Mapping out a One Health model of antimicrobial resistance in the context of the Swedish food system: A literature scan

doi:10.21203/rs.3.rs-2248339/v1

Download PDF

Research Article

Mapping out a One Health model of antimicrobial resistance in the context of the Swedish food system: A literature scan

https://doi.org/10.21203/rs.3.rs-2248339/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Antimicrobial resistance (AMR) causes worsening health, environmental, and financial burdens. Modeling complex issues such as AMR can help clarify the behaviour of the system and assess the impacts of interventions. While models exist for specific AMR contexts (e.g. on-farm, in hospital), due to inadequate collaboration and data availability, how well such models cover the broader One Health system is unknown. Our study aimed to identify models of AMR across the One Health system with a focus on the Swedish food system (objective 1), and data to parameterize the models (objective 2), to ultimately inform future development of a comprehensive model of possible AMR emergence and transmission across the entire system.

Methods: Using a previously developed causal loop diagram (CLD) of factors identified as important in the emergence and transmission of AMR in the Swedish food system, an extensive literature scan was performed to identify models and data from peer-reviewed and grey literature sources. Articles were searched using Google, Google Scholar, and Pubmed, screened for relevance, and the models and data were extracted and categorized in an Excel database. Visual representations of the models and data were overlayed on the existing CLD to illustrate coverage.

Results: A total of 126 articles were identified, describing 106 models in various parts of the One Health system; 54 were AMR specific. Four articles described models with an economic component (e.g. cost-effectiveness of interventions, cost-analysis of disease outbreaks). Most models were limited to one sector (n=60, 57%) and were compartmental (n=73, 69%); half were deterministic (n=53, 50%). Few multi-level, multi-sector models, and models of AMR within the animal and environmental sectors, were identified. A total of 414 articles were identified that contained data to parameterize the models. There were major data gaps for factors related to the environment, wildlife, and broad, ill-defined, or abstract ideas (e.g. human experience and knowledge).

Conclusions: There were no models that addressed the entire system and few that addressed the issue of AMR beyond one context or sector. Existing models have the potential to be integrated to create a mixed-methods model, provided that data gaps can be addressed.

Antimicrobial resistance

One Health

literature scan

complex system

systems models

Antimicrobial Resistance (AMR) is one of the largest threats to public health across the globe [1, 2], causing an estimated 4.95 million deaths that were associated with bacterial AMR globally in 2019, and 1.27 million deaths due to bacterial AMR directly [3]. Beyond the burden to human health and wellbeing, AMR also negatively impacts animal health and is a financial burden. Europe loses 1.5 billion US dollars annually from increased healthcare costs and loss of productivity connected to multi-drug resistant bacteria [4]. AMR can also impact the agricultural sector via production losses from animals with resistant infections, and possibly decreased trade due to a fear of resistance [4].

Antimicrobial use (AMU) is known to drive resistance, but other factors impact how and where antimicrobials (AM) are used, and can affect spread and transmission of resistant bacteria [1, 2, 4–6]. AMR can develop in micro-organisms in humans, animals, and the environment and be transmitted between these sectors through a multitude of pathways [1, 2, 4, 6, 7]. While it is known that resistant bacteria can spread across many transmission pathways, how these pathways intersect and interact to affect AMR is less known. This makes it difficult to build a model that captures the entire system of drivers of AMR based on current empirical data and mathematical modelling techniques.

Governing and creating interventions and implementing policy to combat AMR has proven difficult due to the complexity of the system that drives AMR [8–12]. Complex systems are systems composed of many elements with multiple interactions between them, in which the elements adapt and react to the patterns the interactions create [13]. The three main components of complex systems are non-linearity, emergence, and unpredictability [13, 14]. Nonlinearity refers to disproportionate causation, in which small changes in one element do not necessarily cause a small change in another [14]. Alternatively, small shifts in one part of the system can lead to sudden and large changes throughout the system [14]. This is typically due to multiple interactions that interact to create the overall behaviour of the system [14]. For example, multiple factors may be acting in opposing ways on an element in the system, which may also be connected to other factors in the system. Therefore, changes in one element can cause non-linear effects [14]. The second feature is emergence, which refers to the emergence of macroscopic patterns (system-wide change) resulting from microscopic interactions between elements (an association between two elements) [14]. Together, the non-linearity and emergence cause great uncertainty and unpredictability within the system [13, 14]. For example, changing one small part of the system, could have large and unpredictable consequences elsewhere in the system. The system that produces AMR is comprised of many drivers across many sectors and ecological scales and involves many moving parts. The human, animal, and environmental sectors are intimately connected, and AMR can develop and spread across these sectors with ease [2, 4, 6]. Therefore, changes in one sector can have large and unpredictable consequences across of the system [2, 4, 6].

This complexity makes creating models of the system that drives AMR a challenge. To date, many models of disease transmission and AMR exist, but most are oriented towards specific contexts within sectors (e.g. on-farm transmission in a small cattle herd, or transmission in an intensive care unit (ICU) at a hospital). Few have attempted to merge existing models to better account for the inter-connections between the sectors, despite AMR being a One Health issue at the intersection of humans, animals, and the environments they share [7]. There has been growing interest in including the more socio-ecological drivers into dynamic models of AMR [9, 15], but this work is still in its infancy [9]. Therefore, to build a model of AMR that covers a broad and more complex system of drivers across the One Health spectrum, it is useful to identify what types of models currently exist for different sectors or sub-systems and how they could be combined to cover the broader system. It is then necessary to identify data to parameterize such a complex model. We chose to do this for the Swedish food system, as an exemplar case study.

A formal review of the literature is not an essential first step in the model building processes, nor is there a set framework for gathering data to parameterize models. However, due to the vast number of factors to be parameterized, a clear framework to gather and organize the data for use in modelling would improve transparency in the data collection process. With a shift towards more inclusive and integrated models, conducting scoping reviews for parameters has risen in popularity as a primary step [16–18], and may become the norm in One Health modelling of complex systems due to the increased number of factors and connections to be included. The model proposed in this paper aims to incorporate 92 factors a previously created qualitative model of the drivers of AMR [19], and therefore a full scoping review would be extremely resource intensive. Therefore, performing a literature scan inspired by the scoping review methods outline by Arksey & O’Malley [20] would provide a balance of structure and transparency in data collection with the flexibility to explore the literature more freely (e.g. expand searches, incorporate a wider range of data sources or types of data).

Systematic and scoping reviews have been conducted to identify models of the system that drives AMR, however, they have been limited in their scope. For example, systematic reviews have been conducted to find population-level mathematical models of AMR within human populations [21], at the microbial or within-host level [22], and have summarized within-host and population-level models in humans [23]. A more recent scoping review (conducted in 2019) aimed to identify dynamic models of AMR [24]; however, the results were still mainly human-focused. This literature scan aims to be more comprehensive and capture models that extend beyond AMR-specific models, to identify models that represent the broader system of drivers and impacts and have potential to be adapted for the AMR context.

This literature scan also aimed to gather data that could be used to parameterize a broad One Health model (which we defined as a model that include factors from the human, animal, and the environmental sectors that interact to create and perpetuate an issue) of AMR developed during participatory modelling workshops [19]. The qualitative model created through participatory modelling included many factors that were unlikely to have quantitative data available. Therefore, we aimed to find a wider variety of data, both quantitative and qualitative, from a broad set of sources (grey and published literature) to be able to expand the scope of current AMR models and incorporate a wider array of factors from the broader system. The objectives of this study were to identify: (1) the different types of existing dynamic models (e.g., compartmental model, agent-based model, network model) across various parts of the broader AMR One Health system, and (2) the data sources and evidence that could be used to model the different parts of the Swedish food system (further referred to as “the system”).

To set a broad scope for what to consider as part of a One Health model of AMR in the Swedish food system, we used an existing causal loop diagram (“diagram”) from two participatory modelling workshops that were held in Stockholm, Sweden in September 2019 in which participants mapped the wider system of drivers of AMR in the Swedish food system (refer to the workshops carried out by Lambraki et al. for the methods and full results) [19]. The resulting diagram contained 92 nodes and 331 relationships and represented the structure of a hypothetical One Health model of AMR in a European (specifically Swedish) food system context [19]. For a full description of the nodes, refer to Lambraki et al. [19]. Using this diagram to bound the scope and define the search terms, we conducted a literature scan, inspired by Arksey & O’Malley [20] framework for scoping reviews, to address our two objectives. The search took place from September 1^st to December 31^st, 2020.

1.2.1 – Objective 1: Existing Models

A literature search of peer-reviewed publications was performed in Google Scholar and PubMed to identify different types models, including: 1) mathematical models pertaining to the transmission of micro-organisms between humans, animals, and their environments; 2) mathematical models of AMR transmission or emergence; 3) models of AM decay and residue build up in waste, waste-water, and other settings; 4) economic models of agriculture and the food system, and; 5) economic models of health systems. Models were not limited to a specific geographical context, however, the search was conducted in English.

A snowball search approach was used, starting with broad, high-level search terms that aligned with major domains of the diagram (e.g. “model AND antimicrobial resistance AND humans/animals/One Health” or “model AND consumer demand AND food”). Models of E. coli were explicitly searched because E. coli has been identified as one of the major threats for AMR [25], is commonly identified in and transmits between humans and animals [25], and is heavily represented in the existing AMR surveillance data for animals and humans in Sweden [26,27]. In instances where specific sections of the diagram were not well described, search terms were refined and the search was narrowed to capture more specific nodes or parts of the system (e.g. models of AMR and resistant E. coli were further refined to the healthcare or agricultural systems using search strings such as: “infectious disease model AND antimicrobial resistance/resistant E. coli AND hospital/on-farm/abattoir”, or economic models for specific food commodities were searched using a search string such as: “economic/supply-demand/consumer demand model AND chicken/beef/fish”). This approach allowed for a narrowing of the search criteria to ensure key models were captured in the search, that would be of use for our specific purpose. A full list of search criteria is given in Additional file 1, Table A1. We found that the first 100 results generally yielded the best fit given our search criteria, therefore, we focused on the first 100 results for each search. Broad search terms and few exclusion criteria were deliberately used to capture a wide range of models from a variety of sectors. Citations about models that were strictly statistical in nature (e.g. linear and logistic regression models) were excluded. This was because we wanted to obtain models that simulated or encompassed the transmission or emergence of AMR or simulated attempted to encompass other parts of the system (economics) and not models that provided estimates of association between nodes (e.g. AMU is associated with an increased risk of AMR).

The lead author (MC) screened the titles and abstracts for inclusion based on the criteria above, and then reviewed the full text articles and excluded any sources that were not identified as simulation models as described above (e.g. statistical models focussed on associations between nodes). For the sources that were included, the following information was extracted and organized into a database created in Microsoft Excel version 16.60: type and process of model (e.g. agent-based model, network model, compartmental model), the sector(s) it represented (e.g. food-producing animals, humans, crops, environment), the micro-organism and/or the antimicrobial(s) involved (e.g. fluoroquinolone-resistant E. coli), and other model characteristics (see Tables 2.1, 2.2, and 2.3, full database) [28]. We categorized models to the environment sector if they represented all anything that is external to a host [29]. In more detail, for this paper the environment sector That included any area in which a person, animal, or plant is living or operating [30], which was dependent on the setting of the model and the population of interest (e.g. the bed, the light switch, or the keyboard in a hospital, or a river or surrounding landscape).

The identified models were visually depicted in situated within the diagram (Figure 1) to help identify gaps in system coverage, as well as depict the amount of overlap between sectors captured by existing model. For example, to highlight if there models of human health, on-farm, or the environment, as well as models that cross sectors such as zoonotic transmission, food transmission, models including environmental reservoirs, or One Health models.

1.2.2 – Objective 2: Existing Data Sources and Evidence

A second literature search of peer-reviewed publications and grey literature was performed in Google Scholar, PubMed, and Google, to find and compile data to populate the model outlined above using multiple types of data (quantitative and qualitative) from a variety of data sources from 1995-present, prioritizing sources 2000 to present and those that were Sweden-focused. While our interest was to identify data from 2000 to present, we used 1995 as a cut off to account for potential gaps in data collection (e.g. data collected every five years which causes a gap from 1997-2002). Furthermore, if no Sweden-specific data existed, northern Europe was used as a proxy, and if nothing was specific to northern Europe, then all of Europe or a European average was included. The search was not limited by language; all non-English publications were translated with Google Translate.

Search criteria were created based on the 92 nodes identified on the diagram [19]. A search was created for each node, and some were refined further if found necessary during the search. For example, for the node “on-farm production”, a separate search was done for on-farm production of animal-based foods (e.g. chicken, beef, dairy), fruits (e.g. apples), vegetables (e.g. potatoes), and other important crops (e.g. wheat, rice). Examples of search criteria include: “Antimicrobial resistance AND human AND Sweden”, “Imports AND chicken AND Sweden”, “Profits AND Pharmaceuticals AND Sweden”. Some searches were narrowed using specific examples given by the participants within the workshop that led to the diagram [19] when the nodes were too broad to perform an adequate search. For example, for the node “New and emerging food”, specific products that are in development or becoming of interest to the European population were searched, such as: insects, genetically modified foods, three-dimensionally printed foods, and lab-based meat. There were nodes for which it was more difficult to create an adequate search strategy. These nodes were those that were abstract (e.g. diverse experiences and opinions) or broad (e.g. AMU in countries other than of Sweden). The aim was to find search strategies that were specific enough to retrieve relevant information, while still capturing a broad range of data and sources. A full list of the search strategies used can be found in Additional file 1, Table A2. As with objective 1, the searches for the 92 nodes each returned over 750 results, and therefore we reviewed titles from the first 100 results, for each search string, from each database. Citations that were excluded were: 1) those that did not contain information relevant to the list of nodes from the diagram, and 2) those published before 1995.

Due to the number of separate searches performed (at least one per node), the literature scan was time and resource intensive. Therefore, per recommendations by Arksey & O’Malley [20], a three- month cut-off date was used (December 31^st, 2020) in which articles identified as relevant by title and abstract but were not read in full were noted in a separate list (see Additional file 2). These articles did not undergo the proceeding steps of full review for inclusion and data extraction.

MC reviewed the full text for titles and abstracts that were screened in and excluded any sources that did not contain data relevant to the nodes of interest. MC then extracted the data. Due to the variety of sources and types of reported data, an inter-coder reliability check was done with three members of our team (MC, KD, XMYK) to ensure the same data were being extracted from the articles. MC selected three articles that represented the spectrum of the articles identified for data extraction (2 peer-reviewed and 1 grey literature that were quantitative, qualitative, and mixed-method) and a full article review was performed. MC, KD, and XMYK compared and discussed results to refine what needed to be extracted, or what was being missed.

For the sources that were included, the following information was extracted (MC) for each node and organized into an excel database (MNV & MC): the data or parameter extracted (e.g. 1,000,000 prescriptions/year), the year the data was collected, the type of data (e.g. pharmaceutical sales, quantitative), the source (e.g. peer-reviewed or grey literature), what kind of study it was (e.g. survey, cross sectional), the country or countries for which the data are available, and the date and country of publication (see Table 4, full database) [28]. The country of publication was identified by either the affiliation of the first author on a peer-reviewed publication, the country in which the author of an article (magazine, newspaper, blog) was located, or the country of the main headquarters of an organization, magazine, newspaper, or webpage.

The amount of data and an associated level for the node was then initially assigned by MC’s personal judgment, and then verified by the research team through discussion, with disagreements being resolved through consensus. The level of the node refers to the position of that node in the context of the Swedish food system on a scale of the amount, quantity, extent, or quality compared to a referent (e.g. Sweden versus other countries within or outside of Europe, Sweden currently versus historically). For example, there is “high” AMU in agriculture in Sweden in 2020 compared to 2010 or there is “low” AMR in Sweden compared to other countries in Europe. The following levels were assigned: very high, high, medium, low, very low, or none. A node was able to be assigned a level based on the decision criteria found in Figure 2, as follows.

First, a node had to have enough data to create an accurate evaluation of the state of the node. Then it had to satisfy the following two criteria: 1) a good comparator to be able to judge the node against (e.g. historical data or data from another country), and 2) the node could be accurately described by a single level. For example, the node “Consumer choice, demand, and behaviour” could theoretically be split into many different nodes that represent how consumers feel about different commodities (e.g. consumers in Sweden have “high” demand for organic and animal-welfare friendly foods, but “low” demand for genetically modified foods). Although our search identified that this node could be split into multiple nodes, including demand for: animal welfare-friendly products, consumption of meat and other animal-based food vs other (non-food producing animal) food, antimicrobials, new and emerging food, organic vs conventional food (production system), and local vs imported food, doing so was beyond the scope of our study. And so, in spite of abundant data, no level could be assigned to “Consumer choice, demand and behaviour”.

The amount of data to inform each node was also assigned a categorical level (very little, a little, some, a lot, and most), which was assigned using personal judgement based on the number of sources and data points (e.g. many sources or many years of data collected), and the amount of quantitative and qualitative data that existed for a given node (refer to decision tree in Figure 3). A data point was defined as a single piece of qualitative (e.g. a quotation) or quantitative data (e.g. a statistic) referring to a single node.

A visual representation of the existing data was overlayed on the diagram of AMR in the Swedish food system. This included a representation of major data gaps, the amount of data that exists, the quality of the data and the associated level of the nodes.

Although searches were conducted based on the nodes, data pertaining to the relationships between the nodes was also found and extracted into the database, but these were not categorized into categorical levels [28]. New relationships that were identified via the literature search were mapped on top of the original diagram (Additional file 3, Figure A1), and the sources that had data pertaining to the existing relationships were also visually mapped (Additional file 3, Figure A2).

1.3.1 – Objective 1: Existing Models

We identified a total of 146 relevant peer-reviewed articles (Table 1) that provided good coverage of the One Health system (Figure 1). Each individual article and relevant information pertaining to the articles can be found in the database [28]. Most articles were published after 2000 (Figure 4). Most (79/146, 54%) models were either non-geographic context specific or included countries from all over the world (e.g. international travel, global food demand [28]), four of which were situated in multiple countries of in Europe. Many articles in which a country was defined were from high income countries (58/146, 40%), as defined by World Bank (e.g. United States of America (USA), Sweden, Denmark) [31], and only 10 articles (7%) represented models from low- and middle-income countries. Our search yielded 14 articles that described economic models, however upon further analysis we found that these models were not useable for our purpose as they represented statistical models (e.g. econometric, time series) that reflect the associations between nodes, and were not simulation models. These excluded models are described in Additional file 1, Table A3 (details about each model can be found in the database [28]) and are of interest for future research.

There was a total of 126 articles that described models focused on disease transmission and/or AMR emergence and transmission, 102 of which were models of a single system, 4 were models of a single system that included an economic component (e.g. cost-effectiveness of an intervention or cost-analysis of a disease outbreak), and 20 were review articles. Table 2 summarizes the models in terms of the location within the system the model represents (e.g. a hospital, on farm, food), the type of transmission that is described (e.g. person-to-person, animal-to-animal), and whether the model was representing a sensitive microbe, a resistant microbe, or an AM. Table 3 summarizes the microbes and AMs that were modelled and the different sectors or transmission pathways these models encompassed. Overall, less than half of these models addressed two or more sectors (46/106, 43%). Many articles were focused on humans (42/106, 40%). Many of those that dealt with the human-animal interface were concerned with human illness as the main outcome and considered animal exposure as a risk factor. Models that considered the environment as part of the transmission pathway looked at either the immediate surroundings of the host (contaminated pens, hospital equipment, and production equipment; 23/106, 22%) or the natural environment (water sources, soil, plants; 8/106, 8%). The environment as a target area for modelling was very limited (3/106, 3%), and was mostly framed in terms of causing human illness or as a reservoir for pathogens. There were very few models about plant/crop agriculture (5/106, 5%).

*Table 3 shall appear here

Resistant bacteria, specifically E. coli and methicillin-resistant Staphylococcus aureus (MRSA), were the most commonly modelled compared to other resistant microbes, (Table 3). Five models focussed on an AM instead of a target microbe (Table 3), specifically on the decay of antimicrobials in different settings (one in the gastrointestinal tract of a host, and four in water systems) and the build-up of residues in these settings. One model also described the antimicrobial’s effect on the development of AMR in a pathogen. Five models did not have a specified host (Table 3) and were mainly concerned with transmission of resistance genes and the development of resistance in a population of micro-organisms after exposure to an antimicrobial.

Transmission model characteristics related to the type of pathogen (susceptible or resistant) and AM being modelled are shown in Table 4. The majority (42/106; 40%) were human-focused and were either at a population- level (e.g. transfers of patients between hospitals and community members) or at a microscopic or micro-organism level (e.g. within-host models of gene transfer and AMR emergence), with very few including both mechanisms into a multi-level model (Table 4). There were very few animal-focussed models of AMR. Alternatively, models of susceptible microbes (or those that did not state the resistance status) were mainly animal centred and focussed on within-farm or between-farm transmission (Table 2).

*Table 4shall appear here

Finally, most of the models described were compartmental (n=73/106, 69%) [32–34] and deterministic (53/106, 50%) [32,34], and used data from the peer-reviewed literature as inputs for model parameter values (53/106, 50%). This was especially true within the AMR-focused models (Table 2). The use of theoretical models was also widespread (34/106, 32%), especially within the AMR models. We defined theoretical models as models that provide a general structure and are solved mathematically, but are not informed by empirical data. These models are occasionally tested with parameters from the literature or with wide ranges for parameters to capture many possible values. Models rarely included all indications of rigour (sensitivity analysis, validation, and calibration [35–37] but over half of the models (56/106, 53%) included at least one of these features.

1.3.2 – Objective 2: Existing Data Sources and Evidence

A total of 414 articles were read in their entirety that addressed 64 of the 92 nodes identified as important drivers in the Swedish food system for the emergence and transmission of AMR (see full database [28]). Despite the thorough review, there were 28 nodes for which no literature was found. These nodes included: AMU in wildlife; AMU and AMR in other countries; antimicrobial- resistant organisms (AROs) in plant agriculture; cost per unit set by quota; disposal of AMs (e.g. unused, unmetabolized); diverse experiences, education, and training; existing farm infrastructure; existing healthcare infrastructure; exposure to AROs through imported products; good farm practices; host microbiome; level of resistance AMR in other countries; national budgets money, and funding; non-AM infection prevention and control in plant agriculture, by the public, and in other social institutional settings; producer profitability; research, development, and innovation; resistance at the abattoir/processor; restocking with animals/eggs at higher risk of infection; retail availability of meat/eggs in domestic market; science and academia; time to market weight; treatment post-procedure; what is being farmed; and the wider environment microbiome (e.g. water, soil).

The data came from both peer-reviewed (149/414, 40%) and grey literature (228/414, 60%). Full details on the types of sources identified in the grey literature are depicted in Figure 5. These sources came from multiple countries (Figure 6) with Sweden being the predominant country of origin as it was prioritized in the search (117/414, 28%). The sources were mainly published after 2017 (205/414, 50%; Figure 7). Many databases reported surveillance data and national and regional statistics or indicators which were particularly useful, including: Food and Agriculture Organization of the United Nations (FAO), Worldbank, Eurostat, Migration Data Portal, Our World in Data. Similarly, other useful statistical webpages depicted or combined information found in common databases: Statista, Knoema, TrendEconomy, and Indexmundi.

There was a combination of qualitative and quantitative data reported, the majority being quantitative in nature (4445/5432 data points, 82%). However, more qualitative data were available to parameterize some nodes, due to either a lack of available quantitative data (e.g. development of alternatives to AM), or due to the nature of the node being described (e.g. animal welfare/low stress, consumer demand and behaviour). The data described in Table 5 were used to determine the amount and quality of the data for each node (see Additional file 3, Figure A2). Ten of the nodes had only one source and limited very limited sources (1 source) and data (1-15 data points) to inform the node; they and were assigned the category “very little” for amount and quality of data. Six of the nodes had a “most” for the amount of data. These nodes had 16 to 30 sources and 234 to 413 data points per node, with an average of 94% (88-97%) being quantitative data points.

*Table 5 shall appear here

The literature search was targeted towards the nodes, however, as a result of reading and analyzing the extracted literature, there was valuable information identified pertaining to the relationships (arrows) in the diagram. There were a total of 325 relationships found in the literature, 86 were already in the diagram, and 239 were newly identified as a result of the data extraction (see Additional file 3, Figure A1).

Two new nodes, mentioned as a part of the newly identified relationships, were also added to the diagram: “Access to healthcare (doctors, hospitals, veterinarians, etc.)” and “Number of abattoirs”.

The primary purpose of this literature scan was to map out the existing data landscape, specifically models and data related to AMR that exist across a wider One Health system, and specifically for the Swedish food system context. We found a wide variety of dynamic models that described many of the transmission pathways within the system. These models however were segregated and had limited connection across sectors. We also found quantitative and qualitative data to help understand the current and past states of many of the nodes representing the system. The data were highly variable in terms of quantity, source, and type but overall, when combined together, helped extend our understanding of the overall system and have the potential to be incorporated into future models of AMR.

1.4.1 – Literature scan of models of, and factors associated with AMR

Scoping reviews are useful tools for summarizing and disseminating existing literature in a useable and concise format for use by other researchers, policy makers, and other stakeholders who may not have the skills or resources to perform such a task [38]. Therefore, by basing this literature scan in scoping review methodology, this literature scan: 1) provides a framework to improve the transparency of data collection for parameterization of a model of a complex system; 2) can support the building of a future comprehensive One Health model of AMR related to the Swedish food system; and 3) can assist other modellers by providing a comprehensive list of existing models of AMR in various sectors, where data exist within those sectors, and how to gain access to these data. It also provides insight into the areas in which more empirical data is are needed before comprehensive quantitative models can be created. For example, we found that more data are needed about AMR in the environmental sector (e.g. levels of resistance within the soil and waterways), in wildlife populations (e.g. the level of resistance in birds, rodents, and other wild animals that have intimate contact with food-producing animals and humans and the relative contribution these pathways have to overall transmission to humans or food-producing animals), in crop agriculture, and for nodes referring to economics, resources, and practices on farm and in healthcare (e.g. healthcare and farm infrastructure, good farm practices, feed quality and feed efficiency). Furthermore, because most models were only created to encompass a single sector, we also found that more research is needed into modelling methods that can be used to connect existing models from the various sectors.

1.4.1.1 – Disease transmission models embedded within the food system

We found 106 existing models that provided good coverage of the main parts of the One Health system (human, animal, environment), especially for specific human populations (e.g. within hospital) and agriculture (e.g. on-farm transmission). This review included models that are not directly to AMR (e.g. Foot and Mouth Disease, African Swine Fever, Influenza) as these models can further our understanding of the system that drives AMR and have modelled parts of the system that have not yet been explored in AMR-specific models (e.g. importation and global spread of Influenza). However, even with the inclusion of these models, the models had limited connection between the sectors, and did not cover nodes less directly related to human health and agriculture (e.g. development of new AMs, research and development, what is being farmed). Similarly, human-focused models and limited connection between sectors was also identified as a major gap in models of AMR in previous reviews [38,39]. Because many microbes and resistant organisms easily spread between humans, animals, and the environment (e.g. resistant E. coli can be shed by animals into the environment and contaminate the watershed which can then infect humans and vice versa [1,2,25,27] and because AMU in one of these sectors has the ability to select for resistance in the other sectors (e.g. humans excrete antimicrobial residues in their faeces which can then make their way to the environment through wastewater and select for resistance in pathogens in the environment) [1,2], cross-sector connections are important to capture in models if we want to model the complex One Health dynamics of AMR. However, the current research design for some microbes have shown limited evidence of transmission between sectors that pose a risk to human health [24,39]. Furthermore, the identified models typically took a humanistic perspective, looking at how animals, crops, or the environment can lead to resistance in humans, however humans also have the potential to introduce and spread resistant microbes to animals and into the environment [1,2,7]. Therefore, looking at this issue from a more One Health perspective, including how humans influence the system and drive AMR in animals and the environment, within models is another important step into a more integrative modelling approach.

Here, smaller scale models, such as those in a single hospital or farm included the environment as a source of transmission (e.g. shedding feces, contaminated spaces and workers in hospital). In contrast, larger community models of AMR were simplified to just human-human or animal-animal transmission. Therefore, it is becoming increasingly clear that models need to account for these different transmission pathways to better understand and address the complex issue of AMR. Compartmental models provide a good foundation, and are typically the starting point for modelling new systems [32]. Therefore, by creating a set of interlinked compartmental models from multiple areas of the system, we could create a more comprehensive model of the system. However, models of this level of complexity are difficult to create due to lack of data and knowledge about some of the associations or dynamics, which can lead to larger uncertainty in results and difficulty in interpreting results. Compartmental models provide a good foundation and are typically the starting point for modelling new systems [32]. Therefore, by creating a set of interlinked compartmental models from multiple areas of the system, we could create a more comprehensive model of the system.

1.4.1.1.1 – AMR-specific disease transmission models

The AMR-specific models identified were mainly quantitative, deterministic, compartmental models, which aligns with past reviews of the literature [24,39] and was expected since AMR is a developing field, and these models are usually the starting point for understanding processes and the first models built when addressing an issue [32]. These compartmental models also lend themselves well to be expanded into integrated or multi-level models (e.g. within-host and population level [32]). Agent-based, or individual-based models allow researchers to add more heterogeneity to the population and include additional population level attributes (e.g. different levels of contact, spatial elements, individual behaviours [32,40,41]). These models are extremely useful but require more data about the populations and the pathogens they model and therefore can be more difficult to parameterize [32,40,41]. Expanding a compartmental model is a good starting point to incorporate more complexity and capture the wider system, but still the data must exist in order to parameterize the model without the need for multiple assumptions and uncertainty. In fact, many of the models were theoretical models, which are models that outline the structure and transmission pathways, but do not use data to inform them. Previous reviews highlighted that although these theoretical models are useful to understand the overall transmission, without data to inform them, their ability to be used to accurately model a system or to assess interventions is limited and untested [24,39]. Although quantitative models dominate the simulation modelling world, the inclusion of qualitative data would greatly expand the information that could be used in modelling and a further search for qualitative or mixed-methods simulation systems models could help bridge the gap.

1.4.1.2 – Economic models embedded within the food system

Economic modelling is becoming more important in making policy change in many fields, including: health interventions such as obesity [42,43], the use of medicine and other healthcare technologies [44,45], hospital interventions for disease control [46,47], and AMR [48], as well as other fields such as energy [49], agriculture [50], and climate change [51]. It has been deemed necessary to identify solutions that can have the most impact with the least amount of associated costs [52–54]. Cost is important to consider when dealing with agricultural and food production as there are many levels at which the costs can be applied [55,56]. For example, if the ultimate goal is for impacts at the human population- level (e.g. reducing AMR), but it is not economical for producers to implement the intervention (e.g. reducing AMU, changing farming practices, or updating farm infrastructure), then it is not going to be adopted as easily as an intervention that allows producers to still make a profit (lack of profitability could put the farmer out of business) [6,7,55,57]. Economics are also important because consumer demand is a strong driver of food availability, cost, and ultimately AMR [17,19]. To provide enough food for the growing population, at a price that consumers can afford, agricultural practices have adapted to high intensity farming which may lead to higher animal density and may drive the need for AMs to for disease prevention, treatment or control combat inevitable diseases [19]. Including economic analysis when modelling the impacts of interventions can be useful to help weigh high initial costs against longer-term pay-offs to determine the overall cost-effectiveness of the intervention. Therefore, engaging economists and economic modelers to help create more cross-over models (models that include economics and AMR dynamics) may be an important next step when it comes to testing interventions that can inform policies that address controlling AMR at an international scale.

1.4.1.3 – The evidence landscape of factors that drive AMR

Overall, we found the evidence landscape was challenging to navigate. The data existed across many formats and sources, and it required substantial interpretation, a lot of searching, deciphering jargon, and sifting through multiple databases, government webpages, and other literature. After the data were found, it was hard to compare between populations and contexts, with many different metrics being used and reported. In some cases, the data were patchy and incomplete, or were not publicly available. Therefore, to create a systems model including all 92 nodes identified by the experts from the participatory modelling workshops [19], a separate literature scan may be needed to address to be conducted for each specific research question (node), especially those for which data were not readily available. We estimated that this could take upwards of 276 person-months (given a cut-off of 3 months per node for 92 nodes), which would require significant time and resources.

From our literature scan, we found that some of the nodes were informed by quantitative data, and these data captured many years and were specific to the Swedish context. However, in some cases, the way in which the data were either collected or reported was not useful for quantitative modelling purposes. For example, some data were only reported monthly or yearly (as opposed to daily or weekly which is required for many quantitative models), were not representative of the entire country, or did not capture the entire context (e.g. surveillance data but only for specific hospitals or farms) and could not be generalized to a population level. The types of quantitative data we found could, however, be reduced to qualitative categories to represent the state of the node within a semi-quantitative model. For example if one country’s human AMU in 2015 is 5 defined daily doses (DDD) per 100,000 population, the human AMU in another country is 10 DDD per 100,000 population, and the human AMU in a third country is 18 DDD per 100,000 population, [58]. These values could be turned into categories. For example, 18 DDD per 1000,000 population in the high category, 5 DDD per 1000,000 population in the low category, and 10 DDD per 100,000 population falling in the middle (medium). We also found that some nodes were mainly represented in qualitative terms. These nodes are still important to the system (as identified by the workshop participants [19]) and therefore finding a way to incorporate this valuable data is necessary to capture the nuances of the One Health system. This provides further evidence of the need for qualitative or mixed-methods simulation modelling when trying to model a system of this complexity and breadth.

1.4.2 – Limitations

This literature scan aimed to identify a breadth of models that fit our context and outcome of interest (One Health model of AMR in a high-income food system), not to identify every possible model of AMR and zoonotic transmission. Therefore, it is likely that some models were missed that could provide additional insight into the system. However, many models were identified that provide wide coverage of the system and provide wide coverage of the system and a good foundation for creating a One Health AMR model.

When performing the literature scan, the search for models were not limited by geographical context. However, the majority of the models identified were from a high-income context. Although this is not an issue for our specific goal of creating a model in Sweden, a high-income country, this limits the generalizability of this review to modelling low- and middle-income contexts and a specific search would be necessary to find these models.

The literature scan of the existing evidence landscape was thorough, in which searches for data to inform all 92 nodes were attempted by creating, testing, and refining of the search strings. However, it was more difficult to create searches that captured some nodes in their entirety or were narrow enough to capture some of the more detailed or niche aspects of the nodes. For example, the node “Consumption of other (non-meat/egg) foods”, is extremely broad and could include fruits, vegetable, and grains, but also things such as pop, snack foods, and alcohol. Therefore, while executing a literature scan of a large and complex system requires significant time and resources, it is an extremely important first step in the research and model building process. Through conducting this literature scan, it became evident that this area of research requires greater attention. This could include a more comprehensive scan of the literature pertaining to the rest of the system by a large inter-disciplinary team to identify all potential data sources that could be used to inform a model or refining the node to be more appropriate from a modelling perspective. However, this literature search provides a preliminary review of the existing models which met our inclusion criteria and data in the realm of AMR which may be applicable to these models and highlights that despite the availability of information, there are substantial challenges to it is not cohesive, accessible, or easy to finding, accessing and compiling. This is also true for the relationships (arrows) in the diagram. There were 331 relationships in the diagram as identified by the workshop participants [19], with 239 new relationships identified through the literature scan. To quantify (or semi-quantify) these relationships, a more in-depth search would need to be performed targeting each of these relationships individually.

Similarly, there were a subset of nodes that were broad and abstract in nature (e.g. diverse experience, knowledge, and training) and although these nodes are important to AMR dynamics, they are too challenging to describe on a national and population level. For example, an individual’s cultural and educational background along with their past experiences can greatly shape how they view different aspects of the system, such as: what they eat , how they access the healthcare system [19,59,60], their trust in doctors and medicine [19,61]. This can then shape their exposures and risk of AMR. However, it is challenging to put this into a compartmental model and capture these differences amongst the population at a national level to determine as a collective how AMR may emerge and spread through the country. There are certain aspects that are important and would be interesting to address when modelling but some nodes may be too complex or abstract, and some relationships may exist that quantitative modelling cannot capture. Therefore, further engagement with experts would be required to better define/refine these nodes, for example, by discussing how best to create more defined and quantifiable nodes (e.g. population averages for level of education, dietary preferences, or knowledge of AMR [62]). However, more sophisticated modelling techniques (e.g. individual level models that incorporated decision making based on past experiences) or other approaches beyond quantitative simulation modelling (e.g. scenario planning [63]) could also help address the issue of incorporating individual level characteristics, and future research should include interdisciplinarity and One Health approaches.

Finally, this literature scan was conducted as a pragmatic way of collecting data to outline and parameterize a large-scale model of AMR across a One Health system. Therefore, validating the quality of evidence to support the data or validity of the sources was not a primary aim of this review and not inherently part of the literature scan process. However, most sources were from peer-reviewed, government, or major newspaper source and therefore, we have confidence in the data that were collected can be representative of the situation within a high-income country such as Sweden.

This literature scan identified many models that addressed different aspects of the Swedish food system. However, models they were disparate and primarily quantitative, and have not been integrated into models that represent the One Health aspects of this issue. Furthermore, there are many data gaps that exist for multiple nodes, making it difficult to model the entire system using empirical data, with some nodes being too broad or abstract to include in a population or national level model, though they were deemed important to the overall system. It is clear from the literature that current quantitative models are unable to capture the complexity, nonlinear relationships, and interconnectedness of the system that drives AMR. Therefore, given the existing models found from this literature search along with the data requirements for the models and the data availability, it is not possible to create a fully quantitative model of the complex systems of drivers of AMR in the Swedish food system context without including overtly simplifying assumptions. This study shows that there is a base of knowledge that exists in the literature, however further work is needed to determine how to integrate this knowledge and put the different pieces together to create a comprehensive model and understanding of the food system that drives AMR emergence and transmission within the Swedish context.

One way in which we could use the plethora of valuable data we found through our review, including both the quantitative and qualitative data, is through the creation of a mixed-methods or semi-quantitative compartmental model of the 64 nodes for which there were data. These alternative methods may be able to address the complexity of the system and help provide evidence to inform the factors and the nonlinear relationships that are hard to quantify through traditional quantitative methods. In the interim, the data gaps found through this review can be used to advocate for further evidence that could inform an empirically driven quantitative model of the broader One Health system that drives AMR in a high-income context. To accomplish this, there is a need for more interdisciplinarity and cross-sector collaboration to help bring different perspectives, expertise, and knowledge of existing or novel models or where and how data are being collected and utilized, thus fostering communication and sharing of information to gain a more holistic and One Health view of the system of AMR.

AM – Antimicrobial

AmpC - AmpC-β-lactamase

AMR – Antimicrobial resistance

AMU – Antimicrobial use

ARO – Antimicrobial- resistant organism

ASF – African swine fever

CLD – Causal loop diagram

CRE - Carbapenem-resistant Enterobacterales

EEA - European Economic Area

EU – European Union

ESBL - extended-spectrum-β-lactamase

FMD - Foot and Mouth Disease

ICU – Intensive care unit

MRSA - Methicillin-resistant Staphylococcus aureus

OH – One Health

STEC – Shiga toxin-producing Escherichia coli

USA – United States of America

UK – United Kingdom

VRE - Vancomycin-resistant enterococci

VTEC - Verocytotoxin-producing Escherichia coli

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This study is funded through an operating grant of the 5^th Joint Programming Initiative on Antimicrobial Resistance (JPIAMR 2017). Funding was provided by an operating grant from the Canadian Institutes for Health Research (Institute of Infection and Immunity, Institute of Population and Public Health, grant number 155210, PI SEM), a Swedish Research Council grant (grant number 2017-05981, PI and Project Coordinator PSJ); and an operating grant from the Swiss National Science Foundation (grant number 40AR40_180189, PI DW). The funders had no role in the design analysis or writing of this article.

Authors’ contributions

All authors were involved in the study conception and design, analysis and interpretation of data, and drafting of the manuscript. MC conducted the literature scan, analyzed and interpreted the results and drafted the original manuscript. MNV organized the data and sources for Objective 2. All authors contributed to the revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.

Acknowledgements

The author’s would like to thank Kim D’Mello and Xenia Man Yuk Kwan for conducting the inter-coder reliability check.

WHO (World Health Organization). Antibiotic Resistance [Internet]. 2018. Available from: https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance
Government of Canada. Tackling Antimicrobial Resistance and Antimicrobial Use: A Pan-Canadian Framework for Action. 2017; Available from: https://www.canada.ca/content/dam/hc-sc/documents/services/publications/drugs-health-products/tackling-antimicrobial-resistance-use-pan-canadian-framework-action/tackling-antimicrobial-resistance-use-pan-canadian-framework-action.pdf
Murray CJ, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399(10325):629–55.
Commission E. AMR: a major European and global challenge [Internet]. 2019. Available from: https://ec.europa.eu/health/amr/sites/amr/files/amr_factsheet_en.pdf
O’Neill J. Tackling drug-resistant infections globally: Final report and recommendations [Internet]. 2016. Available from: https://amr-review.org/sites/default/files/160518_Final paper_with cover.pdf
Aarestrup FM, Wegener HC, Collignon P. Resistance in bacteria of the food chain: epidemiology and control strategies. Expert Rev Anti Infect Ther [Internet]. 2008 Oct 1;6(5):733–50. Available from: https://doi.org/10.1586/14787210.6.5.733
Mcewen SA, Collignon PJ. Antimicrobial Resistance: A One Health Perspective. Microbiol Spectr [Internet]. 2017;6(2):1–26. Available from: www.chathamhouse.org
Søgaard Jørgensen P, Folke C, Henriksson PJG, Malmros K, Troell M, Zorzet A. Coevolutionary Governance of Antibiotic and Pesticide Resistance. Trends Ecol Evol [Internet]. 2020;35(6):484–94. Available from: https://doi.org/10.1016/j.tree.2020.01.011
Malik B, Bhattacharyya S. Antibiotic drug-resistance as a complex system driven by socio-economic growth and antibiotic misuse. Sci Rep. 2019;9(1):1–12.
Islam S, Aldstadt J, Aga D. Global antimicrobial resistance: a complex and dire threat with few definite answers. Trop Med Int Heal. 2019;24(6):658–62.
Wang L, Zhang X, Liang X, Bloom G. Addressing antimicrobial resistance in China: Policy implementation in a complex context. Global Health [Internet]. 2016;12(1):1–9. Available from: http://dx.doi.org/10.1186/s12992-016-0167-7
Acar JF, Moulin G. Antimicrobial resistance : a complex issue Pre-existence of antimicrobial resistance determinants Origin of antimicrobial resistance determinants. Rev Sci Tech. 2012;31(1):23–31.
Ladyman J, Lambert J, Wiesner K. What is a complex system? Vol. 3, European Journal for Philosophy of Science. 2013. 33–67 p.
Waterloo Institute of Complexity and Innovation. What are complex systems? [Internet]. Univeristy of Waterloo. 2022. Available from: https://uwaterloo.ca/complexity-innovation/about/what-are-complex-systems
Chen X, Fu F. Social learning of prescribing behavior can promote population optimum of antibiotic use. Front Phys. 2018;6(DEC):22–5.
Murphy CP, Carson C, Smith BA, Chapman B, Marrotte J, McCann M, et al. Factors potentially linked with the occurrence of antimicrobial resistance in selected bacteria from cattle, chickens and pigs: A scoping review of publications for use in modelling of antimicrobial resistance (IAM.AMR Project). Zoonoses Public Health. 2018;65(8):957–71.
Primeau C. Exploring the contributions of genotypic, phenotypic, social and qualitative data sources to our understanding of antimicrobial resistance in Canada [Internet]. Univeristy of Guelph; 2020. Available from: http://hdl.handle.net/10214/17935
Goltz J. Investigating Factors Associated with Extended-Spectrum Beta- Lactamase-Producing Enterobacteriaceae Colonization and / or Infection in Humans : A One Health Approach by [Internet]. University of Guelph; 2022. Available from: https://hdl.handle.net/10214/26906
Lambraki IA, Cousins M, Graells T, Leger A, Henriksson P, Harbarth S, et al. Factors influencing antimicrobial resistance in the European food system and potential leverage points for intervention: A participatory, One Health study. PLoS One. 2022;17(2 February):1–19.
Arksey H, O’Malley L. Scoping studies: Towards a methodological framework. Int J Soc Res Methodol Theory Pract. 2005;8(1):19–32.
Niewiadomska AM, Jayabalasingham B, Seidman JC, Willem L, Grenfell B, Spiro D, et al. Population-level mathematical modeling of antimicrobial resistance: A systematic review. BMC Med. 2019;17(1):1–20.
Spicknall IH, Foxman B, Marrs CF, Eisenberg JNS. A modeling framework for the evolution and spread of antibiotic resistance: Literature review and model categorization. Am J Epidemiol. 2013;178(4):508–20.
Opatowski L, Guillemot D, Boëlle PY, Temime L. Contribution of mathematical modeling to the fight against bacterial antibiotic resistance. Curr Opin Infect Dis. 2011;24(3):279–87.
Ramsay DE, Invik J, Checkley SL, Gow SP, Osgood ND, Waldner CL. Application of dynamic modelling techniques to the problem of antibacterial use and resistance: A scoping review. Epidemiol Infect. 2018;146(16):2014–27.
Léger A, Lambraki I, Graells T, Cousins M, Henriksson PJG, Harbarth S, et al. Characterizing social-ecological context and success factors of antimicrobial resistance interventions across the One Health spectrum: analysis of 42 interventions targeting E. coli. BMC Infect Dis [Internet]. 2021;21(1):1–13. Available from: https://doi.org/10.1186/s12879-021-06483-z
European Centre for Disease Prevention and Control. Surveillance and disease data for Escherichia coli [Internet]. 2021. Available from: https://www.ecdc.europa.eu/en/escherichia-coli-ecoli/surveillance-and-disease-data
WHO Regional office for Europe. The burden of food borne diseases in the WHO European region [Internet]. The Regional Office for Europe of the World Health Organization. 2017. Available from: https://www.euro.who.int/__data/assets/pdf_file/0005/402989/50607-WHO-Food-Safety-publicationV4_Web.pdf
Cousins M. Mapping out a One Health model in the context of the Swedish food system using a modified scoping review methodology: Scoping Review Database. Borealis. 2022. p. V1, UNF:6:Scmti+htcN4SQ9/sgkUwFQ== [fileUNF].
Porta M. Environment. In: Dictionary of epidemiology. 6th ed. Oxford University Press; 2016.
Oxford Learner’s Dictionary. Environment [Internet]. Available from: https://www.oxfordlearnersdictionaries.com/us/definition/english/environment
World Bank Group. The world by income and region. WDI - The World by Income and Region.
Hohle M. Infectious Disease Modelling 1. In: Hand-book on Spatial Epidemiology. 2015. p. 1–19.
Blackwood JC, Childs LM. An introduction to compartmental modeling for the budding infectious disease modeler. Lett Biomath. 2018;5(1):195–221.
Grassly NC, Fraser C. Mathematical models of infectious disease transmission. Nat Rev Microbiol [Internet]. 2008;6(of Publication: June 2008):6 (6) (pp 477-487), 2008. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&NEWS=N&PAGE=fulltext&D=emed8&AN=2008333498
Barlas Y. Formal aspects of model validity and validation in system dynamics. Syst Dyn Rev Vol. 1996;12(3):183–210.
Martis MS. Validation of simulation based models: A theoretical outlook. Electron J Bus Res Methods. 2006;4(1):39–46.
van Kleef E, Robotham J V., Jit M, Deeny SR, Edmunds WJ. Modelling the transmission of healthcare associated infections: A systematic review. BMC Infect Dis. 2013;13(1).
Antman E, Lau J, Kupeinick B, Mosteller F, Chalmers T. A comparison of results of meta-analysis of RCTs and recommendations of clinical experts. J Am Med Assoc. 1992;268(2):240–8.
Birkegård AC, Halasa T, Toft N, Folkesson A, Græsbøll K. Send more data: A systematic review of mathematical models of antimicrobial resistance. Antimicrob Resist Infect Control. 2018;7(1):1–12.
Bonabeau E. Agent-based modeling: Methods and techniques for simulating human systems. Proc Natl Acad Sci U S A. 2002;99(SUPPL. 3):7280–7.
Maglio P, Mabry P. Agent-Based Models and Systems Science Approaches to Public Health. Am J Prev Med [Internet]. 2011;40(30):392–4. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3624763/pdf/nihms412728.pdf
Sacks G, Veerman JL, Moodie M, Swinburn B. Traffic-light nutrition labelling and junk-food tax: A modelled comparison of cost-effectiveness for obesity prevention. Int J Obes. 2011;35(7):1001–9.
Gortmaker SL, Swinburn B, Levy D, Carter R, Mabry PL, Finegood D, et al. Changing the future of obesity: science, policy, and ... [Lancet. 2011] - PubMed - NCBI. Lancet [Internet]. 2011;378(9793):838–47. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21872752
Rittenhouse B. Uses of Models in Economic Evaluations of Medicines and other Health Technologies. London; 1996.
Luce BR. Policy implications of modeling the cost-effectiveness of health care technologies. Drug Inf Journal,. 1995;29:1469–75.
Schechner V, Carmeli Y, Leshno M. A mathematical model of Clostridium difficile transmission in medical wards and a costeffectiveness analysis comparing different strategies for laboratory diagnosis and patient isolation. PLoS One. 2017;12(2):1–12.
Assimakopoulos N. A network interdiction model for hospital infection control. Comput Biol Med. 1987;17(6):413–22.
Smith DL, Levin SA, Laxminarayan R. Strategic interactions in multi-institutional epidemics of antibiotic resistance. Proc Natl Acad Sci U S A. 2005;102(8):3153–8.
Hafezalkotob A. Modelling intervention policies of government in price-energy saving competition of Green Supply Chains. Comput Ind Eng. 2018;119:247–261.
Martínez-López B, Ivorra B, Fernández-Carrión E, Perez AM, Medel-Herrero A, Sánchez-Vizcaíno F, et al. A multi-analysis approach for space-time and economic evaluation of risks related with livestock diseases: The example of FMD in Peru. Prev Vet Med. 2014;114(1):47–63.
Kahil MT, Connor JD, Albiac J. Efficient water management policies for irrigation adaptation to climate change in Southern Europe. Ecol Econ [Internet]. 2015;120:226–33. Available from: http://dx.doi.org/10.1016/j.ecolecon.2015.11.004
Jamison DT. Intervention Cost-Effectiveness: Overview of Main Messages. In: DT J, JG B, AR M, Al E, editors. Disease Control Priorities in Developing Countries. 2nd Ed. Washington, DC: The International Bank for Reconstruction and Development / The World Bank; 2006.
Levin C, Chisholm D. Cost-Effectiveness and Affordability of Interventions, Policies, and Platforms for the Prevention and Treatment of Mental, Neurological, and Substance Use Disorders. In: Patel V, Chisholm D, Dua T, et al, editors. Mental, Neurological, and Substance Use Disorders: Disease Control Priorities. 3th Ed (4t. Washington, DC: International Bank for Reconstruction and Development / The World Bank; 2016.
Marseille E, Larson B, Kazi DS, Kahn JG, Rosen S. Thresholds for the cost–effectiveness of interventions: Alternative approaches. Bull World Health Organ. 2015;93(2):118–24.
FAO. The future of food and agriculture – Trends and challenges [Internet]. 2017. Available from: https://www.fao.org/3/i6583e/i6583e.pdf
Animals C on DU in F. The Use of Drugs in Food Animals: Benefits and Risks. Washingtotn, DC: National Academy Press; 1999.
Wernli D, Jørgensen PS, Morel CM, Carroll S, Harbarth S, Levrat N, et al. Mapping global policy discourse on antimicrobial resistance. BMJ Glob Heal. 2017;2(2):e000378.
The Centre for Disease Dynamics, Economics & Policy. ResistanceMap: Antibiotic use.
Lohm D, Davis M, Whittaker A, Flowers P. Role crisis, risk and trust in Australian general public narratives about antibiotic use and antimicrobial resistance. Heal Risk Soc [Internet]. 2020;22(3–4):231–48. Available from: https://doi.org/10.1080/13698575.2020.1783436
Haenssgen MJ, Charoenboon N, Xayavong T, Althaus T. Precarity and clinical determinants of healthcare-seeking behaviour and antibiotic use in rural Laos and Thailand. BMJ Glob Heal. 2020;5(12):1–12.
Blommaert A, Marais C, Hens N, Coenen S, Muller A, Goossens H, et al. Determinants of between-country differences in ambulatory antibiotic use and antibiotic resistance in Europe: A longitudinal observational study. J Antimicrob Chemother. 2014;69(2):535–47.
Collignon P, Beggs JJ, Walsh TR, Gandra S, Laxminarayan R. Anthropological and socioeconomic factors contributing to global antimicrobial resistance: a univariate and multivariable analysis. Lancet Planet Heal [Internet]. 2018;2(9):e398–405. Available from: http://dx.doi.org/10.1016/S2542-5196(18)30186-4
Amer M, Daim TU, Jetter A. A review of scenario planning. Futures. 2013;46:23–40.

Table 1: Breakdown of the types of models found in the sources identified in the literature (n=146).

	Number of articles
Disease transmission models	122
Single model or system	102
Review articles	20
Economic models	14
Single model or system	10
Review articles	4
Combination models	4
Single model or system	4
Review articles	0
Total	146

Table 2: Distribution of the 106 articles* referring to models of disease transmission (n=102) and models of disease transmission with an economic component (n=4) according to model processes, type of transmission, and type of model system (sensitive microbes, resistant microbes, and AMs**).

Location within the system that drives antimicrobial resistance	Type of transmission	Total Number of Articles	Sensitive microbes	Resistant microbes	AM only**
Community	person-person, person-environment	10	8	2	-
Hospital/healthcare facility	person-person, person-environment	11	2^†	9^†	-
Community-healthcare facility	person-person, person-environment	13	-	13	-
On-farm transmission	animal-animal, animal-environment	18	17^†	1	-
Between/introduction to farm transmission	animal-animal	10	9	1	-
Farm-to-slaughter	animal-animal	2	1	1	-
Crop agriculture	plant-environment	3	-	3	-
Processing plant/Slaughter house	animal-person, plant-person, cross-contamination	4	4	-	-
Zoonotic transmission	animal-person	5	4	1	-
Veterinary clinic	animal-person	1	-	1	-
Foodborne	animal-person, plant-person	5	2	3	-
Waterborne	environment-person	2	2	-	-
One health model	human-animal-environment	1	-	1	-
Pathogen level	with-in host, AMR emergence	14	-	14	-
Pharmacokinetic model	AM levels and decay	2	-	1	1
Hydrological model	AM residue and genetic element levels and decay	5	-	1	4
*TOTAL*		106	49	52	5

^*To find more details for each model that falls under the given category please refer to the complete database [28]. The complete list of articles are categorized with the same categories as appear in the table for easy cross-reference.

^**AM – Antimicrobial

^†At least one of these models contain an economic component

Table 3: Distribution of the 106 articles* referring to models of disease transmission of a single system (n=102) and models of disease transmission with an economic component (n=4) according to microbe and/or antimicrobial class and sector/population involved.

	Total number of articles	Unspecified host/ Microbe	Human	Animal	Crops	Environment	Human-Animal	Human-Environment	Animal-Environment	Crop-Human	Human-Animal-Crops-Environment (any 3)
Sensitive microbe
African swine fever (ASF)	1	-	-	1	-	-	-	-	-	-	-
Campylobacter	1	-	-	1	-	-	-	-	-	-	-
Vibrio cholerae	1	-	-	-	-	-	-	1	-	-	-
Clostridium difficile	1	-	1	-	-	-	-	-	-	-	-
Shiga-toxin producing Escherichia coli^**	18	-	-	3	1	-	-	-	13	1	-
Foot and Mouth Disease (FMD)	2	-	-	1	-	-	1	-	-	-	-
Influenza	6	-	3	-	-	-	3	-	-	-	-
Listeria monocytogenes	1	-	-	-	-	-	-	-	-	-	1
Neisseria gonorrhoea	2	-	2	-	-	-	-	-	-	-	-
Salmonella	3	-	-	1	-	-	-	-	2	-	-
Not defined (multiple)	13	-	3	6	-	-	2	1	-	-	1
TOTAL	49	-	9	13	1	-	4	2	15	1	2
						-
Antimicrobial only
Chlortetracycline	1	-	-	1	-	-	-	-	-	-	-
Fluoroquinolones	1	-	-	-	-	1	-	-	-	-	-
Tetracycline	2	-	-	-	-	1	-	-	1	-	-
Unspecified (multiple)	1	-	-	-	-	-	-	1	-	-	-
TOTAL	5	-	-	1	-	2	-	1	1	-	-

AMR (Microbe/AM)
CRE (Carbapenem-resistant Enterobacteriaceae)	1	-	1	-	-	-	-	-	-	-	-
Chlorotetrecycline- resistant E. coli**	1	-	-	1	-	-	-	-	-	-	-
ESBL (extended-spectrum-β-lactamase )-producing ST131 E. coli**	1	-	1	-	-	-	-	-	-	-	-
ESBL (extended-spectrum-β-lactamase) and AmpC (AmpC-β-lactamase)-producing E. coli**	1	-	-	1	-	-	-	-	-	-	-
Fluoroquinolone-resistant Campylobacter	1	-	-	-	-	-	1	-	-	-	-
Tetracycline-resistant E. coli**	1	1	-	-	-	-	-	-	-	-	-
MRSA (methicillin-resistant Staphylococcus aureus)	10	-	6	1	-	-	-	3	-	-	-
VRE (Vancomycin-resistant enterococci)	1	-	-	-	-	-	1	1	-	-	-
Resistant Klebsiella pneumoniae	1	-	1	-	-	-	-	-	-	-	-
Resistant Campylobacter	1	-	-	-	-	-	1	-	-	-	-
Resistant E. coli**	12	1	3	2	-	-	-	2	1	-	1
Resistance in commensal bacteria	4	1	3	-	-	-	-	-	-	-	-
Resistance in foodborne pathogens	1	-	-	-	-	-	-	-	1	-	-
Resistant bacteria (general)	10	1	3	1	-	1	1	2	-	-	1
Resistant fungi (general)	2	-	-	-	2	-	-	-	-	-	-
Resistant parasites (general)	1	-	-	-	1	-	-	1	-	-	-
Not defined (multiple)	3	1	1	-	-	-	-	1	-	-	-
TOTAL	52	5	19	6	3	1	4	10	2	-	2
					-
*TOTAL*	106	5	28	20	4	3	10	13	18	1	4

^*To find more details for each model that falls under the given category please refer to the complete database [28].

^**E. coli - Escherichia coli

^†Shiga toxin-producing E. coli includes: Shiga toxin-producing E. coli (STEC), Verocytotoxin-producing E. coli (VTEC)

Table 4: Distribution of the 106 articles* referring to models of disease transmission of a single system (n=102) and models of disease transmission with an economic component (n=4) according to study characteristics including the main population of interest, model type and specific model features, and the type of model system (sensitive microbes, resistant microbes, and antimicrobials (AMs**).

	Total number of articles	Sensitive microbes	Resistant microbes	AM^ only**
Main population of interest
Human	28	9	19	-
Companion animal	-	-	-	-
Cattle	7	4	2	1
Pigs	3	1	2	-
Poultry	3	1	2	-
Sheep	1	1	-	-
Fish/Aquaculture	-	-	-	-
Crops	2	-	2	-
Wildlife	-	-	-	-
Unspecified host (bacterial level)	5	-	5	-
Environment	3	-	1	2
2 populations	43	28	13	2
>3 populations	11	5	6	-

Model class
Compartmental	73	31	39	3
Agent-based	9	3	6	-
Network	8	5	3	-
Risk analysis	11	9	2	-
Combination/multiple	3	1	2	-
Other	2	-	-	2

Model type
Deterministic	53	21	27	5
Stochastic	43	23	20	-
Both	6	1	5	-
Hybrid (e.g. semi-stochastic)	4	3	1	-


Specific Model features^†
Multi-strain/co-infections (e.g. competition and gene transfer)	30	2	28	1
AMR emergence from AMU	23	-	23	1
Spatial (e.g. patch models, position or GPS data)	17	15	2	-
Multi-level model (e.g. within host and population transmission)	8	4	4	-
Vector-borne (e.g. healthcare workers)	11	3	8	-
Seasonal factors	5	4	-	1
Super-shedders/shedding rate	9	8	1	-
Pathogen survival	11	6	5	-
Vertical and pseudo-vertical transmission	3	2	1	-
Importation/Migration	36	22	14	-
AM effect on gut microbiome	6	-	6	-
AM residue levels and decay	7	-	2	5
Cost-benefit/cost analysis	4	3	1	-
Participatory modelling	1	-	1	-

Model parameters
Referenced data	53	28	23	2
Primary data	12	4	8	-
Both	7	2	2	3
Reference/Theoretical model	34	16	18	-

Publicly available dataset
Yes	7	1	4	2
No	99	48	48	3

Type of data
Quantitative	103	47	51	5
Qualitative	-	-	-	-
Mixed-methods	3	2	1	-

Model rigor
Calibration	52	26	23	3
Validation	56	30	21	5
Sensitivity analysis	56	24	31	1

^*To find more details for each model that falls under the given category please refer to the complete database [28].

^**AM – Antimicrobial

^†The subsection "Specific model features” provides details about specific components that were addressed within the models, in which a model could address multiple components or none of the components, therefore this subsection does not equate to the total number of models analyzed (n=106).

Table 5: Distribution of the 414 sources of data relating to the nodes (n=64) according to the study characteristics including the number of sources, number of data points, type of data, and regions and years covered within the data.

Node	Number of Sources	Number of data points	Qualitative data	Quantitative data	Region(s) covered	Year(s) covered
*(Terrestrial) On-farm AM^ use**	3	31	1	30	Denmark, Sweden, EU/EEA and Switzerland	2000-2018
*Access to AMs^ outside of the system**	5	32	16	16	Global, EU, Europe, Northern Europe, Southern Europe, Eastern Europe, Sweden, Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Georgia, Kazakhstan, Kyrgyzstan, Montenegro, North Macedonia, Republic of Moldova, Russian Federation, Serbia, Tajikistan, Turkey, Ukraine, Uzbekistan, Kosovo, USA	2002-2019
*AM^ use in companion animals**	5	20	6	14	Europe, EU/EEA and Switzerland, Sweden, Italy	2000-2017
*AM^ use in plant agriculture**	5	18	7	11	Global, EU, Sweden, The Netherlands, USA	2007-2017
Amount of imported product	16	234	8	226	Global, Europe, Sweden, Norway, Thailand	1990-2019
Amount of product in the domestic market	3	31	3	28	Sweden	2000-2018
Animal density	3	7	2	5	Sweden	1980-2005
Animal welfare/stress	8	58	46	12	Sweden	2001-2020
Aquaculture AM use	2	4	1	3	Global, EU/EEA and Switzerland, Sweden, Norway	2008-2017
AROs^† in companion animals	7	54	5	50	Europe, Austria, Belgium, Denmark, France, Germany, Greece, Italy, the Netherlands, Portugal, Serbia, Spain, Sweden, Switzerland, United Kingdom	2005-2018
AROs^† in food products	5	35	2	33	Sweden, Switzerland, UK	2004-2018
AROs^† in food-producing animals	9	84	6	78	EU, Belgium, Denmark, Spain, Sweden, Switzerland, The Netherlands, UK	2007-2019
AROs^† in humans	4	7	1	6	Sweden	2014-2018
AROs^† in imported food products	1	1	0	1	Sweden	2004
AROs^† in wildlife	1	3	0	3	Sweden	2009-2019
Chronic, non-communicable diseases	3	14	0	14	EU, Poland, Sweden	1986-2017
Companion animal illness	5	22	3	19	Europe, Austria, Belgium, Denmark, France, Germany, Greece, Italy, the Netherlands, Portugal, Serbia, Spain, Sweden, Switzerland, United Kingdom	1999-2020
Consumer choice, demand, and behaviour	57	365	289	76	Global, Europe, EU, North America, Nordic countries, Austria, Belgium, China, Denmark, England, Finland, France, Germany, India, Italy, Leitchenstein, Spain, Sweden, Switzerland, Thailand, The Netherlands, UK, USA	1960-2020
Consumption of other (non-meat/egg) foods	20	312	11	301	Europe, Finland, France, Iceland, Norway, Russia, Spain, Sweden, UK	1960-2018
Corporate profits from AM^*	12	61	12	49	Global, Europe, EU/EEA and Switzerland, US	1980-2018
Death (Human)	16	60	6	54	Global, Europe, EU, Poland, Sweden	1990-2019
Development of alternatives to AM^*	14	97	82	15	Global, Europe, EU, Australia, Brazil, Canada, Denmark, Japan, Singapore, The Netherlands, UK, USA	1997-2020
Development of new AMs^*	35	175	84	91	Global, Europe, EU, Canada, England, India, Scotland, UK, USA	1911-2020
Diagnostics	8	30	18	12	Global, Europe, Africa, Australia, Canada, Norway, Sweden, UK, USA	2003-2018
Digital health	20	77	18	59	Global, Europe, EU/EEA and Switzerland, Austria, Belgium, Denmark, Finland, Estonia, Ireland, Sweden, Denmark, Portugal, Spain, Germany, UK, France, Italy, Russia, Poland, Switzerland, The Netherlands, UK, USA	2002-2023
Disease in plant agriculture (crops, horticulture)	2	9	0	9	Sweden	2006-2011
Domestic and international trade	12	184	18	166	Global, EU, Sweden	1990-2019
Feed efficiency	1	1	0	1	Global	2019
Feed quality	1	1	1	0	Sweden	2020
Food and water security (personal, national)	4	27	0	27	Global, EU, Sweden	2000-2019
Food-producing animal illness	11	44	1	43	Denmark, Norway, Sweden, UK	1998-2020
Healthcare costs	12	53	2	51	Global, Europe, EU, Canada, China, Germany, Malaysia, Norway, Sweden, Switzerland, Serbia, Thailand, USA	2000-2050
Healthcare resources	16	314	8	306	Global, Europe, Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, The Netherlands, Portugal, Spain, Sweden, United Kingdom	1999-2019
*Human AM^ use**	17	51	3	48	Global, Europe, Southern Europe, Northern Europe, Eastern Europe, Central Europe, Austria, Belgium, Denmark, England, Finland, France, Germany, Greece, Italy, The Netherlands, Portugal, Spain, Sweden, United Kingdom	1994-2020
Human illness	15	106	3	103	Global, High-Income countries, Europe, Nordic region, Albania, Bosnia, Bulgaria, Croatia, Czech Republic, Hungary, North Macedonia, Montenegro, Poland, Romania, Serbia, Slovakia, Slovenia, Denmark, Finland, Iceland, Norway, Sweden, Greenland, Poland, Spain, The Netherlands, UK	1970-2018
Human vaccination	8	41	5	36	Europe, EU, EU/EEA, Canada, Italy, Sweden	2000-2019
Market price per production unit	5	187	1	186	EU, Sweden	1980-2018
Meat/egg consumption	30	296	20	276	Global, Europe, EU, Denmark, Finland, France, Germany, Norway, Italy, Poland, Russia, Spain, Sweden, UK, USA	1960-2020
Movement of animals	15	95	16	79	Europe, EU, Sweden	1999-2020
Movement of people	10	127	18	109	Europe, EU, Austria, France, Germany, Italy, Sweden, Poland, Hungary	2000-2018
New and emerging foods	36	335	60	275	Global, Asia, Africa, Europe, North America, Australia, Belgium, Brazil, Canada, Czech Republic, The Netherlands, France, United Kingdom, Denmark, Sweden, Norway, England, Italy, Poland, Portugal, Slovakia, Spain, Switzerland, United States, UK	1965-2020
*Non-AM^ disease prevention and infection control in health and social care settings**	16	140	26	114	Global, EU, Northern Europe, Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, The Netherlands, Norway, Portugal, Spain, Sweden, United Kingdom,	2002-2020
*Non-AM^ infection control in food-producing animal agriculture**	1	1	1	0	EU	2016
Number of units set by quota	1	3	0	3	Sweden	2002
Nutritional quality of diet	2	8	3	5	Sweden	2002-2020
On-farm production level	17	341	15	326	Denmark, Norway, Sweden	1984-2020
Pharmaceutical market, sales, and PR^‡	6	63	7	56	Global, Europe, EU, Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Georgia, Kazakhstan, Kyrgyzstan, Montenegro, North Macedonia, Republic of Moldova, Russian Federation, Serbia, Tajikistan, Turkey, Ukraine, Uzbekistan, Kosovo, Canada, France, Ireland, Sweden, UK	1995-2018
Population vulnerabilities	12	156	22	134	Global, OECD countries, Europe, EU, Nordic countries, Austria, Bosnia, Belgium, Romania, Bulgaria, Croatia, Czech Republic, Denmark, England, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Lithuania, Norway, Poland, Portugal, Romania, Russia, Scotland, Slovania, Spain, Sweden The Netherlands, United States, Wales, Switzerland	1992-2019
Prescribing, diagnosing, treatment practices (appropriateness)	35	186	34	152	Global, Europe, EU**, Austria, Belgium, Canada, Croatia, Denmark, England, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Lativia, Lithuania, Norway, The Netherlands, Spain, Sweden, United Kingdom, USA	1995-2019
Production costs	1	3	0	3	Sweden	2018
Production systems	28	413	48	365	Global, Europe, EU, Austria, Bosnia, Canada, Czech Republic, Denmark, Estonia, Finland, France, Germany, Latvia, Lithuania, Italy, Malta, Macedonia, Norway, Romania, Russia, Spain, Sweden, Switzerland, Thailand, UK, USA	1991-2030
Psychological health	8	60	0	60	Europe, EU, Belgium, Bulgaria, Czechia, Cyprus, Finland, France, Germany, Iceland, Italy, Lithuania, Luxemburg, Netherlands, Norway, Portugal, Romania, Spain, Sweden, Switzerland, The Netherlands, Turkey	2000-2018
Resistance in the wider environment	1	1	1	0	Global	2018
Retail cost of food	16	245	16	229	Europe, EU, Austria, Czech Republic, Belgium, Bulgaria, Denmark, Finland, France, Hungary, Iceland, Ireland, Lithuatia, Luxemburg, Monico, Norway, Poland, Romania, Sweden, Switzerland	1999-2020
Retailer demand for product	8	9	7	2	Europe, Sweden	2005-2019
Treatment of waste and waste-water	2	12	2	10	Global, EU, Bangladesh, Ghana, India	2000-2018
Understanding and awareness	2	11	1	10	Global, EU, Sweden, United states	2010
Unregulated meat sales	1	1	0	1	Sweden	2002
Use for prevention in humans	3	46	0	46	EU/EEA, Sweden	2002-2012
Use for preventive purposes	1	2	0	2	Italy	2000-2007
Use for treatment	1	1	0	1	Denmark	1995-2008
Use for treatment in humans	1	15	0	0	Sweden	2003-2010
Viability of domestic meat production	3	10	5	5	Europe, Sweden	2002-2019

*AM: Antimicrobial

**EU/EEA: European Union/ European Economic Area

^†ARO: Antimicrobial resistant organism

^‡PR: Public relations

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Mapping out a One Health model of antimicrobial resistance in the context of the Swedish food system: A literature scan

Status:

Version 1

Abstract

Figures

1.1. Background

1.2. Methods

1.2.1 – Objective 1: Existing Models

1.2.2 – Objective 2: Existing Data Sources and Evidence

1.3. Results

1.3.1 – Objective 1: Existing Models

1.3.2 – Objective 2: Existing Data Sources and Evidence

1.4. Discussion

1.4.1 – Literature scan of models of, and factors associated with AMR

1.4.1.1 – Disease transmission models embedded within the food system

1.4.1.1.1 – AMR-specific disease transmission models

1.4.1.2 – Economic models embedded within the food system

1.4.1.3 – The evidence landscape of factors that drive AMR

1.4.2 – Limitations

1.5. Conclusion And Recommendations For Future Research

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1