landscape and areas of clinical importance in order to focus on identified areas of key importance to research and clinical practice.

Literature review: an extensive literature review examining (i) complex intervention design theory; (ii) antimicrobial prescribing in agriculture; and (iii) AMS interventions was conducted. This review provided an evidence base for the intervention design, identified potential barriers, and enablers to stewardship in the veterinary context and highlighted known areas of high antimicrobial use for specific focus.

#### 2.4.2. Phase Two: Integrating Theory

The second phase of AVC intervention design aimed to foster the motivational internalisation of AVC activities and AMS change for participating VPCs. To ensure VPCs' motivation was cultivated in this internalised, agentic form, active integration of the psychological needs highlighted within SDT was critical. Namely, the need for VPCs to feel AVC activities and selected AMS change(s) (i) enhanced their competence (perceived selfefficacy); (ii) supported their autonomy (a sense of choice, being the origin of one's own behaviour); and (iii) promoted their sense of relatedness (feeling understood and cared for by others) [26]. Operational conditions for these psychological needs have been detailed for consideration in the design of SDT-informed interventions of this kind [27]. These conditions were adapted to create guiding principles for the AVC intervention design (Table 1) informing the selection, content, and thoughtful delivery of activities within the AVC training schedule, as highlighted in Results.

#### **3. Results**

#### *3.1. Participation of VPCs*

A total of 43 farm animal veterinary surgeons were recruited to the AVC project from March 2020, representing 41 veterinary practices across Wales. Participants were offered no incentives for taking part in this study, although the training could be counted towards mandatory continuing professional development requirements of UK practicing veterinary surgeons. Out of the 50 Welsh practices involved in farm work (defined as practices with Official Veterinarians registered with the Veterinary Delivery Partners), nine did not take part in the programme. Of these, five stated that they did not do sufficient farm work within Wales to make participation worthwhile, one practice withdrew from the programme due

to increased workload relating to the Coronavirus Disease (COVID-19) pandemic and no response was received from three practices. Demographics of participants can be seen in Table 2.

**Table 2.** Participant demographics of Veterinary Prescribing Champions (VPCs) enrolled on the Arwain Vet Cymru project in Wales.


Fifty-eight percent of the veterinary surgeons participating were either business partners or directors, consultants, or clinical directors, with the remaining 42% identifying as salaried assistants. Half of the participants had been graduated for >20 years, with only 5% having graduated fewer than five years prior to the programme beginning. Participants had a diverse range of interests across the spectrum of farm animal clinical work, with similar proportions interested in dairy, sheep, beef, mixed practice, and smallholder work. Twenty-six practices belonged to the South Wales Network, and 17 practices belonged to the North Wales Network. Eight practices were based over the Wales–England border, but served a significant number of Welsh farms. Nineteen participants (44%) were female and the remaining participants male.

The key barriers to implementation found so far in the AVC project can be best characterized as time constraints for participants and concern that restricting antimicrobial prescribing may lead to farming clients sourcing medicines elsewhere. However, despite a focus on these barriers during group discussions, they have not impacted significantly on participation in the programme to this point.

#### *3.2. Defining the AVC Intervention Structure*

Four knowledge pathways, as outlined in Methods, determined the structure, and focus of the overall intervention:

Stakeholder engagement: stakeholder response was positive, with all those contacted recognising the need for an AMS programme in Wales. Topics that emerged as important from the stakeholder engagement included a focus on responsible antimicrobial sales practices, the need for greater communication and collaboration between veterinary practices within a region, supporting veterinary surgeons to make responsible prescribing decisions and improving knowledge of relevant legislation and guidance.

Practicing veterinary surgeons: an informal survey of the needs and desires of farm animal veterinary surgeons in Wales indicated that they had similar areas of concern and focus as those identified by stakeholders. Of particular importance was the issue of responsible antimicrobial sales practices and improving communication between practices. Practising veterinary surgeons also outlined an interest in behaviour change principles, and how they could be applied when encouraging farmers to use medicines responsibly.

Expert input: interdisciplinary research group meetings outlined several key areas that were viewed as important in the design of this intervention. These included improving knowledge of the legal aspects, professional regulations, and industry guidelines surrounding prescribing, the principles of evidence-based veterinary medicine and the importance of participatory approaches to change.

Literature review: grounding the design in the theory of behaviour change and complex interventions in healthcare was identified as very important to the programme's success. Reviewing the literature indicated that the use of so-called "Champions" in health care interventions had been successful in other settings. The literature also highlighted the benefit of building sustainable communities of practice for complex healthcare interventions, and of combining education and training resources with reflective exercises and goal setting.

By combining the results of these four knowledge pathways, AVC's design was focussed around addressing the following key areas:


#### *3.3. Enactment of the AVC Network: Combining Intervention Goals and Theoretical Drivers*

The overall design of the implementation can be seen in Figure 2, and the training schedule can be seen in Table 3. Initially, implementation of the programme was designed to consist of several in-person meetings of all VPCs over the course of the first year. However, following the COVID-19 global pandemic and subsequent lockdown in the UK in March 2020, combined with the uncertain future of large gatherings, it was necessary to reimagine the AVC process in an entirely online format in early 2020. Each element of this online format within the AVC process will be discussed with reference to the operational conditions of SDT (Table 1) identified as critical guiding principles of the AVC intervention design.


**Table 3.** Arwain Vet Cymru Veterinary Prescribing Champion (VPC) training schedule developed as outlined in Figure 1.

#### *3.4. Webinars*

In order to address the goal of improving VPCs' knowledge of the key areas of AMS identified in the knowledge pathways outlined above, an educational programme of six webinars was included in the overall design. Expert speakers were invited from a range of academic institutions, with content informed by the literature review and expert panel meetings along with veterinary and stakeholder engagement. Six one-hour webinars were co-designed with the speakers. These webinars were broadcast weekly on Wednesday afternoons, the day identified during recruitment as the best time for VPCs as routine tuberculosis testing does not usually occur on this day. Participants were given the opportunity to attend webinars during the live broadcast or to watch recordings asynchronously, at their convenience.

A brief description of the content of each webinar is outlined in Table 3. Briefly, webinars covered topics such as Welsh AMR policy, the concept of AMS, legislation and guidelines relevant to prescribing, behaviour change theory, evidence-based veterinary medicine, antimicrobial use benchmarking across the different species and a selection of case examples from practices that had successfully implemented various AMS schemes. These topics were selected based on the key areas identified in the literature review, stakeholder engagement, expert input, and informal survey of practicing veterinary surgeons, as outlined in Section 3.2.

The format and delivery of these webinars was chosen to actively promote operational conditions within SDT to enhance VPC engagement (Table 1). The provision of instrumental and practically relevant AMS training—in addition to clarifying VPCs' expectations of their involvement in the AVC Network—promoted support for VPC-perceived competence. Additionally, focusing on promoting VPC self-endorsement of AVC activities through provision of a variety of rationales from well-respected, expert speakers whilst providing choice in how webinars were accessed by participants (i.e., synchronous or asynchronous) embedded key attributes of autonomy support.

#### *3.5. Discussion Groups*

To develop a sense of community, collaboration, and group identity, informal online discussion sessions were held every third week of the nine-week training timeline (Table 3). VPCs were divided by region into North Wales and South Wales groups. Participants were given a choice of which group they wished to belong to, since those working in mid-Wales may have identified more strongly with a different region than might have been suggested geographically. Discussion sessions were hosted using online videoconferencing software and were facilitated by two researchers experienced in group facilitation (G.R. and A.B.).

Topics of discussion in each session were iterative and informed in part by the content of the previous webinars, in addition to topics raised during informal feedback and webinar question-and-answer sessions. These topics were guided by the facilitators, but were

semi-structured in nature, allowing some freedom for participants to discuss issues they felt to be important at the time. Utilising interactive polling and small group breakout rooms, participants were asked to focus on and discuss specific areas related to veterinary prescribing before joining plenary discussion sessions where participants were able to share their views and discuss further with the whole group. Discussion group size varied from a minimum of five participants to a maximum of 17 participants between group meetings, and the facilitation of these groups was flexible in order to account for varying group size. Where discussion group size was greater than five, virtual "break-out rooms" were used, and participants were asked to discuss in small groups before returning to the main plenary discussion to report back on their discussions. This flexibility was important because the availability of participants to join discussion groups would vary depending on clinical veterinary duties on the day.

Discussion sessions enabled participants to outline the main challenges they perceived when considering implementing AMS programmes, along with exploring opportunities for change(s), and developing a sense of shared ownership over the outcomes of the project. Discussion sessions were also an opportunity to prepare Champions for the subsequent workshops. Promoting congruence with the tenets of SDT underpinned discussion session design. To cultivate a sense of autonomy for VPCs, attendance was made non-compulsory and VPCs chose their regional group allocation. The sessions were also opportunities for AVC facilitators to actively evoke and acknowledge VPCs' feelings and agendas with regards to the breadth of potential interventions covered in the webinar sessions. This, in turn, further promoted autonomy through respect for the VPCs' unique choices and intentions with regards to the AMS foci. To support VPCs' competence, the discussion groups offered facilitators the chance to provide relevant and non-judgemental feedback on VPC perspectives on AMS foci, whilst allowing facilitators to shape participatory activities to also encourage positive peer-to-peer feedback. Finally, enhanced relatedness was achieved through a focus on evoking, exploring and understanding VPC perspectives, both by the facilitators and through targeted peer-to-peer activities, creating opportunities for promoting group empathy and attunement (a felt sense of union) between AVC participants.

#### *3.6. Workshops*

Two three-hour facilitated workshops were included in the design of the programme and followed on from the webinars and discussion groups in order to enable goal setting and the creation of action plans by VPCs, as outlined below. The first workshop was intended to allow VPCs to develop the knowledge and ideas gained during the webinars and discussion groups and distil these into actionable goals designed specifically for their practice context. VPCs were responsible for designing their own context-specific AMS intervention, relevant to their veterinary practice's prescribing context. A second workshop, designed to inform policy, was included in order to allow VPCs the opportunity to contribute to the wider professional context with regards to matters of AMS.

#### 3.6.1. Stewardship Intervention Design Workshop

The stewardship intervention design workshop aimed to enable each participating VPC to design and develop their own personal action plan, as well as a stewardship intervention for their practice. Examples of the kind of action plans discussed include:


This workshop was run by an experienced participatory action research facilitator and co-facilitated by two experienced facilitators familiar to the VPCs (G.R. and A.B.). Participants were asked to set goals and create action plans outlining how they would implement their stewardship intervention according to the SMART framework (Specific, Measurable, Attainable, Relevant, Timely) [30]. Structured discussions utilising online fora and breakout rooms allowed VPCs to consider their plans with their peers, helping to identify potential barriers to implementation and possible solutions by drawing on their collective experiences.

Central to the concept and design of this workshop was fostering VPCs' sense of autonomy in their AMS roles. Workshop activities consolidated VPCs' own ideas for an AMS intervention strategy depending on what they envisaged for their own practice context, whilst a primary facilitator experienced in non-directive, participant-led workshops emphasised the ethos of VPC choice and self-endorsement throughout. Workshop activities also aimed to encourage VPCs toward the choice of an optimal AMS challenge (i.e., not too easy nor too difficult) for their circumstances and skill set, to drive competence-infused practice change. Relatedness was embedded within workshop activities through, for example, informal and personal introductions in each workshop to foster rapport, by offering attending VPCs opportunities to vocalise fears, concerns, and thoughts on intervention interests for peer validation, and facilitating peer-to-peer exploration, reflection, and group feedback on personal perspectives of AMS activities and policy within Wales. Together, activities of this kind sought to foster empathy and union (attunement).

#### 3.6.2. Policy Workshop

This workshop was designed to allow VPCs the opportunity to inform AMR policy at the national level. Participants were encouraged to identify important areas of focus, outline the policy support required to enable them to be responsible prescribers, and construct practical solutions to help address some of the barriers identified in the AVC project. This workshop was created with the support of Welsh government, who agreed that outcomes would be presented to the Welsh government's Animals and the Environment AMR Delivery Group.

This policy workshop offered VPCs the chance to develop a sense of personal influence over policy decisions impacting their profession; thus, engendering a feeling of self-endorsement critical to autonomous engagement. The premise of the workshop highlighting the unique role of the AVC network as a valued voice in determining AMR strategy in Wales—emphasised the importance of this group influence of AVC Champions thus forging the relatedness of group members further. Providing another practical opportunity for VPCs to explore and construct solutions to AMS challenges, elevated to the national perspective within Wales, was a final training opportunity for VPC competence development.

#### *3.7. Stewardship Intervention Implementation*

The initial AVC programme outlined above required a time commitment of around 15 h by the VPCs. Following on from the workshops, VPCs are expected to disseminate the AMS messages to and implement their co-designed, individual AMS plans in their respective practices. This will lead to 41 different AMS schemes being implemented—one at each participating practice—beginning in January 2021. Participants will be asked to complete monthly reports outlining the implementation of their stewardship scheme as well as provide feedback on ease of implementation, relevant actors involved, scope of the changes, outcomes observed, and barriers encountered. Overall practice prescribing behaviours will be evaluated through a longitudinal prescribing audit.

Throughout implementation, AVC facilitators will manage VPCs through a continued ethos of avoiding directive, coercive, or authoritarian approaches, first and foremost emphasising VPC autonomy in the enactment of practice-based interventions and how VPCs choose to engage with the AVC Network, and support staff throughout this process.

Practices will be supported by the project as required, with AVC facilitators prioritising trust in VPC competencies in overseeing the implementation of these intervention choices, with a focus on responsive management determined by VPCs themselves. It is hoped that the attunement developed from peer-to-peer activities within the AVC Network (through discussion groups and workshops) will also be an avenue of relatedness support for participating veterinary surgeons during this implementation process.

#### **4. Discussion**

In a complex healthcare intervention such as this, multiple interconnected elements all inform and affect each other. The AVC design process included stakeholder engagement, reviewing current literature, drawing on theory and understanding context as laid out by O'Cathain et al. [18]. As such, it is difficult to appraise each individual element of such a programme, and the entire intervention must be considered as a whole. However, by examining some of the principal domains of the intervention design (Figure 2), it is possible to explore how they informed—and became intrinsic to—the programme structure. By attempting to understand the context, theoretical basis, and implementation of the intervention, we can examine how and why VPCs within the AVC network might promote change through their AMS interventions and lead to more responsible use of antimicrobials.

Considerable effort was spent throughout the intervention process in engaging and communicating with farm animal veterinary surgeons in Wales. By grounding the design in the available literature and accessing informal feedback, key barriers to implementation could be identified and attempts to overcome them could be incorporated into the design from the outset. A recent scoping review found that knowledge, responsibility (the influence of peer behaviour) and the veterinary surgeon–client relationship represented significant barriers to AMS for cattle veterinary surgeons [31]. The AVC implementation design sought to address these barriers through education and building communities of practice. In Golding et al.'s exploration of veterinary surgeons' beliefs about AMS, one perceived barrier to implementation was the concern that farming clients might simply change to a rival practice if denied the antimicrobial of their choice [32]. This was also identified by stakeholders and participants as a barrier to change. In response to this, the design of the AVC intervention included an emphasis on building a sense of common purpose between practices, encouraging open communication, and creating a community by incorporating informal discussion groups.

The demographic characteristics of the participants was hypothesised to play an important role in the likely success of the program. Experienced veterinary surgeons with a senior role in their practices were thought to have a greater degree of autonomy and authority with which to implement AMS interventions. The relatively few female participants (44%) compared with the 57% of females who make up the UK's veterinary workforce was statistically significant (*p* = 0.0472) using the N-1 Chi-Squared Test for two proportions and may be explained in part by the increasing "feminisation" of the profession and the under-representation of female veterinary surgeons in senior roles [33]; recruitment of older, senior veterinary surgeons meant they were more likely to be male. It would be interesting to understand whether the gender ratio of such a participant group and its representativeness of the wider study population—influences the effectiveness of the intervention. Further research into the role of gender in complex intervention implementation through realist evaluation principles is required [34].

In identifying a theoretical driver for AVC, full consideration of the intervention context was essential. Following the recommendations within COM-B [23], AVC design considered (i) the target behaviour; (ii) intervention options; and (iii) content and implementation options. The target behaviour within AVC was complex, with the aim of participants cultivating a prescribing "Champion" mindset and cementing their intention to design and implement an AMS intervention within their own professional environments. Fundamentally, the AVC goal was therefore to create the facilitative conditions for this mindset and practice change to occur. Intervention options from the Behaviour Change

Wheel targeting training, environmental restructuring, and enablement appeared most appropriate for this purpose, influencing the multifaceted intervention design and the inclusion of webinars, discussion groups, and workshops informed by the four knowledge pathways [23]. Through consideration of how best to integrate these intervention foci effectively as drivers of capability, opportunity, and motivation with regards to a "Champion" mindset, a theoretical underpinning was sought to foster VPC engagement throughout.

Key to the aim of this intervention was the need for the "Champion" mindset to be sufficiently salient, psychologically, to drive VPC self-directed behaviour as VPCs are expected to implement their own AMS intervention in January 2021, following active engagement in the AVC scheme. Understanding the motivational factors that facilitate or undermine a sense of initiative and volition, in addition to the quality of performance, is central to SDT [19,23]. This theory therefore appeared uniquely adapted to the demands of the AVC scheme. The psychological conditions posited to encourage individuals to value, self-regulate and (without external pressure) carry out and maintain behaviour competence, autonomy, and relatedness [26]—were adopted as guiding principles in the practical realisation of the intervention design. The strength of this intervention lies in having conducted a thorough assessment of: the behaviour change target in question, what might be needed to achieve this change, and where a theoretical underpinning resonating with the project aims might enrich implementation [35].

In the pursuit of forging individual identities within health interventions, the concept of using "Champions" as a means of motivating change in healthcare settings is not new. In Australia, Antibiotic Champions have been used to support an AMS campaign within Children's Health teams [36]. Medical, veterinary and dental students in the UK can register to become Antibiotic Guardian Champions [37], and the UK's National Health Service (NHS) has several Champion schemes, addressing such issues as social prescribing [38], diabetes [39], physical activity [40], perinatal metal health [41], and digital health [42]. In this programme, giving the participants an identity, as a VPC was a crucial part in developing a sense of community and leadership. Champions were representing the programme within their practices but were also representing their practice within the network.

Complementing this individual shift in perspective was the hypothesized creation of communities of practice, forging a group identity for AVC participants. The Situated Learning Theory [43], whereby professional learning occurs through interaction with peers and participation in practice, forms the basis for the concept of communities of practice. These are groups of people who interact on an ongoing basis in order to share expertise and deepen knowledge on an area of concern [44]. They have been utilised in healthcare settings as a means of improving performance and sharing knowledge "in response to the challenges of complex systems" [45,46]. By encouraging the development of a superordinate identity—in this case that of a national Prescribing Champion—alongside their professional identities as veterinary surgeons working in discrete private practice, it was intended that VPCs could overcome the professional barriers to AMS identified in the literature, as suggested by Bartunek et al. [47].

The inclusion of goal setting and action planning in this programme, through the intervention design workshop, allowed VPCs the opportunity to translate the knowledge and ideas gained during the initial training into defined, outcome-driven actions. By using the SMART framework [30], creating individual action plans based on overarching goals was expected to help narrow the intention–behaviour gap [48]. Literature establishes that planning within a particular context of who, when, where, and how is important when considering behaviour change [4]; indeed, it is at the heart of the theoretical underpinnings of the COM-B model [23]. Encouraging VPCs to consider these elements in their individual intervention designs will ideally facilitate AMS plans that appropriately echo tenets of the COM-B model, even in the absence of direct training on the intricacies of this model. A recent paper by Atkins et al. specifically called on National AMS intervention design

to include goal setting and action planning, as they were areas identified as being underrepresented in current AMS programmes [49].

Educational interventions have been shown to improve knowledge of pharmacovigilance [50] and prescribing competency [51] as well as to strengthen AMS [52–54], although the effects may be short-lived. Online learning as part of AMS programmes has been playing an increasing role [55] and online training of GPs has been shown to reduce antimicrobial prescribing for respiratory disease [56]. An online process also enabled the inclusion of a diverse range of external expert speakers who may not have been able to attend in more traditional in person provision, potentially improving the overall content. Another unintended but positive effect of online provision was the distribution of training sessions over the course of several weeks, interspersed with other activities, thus potentially consolidating the VPCs' participation in the programme.

The Medical Research Council's new guidance for developing and analysing complex interventions [57] highlights the importance of practical effectiveness—that is, whether the intervention works "in the real world"—as a key measure when evaluating complex interventions. In order to answer this question, process evaluation can use ethnographic and qualitative methodology in order to explore the impact of the intervention, identify any unintended consequences and be able to describe the experience of the participants who take part in any intervention programme. Through an ongoing process evaluation combining ethnographic exploration of implementation and quantitative measures of prescribing, the implementation of the AVC programme will be under continuous appraisal until completion (September 2021). Results of this evaluation will be published separately.

It remains to be seen whether the Arwain Vet Cymru project produces workable AMS interventions in clinical veterinary prescribing practice as predicted. While it is hoped that this complex intervention is successful in improving responsible prescribing practices, further empirical evidence is currently being collected in order to enable full conclusions to be drawn. Any unforeseen negative consequences are of course also important, and all outcomes are meaningful when informing future development of similar programmes, both in Wales and further afield.

#### *Limitations*

While in an ideal world the design–evaluation–implementation process would occur in a relatively linear fashion and follow best-practice study design, practically this is not always possible. In this instance, the intervention took place in the context of political and industry-led pressure on the veterinary profession to improve prescribing, with an impactled rather than research-led funding focus. As such, design, evaluation and implementation occurred in a more cyclical and iterative process in this study.

The establishment of this national Network of Prescribing Champions has been relatively labour-intensive, requiring high levels of ongoing engagement with key actors across many stakeholder groups. Participation in the project has involved around 15 h of time investment from participating VPCs, and the ongoing time commitment required to implement their action plans will be dependent on the complexity of the intervention each VPC has designed. The other stakeholders involved in the development of the AVC stewardship programme were not compensated for their time, and we believe their involvement to be motivated by a desire across the veterinary profession to improve antimicrobial prescribing both for the "greater good" and to improve the image of UK agriculture. Given the economic and political sensitivities of bringing individuals from separate, competing interests together to tackle a common concern, the very high level of recruitment to the programme is both surprising and encouraging. Establishing this pan-Wales network of highly motivated clinicians may make it possible to overcome some of the perceived barriers to change. The ongoing sustainability of the network—and its legacy after the end of the funded project—is an important area for development in the next stage of the programme. By moving to a self-sufficient model of participant-led network maintenance, it may be possible to continue the network beyond the lifespan of the project.

#### **5. Conclusions**

Designing a novel national AMS programme for farm animal veterinary surgeons requires several supporting factors. The applicability of this programme design to other parts of the UK and the rest of the world is difficult to predict; however, we believe that by focusing on a robust theoretical grounding and giving full consideration to the context of the intervention as evidenced in this paper, stewardship interventions can be improved worldwide. A favourable policy background, collaboration with key actors within the profession, stakeholder consultation, an emphasis on autonomy, and commitment to developing a sense of community have all helped to promote high levels of engagement in this voluntary national network of VPCs. Empirical data from both qualitative and quantitative process evaluations will help reveal the impact this type of complex intervention may have on AMS in rural veterinary medicine.

**Author Contributions:** Conceptualization, G.M.R.; methodology, G.M.R. and A.B.; investigation, G.M.R. and A.B.; writing—original draft preparation, G.M.R. and A.B.; writing—review and editing, K.K.R.; visualization, G.M.R.; supervision, K.K.R.; project administration, G.M.R.; funding acquisition, G.M.R. and K.K.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Welsh government through the Rural Development Programme.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the University of Bristol Faculty of Health Sciences (protocol code 99522, 18 March 2020).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** The authors would like to thank the collaborative partners in this project, Iechyd Da and Welsh Lamb and Beef producers, along with all of the participants and stakeholders involved in development of this project.

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

#### **References**


*Article*

#### **Knowledge, Attitudes and Practices of Veterinarians Towards Antimicrobial Resistance and Stewardship in Nigeria**

**Usman O. Adekanye <sup>1</sup> , Abel B. Ekiri 2,\* , Erika Galipó 2 , Abubakar Bala Muhammad <sup>3</sup> , Ana Mateus <sup>4</sup> , Roberto M. La Ragione <sup>2</sup> , Aliyu Wakawa <sup>5</sup> , Bryony Armson <sup>2</sup> , Erik Mijten <sup>6</sup> , Ruth Alafiatayo <sup>2</sup> , Gabriel Varga <sup>6</sup> and Alasdair J. C. Cook <sup>2</sup>**


Received: 7 July 2020; Accepted: 24 July 2020; Published: 28 July 2020

**Abstract:** Antimicrobial resistance (AMR) is a global health concern and the inappropriate use of antibiotics in animals and humans is considered a contributing factor. A cross-sectional survey to assess the knowledge, attitudes and practices of veterinarians regarding AMR and antimicrobial stewardship was conducted in Nigeria. A total of 241 respondents completed an online survey. Only 21% of respondents correctly defined the term antimicrobial stewardship and 59.8% were unaware of the guidelines provided by the Nigeria AMR National Action Plan. Over half (51%) of the respondents indicated that prophylactic antibiotic use was appropriate when farm biosecurity was poor. Only 20% of the respondents conducted antimicrobial susceptibility testing (AST) frequently, and the unavailability of veterinary laboratory services (82%) and the owner's inability to pay (72%) were reported as key barriers to conducting AST. The study findings suggest strategies focusing on the following areas may be useful in improving appropriate antibiotic use and antimicrobial stewardship among veterinarians in Nigeria: increased awareness of responsible antimicrobial use among practicing and newly graduated veterinarians, increased dissemination of regularly updated antibiotic use guidelines, increased understanding of the role of good biosecurity and vaccination practices in disease prevention, and increased provision of laboratory services and AST at affordable costs.

**Keywords:** antibiotic; antimicrobial resistance; veterinary; animal health; Africa

#### **1. Introduction**

Resistance to antimicrobials is rising worldwide, threatening our ability to treat common infectious diseases of humans and animals [1]. The direct consequences of infection with resistant microorganisms can be severe, including longer and more severe illnesses, increased mortality, prolonged stays in hospital, increased rates of therapeutic failure resulting in loss of protection for patients undergoing operations and other medical procedures, and increased healthcare costs [2]. The overuse and misuse of antibiotics in humans and animals has been linked to the emergence of antimicrobial resistance (AMR) in animals and the environment [1,3–7]. In livestock farming, antibiotics are used for prophylaxis, metaphylaxis (the treatment of a group of animals after the diagnosis of infection and/or clinical disease in part of the group), growth promotion, or are used therapeutically, to maintain health and increase productivity. Interaction between animals, humans and the environment promotes the transfer of resistant genes across different species, making AMR an important One Health challenge emerging on a global scale [3–6,8]. With the dwindling repertoire of antibiotic options available for the control of emerging, life-threatening and multi-drug resistant bacteria, there is a need for proper antibiotic stewardship to preserve the efficacy of existing antibiotics [9].

Antibiotic stewardship programs can play an important role in improving, prescribing and optimizing the use of antibiotics [10–14]. In the human sector, which has made significant strides in this area, antibiotic stewardship programs are defined as hospital-based programs dedicated to improving antibiotic use [15]. These programs increase infection cure rates while reducing treatment failures, adverse effects, hospital costs and lengths of stay, and antibiotic resistance [14,16,17]. Considering the potential benefits of antibiotic stewardship programs, the World Health Organization (WHO) strongly recommends that governments implement them for the containment of AMR [18]. Therefore, it is imperative that governments implement tailored interventions to encourage antimicrobial stewardship among healthcare professionals [19]. Beyond stewardship programs, strategies to tackle other related challenges also need to be considered.

In most sub-Saharan African countries, surveillance programs for antibiotic use and AMR in humans and animals are either lacking or are in their infancy, and the human and animal healthcare sectors at the government or ministry levels tend to work in silos, resulting in a lack of intersectoral collaboration. Furthermore, in Nigeria, the lack of regulation of existing veterinary drug markets and low involvement of pharmacists [20] and veterinarians in the formal drug distribution market [21] may contribute to the issue of substandard drugs in the marketplace. In 2018, the Director General of the Nigeria National Agency for Food and Drug Administration and Control (NAFDAC) held a town hall meeting with all players in the livestock industry, including practicing veterinarians and manufacturers and suppliers of veterinary medicines, and at the meeting announced the ban of use of some antimicrobials and growth promoters in livestock as part of efforts to control AMR and to support the One Health triad [22]. The banned antimicrobials and growth promoters included chloramphenicol, furazolidone, metronidazole, nitrofuran, and carbadox [22]. To address AMR in both humans and animals in sub-Saharan African countries, strong multidisciplinary collaborations are needed; however, these are lacking because of the poor One Health coordination of animal and human national disease surveillance systems [23].

In 2017, the Nigerian Federal Ministry of Agriculture and Rural Development (FMARD), Federal Ministry of Environment and Federal Ministry of Health developed a National Action Plan (NAP) for antimicrobial resistance (the Nigeria Center for Disease Control's five-point action plan) as part of the country's efforts to address the problem of AMR and to promote the responsible use of antimicrobials through a One Health approach [23]. Veterinary surgeons are typically responsible for prescribing and overseeing antimicrobial use in animals. Therefore, the role of the veterinarian in tackling AMR cannot be over-emphasized as they are the custodians of antimicrobials used in animal health [24] and food production. In Nigeria, regulatory authorities like NAFDAC and the Nigeria Centre for Disease Control (NCDC) are involved in creating awareness among veterinarians and veterinary students on the challenge of AMR through campaigns convened by the Veterinary Council of Nigeria (VCN) and the umbrella association for veterinarians in Nigeria, the Nigeria Veterinary Medical Association (NVMA).

Despite the potential negative impact of AMR on animal and public health, there remains a paucity of data concerning the awareness of this problem in sub-Saharan countries [25]. The attitudes of veterinarians towards antibiotic use and determinants influencing prescribing behavior of veterinarians have been investigated elsewhere [26–29]. In Nigeria, a few studies have explored the knowledge, attitude and practices of veterinarians towards antibiotic use, resistance and stewardship. A previous study by Anyanwu and Kolade reported that knowledge about antibiotic stewardship among veterinarians was as low as 21.4% [30]. However, this study was limited to Enugu State and used a non-probability sampling technique, which affected the generalizability of the study findings [30]. A recent study that involved only veterinary students reported that 60% of respondents had unsatisfactory knowledge scores for AMR [31]. To expand on the knowledge base in this area and to inform the development of interventions to promote responsible antibiotic use, a nationwide study of veterinarians in Nigeria was conducted. The objective of this study was to assess the knowledge, attitudes and practices towards AMR and antimicrobial stewardship and to identify factors that influence antibiotic prescription practices of veterinarians in Nigeria.

#### **2. Results**

The survey was sent to 5603 participants; there were 488 responses, corresponding to a response rate of 13%. Out of the 488, six did not consent, 59 consented but did not start or attempt the survey, while 128 consented and attempted but did not complete the survey. Thus, 241 respondents consented and completed the survey.

#### *2.1. Demographic Information*

Most of the respondents were male (79.7%, 192/241) and almost half were aged 25-34 years (48.1%, 116/241) (Table 1). A majority of respondents reported having a veterinary degree (63.1%, 152/241) as the highest educational qualification. More than a third of respondents had been registered as a veterinarian for 0–5 years (36.1%, 87/241). Mixed practice (defined as any combination of small, large, poultry or other type of practice) was the most frequently reported type of practice (63.5%, 153/241). Five percent of respondents (5.4%, 13/241) did not practice. Most of the respondents were employed in private practice (38.2%, 92/241).


#### **Table 1.** Demographics of respondents


**Table 1.** *Cont*.

\* The results for this variable are presented as row percentages instead of column percentages, as such the column percentages for this variable do not add up to 100%.

The distribution of respondents based on the location of the vet practice within Nigeria's six geopolitical zones showed that almost half of the respondents were located in the North Central (29%, 69/241) and the North West zones (20%, 49/241) (Table 1). The lowest number of respondents was recorded in the South South zone (7%, 17/241). The distribution of respondents based on the location of the vet practice within Nigeria's 37 states indicated that the Federal Capital Territory, Abuja, had the highest number of respondents (13.7%, 33/241) followed by Kaduna State (10%, 24/241) and Borno State (7.1%, 17/241) (Figure 1). Yobe and Nasarawa States were the only two states without any respondents.

#### *2.2. Knowledge*

Eighty-nine (36%) of the 241 respondents had heard of the term antimicrobial stewardship and of these, 69% (61/89) were able to correctly define antimicrobial stewardship (Table S1). Most of the respondents (81.7%, 197/241) were able to differentiate between an antibiotic and an antimicrobial agent. Most of the respondents were aware that antibiotics kill both commensal and pathogenic bacteria (91.3%, 220/241) and 94.6% (228/241) knew that overuse of antibiotics renders them ineffective. Most respondents were aware that antibiotics do not kill viruses (93.4%, 225/241) and all respondents were aware that bacteria can become resistant to antibiotics. Many respondents (93.4%, 225/241) were aware that there was a need to observe withdrawal periods before consuming milk from cows treated with antibiotics, and 97.9% (236/241) were aware that a withdrawal period is necessary before treated poultry can be considered fit for human consumption.

More than half of the respondents (59.8%, 144/241) had not heard of or read the Nigeria Center for Disease Control's five-point action plan for responsible use of antimicrobials. When asked what topics they would like to receive more information on, 75.5% (182/241) of the respondents selected "links between the health of humans, animals and the environment", and 55.6% (134/241) chose "microbial culture and sensitivity testing" (Table S1).

*Antibiotics* **2020**, *9*, x FOR PEER REVIEW 5 of 16

**Figure 1.** Distribution of survey respondents based on reported location of veterinary practice by state in Nigeria. **Figure 1.** Distribution of survey respondents based on reported location of veterinary practice by state in Nigeria.

#### *2.2. Knowledge 2.3. Attitude*

Eighty-nine (36%) of the 241 respondents had heard of the term antimicrobial stewardship and of these, 69% (61/89) were able to correctly define antimicrobial stewardship (Table S1). Most of the respondents (81.7%, 197/241) were able to differentiate between an antibiotic and an antimicrobial agent. Most of the respondents were aware that antibiotics kill both commensal and pathogenic bacteria (91.3%, 220/241) and 94.6% (228/241) knew that overuse of antibiotics renders them ineffective. Most respondents were aware that antibiotics do not kill viruses (93.4%, 225/241) and all respondents were aware that bacteria can become resistant to antibiotics. Many respondents (93.4%, 225/241) were aware that there was a need to observe withdrawal periods before consuming milk All but one (99.6%, 240/241) of the respondents believed that veterinarians have a role to play in preventing public health threats posed by AMR (Table S2). Respondents were asked to indicate if they considered the following as important global challenges and the most frequently reported challenges were: AMR (80.9%, 195/241), food security (74.5%, 180/241) and climate change (60.9%, 146/241). Most respondents (97.9%, 233/238) thought AMR was a national problem in Nigeria, and 96.1% (223/232) of respondents believed AMR will be a greater problem in veterinary practice in the future than it is today. Note the denominators used to calculate the above two percentages, and in the sections below (where relevant), do not add up to 241 because they exclude "unknown" responses (Table S2).

from cows treated with antibiotics, and 97.9% (236/241) were aware that a withdrawal period is necessary before treated poultry can be considered fit for human consumption. More than half of the respondents (59.8%, 144/241) had not heard of or read the Nigeria Center for Disease Control's five-point action plan for responsible use of antimicrobials. When asked what topics they would like to receive more information on, 75.5% (182/241) of the respondents selected Many of the respondents considered the excessive use of antibiotics in livestock (83.8%, 202/241) and under dosing of antibiotics (78.8%, 185/241) as the most important potential contributors to the development of AMR (Table S3). Most of the respondents agreed that prescribing unnecessary antibiotics was professionally unethical (97.5%, 234/240) and 78.2% (172/220) believed the antibiotics they prescribe may contribute to AMR (Table S2). Most respondents (99.2%, 237/239) agreed with

"links between the health of humans, animals and the environment", and 55.6% (134/241) chose

"microbial culture and sensitivity testing" (Table S1).

the statement "biosecurity was important in food production" and 28.2% (68/241) considered poor biosecurity practices as a contributor to AMR development. Of concern was that over half of the respondents (56%, 112/200) agreed with the statement "Prophylactic antibiotics are an appropriate alternative to protect animal health when there is poor biosecurity". Almost half of the respondents (42.3%, 102/219) indicated that they lacked enough knowledge on antibiotic use, while 28.9% (59/204) believed there were not enough antibiotics under development to combat the problem of resistance (Table S2).

#### *2.4. Practices Influencing Antibiotic Use*

A total of 132/220 (60%) of the respondents reported that they frequently encountered animal owners who had already initiated antibiotic treatment without veterinary supervision (Table S4). Note the denominators used to calculate the percentages for some variables in this section (where relevant) do not add up to 241 because they exclude "don't know" responses (Table S4).

More than half of the respondents (68.4%, 154/225) reported that their practice had a standardized protocol for the treatment of sick animals (Table S4). When asked what guidelines were followed to help select the appropriate antibiotic when a patient was presented for the first time, 53.1% (128/241) of respondents reported using microbiological culture and antimicrobial susceptibility testing (AST) for guidance. More than a half of the respondents (59.8%, 144/241) indicated that they administered empirical treatment while awaiting AST results. A small proportion (14.5%, 35/241) reported selecting antibiotics based on what the client could afford.

With regards to the frequency of AST use before starting antibiotic treatment, 48.7% (112/230) of respondents used AST at a frequency of one to three times in a month while only 21.3% (49/230) used AST more than three times a month. Over a quarter of the respondents (30%, 69/230) never conducted AST (Table S5).

Most respondents (75.5%, 182/241) indicated that poor response to initial antibiotic treatment or treatment failure influenced the veterinarians' decision to request AST. Other reported drivers for AST use included recurrent health conditions (70.5%, 170/241), having no knowledge of the animal or farm's health history (26.6%, 64/241) and owner request (18.3%, 44/241) (Table S5). When asked what were the most important barriers to the use of AST, the majority of respondents selected unavailability of laboratory services (82.2%, 198/241), followed by owners' inability to pay for AST tests (71.8%, 173/241), urgent need for antibiotic therapy (56.8%, 137/241), long waiting time for AST results (35.3%, 85/241) and uncertainty of what to request from the lab to guide antibiotic selection (3.3%, 8/241) (Table S5).

The cost of antibiotics (80.9%, 195/241) and owners' ability to pay (81.3%, 196/241) were reported to influence the respondent's decision when selecting antibiotics. Other cost related influences included: expected profit margin to the veterinarian (27.4%, 66/241), marketing offers (16.6%, 40/241), adverts by pharmaceutical company representatives (12.9%, 31/241) and medicine sellers (8.3%, 20/241) (Table S6).

When asked what antibiotic characteristics had the most influence upon the veterinarian's selection of antibiotics, 85.5% (206/241) of respondents reported the antibiotic's spectrum of activity, 63.1% (152/241) reported AST results, 50.6% (122/241) reported withdrawal period, 44.4% (107/241) reported the ease of administration and 43.2% (104/241) reported the risk of development of AMR (Table S6). Other reported factors that influenced veterinarians' decision to select antibiotics were veterinarians' previous experience (96.3%, 232/241), advice from colleagues (68%, 164/241) and owner preference for a specific antibiotic (9.5%, 23/241) (Table S6).

Finally, when asked what sources of information influenced the veterinarians' decision the most when selecting an antibiotic to use, 79.3% (191/241) of respondents indicated veterinary education and training, followed by prescription guidelines or policies supplied by veterinary hospital or bodies (68.9%, 166/241), product labels or leaflets (64.3%, 155/241), legal restriction of drug to a defined species (38.6%, 93/241) and published scientific literature (35.3%, 85/241) (Table S6).

#### *2.5. Relationship between the Use of AST before Antibiotic Treatment and Select Investigated Parameters*

The relationship between the "use of AST before antibiotic treatment" and nine selected variables was assessed (Table S7). The analyses included 230 respondents that responded to the variable "use of AST before antibiotic treatment" and excluded respondents that answered "don't know" (n = 11). The proportion of respondents that reported the "use of AST before antibiotic treatment" was significantly different across the response levels for the following variables: years in practice (P = 0.049), knowledge of correct definition of antimicrobial stewardship (P = 0.032), knowledge of NCDC five points (P = 0.003), agreement with the statement that prophylactic use of antibiotics when farm biosecurity is poor is inappropriate (P = 0.029), and having a standard antibiotic treatment protocol in the veterinary practice (P ≤ 0.001) (Table S7).

#### *2.6. Relationship between Knowledge Level of Appropriate Antibiotic Use and AMR and Select Investigated Parameters*

The relationship between "knowledge level on appropriate antibiotic use and AMR" and selected variables was assessed (Table S8). The analyses included 240 respondents that responded to the variable "knowledge level on appropriate antibiotic use and AMR" (assigned to the category "high knowledge" or "low to moderate knowledge") and excluded one respondent that was assigned a knowledge score of zero because they had provided no correct answer. The proportion of respondents with high knowledge was significantly different across the response levels for the following variables: age group (P = 0.024), education level (P = 0.024), and agreement with the statement that prophylactic use of antibiotics when farm biosecurity is poor is inappropriate (P = < 0.001) (Table S8).

#### **3. Discussion**

The current study assessed the knowledge, attitudes, and practices towards AMR and antimicrobial stewardship of veterinary professionals in Nigeria. To the best of our knowledge, this is the first nationwide baseline study on the subject. Most of the veterinarians in the current study were between 25 and 44 years old, which probably explains why over 67% of respondents were within 10 years of being registered veterinary surgeons. Most of the respondents were private or government practitioners (69%), followed by teaching (10.4%) and non-governmental organization employees (9.1%); these results likely reflect the distribution of veterinarian employment in Nigeria.

Sixty three percent of the respondents were familiar with the term antimicrobial stewardship, compared to 17% reported in a similar study conducted in Enugu State, Nigeria [30]. The reasons for the observed differences are not clear but the current study can be considered more representative because it targeted participants across the country. However, our study highlights that there is still inadequate awareness of the concept of antimicrobial stewardship among veterinarians. Our study findings also revealed that the proportion of respondents who used AST before antibiotics administration varied among respondents who correctly or incorrectly defined antimicrobial stewardship. This finding suggests that educational strategies aimed at increasing awareness of antimicrobial stewardship among practicing and new veterinarians both at the practice and veterinary school levels may be helpful in promoting the responsible use of antibiotics.

Although most respondents reported that antibiotic resistance occurred in bacteria (98%) and could differentiate between an antibiotic and an antimicrobial (82%), a small percentage of respondents (4%) reported that antibiotics kill viruses, suggesting this proportion of respondents may prescribe antibiotics for viral infections. In comparison, 1% of student healthcare professionals (human and animal health students) surveyed in the United Kingdom thought antibiotics killed viruses [32], suggesting that study population may have been more knowledgeable on this aspect compared to our study population. Additionally, in our study, a small percentage of respondents (6%) reported that a withdrawal period of antibiotics-treated animals is not necessary before milk consumption. Failure to observe appropriate withdrawal periods following antibiotic treatment may result in the introduction of antibiotic residues in animal foods consumed by humans [33,34]. The failure to observe

appropriate withdrawal periods has potential human health implications including the development of drug-related allergies and hypersensitivity reactions, especially with beta lactam antibiotics and penicillin [35], and the risk of development of AMR [36,37].

Although most respondents knew that biosecurity is important in food production, this study highlighted a misconception regarding the link between biosecurity and antibiotic use. Over half of the respondents thought that prophylactic antibiotic use was appropriate in situations where biosecurity was poor. The reliance on antibiotics when biosecurity practices are poor has been reported in other studies [38]. Antibiotics can be an integral part of disease preventive methods but should be used only when indicated; they should not be used as the first line of action [39] or as a substitute for poor biosecurity practices [40]. Prophylactic and metaphylactic use of antibiotics administered to animal groups through water and feed may lead to increased environmental concentration of antibiotic residues which can in turn result in exposure of animals and humans alike and elevate the risk of AMR development [41,42]. Based on the current study findings there is a need for an improved understanding among veterinarians of the role of biosecurity practices in preventing and minimizing the risk of infections and reducing the overuse of antibiotics. Biosecurity practices are a key component of animal husbandry and disease prevention measures that can be implemented to improve animal health and welfare, and to reduce the need to use antimicrobials [43].

This study showed that the proportion of respondents with a high knowledge score varied with age group and education level. Although not conclusively established in the current study, this suggests that it is possible that older respondents and those with additional training to a veterinary degree may have a higher knowledge of appropriate antibiotic use and AMR or that the observed differences may be linked to work experiences accrued over time. Further investigation and understanding of the perceptions and barriers to responsible use of antibiotics among this subpopulation would help inform educational efforts.

More than half of the respondents' practices had a standard protocol for the treatment of sick animals to ensure the correct dosages and regimes are administered and to reduce the risk of adverse drug reactions or drug toxicity in the animals. Additionally, the proportion of respondents that reported the use of AST before antibiotic treatment was higher among those with a standard treatment protocol compared to those without. The observed relationship suggests having a standard treatment protocol may contribute to good antimicrobial stewardship. In the human sector, the use of standard treatment guidelines is considered an effective means of improving patient care while enhancing cost savings and changing behavior [44]. The treatment guidelines also reflect data on antimicrobial resistance, recognizing that local patterns of resistance often differ across geographical regions [44]. In the context of the animal health sector in Nigeria, strategies that consider up-to-date antimicrobial stewardship guidelines may be helpful in promoting appropriate antibiotic use among veterinary professionals.

Over half of the respondents (53%) reported conducting microbiological culture and AST before starting treatment, which is higher than the 24% reported in a study conducted in Enugu State, Nigeria [30], and the 38% reported among veterinarians in Europe [45]. The reason for these differences is not clear but may be related to regional differences in the level of awareness of the need to conduct microbiological culture and AST before starting antibiotic treatment or reporting bias which can occur in questionnaire-based studies. However, when asked specifically how often AST was requested before starting antibiotic treatment, only 20% of respondents reported requesting AST frequently (more than three times a month) and a third of the respondents never conducted or requested AST before commencing treatment. The observed drop from 53% to 20% may also be explained by reporting bias. Nevertheless, these findings suggest a considerable number of veterinary professionals do not use AST. There is a need to examine the reasons for the low AST use and identify appropriate interventions.

The respondents in the current study reported that poor response to initial antibiotic treatment (76%) and conditions that recur (71%) were the main factors that influenced their decision to conduct AST, but a lack of information on the animal or farm's health status, and owner requests were also influences, consistent with the findings from a study conducted in Europe [45]. The findings suggest having structured and up-to-date antibiotic use guidelines at the practice level, having access to rapid, cheap diagnostic tools and being able to handle clients' expectations through effective communication, may be helpful in providing guidance to veterinary professionals.

Several barriers to the use of AST were reported in this study. The unavailability of laboratory services and owner's inability to pay were reported as key barriers to the use of AST. The owner's inability to pay is a major concern because it limits the options available for a veterinarian in making decisions that allow for an appropriate diagnosis and treatment and may subsequently negatively influence the veterinarian's decisions and choices regarding antibiotic use. Strategies that explore ways to increase availability of veterinary laboratory services across the country and the provision of AST at affordable costs are necessary. Additionally, respondents reported that there is often an urgent need to administer antibiotics due to the acute onset of severe clinical signs. This is an accepted practice when antibiotics are urgently needed to counteract the disease progression and when delays in administering the therapy can lead to a poor outcome [46]. Nonetheless, veterinarians should take several factors into account to inform their decision on a sound empirical therapy: records from previous AST results in the local area if available, information on local patterns of bacterial resistance if available, previous patient cultural and AST results, the suspected anatomic site affected by infection and etiologic pathogen [47]. The records from previous AST results in the local area and information on local patterns of bacterial resistance were unlikely to be available in the context of Nigeria, emphasizing the need for a national integrated AMR surveillance program, as identified by the Nigeria National Action Plan for Antimicrobial Resistance [23]. The empirical antibiotic treatment should not prevent veterinarians from submitting samples for AST but instead be considered a temporary intervention while waiting for AST results that will inform the final, targeted, antibiotic treatment [46].

Veterinary education or training followed by prescription guidelines and policies were the most frequently selected parameters that influence a veterinarian's decision to select antibiotics. These findings suggest the veterinary curriculum may be a useful means to provide training on appropriate antibiotic use and selection. For example, in the fourth year of study, veterinary students in Nigeria undertake a course in pharmacology and therapeutics, which involves instruction on the types of veterinary pharmacological products and prescription practices. Furthermore, a relationship was observed between having additional education or training on AMR and knowledge of appropriate antibiotic use. Veterinary education or training may present an opportunity to expand and strengthen knowledge on appropriate antibiotic use practices and antimicrobial stewardship, if included in the training curriculum for new animal health professionals. Veterinary education or training also provides an opportunity for practicing veterinarians to update their knowledge, as prescription practices and protocols change over time.

The finding of antibiotic prescription guidelines as one of the most frequently selected parameters that influence a veterinarian's decision to select antibiotics suggests that there is a need for updated and increased dissemination and uptake of antibiotic guidelines. The Nigerian Veterinary formulary [48], provided by the Veterinary Council of Nigeria (VCN) to guide the prescription and administration of veterinary pharmaceuticals in different animal species was produced in 2007 and has not been updated since. In the present study, less than half of the respondents were aware of the Nigeria Center for Disease Control (NCDC) five-point agenda for AMR control, and a relationship was observed between awareness of the NCDC five-point action plan and AST use. These findings suggest regular updating of the VCN guidelines combined with increased awareness of the NCDC five-point action plan may be helpful in promoting antimicrobial stewardship among veterinary professionals.

A concerning, but not surprising, observation was that 86% of respondents reported encountering client-initiated antibiotic therapy without veterinary supervision. Even though the current regulatory policies in Nigeria require that only qualified veterinarians and para-vets can administer medications and treatment, livestock farmers can obtain and administer antibiotics without the requirement of a veterinarian's prescription and this is most likely to occur without regard to antibiotic indication guidelines [49]. This highlights the need for education of not just the veterinary professionals but also

the clients, veterinary drug sellers or shop keepers, pharmacies and farmers on appropriate antibiotic use and the risk of antibiotic misuse and AMR. In addition, government interventions such as the formulation and implementation of relevant policies and regulations may also be useful in improving appropriate antibiotic use and stewardship.

In the present study, the proportion of respondents that used AST before antibiotic treatment varied with knowledge of antimicrobial stewardship, knowledge of NCDC's five points, and with agreement with the statement that prophylactic use of antibiotics is inappropriate when biosecurity is poor. These findings suggest that knowledge of antimicrobial stewardship, NCDC's five points and prophylactic use of antibiotics may be related to appropriate antibiotic use. These areas could be targeted when developing strategies to improve antimicrobial stewardship and reduce AMR in veterinary practice in Nigeria.

A few limitations were observed during the conduct of the current study. The response rate of the survey was low, and as such, the study's findings may not be generalizable to the whole of the country. Nevertheless, the gaps identified can still be used to inform discussions by policy makers involved in the development of interventions targeting all veterinarians in Nigeria. The low response rate may have been due to several factors such as unwillingness to participate or lack of internet access in some parts of the country. There may also have been selection bias in the respondent population. For example, it is possible that respondents that completed the survey were more technologically astute or inclined. It could also be that these respondents had a special interest in the subject, hence their participation. Another potential limitation was social desirability bias which might have affected the nature of the responses provided; it is possible that some respondents may have declined to share information they considered inappropriate or erroneous, resulting in an under-reporting of certain aspects on antibiotics and AMR knowledge and practices. Finally, it is important to note that the data analyses performed to assess relationships between selected variables in the current study were exploratory in nature and were not intended to be exhaustive, therefore, no additional inferential analyses such as logistics regression were conducted.

#### **4. Materials and Methods**

Ethical review and approval were granted by the Nigeria Ministry of Defense Health Research Ethics Committee, Abuja, via an ethics review application (Ethics approval number: MODHREC/APP/20/12/11/20/1/8/) and by the Research Integrity and Governance Office at the University of Surrey, United Kingdom (Response ID: 353003-352994-41119696).

#### *4.1. Study Area*

Nigeria is the most populous nation located in West Africa, with an estimated population of approximately 202 million people [50]. Crude oil, agriculture and solid minerals are the mainstay of the economy. It has 37 states including the Federal Capital Territory, Abuja. There are three major tribes and about 250 ethnic groups. Nigeria has a tropical climate and two distinct weather seasons (rainy and dry seasons). To the north of the country is the Sahel climate, to the west is the tropical savannah while the south and east are characterized by tropical monsoon climates.

#### *4.2. Study Population*

There are circa 8000 registered veterinarians in Nigeria involved in livestock/large animal practice, small animal/companion animal practice, poultry practice, public health and academia (personal communication with Interim College secretary, College of Veterinary Surgeons, Nigeria). This study involved veterinarians registered with the Veterinary Council of Nigeria (VCN). For this study, a registered veterinarian was an individual who obtained certification from the VCN as a Doctor of Veterinary Medicine (DVM) upon completion of the six-year university degree program in Nigeria. All registered veterinarians were pooled together irrespective of the type of practice.

#### *4.3. Study Design*

This was a cross-sectional study; a questionnaire survey was designed and used to collect data on the knowledge and attitudes towards AMR and antimicrobial stewardship of veterinarians during the period January to February 2019.

#### *4.4. Sample Size*

Sample size was estimated as described by Lwanga and Lemeshow (1990) [51] with the prevalence of knowledge of antibiotic resistance set at 50% based on a previous study that investigated the veterinary drugs market in Nigeria [21] and the desired level of precision set at 0.02. The estimated calculated sample size was 2400. A non-response rate of 30% was estimated because studies have shown response rates to web-based surveys are generally low [52] and assumed some respondents would have poor internet access or did not consent. The final estimated minimum sample size considering the non-response rate was 3120 respondents (2400 × 1.3).

#### *4.5. Data Collection*

A questionnaire was developed and select questions from previous studies [32,45] were adapted to collect data on demographics, knowledge, attitudes, practices towards antibiotic use and resistance, and awareness of antimicrobial stewardship (Supplementary material S1). The tool was pretested among ten veterinarians in Abuja and thereafter questions were further refined to produce the final survey. The final questionnaire was administered electronically using the Qualtrics® survey platform.

The available list of 5800 registered veterinarians in Nigeria was obtained from the VCN, which was more than the calculated minimum sample size of 3120 respondents. A total of 5603 with phone numbers were randomly selected from this pool of 5800 using a table of random numbers. A link to the survey was sent via a text message to all the 5603 contacts. Of the 5603 contacts, 2662 had email addresses on the VCN register, and the survey link was sent to them via email (in addition to the text message sent to all 5603 contacts). The message inviting contacts to participate in the survey was endorsed by the Nigeria Center for Disease Control (NCDC) and Federal Ministry of Agriculture and Rural Development (FMARD). Further emails and SMS reminders were sent 2, 4 and 6 weeks after the initial message. The survey was made available online for 8 weeks between 2nd January and 28th February 2019.

#### *4.6. Data Analysis*

Survey results were downloaded from Qualtrics® to Microsoft Excel. Data collected during the piloting of the survey were excluded from the final analysis. Descriptive statistics were used to summarize the data using R-Studio 1.2.1335.0.

As part of the descriptive analysis, a scoring system was used to assess the knowledge level on antibiotic use and AMR. A set of nine survey questions on knowledge was selected, and for each question, a score of one point was assigned for each correct answer, with a maximum of nine points allowed (Table S9). One respondent that provided no correct answer was assigned a score of zero. Respondents were further regrouped into two categories based on knowledge score; respondents scoring ≥ 7/9 points were assigned to the category "high knowledge" and those scoring < 7/9 points were assigned to the category "low to moderate knowledge".

Following the descriptive analysis, a bivariate analysis was performed using the Chi-square test or Fisher's exact test, as appropriate, to explore the relationship between two selected outcome variables and eleven selected variables. The outcome variables were: "the use of AST before antibiotic treatment" and "knowledge level on appropriate antibiotic use and AMR". The relationships between these two outcome variables and the following 9 variables were assessed: age group of participants, gender, educational level, years in practice, type of practice, practice location, knowledge of the NCDC five-point action plan for correct antibiotic use, prophylactic use of antibiotic when biosecurity is poor,

and existing antibiotic treatment protocol in practice. Two further variables, the correct definition of antimicrobial stewardship and owner-initiated treatment, were investigated for the outcome variable "the use of AST before antibiotic treatment". Two additional variables were also investigated for the outcome variable "knowledge level on appropriate antibiotic use and AMR": type of employment and knowledge of antimicrobial stewardship. A two-tailed P-value of ≤0.05 was considered statistically significant. It is important to note that the association analyses were exploratory in nature, used selected variables and were not intended to be exhaustive, therefore, no advanced additional inferential analyses, such as logistics regression, were conducted.

#### **5. Conclusions**

Findings from this study provided baseline evidence on the knowledge, attitudes and practices regarding antibiotic use, AMR and antimicrobial stewardship among veterinarians across Nigeria. With respect to knowledge and attitudes on appropriate antibiotic use and AMR, there was little awareness of the concept of antimicrobial stewardship among veterinarians, and the role and use of biosecurity, as well as the prophylactic antibiotic use in the prevention of infection, were not well understood. There is a need for an increased understanding among veterinarians for how the use of biosecurity practices plays a role in the prevention of infection, reducing the burden of disease in animal populations and, therefore, in reducing the need for and use of antibiotics. Education or training strategies aimed at increasing awareness of antimicrobial stewardship among practicing and new veterinarians at the practice and veterinary school levels may be helpful in promoting antimicrobial stewardship.

Regarding practices and factors influencing antibiotic use, the use of AST to inform antibiotic treatment was low, suggesting a need to further examine the reasons for this and identify appropriate interventions. The unavailability of laboratory services and the client's inability to pay were reported as key barriers to AST use. Strategies that explore ways to increase the availability of veterinary laboratory services across the country and the provision of AST at affordable costs are necessary.

Veterinary education or training followed by prescription guidelines and policies were the most frequently selected parameters that influence a veterinarian's decision to select antibiotics. The regular updating of the antibiotic prescription and use guidelines combined with increased awareness and dissemination among veterinarians may be helpful in promoting antimicrobial stewardship. Finally, the reported client-initiated antibiotic therapy was also a concern highlighting the need for education of not just the veterinary professionals, but also their clients and drug shop keepers on appropriate antibiotic use and stewardship.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2079-6382/9/8/453/s1, Supplementary material S1: Questionnaire. Table S1. Knowledge of appropriate antibiotics use and AMR among respondents; Table S2. Attitudes of respondents towards antibiotic use and antimicrobial resistance (AMR); Table S3. Contributors to antibiotic resistance, as reported by respondents; Table S4. Factors influencing antibiotic prescription practices and choice; Table S5. Frequency of antimicrobial susceptibility testing (AST) use and barriers to AST use; Table S6. Practices and factors influencing respondents' decisions on selecting antibiotics and antimicrobial susceptibility testing (AST) use; Table S7. Relationship between the use of antimicrobial susceptibility testing (AST) before antibiotic treatment and select investigated variables. Table S8. Relationship between knowledge level on appropriate antibiotic (AB) use and antimicrobial resistance (AMR) and select investigated variables; Table S9. Scoring of knowledge level on appropriate antibiotic (AB) use and antimicrobial resistance (AMR). For each question, a score was assigned based on the response with 0 given for an incorrect response and 1 for a correct response.

**Author Contributions:** Conceptualization, U.O.A., A.B.M., A.B.E., E.M., G.V.; methodology, U.O.A., E.G., A.B.E., A.M., R.M.L.R.; software, E.G., A.B.E; formal analysis, U.O.A., E.G., A.B.E.; investigation, U.O.A., A.W.; writing—original draft preparation, U.O.A., E.G., A.E.; writing—review and editing, U.O.A., E.G., A.B.E., A.B.M., A.M., R.M.L.R., B.A., A.W., R.A., E.M., G.V., A.J.C.C.; supervision, A.B.E, A.J.C.C.; project administration, U.O.A., A.B.E; funding acquisition, A.J.C.C., G.V. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research work was delivered by the University of Surrey and funded by Zoetis' Africa Livestock Productivity and Health Advancement (ALPHA) Initiative, co-funded by the Bill and Melinda Gates Foundation (BMGF) and Zoetis. Funding from Zoetis was an unrestricted grant. BMGF Grant number: OPP1165393.

**Acknowledgments:** The authors would like to express their sincere gratitude to the following people for their support towards this research work. First, thank you to Tetiana Miroshnychenko of Zoetis' ALPHA Initiative Zaventem team and Isaac Odeyemi from Zoetis Outcomes Research team for guidance. We are grateful to the Veterinary Council of Nigeria who assisted with the sharing of the survey and the Nigerian Center for Disease Control for supporting the study. We are also grateful to Adam Trish of the vHive team at the School of Veterinary Medicine, University of Surrey for administrative assistance, and to Ikenna Onoh of the Nigerian Field Epidemiology and Laboratory Training Program and Pharmacist Estelle Mbadiwe of Ducit Blue Solutions for sharing insights on study implementation. Finally, the authors thank the veterinarians who spared their time to complete the questionnaire, without them, this study would not have been possible.

**Conflicts of Interest:** This study was supported by the Zoetis' Africa Livestock Productivity and Health Advancement (ALPHA) Initiative, co-funded by the Bill and Melinda Gates Foundation and Zoetis. Funding from Zoetis was an unrestricted grant. The Zoetis team (E.M., G.V.) was involved in study conceptualization and the writing of the manuscript. There are no potential areas of bias which may confer any advantage to Zoetis.

#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

#### *Article* **Long-Term Impact of an Educational Antimicrobial Stewardship Program on Management of Patients with Hematological Diseases**

**Ana Belén Guisado-Gil 1,2 , Manuela Aguilar-Guisado 1,\*, Germán Peñalva <sup>1</sup> , José Antonio Lepe <sup>1</sup> , Ildefonso Espigado <sup>3</sup> , Eduardo Rodríguez-Arbolí 3 , José González-Campos <sup>3</sup> , Nancy Rodríguez-Torres <sup>3</sup> , María Isabel Montero-Cuadrado <sup>3</sup> , José Francisco Falantes-González <sup>3</sup> , Juan Luis Reguera-Ortega <sup>3</sup> , María Victoria Gil-Navarro <sup>2</sup> , José Molina <sup>1</sup> , José-Antonio Pérez-Simón <sup>3</sup> and José Miguel Cisneros <sup>1</sup>**

	- <sup>2</sup> Department of Pharmacy, University Hospital Virgen del Rocio, 41013 Seville, Spain; mariav.gil.sspa@juntadeandalucia.es
	- <sup>3</sup> Department of Hematology, Institute of Biomedicine of Seville (IBiS/CSIC/CIBERONC), University Hospital Virgen del Rocio, University of Seville, 41013 Seville, Spain; ildefonso.espigado.sspa@juntadeandalucia.es (I.E.); edurodarb@gmail.com (E.R.-A.); jose.gonzalez.sspa@juntadeandalucia.es (J.G.-C.); nanarotor@hotmail.com (N.R.-T.); mariai.montero.sspa@juntadeandalucia.es (M.I.M.-C.); josef.falantes.sspa@juntadeandalucia.es (J.F.F.-G.); juanlu\_jlr@hotmail.com (J.L.R.-O.); josea.perez.simon.sspa@juntadeandalucia.es (J.-A.P.-S.)
	- **\*** Correspondence: maguilarguisado@yahoo.es; Tel.: +34-670943816

**Abstract:** Antimicrobial stewardship programs (ASPs) in hematological patients are especially relevant. However, information about ASPs in this population is scarce. For 11 years, we quarterly assessed antimicrobial consumption and incidence and death rates of multidrug-resistant (MDR) bloodstream infections (BSI) in the hematology Department. Healthcare activity indicators were also monitored yearly. We performed an interrupted time-series analysis. Antimicrobials showed a sustained reduction with a relative effect of −62.3% (95% CI −84.5 to −40.1) nine years after the inception of the ASP, being especially relevant for antifungals (relative effect −80.4%, −90.9 to −69.9), quinolones (relative effect −85.0%, −102.0 to −68.1), and carbapenems (relative effect −68.8%, −126.0 to −10.6). Incidence density of MDR BSI remained low and stable (mean 1.10 vs. 0.82 episodes per 1000 occupied bed days for the pre-intervention and the ASP period, respectively) with a quarterly percentage of change of −0.3% (95% CI −2.0 to 1.4). Early and late mortality of MDR BSI presented a steady trend (quarterly percentage of change −0.7%, 95% CI −1.7 to 0.3 and −0.6%, 95% CI −1.5 to 0.3, respectively). Volume and complexity of healthcare activity increased over the years. The ASP effectively achieved long-term reductions in antimicrobial consumption and improvements in the prescription profile, without increasing the mortality of MDR BSI.

**Keywords:** antimicrobial stewardship; anti-infective agents; bacteremia; candidemia; hematologic diseases

#### **1. Introduction**

Antimicrobial stewardship programs (ASPs) have been identified as a valuable tool to optimize the antimicrobial use in healthcare centers, improving patient outcomes and reducing adverse events and the selection pressure related to the use of antimicrobial agents [1].

In hematological patients receiving immunosuppressive therapy, collateral damages of antimicrobial consumption, especially broad-spectrum antibiotic therapy, include the selection of multidrug-resistant (MDR) microorganisms [2], an increased propensity to

**Citation:** Guisado-Gil, A.B.; Aguilar-Guisado, M.; Peñalva, G.; Lepe, J.A.; Espigado, I.; Rodríguez-Arbolí, E.; González-Campos, J.; Rodríguez-Torres, N.; Montero-Cuadrado, M.I.; Falantes-González, J.F.; et al. Long-Term Impact of an Educational Antimicrobial Stewardship Program on Management of Patients with Hematological Diseases. *Antibiotics* **2021**, *10*, 136. https://doi.org/ 10.3390/antibiotics10020136

Academic Editors: Albert Figueras and Diane Ashiru-Oredope Received: 16 December 2020 Accepted: 27 January 2021 Published: 30 January 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/).

fungal infections [3], and microbiota dysbiosis [4]. Although, due to these reasons the impact of ASPs in patients with hematological diseases might be especially relevant, information regarding the development of antimicrobial stewardship strategies in these patients is scarce [5–7].

An ASP named Institutional Program for the Optimization of Antimicrobial Treatment (PRIOAM) started in our institution in January 2011. Since then, assessments of antimicrobial use, in-hospital bacterial resistance, and mortality rates associated with nosocomial bloodstream infections (BSI) have decreased significantly [8,9]. This program covers the entire hospital and presents specific interventions focused on hematological patients.

We hypothesized that a comprehensive ASP in hematological patients could also optimize antimicrobial use, reducing the overall consumption and improving the prescription profile without increasing the incidence and mortality rates of BSI produced by MDR microorganisms. Thus, the objective of the present study was to assess the impact of the PRIOAM on antimicrobial consumption and the incidence and death rates caused by MDR BSI in hospitalized adult patients with hematological diseases.

#### **2. Results**

Since the inception of the ASP and as part of PRIOAM educative measures (see "Intervention" at the Materials and Methods section), a total number of 218 face-to-face structured educational interviews (EI) were performed (mean 24 ± 19 EI per year). The main reasons for inappropriate antimicrobial therapy were: an incorrect selection of the drug according to the suspected diagnosis (28.4%) or inappropriate duration (28.0%) in the case of empiric treatments, and failing to de-escalate (11.0%) in the case of targeted therapies.

Moreover, 18 clinical sessions (two per year) were performed about practical aspects of common infections in hematological malignancy patients and 45 reports were produced, including one per quarter and an additional annual report to the head of the department, on the level of attainment of pre-agreed objectives.

#### *2.1. Antimicrobial Consumption*

The mean consumption of all antimicrobials decreased from 148.2 ± 16.2 defined daily doses (DDD) per 100 occupied bed days (OBD) in the pre-intervention period to 112.0 ± 21.7 DDD per 100 OBD in the ASP period (*p* < 0.001). Detailed data from the pre-post analysis are included in the Supplementary material (Table S1).

The interrupted time-series (ITS) analysis (Table 1, Figures 1–3) showed a sustained reduction in favor of the intervention with a relative effect of −62.3% (95% confidence interval [CI] −84.5 to −40.1) nine years after the inception of the ASP, when compared with the expected antimicrobial consumption based on the pre-intervention trend. As for antibiotics, a prompt change in the level after the inception of the ASP of −17.22 DDD per 100 OBD (95% CI −29.17 to −5.28) was found. Regarding antifungal consumption, a decreasing trend with a change in slope of −3.32 DDD per 100 OBD (95% CI −6.04 to −0.60) and a relative effect −80.4% (95% CI −90.9 to −69.9) was obtained with the intervention. Quinolones were the agents that showed the highest reduction with a change in the level of −18.45 DDD per 100 OBD (95% CI −25.29 to −11.62) after the start of the intervention that led to a relative effect of −85.0% (95% CI −102.0 to −68.1) at the end of the study period. Broad-spectrum antibiotics such as carbapenems and glycopeptides presented significant relative effects of −68.8% (95% CI −126.0 to −10.6) and −70.5% (95% CI −138.9 to −2.1), respectively, compared with the expected consumption based on the pre-intervention trend. The global trend is described in Table S2.


**Table 1.** Interrupted time-series analysis of changes in trends of antimicrobial consumption. **Table 1.** Interrupted time-series analysis of changes in trends of antimicrobial consumption.

*Antibiotics* **2021**, *10*, x 3 of 14

Data are presented as quarterly defined daily doses per 100 occupied bed days with a 95% confidence interval, unless otherwise specified. a Increase or decrease in the first quarter after the start of the antimicrobial stewardship program (ASP) period with respect to the expected value. <sup>b</sup> Change in slope for the ASP period. <sup>c</sup> Percentage difference between the expected value according to the pre-intervention trend and the trend nine years after the start of the ASP. otherwise specified. <sup>a</sup> Increase or decrease in the first quarter after the start of the antimicrobial stewardship program (ASP) period with respect to the expected value. <sup>b</sup>Change in slope for the ASP period. <sup>c</sup>Percentage difference between the expected value according to the pre-intervention trend and the trend nine years after the start of the ASP.

**Figure 1.** Interrupted time-series analysis of the trends in antimicrobial consumption (antibacterials for systemic use and antifungals) observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter. **Figure 1.** Interrupted time-series analysis of the trends in antimicrobial consumption (antibacterials for systemic use and antifungals) observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter.

**Figure 2.** Interrupted time-series analysis of the trends in consumption for antibacterials for systemic use (**J01**) and antifungals (**J02**) observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter. **Figure 2.** Interrupted time-series analysis of the trends in consumption for antibacterials for systemic use (**J01**) and antifungals (**J02**) observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter.

**Figure 3.** Interrupted time-series analysis of the trends in consumption for carbapenems and quinolones observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter. **Figure 3.** Interrupted time-series analysis of the trends in consumption for carbapenems and quinolones observed before and after the implementation of the antimicrobial stewardship program. Solid lines show the observed trend during the pre-intervention and intervention periods. Dashed lines show the expected trend after the intervention according to the pre-intervention values. DDD, defined daily doses. OBD, occupied bed days. Q, quarter.

#### *2.2. Clinical Outcomes*

For the entire study period, the most common gram-negative microorganism causing BSI was non-extended-spectrum beta-lactamase (ESBL) producing *Escherichia coli* (48.1%). MDR gram-negative bacteria and *Candida* spp. caused 14.4% and 5.6% of BSI that were monitored, respectively (Table S3). BSI produced by MDR *Pseudomonas aeruginosa* and *Candida* spp. were responsible for the highest values of early and late mortality rates (Table S4).

For incidence density (ID) and mortality rate, the ITS analysis is shown in Table 2. The pre-post analysis and the trend analysis can be found in Tables S5 and S6, respectively.

**Table 2.** Interrupted time-series analysis of changes in trends of incidence and mortality rate of multidrug-resistant bloodstream infections.


Data are presented as quarterly incidence density and all-cause crude death rate per 1000 occupied bed days with a 95% confidence interval, unless otherwise specified. <sup>a</sup> Increase or decrease in the first quarter after the start of the antimicrobial stewardship program (ASP) period with respect to the expected value. <sup>b</sup> Change in slope for the ASP period. <sup>c</sup> Percentage difference between the expected value according to the pre-intervention trend and the trend nine years after the start of the ASP.

> The ID of BSI caused by MDR organisms, which kept low during the entire study period, remained stable (mean incidence 1.11 episodes per 1000 OBD for the pre-intervention period and 0.82 episodes per 1000 OBD for the ASP period) with a quarterly percentage change (QPC) of −0.3% (95% CI −2.0 to 1.4, *p* = 0.709). Early and late mortality of MDR BSI presented a steady trend with a QPC of −0.7% (95% CI −1.7 to 0.3, *p* = 0.154) and −0.6% (95% CI −1.5 to 0.3, *p* = 0.201), respectively.

#### *2.3. Changes in Healthcare during the Study Period*

Activity indicators related to the volume and complexity of the hematology department such as the number of blood cultures per 1000 OBD, total admissions, OBD, and the number of allogeneic hematopoietic stem-cell transplantation (HSCT) increased during the study period. Other indicators of the department's activity remained stable (Table 3).


**Table 3.** Indicators related to the volume and complexity of the activity at the hematology department.

For each year, data are presented as the number of events, unless otherwise specified. In the last column, the annual percentage of change (APC) obtained from joinpoint regression analysis with a 95% confidence interval (CI) is included. OBD, occupied bed days. AML, acute myeloid leukemia. HSCT-related mortality, hematopoietic stem-cell transplantation (HSCT)-related mortality within the first 100 days after allogeneic HSCT from human leukocyte antigen (HLA)-identical siblings.

#### **3. Discussion**

The results of our study show that an education-based ASP in the hematology department was able to achieve long-term reductions in overall antimicrobial consumption and improvements in the prescription profile, especially relevant in broad-spectrum antibiotics such as carbapenems, quinolones, and antifungals, without increasing the mortality rates and maintaining a low incidence of MDR BSI. This positive impact was observed in a tertiary care hospital where infectious diseases consultation (IDC) was performed for more than 25 years and up to the PRIOAM implementation. To the best of our knowledge, this is the first study proving nine years' data on the benefits of ASPs in the setting of hematological patients.

Very few previous studies have evaluated the effect of ASPs on antimicrobial consumption in hematological patients with most of them limited by sample size, study period (<2 years before or after intervention), and the absence of data about specific groups of antibiotics. One of the most rigorous is the study performed by So et al. [10] in leukemia units with audit and feedback as the core measures of the ASP. In contrast to our results, the intervention was associated with a significant decrease in antibiotic use (−35.1 DDD per 100 patient-days), but no significant trend in antifungal prescription was observed during a two-year period (−4.0 DDD per 100 patient-days). Two other research works with a one-year evaluation time and including solid and hematological malignancy patients showed favorable results after the beginning of antimicrobial stewardship strategies in terms of global antibiotic consumption [11] and meropenem prescription [7].

The increase of infections caused by MDR bacteria is a major health problem worldwide [12]. This challenge also affects hematological patients [13–15]. However, the percentage of MDR bacteria in our center was lower than previously reported by others [15–18], and notably, remained stable throughout the study period. The low prevalence of ESBL producing *E. coli* and MDR *P. aeruginosa* and, particularly, the absence of carbapenemaseproducing *Enterobacteriaceae* was especially important. The sustained reduction of the use of all-class antibiotics associated with the intervention has likely contributed to preventing the generalized increase in MDR infections described in other centers.

In our study, the early and late mortality rates from MDR BSI remained stable during the intervention, showing the absence of deleterious effects for reducing antimicrobial use in these patients. Death rates were higher for BSI caused by MDR organisms as

described before [12,18]. Late mortality for ESBL producing *Enterobacteriaceae* was inferior to the 21-day mortality rate reported by Trecarichi et al. [18] for non-susceptible strains (26.2%) and comparable for MDR *P. aeruginosa* (42.4%). Despite the differences in study design, population and antimicrobial utilization due to different local treatment protocols and colonization rates by MDR bacteria, preceding results, similar to ours, illustrated the potential benefits of antimicrobial stewardship approaches. The adherence to ASP recommendation has demonstrated to be an effective and safe strategy with a 64% relative risk reduction in 28-day mortality [19] and a significant decrease in the fatality rate (from 30% to 11%) [5] both in patients with febrile neutropenia and hematological or solid tumors. In patients with hematological diseases and HSCT recipients, stopping antimicrobial therapy early did not significantly increase the incidence of fever relapse and positive blood cultures or the mortality rate, with the advantage of the reduction in the use of antibiotics [20–22].

The results of the current study are even more remarkable if it is taken into consideration that most indicators related to the volume and complexity of the activity at the hematology department increased considerably during the study period. A fact that, in general, is related to a higher frequency of infectious complications and, as a consequence, higher consumption of antibiotics. Only the number of patients diagnosed with acute myeloid leukemia (AML), the length of hospital stay, and the transplant-related mortality within the first 100 days after allogeneic HSCT from human leukocyte antigen (HLA) identical siblings remained stable. The monitoring of changes in healthcare as an internal control as well as the largest period of study, spanning 11 years in total, are some of the strengths of this work. Additionally, the employment of ITS analyses, the preferential method to assess the impact of health interventions over time [23], and the consistent results throughout the variables evaluated, support a potential causality relation between the ASP implementation and the progressive reduction in the antimicrobial pressure. The stable trend in the mortality by MDR BSI supports the safety of the intervention.

PRIOAM's methods diverge from those ASP performed previously in patients with onco-hematological diseases in which educational initiatives were not incorporated as the core element of the program [11], or they were based on a sole recommendation (deescalation, discontinuation, antibiotic cycling, etc.) [6,20,21] and/or a specific diagnosis [10], commonly febrile neutropenia [17,19–21]. The educational nature combined with real-time intervention(s) and the inclusion of patients with all types of hematological diseases comprise a differentiating feature of this work.

In patients with hematological diseases, post-chemotherapy febrile neutropenia was one of the most frequent infectious syndromes requiring antimicrobial courses. In this sense, the contribution of the results of the How Long clinical trial [24], led by investigators from our institution, to change the clinical practice and to decrease antibiotic overpressure in hematological patients was considerable. According to the main findings, in high-risk patients with hematological malignancies and febrile neutropenia, empirical antimicrobial therapy can be safely discontinued after 72 h of apyrexia and clinical recovery irrespective of their neutrophil count. It reinforced the previously published recommendation from the 4th European Conference on Infections in Leukaemia (ECIL-4) about empirical treatment of febrile neutropenia [25].

In our center, quinolones were agents commonly selected as an empirical combination therapy in patients with febrile neutropenia, especially in those with a suspected respiratory infection. Quinolones showed the highest reduction after the start of the intervention and the greatest decrease in the relative effect at the end of the study period. It could be explained by the fact that the ASP guidelines highly recommended the withdrawal of combination therapy 48 h after the start if an infection was not confirmed, or if it was presented and narrower spectrum antibiotics could be employed instead of quinolones. The implementation of this recommendation through the ASP has likely contributed to achieving this result.

However, for this study, some limitations should be noted. First, the study design is not exempt from the possibility of ecologic bias, and, consequently, we could not unequivocally associate the results of incidence and mortality of MDR organisms to the ASP implementation. Although the volume and complexity of the activity in the hematology department were monitored, other potential confounding factors such as those related to patients and the center could possibly interfere with the outcomes. In addition, the single-center design limits the external validity of our results and makes it necessary to confirm the reproducibility of the findings in different settings. Second, the close relationship between the IDC in the pre-intervention period and the ASP made it difficult to elucidate the precise weight of each one on the outcomes achieved. Nonetheless, regarding the use of antimicrobials, the stable trends during the pre-intervention period suggest that the implementation of the ASP was necessary to achieve the goals. The sole IDC was insufficient to promote a change in the entire department, as reported in previous studies [26]. Finally, non-MDR BSI, invasive candidiasis (other than candidemia) and aspergillosis have not been examined in this study. The decreasing trend in overall antimicrobial consumption, including voriconazole and liposomal amphotericin B as common treatments for invasive infections caused by molds [27], suggests, at least, a steady frequency of these infections.

#### **4. Materials and Methods**

#### *4.1. Study Design and Period*

A quasi-experimental before-after study of ITS was performed. The PRIOAM implementation started in January 2011, and, since then, data were prospectively registered for a nine-year period. For the ITS analyses, the study period spanned 44 quarters (11 years) from January 2009 to December 2019.

#### *4.2. Setting*

The program was performed at the 39-bed hematology department of the University Hospital Virgen del Rocio (Seville, Spain), which is a teaching hospital providing a tertiarycare service in Southwest Spain. The hospital, with 1177 beds and 72 intensive care unit beds, is a referral-center for solid-organ and HSCT. Adult patients (aged ≥ 18 years) receiving treatment for hematological malignancies or undergoing HSCT are treated in this unit. Throughout the last nine years, the hematology department has admitted a mean number of 1134 adult patients per year and has performed a mean of 108 autologous and allogeneic HSCT per year in adults.

#### *4.3. Intervention*

The PRIOAM's methods and global outcomes have already been published [8,9]. In brief, it comprises a bundle of educational strategies performed by a multidisciplinary team including infectious diseases physicians, microbiologists, pharmacists, intensive care physicians, pediatricians, and preventivists. The core elements of PRIOAM are summarized in the Supplementary material (Figure S1).

Because most EI were performed when a potentially inappropriate prescription was detected (i.e., use of carbapenems or combination therapy for >48 h, antibiotic duration >7 days or targeted therapies), the main messages tackled in EI were: early identification and management of severe infections, interpretation of microbiologic results, de-escalation and sequential oral treatments whenever possible, diversification of antimicrobial prescriptions, and training in the optimal duration of antimicrobial courses. The form employed for EI is included as Figure S2. No other interventions concerning antimicrobial use (i.e., antimicrobial policies, restrictions, etc.) were performed during the study period. The infection control program in the hematology ward consisted of the isolation in high-efficiency particulate air (HEPA) filters conditioned rooms of neutropenic and HSCT patients, and contact isolation of patients with MDR bacteria or respiratory viruses recovered from clinical samples. Local guidelines for antifungal prophylaxis did not change substantially during the intervention period. No additional measures were implemented regarding

infection prevention, and antibiotic prophylaxis was not recommended for hematological patients in our center since 2005.

Before the start of the PRIOAM, a stable IDC program was running at the hematology department, consisting of bedside advice for the management of complex infections, quick report of all BSI, the production and application of local guidelines, updated every two years, for the prevention, diagnosis, and management of infections, and surveillance and analyses of MDR outbreaks. The usual IDC led the implementation of antimicrobial stewardship tasks in the hematology department.

#### *4.4. Study Measures*

Antimicrobial use was evaluated through quarterly measures of the antibiotic consumption of the Anatomical Therapeutical Chemical (ATC) group J01 (antibacterials for systemic use) and antifungals ATC group J02. Data about antimicrobial consumption were automatically generated by the electronic prescribing system, which provided information about the units (capsules, injection vials, etc.) used by each department. Consumption was calculated as DDD per 100 OBD, according to the ATC Classification methodology and the 2019 World Health Organization DDD values [28]. Because no DDD was suggested for liposomal amphotericin B, we considered the 210 mg dose as the unit.

For the study period, BSI caused by the most relevant microorganisms in patients with hematological diseases (*E. coli*, *Klebsiella pneumoniae*, *P. aeruginosa*, and *Candida* spp.) were registered. The effect of the intervention on the number of BSI produced by MDR microorganisms (ESBL-producing *Enterobacteriaceae*, MDR *P. aeruginosa*, and *Candida* spp.) was monitored quarterly and presented as ID per 1000 OBD. The German Society for Hygiene and Microbiology criteria [29] was taken into account for MDR categorization. The analysis of antibiotic susceptibility and resistance mechanisms was performed following the European Committee on Antimicrobial Susceptibility Testing (EUCAST) criteria [30,31].

The effect on the mortality rates was assessed as the all-cause crude death rate [9,32] (deaths per 1000 OBD per quarter) on day +7 (early mortality) and +30 (late mortality) after the diagnosis of BSI. Patients dying in less than 24 h after blood sample collection were not considered for the mortality analysis, as previously proposed [26,33,34], for a better selection of patients benefitting from the intervention targeting an optimized use of antimicrobials.

To analyze the effect of changes in the hematology department during the 11-year study period, we monitored yearly indicators related to the volume and complexity of the activity at the department that may influence the antimicrobial use, such as the number of blood cultures per 1000 OBD, new patients diagnosed with AML, admissions, and OBD, as well as the mean length of stay. We also monitored the number of allogeneic HSCT and the transplantation-related mortality within the first 100 days after allogeneic HSCT from HLA-identical siblings.

Because presentation, dissemination, and introduction activities of PRIOAM took place in the different departments of the hospital from January to 31 March 2011, we considered 1 April as the beginning of the intervention period for the analysis.

#### *4.5. Statistical Analysis*

For descriptive aims, categorical variables were presented as frequency distribution and percentages, and continuous variables were presented as means ± standard deviations (SD). The Student's *t*-test or the Mann-Whitney U test were employed for univariate pre-post analyses, after checking for normality using the Kolmogorov-Smirnov test.

To assess the effect of the ASP, an ITS analysis was performed to estimate changes in the level and trends before and after the inception of the program. We used a generalized least squares regression approach accounting for autocorrelation by autoregressive moving-average (ARMA) models. The final model selection for each variable was based on the Akaike Information Criterion with validation of the autocorrelation structures by likelihood ratio tests [35]. The long-term effect attributable to the ASP for each outcome

was estimated by calculating the relative effect, as the percentage difference between the values of the expected pre-intervention trend and the modeled trend at the end of the study. Alternatively, a joinpoint regression analysis was conducted to explore the trends of the time-series [36], calculating the QPC during the 11-year study period by using the Joinpoint software modeling annual percentage change calculation to our log-transformed quarterly data with autocorrelated error models.

Confidence intervals or *p-*values (*p*) were included to show statistical significance. Differences were considered statistically significant at *p* < 0.05 (2-tailed tests). Statistical analyses were performed with IBM SPSS Statistics software v. 23.0, R software v. 3.5.2 and Joinpoint Regression Program v. 4.6.0.0.

#### *4.6. Ethics Approval*

The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of the University Hospital Virgen del Rocio (Project identification code: PI-0361-2010).

#### **5. Conclusions**

These results allow us to state that an education-based ASP contributed significantly to the decreasing trend in the use of antimicrobials and, possibly, to maintain the low incidence of MDR BSI despite the increase in the volume and complexity of the activity at the hematology department over the study period. Death rates of BSI caused by MDR organisms were stable, showing that these interventions are safe in this vulnerable population.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2079-6 382/10/2/136/s1. Table S1: Differences between the pre-intervention period and the antimicrobial stewardship program period regarding pre-post analysis of antimicrobial consumption. Table S2: Trend analysis of antimicrobial consumption (2009–2019). Table S3: Frequency of most relevant gramnegative microorganisms and Candida spp. as causative agents of bloodstream infections (2009–2019). Table S4: Mortality of patients with the most relevant gram-negative microorganisms and Candida spp. causing bloodstream infections (2009–2019). Table S5: Differences between the pre-intervention period and the antimicrobial stewardship program period regarding pre-post analysis of incidence and mortality rate of multidrug-resistant bloodstream infections. Table S6: Trend analysis of the incidence and mortality rate of multidrug-resistant bloodstream infections (2009–2019). Figure S1: Description of the core elements of PRIOAM. Figure S2: Form for PRIOAM educational interviews.

**Author Contributions:** Conceptualization, M.A.-G. and J.M.C. Methodology, A.B.G.-G. and G.P. Formal analysis, A.B.G.-G. and G.P. Investigation, A.B.G.-G., M.A.-G., G.P., J.A.L., I.E., and M.V.G.-N. Writing—original draft preparation, A.B.G.-G. Writing—review and editing, M.A.-G., G.P., J.A.L., I.E., E.R.-A., J.G.-C., N.R.-T., M.I.M.-C., J.F.F.-G., J.L.R.-O., M.V.G.-N., J.M., J.-A.P.-S., and J.M.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by public funding from the Regional Health Ministry of Andalucia (grant number PI-0361-2010), which did not participate in the development of the program or the analysis of its results.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the University Hospital Virgen del Rocio (Project identification code: PI-0361-2010).

**Informed Consent Statement:** Taking into consideration the risks and potential harms involved in the research, the Ethics Committee approved the exemption of informed consent.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The whole hematology department of the University Hospital Virgen del Rocio, whose close commitment to the program made its results possible. We also acknowledge the invaluable contribution of all the PRIOAM professionals: physicians, clinical microbiologists, pharmacists, nurses, and other members of the hospital. We thank the hospital manager and medical director, and the Andalusian Health Service of the Regional Ministry of Health of Andalucia (Spain) for supporting the ASP.

**Conflicts of Interest:** J.M.C. has received honoraria as a speaker from Novartis, Astellas Pharma, Pfizer, MSD, Janssen Pharmaceuticals, and AstraZeneca, outside the submitted work. He has also received report grants from Instituto de Salud Carlos III, Spanish Government, co-financed by the European Development Regional Fund "A way to achieve Europe", during the conduct of the study. No potential conflict of interest was reported by all other authors.

#### **References**


#### *Article* **A Qualitative Investigation of the Acceptability and Feasibility of a Urinary Tract Infection Patient Information Leaflet for Older Adults and Their Carers**

**Leah F. Jones <sup>1</sup> , Heidi Williamson <sup>2</sup> , Petronella Downing <sup>1</sup> , Donna M. Lecky <sup>1</sup> , Diana Harcourt <sup>2</sup> and Cliodna McNulty 1,\***


**Abstract:** Urinary tract infections (UTIs) can be life threatening in older adults. The aim of this study was to primarily understand the acceptability and feasibility of using a UTI leaflet for older adults in care homes and the community. Qualitative interviews and focus groups informed by the Theoretical Domains Framework were conducted in 2019 with 93 participants from two English areas where a UTI leaflet for older adults had been introduced to improve self-care advice. Discussions were conducted with care staff (carers and nurses), older adults, general practice staff (GPs, nurses and health care assistants), and other relevant stakeholders and covered experiences of using the leaflet; its implementation; and barriers and facilitators to use. Participants deemed the leaflet an acceptable tool. Clinicians and care staff believed that having information in writing would reinforce their messages to older adults. Care staff reported that some older adults may find the information overwhelming. Where implemented, care staff used the leaflet as an educational guide. Clinicians requested the leaflet in electronic and paper formats to suit preferences. Implementation barriers included lack of awareness of the leaflet, lack of staffing and resource, and weak working relationships between care homes and general practices. It is recommended that regional strategies must include plans for dissemination to care homes, training, promotion and easy access to the leaflet. Improvements to the leaflet consisted of inclusion of antibiotic course length, D-mannose, atrophic vaginitis and replacement of less alarmist terminology such as 'life threatening'.

**Keywords:** urinary tract infections (UTI); older adults; qualitative; leaflet; Theoretical Domains Framework; antimicrobial resistance; antibiotics

#### **1. Introduction**

Urinary tract infections (UTIs) are one of the most common causes of hospitalisation in care home residents, posing a significant threat to life in this age group [1]. Most UTIs are caused by the bacterium *Escherichia coli* (*E. coli*). *E. coli* bloodstream infections (BSIs) rates in the UK have increased by 33.8% since 2012/2013 [2], with the highest rates of *E. coli* bacteraemia observed amongst older adults over the age of 75 [3,4].

The focus of any suspected infection is often difficult to determine in older adults, especially if they have dementia, and therefore clinicians may use point-of-care tests to help determine the focus of infection or use empirical broad-spectrum antibiotics. Despite strong evidence to suggest that antimicrobial therapy to treat asymptomatic bacteriuria (ASB) is unnecessary and potentially harmful [5], urine dipsticks are often used by primary care and care home staff [6]. As a result, a proportion of older adults are misdiagnosed with UTIs rather than ASB and may receive unnecessary antibiotics [7,8].

Combined with improved diagnostic pathways for health care workers, information leaflets can help explain to older adults in the community and in care, clinicians' diagnostic

**Citation:** Jones, L.F.; Williamson, H.; Downing, P.; Lecky, D.M.; Harcourt, D.; McNulty, C. A Qualitative Investigation of the Acceptability and Feasibility of a Urinary Tract Infection Patient Information Leaflet for Older Adults and Their Carers. *Antibiotics* **2021**, *10*, 83. https://doi.org/ 10.3390/antibiotics10010083

Received: 14 December 2020 Accepted: 14 January 2021 Published: 16 January 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/).

decisions and management plans, providing self-care and prevention advice to patients to improve patients' understanding, and self-care skills. Leaflets can be highly valued by patients and can help reduce unnecessary antibiotic prescribing [9–11]. In 2017, Public Health England (PHE) developed an evidence-based UTI leaflet for older adults and their carers (community and care home carers) to increase patients' and carers' knowledge about ASB, UTIs and antimicrobial resistance (AMR) and increase their skills to recognise, help prevent and self-care for UTIs. The pictorial leaflet can be found in Supplementary Material 1, [12] and provides an anatomical illustration of the urinary system, prevention information, signs and symptoms of UTI, other causes of confusion, self-care advice, what to expect from a clinical consultation, a section on AMR, and safety netting for pyelonephritis and sepsis.

The aim of this study was to:


The Theoretical Domains Framework (TDF) [13] is a behavioural model designed to help understand implementation. The TDF was used in the development phase of the leaflet by informing the interview schedules and was used in this study to structure the interview schedules to ensure all behavioural determinants were explored including knowledge, skills, and environmental context.

#### **2. Results**

Ninety-three participants took part in either focus groups or interviews from March to September 2019, from a range of urban and rural locations across Gloucestershire and East Kent. For a detailed figure of the recruitment figures and strategy, please see Supplementary Material 2

Of the 93 participants, 53 were carers working in care homes (3 nurses, 2 administrators, and 48 carers), 4 care home residents, 25 general practice staff servicing care homes and the community (13 GPs, 10 nurses and 2 health care assistants), and 8 stakeholders. Stakeholders included representatives from the Royal College of General Practitioners (RCGP; the professional body for general practitioners in the United Kingdom), the National Health Service Improvement (NHSI; responsible for overseeing foundation trusts and NHS trusts, as well as independent providers that provide NHS-funded care), the Care Association Alliance (CAA; a membership association for local Care Associations to exchange best practice), an academic pharmacist, and members of four Clinical Commissioning Groups (CCGs, clinically-led statutory NHS bodies responsible for the planning and commissioning of health care services for their local area), in East Kent (3) and Gloucestershire (1).

#### *2.1. Key Findings*

2.1.1. The Acceptability and Feasibility of Leaflet Use in Primary Care and Care Home Settings

All participants reported that the leaflet is a suitable tool for care homes and general practice and that they would like the leaflet to be available in both electronic and hard-copy formats. Suggestions for dissemination included giving the leaflet to residents' families and friends as well as to the residents themselves; displaying the leaflet as a waiting room resource; giving the leaflet to patients during consultations for suspected UTI and providing it at the reception desk or next to urine submission boxes when urine samples are submitted. Most older adults would be happy to receive the leaflet, although some concern was raised by clinicians as to the leaflet's acceptability for older adults who are coping with multiple health issues and who may find the information overwhelming.

#### 2.1.2. Value of the Leaflet

All participants valued the leaflet for various reasons, including


Some participants suggested that the leaflet would be suitable for community pharmacy and out-of-hours (OOH) settings, although one nurse practitioner believed that implementation in OOH settings would be very difficult due to transient staff.

#### 2.1.3. Barriers and Facilitators

Despite reporting local dissemination efforts through the provision of the leaflet, local champions and local hydration campaigns, lack of awareness by GP staff and care staff was the biggest barrier to leaflet use. Although some clinicians believed it was their role to cascade information to care homes, most in this study did not, and therefore weak working relationships between care homes and general practices could contribute to lack of implementation, although this will vary across regions and between facilities. Commissioning teams reported that high turnover of care staff, lack of resources and staffing issues in the CCG meant they are unable to visit every general practice and care home to promote the leaflet and conduct training around UTI diagnosis and management.

In one region, the IPC lead reported utilising 'links practitioners' in every general practice and care facility to promote and disseminate their training and resources, but also believed further work is needed to establish whether this approach is effective, as one link nurse reported attending training but not feeding back about materials.

#### 2.1.4. Comments on the Leaflet

Research findings are represented in Table 1. However, many of the leaflet findings do not directly fit into the TDF domains [14] and are therefore represented below with quotes.

Participants suggested minor improvements to the leaflet content including the use of less alarmist terminology. One participant said *"It might get people panicking about lifethreatening* . . . *we currently have two residents with full capacity and the one would be straight on the phone."* Care home staff 3.

Others suggested inclusion of the NICE recommendation of three-day antibiotic courses for proven UTI to address patient expectations for longer antibiotic courses, inclusion of the NICE recommendation of D-mannose (a type of sugar) dietary supplement as a self-care preventative option in recurrent UTI, and mention of atrophic vaginitis as an alternative cause of urinary symptoms in post-menopausal women.

Staff sharing the leaflet liked all sections, reporting that they reinforced the information that they gave to patients. One stated: *"I find them helpful if I'm having a discussion with a patient and they're not really buying into what I'm saying* . . . *it's a little bit of extra evidence that I'm not some weird doctor trying to make up stuff."* General practice staff 3.

A CAA representative recommended increasing leaflet dissemination via the CAA to ensure delivery directly to care homes. The RCGP representative recommended a short RCGP screencast to raise awareness amongst general practitioners (GPs).


**Table 1.** Key findings and corresponding TDF domains covering use and implementation of the UTI leaflet, UTI diagnosis/identification and UTI management in older adults.
