Assessment of Evaluation Tools for Integrated Surveillance of Antimicrobial Use and Resistance Based on Selected Case Studies

Introduction

The importance of combatting antimicrobial resistance (AMR) was highlighted in the Global Action Plan (GAP) released by the World Health Organization (WHO) in 2015 (1). It was further adopted by the Tripartite Collaboration consisting of the members of the WHO, Food and Agriculture Organization of the United Nations (FAO), and the World Organization for Animal Health (OIE) and endorsed by political leaders and the United Nations (UN) General Assembly (2). The Tripartite Collaboration acknowledges that the AMR challenge needs to be addressed using a One Health (OH) approach to reflect that the development and spread of AMR do not respect boundaries between sectors and, therefore, require cross-sectoral collaboration and prevention activities. One of the main objectives of the GAP is to initiate and maintain cost-effective integrated surveillance of antimicrobial use (AMU) and AMR at the global and national levels (1).

Ideally, combatting AMR requires engagement from actors within all sectors of animal health, food safety, environmental protection, plant health, and human health (3). All sectors need to be involved in surveillance to identify emerging resistance, understand the AMR epidemiology, and develop effective policies for AMU and AMR reduction. In short, the integration of sector activities and robust collaboration are essential for successful surveillance and control of AMU and AMR. According to Stärk et al. (4), OH surveillance describes the systematic collection, validation, analysis, interpretation of data, and dissemination of information collected in humans, animals, and the environment to inform decisions for more effective, evidence-based interventions. AMR genes are present in bacteria and spread among humans, animals, and the environment. A program of integrated surveillance of AMR in foodborne bacteria includes coordinated sampling and testing of antimicrobial susceptibility of bacteria from food-producing animals, food, and humans using epidemiological (including sampling) and microbiological methods that enable comparisons of results. The use of comparable methods is necessary to allow comparison of antimicrobial susceptibility results between different areas, countries, and regions (5, 6). Currently integrated OH AMU and AMR surveillance and monitoring systems exist or are under development in many countries (4). However, the surveillance programs do not always address all necessary sectors and they are rarely fully integrated (7). An integrated approach provides a better understanding of the epidemiology of AMR and an easier identification of the best intervention points and enhances the timeliness of surveillance by providing early warning of emergence of new resistant strains from one sector to another. Furthermore, a cross-sectoral collaboration may lead to knowledge/resource sharing, expertise exchange, and capacity building (8), which may result in cost savings and create more efficient and effective systems (9). Full integration might not be necessary to achieve the wanted outputs, and integration and collaboration in itself can be costly without always improving outputs (7, 10). A surveillance approach implies planning, data collection, analysis, interpretation, and dissemination of a given activity. It is useful to apply collaboration across different surveillance activities and integration in all or some of the activities. Identification of the optimal levels of integration to obtain the information needed for decision making is an important task in OH surveillance systems (7, 10).

Aenishaenslin et al. (7) suggested that the value of OH surveillance for AMR can be conceptualized and measured across a selection of different outcomes that can be classified in three dimensions, namely, (i) immediate, (ii) intermediate, or (iii) ultimate. Immediate outcomes include increased understanding of the AMR epidemiology at the human, animal, and environment health interface, and the value would lie in the intellectual or social capital generated. Intermediate outcomes include changes in policy or behaviors, and the expected value is the reduction in AMU and AMR that results from these changes. Ultimate outcomes include tangible benefits such as improved animal, human, and environmental health and associated socioeconomic benefits.

Apart from appropriate planning and designing, surveillance programs also need regular evaluation to remain operational, efficient, and cost-effective. Moreover, evaluation is needed to ensure that the goal is underpinned by the ongoing activities and shared with the essential stakeholders (11). Evaluation is complex and requires agreement on an evaluation objective, a process usually led by food safety/health authorities in consultation with other stakeholders. Secondly, an appropriate evaluation tool should be selected, which requires expertise and knowledge of surveillance evaluation.

Existing tools for evaluation of surveillance [e.g., (12, 13)] are not necessarily appropriate for integrated surveillance as they might not address aspects such as collaboration across sectors (12, 14). Characteristics of OH surveillance programs have been described, and recently, tools to evaluate integrated surveillance systems have emerged, targeting different aspects of the OH or other integrated surveillance activities (7, 11, 15–19). A tool may have been made for evaluation of a particular type of surveillance system, such as animal health surveillance. Still, it might also be used to assess other types of surveillance systems such as AMR surveillance, covering aspects such as sampling strategies and sample sizes of surveillance protocols. The latter may not be covered in details by the tools developed specifically for AMR surveillance evaluation. The different tools vary in their approaches, layouts and user-friendliness, comprehensiveness, terminology, aspects covered, capacity, training, and resources required to use them, as well as their specific usefulness for the evaluation of AMU and AMR surveillance. Hence, a characterization and meta-evaluation of the existing evaluation tools are called for to provide guidance on how to identify the best match between the evaluation objective, the resources available, and the selected evaluation tool.

During 2019–2020, an international network of scientists in the project “Co-Eval-AMR—Convergence in evaluation frameworks for integrated surveillance of AMR” (20) developed guidance for choosing an assessment approach from an inventory of tools suitable for evaluating integrated AMU and AMR surveillance systems, according to the needs of the users. The results presented here originate from the Co-Eval-AMR network aiming to guide assessors in their future selection of evaluation tools. A pilot version of the present study, using one surveillance system case and the first version of the assessment criteria, was published by Nielsen et al. in 2019 (21). The objective of the present study was to describe and assess the characteristics, functionalities, and suitability of tools that might be used for evaluation of integrated AMU and AMR surveillance.

Materials and Methods

Overview of the Evaluation Tools

In the following section, the six tools used are presented in brief.

Assessment Tool for Laboratories and AMR Surveillance Systems

The Assessment Tool for Laboratories and AMR Surveillance Systems (ATLASS) is a tool designed by the FAO for assessing and defining targets to improve national AMR surveillance systems in the food and agriculture sectors (18). It is composed of two modules: a surveillance module and a laboratory module. Each module includes two standardized questionnaires, which are to be completed by the assessors. The assessments generate a baseline and classify a “stage” for AMR laboratory capacity detection, AMR surveillance, and dissemination of information.

Evaluation of Collaboration for Surveillance

The Evaluation of Collaboration for Surveillance (ECoSur) tool aims at evaluating the organization, functioning, and functionalities of collaboration taking place in a multi-sectoral surveillance system (11). The final purpose is to assess whether collaboration as planned and implemented is relevant and functional to produce the expected collaborative outputs. The tool relies on the scoring of 22 attributes and three indexes characterizing the organization of collaboration at the governance and operation level and nine attributes referring to core functions of collaboration to ensure the sustainable operation of an effective multi-sectoral surveillance system. Three automatically generated outputs display the evaluation results for attributes and indexes and support the identification of strengths and weaknesses of collaboration and the formulation of recommendations for its amelioration.

Integrated Surveillance System Evaluation Project

The AMR integrated surveillance system evaluation project (ISSEP) tool is a conceptual tool developed in Canada with the aim to structure an evaluation of the added value of integrated surveillance systems for AMR (7). It comprises five evaluation levels that target the evaluation of OH integration in the surveillance system; its capacity to produce integrated information and expertise, to generate actionable knowledge, and to influence decision making; and health and economic impacts. For each level, a set of evaluation questions are defined, and links are made with existing evaluation tools. A semi-quantitative scale is applied to show the level of integration of the surveillance system (19).

Network for Evaluation of One Health

The Network for Evaluation of One Health (NEOH) tool is part of a framework resulting from the EU COST Action “Network for Evaluation of One Health” to provide science-based guidance for the evaluation of One Health and other integrated approaches to health (16, 20, 21). There are four elements, namely, “system definition and description of OH initiative within the system,” “theory of change” (ToC), “assessment of OH-ness,” and “outcome evaluation.” Qualitative assessment as well as semi-quantitative scorings are used for the evaluation of the degree and of the “OH-ness” (OH index and OH ratio) and metrics for different outcomes. Illustrative web diagrams of the distribution of scores for gap identification are presented in the Excel tool for assessment of OH-ness (20).

Progressive Management Pathway Tool for AMR

The Progressive Management Pathway tool for AMR (PMP-AMR) tool is a self-assessment tool designed by the FAO to provide guidance to countries for implementation of their National Action Plans (NAP) for AMU and AMR (17, 21). It includes four focus areas for evaluation: awareness, evidence, governance, and practices. For each focus area, specific activities, achievements, and key performance indicators (KPI) are listed. The tool provides a dashboard, showing the progress made for each focus area toward an optimal and sustainable use of antimicrobials.

Survtools

SURVTOOLS was developed as a part of the EU FP7-funded project RISKSUR: risk-based animal health surveillance systems. The evaluation tool (EVA tool) is a support tool for the evaluation of animal health surveillance systems, developed to provide guidance for evaluation of animal health surveillance including economic evaluation (12, 21). When planning an evaluation, the user is guided through three main steps: defining the evaluation context; defining the evaluation question; and selecting the evaluation attributes and the economic criteria. Furthermore, the tool provides additional information and guidance on how to use the evaluation plan to perform the evaluation and how to report on the evaluation outputs. An online web version of the EVA tool is available (12).

Methodology Used to Assess the Tools

The details of the scoring scheme for functional aspects, the SWOT-like approach, and the scoring scheme for the themes describing the scope of the tools are presented below.

The Case Study Approach

A total of eight country-based case studies of AMU and AMR surveillance systems were included in the study (Table 1). Each country-based case study was undertaken by individuals or a group of individuals with expertise on the respective national cases (hereafter called the assessors), making a total of 20 assessors. The choice of case was the NAP on AMR or parts of it in the respective assessor's country. To collect the information needed to carry out the assessment, the assessors reached out to additional experts and other sources.

Table 1. Overview of eight country-based case studies involving six different tools for evaluation of surveillance of antimicrobial use and resistance, 2019.

The assessors met regularly, and initially, there was an assessment methodology developed in collaboration with selected members of the Co-Eval-AMR network group. The methodology included two standardized scoring schemes; a strengths, weaknesses, opportunities, and threats (SWOT)-like analysis scheme; and templates for reporting and instructions. The evaluation tools were applied on the country-based case studies using one or more tools on each case. Overall, the outcome was the users' experience regarding applicability of the tool. Each tool was assessed between one and four times.

Scoring Functional Aspects

A scoring scheme aiming at assessing 11 functional aspects was developed, and answers were scored numerically, where 1 = not covered, 2 = not well-covered, 3 = more or less covered, and 4 = well-covered. With each score, a comment was requested explaining the score. The 11 aspects were as follows: (1) user friendliness, (2) compliance with evaluation objectives, (3) efficiency (number of people and time taken vs. what the evaluation should be used for), (4) use of a step-wise approach to the evaluation, (5) overall appearance, (6) generation of actionable evaluation outputs, (7) evaluation of OH aspects, (8) workability in terms of required data, (9) workability in terms of required people to include, (10) workability in terms of analysis to be done, and (11) time taken for application of the tool.

The combined scores for each tool were presented in a heat map. In the case one assessor/assessor group scored over a range of numbers, averaging was used followed by rounding up if necessary to obtain a whole number for the total score. A crude summary score for each tool was calculated and presented in heat maps. The scores should only be interpreted relatively within this study material. The justification for each score, provided by the individual assessors, was condensed by the first author and checked for correctness by the other authors, and the “condensed results” were then presented.

A SWOT-Like Approach

A SWOT-like scheme was developed asking the assessors to answer four questions: (1) things that I liked or that the tool covered well; (2) things that I struggled with when using this tool; (3) things people should be aware of when using this tool; and (4) things that this tool covers insufficiently. A qualitative synthesis of the result was done in two steps. First, all individual phrases were captured. In a second step, phrases with the similar meaning were reduced into one, implying that a phrase was simplified or made into one word, if possible. It also implied that no phrase or word was repeated for each of the SWOT analyses and tools. The first synthesis was carried out by the assessors for the tools they had applied. The second synthesis was condensed by two of the authors, and the condensed results were checked for correctness by the other authors and subsequently presented.

Scoring Themes for the Scopes

A second scoring scheme consisted of eight themes to describe the scope of the tool: developed specifically for AMU and AMR, collaboration, resources, output and use of information, integration, governance, adaptivity, and technical operations. Seven of the themes included in the scheme were developed in the Co-Eval-AMR project (22). Additionally in this study, the theme governance was added. The objective was to score how well each theme was covered by the specific evaluation tool. A more detailed description of the individual theme scope is given in Table 2. The same scoring scale as in the Scoring Functional Aspects section was used. The combined scores for each tool were presented in a heat map, based on a similar way of estimation as described in the Scoring Functional Aspects section. A crude summary score for each tool was calculated, but this should only be interpreted relatively within this study material. Again, the free text justifications behind the scores provided by the assessors were synthesized by the first author, checked for correctness by the other authors, and subsequently presented.

Table 2. Description of the themes describing the scope of the tool in relation to surveillance identified in the Co-Eval-AMR project and used for the additional assessment of the evaluation tools for surveillance programs/activities.

Results

All detailed answers and justifications from the scoring of the functional aspects and the themes and from using the SWOT-like approach are published on the Co-Eval-AMR project webpage (https://coevalamr.fp7-risksur.eu/) and in Nielsen et al. (21).

Scoring of the Functional Aspects of the Tool

The results from the scoring of the case studies according to the 11 functional aspects of AMU and AMR surveillance systems are shown in Table 3. A summary of the justifications behind the scores is shown in Table 4. A crude summary of the scores showed that ISSEP and NEOH had the lowest scores, 25 and 30 respectively of the total 44 that could have been achieved. ATLASS and PMP-AMR had the highest, 39 of the 44 possible.

Table 3. Result of the scoring of all six tools with respect to the 11 functional aspects, shown as a heat map (the number of times the tool was assessed is given in the bracket). The scoring scale used was where 1 (red) = not covered, 2 = not well covered (orange), 3 = more or less covered (yellow), 4 = well covered (green).

Table 4. Results of synthesis of the underlying reasoning for the scoring according to the 11 functional aspects.

For OH aspects, ATLASS and NEOH scored the highest. PMP-AMR, ATLASS, ECoSur, and NEOH provide semi-quantitative scores for the aspects evaluated, whereas ISSEP and SURVTOOLS will result in a plan for how to conduct evaluation(s). ISSEP, ECoSur, NEOH, and SURVTOOLS allow for in-depth analyses and, therefore, require more complex data, information, and specific training of the evaluator(s). PMP-AMR and ATLASS seemed to be the most user-friendly tools, particularly designed for food safety authorities managing the surveillance system.

The SWOT-Like Approach

The results of the SWOT-like approach applied to assess the tools are shown in Table 5. The variation in answers to the four SWOT-like questions was low among the assessors of each tool, indicating consistency regarding the general impression of the tools. The PMP-AMR and ATLASS were liked for the semi-quantitative scorings that could be made directly and that the tools were particularly made for evaluation of AMR surveillance systems. What is not covered in these two tools is the environmental, plant, and human parts of surveillance.

Table 5. Synthesis of phrases provided in the SWOT analysis of six different evaluation tools used in eight country-based case studies.

The ECoSur was liked because it allowed evaluation of collaboration in detail; however, the level of abstraction in the language in the existing version of the tool was a struggle. ISSEP was liked because it described the relationship between the integrated surveillance activities for AMU and AMR, OH outputs produced, and the different expected outcomes very well.

NEOH was liked for being comprehensive, multi-faceted, and fit for a transversal analysis of OH initiatives. The main struggle related to NEOH was that it was cumbersome and time-consuming to use. Similarly, SURVTOOLS was liked because information for evaluation of all aspects of a surveillance system including the epidemiological part is provided as scientific references. Furthermore, an epidemiological calculator is provided. However, SURVTOOLS is one of the tools that only provide an evaluation plan.

Scoring of the Themes Describing the Scope of the Tool

The results from the scoring of each tool for the eight themes describing the scope of the tool in relation to surveillance are shown in Table 6. A summary of the justifications behind the scores are shown in Table 7. A crude summary of the scores for the tools, regarding which themes they covered, showed a limited variation. ATLASS had the highest crude summary scores of 28 followed by ISSEP and ECoSur both with 25.

Table 6. Results of scoring of six tools for AMR surveillance evaluation according to eight themes describing the scope of the evaluation tool (the number of times the tool was assessed is given in the bracket).

Table 7. Synthesis of the underlying reasoning for the scoring according to the eight themes describing the scope of six AMR surveillance evaluation tools.

PMP-AMR, ATLASS, and ISSEP have been developed specifically for AMR-related activities. NEOH and ISSEP were perceived as the best tools for evaluation of all OH aspects, and ECoSur and ISSEP for evaluation of the quality of collaboration. ATLASS is the only tool evaluating laboratory activities specifically. Only ISSEP produced a direct measure of the “integration” and “impact on decision making.” SURVTOOLS has an epi-sample size calculator and is, hence, the only tool providing a quantitative assessment of the technical operations in surveillance.

Discussion

Tools Developed Specifically for Evaluating AMU and AMR Surveillance

Only PMP-AMR, ATLASS, and ISSEP have been developed especially for evaluating AMU and AMR surveillance. Generally speaking, ISSEP was the only tool assessed that addressed AMU and integration aspects. The strengths of PMP-AMR and ATLASS are governance and, hence, strategic implementation of NAPs. PMP-AMR addresses neither evaluation of design of surveillance nor integration or collaboration. ATLASS is structured in such a way that detailed information about the sectors involved and the laboratories in the surveillance system can be captured. Thereby, it addresses the gaps in a laboratory's capacity to implement surveillance activities. A quantitative evaluation of the epidemiological designs is impossible in ATLASS. Moreover, ATLASS does not provide an output of the level of integration—but all data collated could provide the evaluator with an impression of the level of integration in the system evaluated.

However, the other evaluation tools were also considered suitable for evaluation of AMU and AMR surveillance programs. In fact, several of the tools showed a high degree of flexibility and were applicable to different surveillance evaluation objectives. Still, the most accurate evaluations originated from the tools that match the specific evaluation questions. Generally speaking, an evaluation of integrated AMU and AMR surveillance systems will benefit from using tools developed specifically for evaluating AMR surveillance and OH aspects since specific characteristics are encountered.

User Friendliness and Potential Value

The PMP-AMR and ATLASS tools are to a high extent self-instructive and the questions were, therefore, easy to answer. The structure of PMP-AMR was very easy to understand, whereas ATLASS was more complicated to fill in, since it comprises many questions at all levels of organization. The handbook/guidance/surveillance evaluation wiki to SURVTOOLS was perceived by some of the assessors as very clear and easy to read. It also provides advice on how to cover many of the required aspects of evaluation. The online evaluation tool itself looks very aesthetic but covers less information than the handbook and is not fully self-instructive for all evaluation objectives. NEOH requires knowledge of both the relevant context (in the NEOH framework denoted “the underlying system and its system boundaries”) and the integrated surveillance activities (“the initiative under evaluation”) in question, because the assessor must define all components that form part of the underlying system (the context) included in or affected by the surveillance. NEOH allows the assessor to identify and assess expected outcomes based on the ToC of the initiative. ToC is a specific type of methodology for planning, participation, and evaluation that is used in companies, philanthropy, and not-for-profit and government sectors to promote social change. Further, it defines long-term goals and then maps backward in time to identify the necessary preconditions and actions to be taken. The ToC focus will lead to learning and perhaps a better understanding of the surveillance and its potential societal impacts. It is easy to get lost in the extensive handbook published to assist in using NEOH, and a quick guide is currently missing. The many detailed questions about integration such as OH implementation including systemic organization and level of sharing (infrastructure aspects) and learning (operational aspects) allow for nuances in the answers and, thereby, a better quality of the results. However, the evaluator should be aware that applying this tool requires time investment and training, including specific training in “systems thinking.”

ISSEP, ECoSur, and SURVTOOLS also allow for an in-depth analysis requiring collection of more complex data and information. For SURVTOOLS, a specific training in design of epidemiological studies and a wide spectrum of analytical methods are needed before a full exploitation of the tool can be expected. Many of the tools could also be used to guide the design of AMU and AMR surveillance systems in addition to evaluation of existing systems.

Many of the tools, especially ATLASS, produce intermediate outputs of how well the different parts of the program are integrated and how well the partners collaborate. In contrast, the interpretation of evaluation results of ECoSur supports the identification of strengths and weaknesses of collaboration and the formulation of recommendations. Among the six tools investigated, this tool allows for addressing collaboration in most detail and in different dimensions.

It became clear during this study that adequate resources are needed to perform a full evaluation, sometimes requiring involvement of many assessors and/or stakeholders, and it might take weeks to months to finalize. For all tools, training and instructions would be required to understand the tools sufficiently well to work effectively. Furthermore, the assessor should preferably have a moderate level of understanding of surveillance processes. Moreover, it is important to balance the degree of complexity of the evaluation tool with the available resources in terms of number of people, data, and time.

Output and Use of Information (Impact)

The ISSEP and partly SURVTOOLS approaches provide a conceptual basis for structuring the evaluation of different surveillance outcomes, from the level of integration to the evaluation of the decisions as well as economic efficiency. The outputs of an evaluation may consist of first-level outputs, such as epidemiological performance measures, as well as intermediate outputs, such as how well the system is integrated. For successful AMU and AMR surveillance, the final impact would be that there are antibiotics available to treat future generations of humans and animals against infections. PMP-AMR and ATLASS only produce intermediate outputs through the theme collaboration. It remains unknown whether this and similar themes really reflect what is necessary to implement to reach the final desired impact in the AMU and AMR surveillance. ATLASS and PMP-AMR are contributing to this final impact by providing evaluation of the governance, strategic support, and budgets for surveillance. An evaluation of the impact of surveillance will be further addressed in a phase 2 of the Co-Eval-AMR, just initiated as a follow-up project funded by Joint Programming Initiative for AMR (JPIAMR) (https://www.jpiamr.eu/project/coeval-amr-phase-2/).

The Limitations of the Study

We have presented the experiences of eight country-based case study groups in using six evaluation tools. Due to resource constraints, some tools were only assessed in a limited number of case studies. Some of the tools were only scored by two assessors, by two assessor groups, or by the creator(s) of the tool. For NEOH, ECoSur, ISSEP, and PMP-AMR, co-developers of the tools were involved in the assessment, but the tools were also assessed by other case study groups. The assessments were done by different persons, and the scores were perceived as crude and subjective. The assessors had varying levels of understanding of the evaluation tools; some were involved in the development of one of the tools, whereas others were trained in using a specific tool. The first group of assessors may have had greater insights into the tool(s) that they assessed and may have been biased in some aspects of the assessment, e.g., user friendliness. During the assessment process, there was some convergence in the scoring done by the assessors due to the development of a common understanding of the words and sentences used in the tools. Therefore, the results of the scoring of the functional criteria had a higher variation than the results of the scoring of the attributes that was done later in the process. The qualitative assessments are probably more informative for the pros and cons of each tool than the actual scores. The remaining tools were assessed by “non-developers.”

Monitoring and stewardship of AMU as part of AMR surveillance were not addressed in the assessment. In the second phase of the Co-Eval-AMR, additional assessments using other tools are planned. Moreover, focus will be on how to assess the impact of integrated surveillance systems for AMU and AMR as well as on how to evaluate governance. The online assessment system made by the Co-Eval-AMR project group can also be used by other scientists for doing similar comparisons and hence more experiences will be collected (https://coevalamr.fp7-risksur.eu/). Most of the participants in the case study groups were veterinarians or professionals working within veterinary public health. Persons in human health only participated indirectly when being interviewed, and there was no focus on the environment. In phase 2 of the project, collaboration among others and with social scientists will broaden the scope and the way of looking at surveillance and evaluations.

Development of Assessment Methodology and Reporting the Results to Capture the Variation in the Underlying Reasoning

In the Co-Eval-AMR project, the methodology was developed to capture the usability of the tools for evaluation of AMU and AMR surveillance activities in a systematic way, allowing for comparisons between assessors. The assessment methodologies covered aspects known as contributing to controlling AMR, e.g., evaluation of OH aspects, mentioned by for instance Holmes et al. (3). The 11 functional aspects included elements such as user-friendliness and whether the tool meets evaluation needs/produces actionable outputs and the resource needed related to data, manpower, and time. In the second phase of the Co-Eval-AMR project, improvements in assessment criteria will be considered.

As opposed to the other tools, ISSEP and SURVTOOLS generated only a plan for how to conduct the actual evaluation based on the chosen evaluation questions. Hence, scoring these for some of the 11 functional aspects and the eight themes was difficult. The PMP-AMR, ATLASS, ECoSur, and NEOH tools provide semi-quantitative evaluation outputs. PMP-AMR and ATLASS measure the progress over time and can be used repeatedly. Moreover, PMP-AMR and ATLASS seemed suitable for non-scientists too, since they do not require specific knowledge of epidemiology and surveillance for their application. The tools are not interchangeable—they do not have common scopes and objectives; therefore, one cannot choose a tool only based on the appreciation as assessed only by these case studies. Some lack of consistency exists between the work done in the different working groups of the Co-Eval-AMR project, because some of the development of methodologies was undertaken simultaneously in all working groups, e.g., governance was therefore only assessed by “country case study groups” with a few exceptions. The latter reflected in the missing data given as a footnote in Table 3.

Establishing a Data Capture System for Generation of Assessment Experiences

The developed reporting template enables other assessors to report their experiences using the tools in a comparable way. The template consists of four sections; (1) general information, (2) scoring of 10 functional aspects, (3) SWOT-like approach, and (4) scoring of eight themes describing the scope of the tool. The idea was to develop a kind of user experience scoring overview similar to many internet applications such as TripAdvisor and Google reviews providing the readers with quick, yet detailed, insights of the tools. The template is placed in an online platform on the homepage of Co-Eval-AMR (https://coevalamr.fp7-risksur.eu/). We encourage users of the tools to provide their inputs and expect that over time a growing collection of experiences will help users in choosing more easily among the existing tools.

Conclusion

Evaluation of integrated surveillance is needed at regular intervals using robust tools. It is important to choose a tool that adequately addresses the specific evaluation objectives. We provided a portfolio of the experiences of 20 users representing eight country-based case studies in which six different tools were applied, to highlight their attributes, pros and cons, and requirements.

Only PMP-AMR, ATLASS, and ISSEP have been developed especially for evaluating AMU and AMR surveillance—with ISSEP being the only tool providing a semi-quantitative score of AMU and AMR integration. All six tools demonstrate a high degree of complementarity. Depending on the evaluation questions selected, assessors may choose among the different tools to conduct the evaluation as such, namely, ECoSur for addressing collaboration, NEOH for the OH-ness and the relationship between ToC and expected outcomes of the surveillance, ATLASS for the laboratory capacities, and SURVTOOL for epidemiological and economic performance.

An online platform for reporting of users' experiences will help users interested in conducting an evaluation of AMU and AMR surveillance in choosing the most adequate tools for their specific evaluation needs: https://guidance.fp7-risksur.eu/welcome/decision-support/. Furthermore, this platform could help further extend general user experience of AMU and AMR surveillance evaluation tools.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author Contributions

MS, AH, MB, and LA were substantially involved, and took the lead, in all steps of the study from the conception to the design of the work. AH and UB designed the layout of Table 3 and Table 6. All authors contributed to the assessments and/or to the interpretation of the results of these. Most of the authors initially drafted parts of the paper. All authors approved the final version of the paper. They also agreed to be accountable for all aspects of the work, in ensuring that questions related to the accuracy or integrity of any part of the work were appropriately investigated and resolved.

Funding

Co-Eval-AMR was funded by JPIAMR (www.jpiamr.eu) and led by Dr. Barbara Häsler from grant holder Royal Veterinary College in the UK, receiving funding from the Medical Research Council, grant number MR/S037721/1.

Conflict of Interest

MB was involved in the development of ECoSur and LN was involved in the development of NEOH.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

We thank all participants of Co-Eval-AMR network project for their contribution as well as the stakeholders in the different case studies, who provided us with information to perform the assessments.

References