Chronic wasting disease (CWD) is an infectious and fatal prion disease occurring in the family Cervidae. To update the research community regarding the status quo of CWD epidemic models, we conducted a meta-analysis on CWD research. We collected data from peer-reviewed articles published since 1980, when CWD was first diagnosed, until December 2018. We explored the analytical methods used historically to understand CWD. We used 14 standardized variables to assess overall analytical approaches of CWD research communities, data used, and the modeling methods used. We found that CWD modeling initiated in the early 2000s and has increased since then. Connectivity of the research community was heavily reliant on a cluster of CWD researchers. Studies focused primarily on regression and compartment-based models, population-level approaches, and host species of game management concern. Similarly, CWD research focused on single populations, species, and locations, neglecting modeling using community ecology and biogeographic approaches. Chronic wasting disease detection relied on classic diagnostic methods with limited sensitivity for most stages of infection. Overall, we found that past modeling efforts generated a solid baseline for understanding CWD in wildlife and increased our knowledge on infectious prion ecology. Future analytical efforts should consider more sensitive diagnostic methods to quantify uncertainty and broader scale studies to elucidate CWD transmission beyond population-level approaches. Considering that infectious prions may not follow biological rules of well-known wildlife pathogens (i.e., viruses, bacteria, fungi), assumptions used when modeling other infectious disease may not apply for CWD. Chronic wasting disease is a new challenge in wildlife epidemiology.

Chronic wasting disease (CWD) is an invariably fatal neurologic disorder caused by a misfolded protein called a prion (Prusiner 1982). It has recently received considerable attention from the public and scientific community. Chronic wasting disease is a transmissible spongiform encephalopathy (TSE; Williams and Young 1980). Cervids experience cognitive decline and neurodegeneration as a result of infectious prion (PrPCWD) replication in neurons. Chronic wasting disease was first observed in 1967 in a captive deer facility in Colorado, US, classified as a TSE in 1979 (Williams and Young 1980), and discovered in wild herds in 1981 (Spraker et al. 1997). By late 2019, the disease was detected in wild cervid populations in 24 states (US), two Canadian provinces, and areas of Norway, Sweden (Statens Veterinärmedicinska Anstalt 2019), and Finland, as well as in captive facilities in 17 states (US), four Canadian provinces, and two South Korean provinces (Kim et al. 2005; Dubé et al. 2006; USGS 2019). Naturally susceptible cervid species to date include white-tailed deer (Odocoileus virginianus), mule deer (Odocoileus hemionus), elk (Cervus canadensis), red deer (Cervus elaphus), caribou (Rangifer tarandus), moose (Alces alces), black-tailed deer (Odocoileus hemionus hemionus), and sika deer (Cervus nippon; Spraker et al. 1997; Baeten et al. 2007; Benestad et al. 2016; Centers for Disease Control and Prevention [CDC] 2019a). Because of its highly contagious nature, CWD prevalence can reach 40–50% in wild cervid populations (Edmunds et al. 2016; Carlson et al. 2018) and 80–100% in captive populations (Keane et al. 2008). These epidemiologic characteristics make CWD management a priority for wildlife managers and conservationists responsible for species of economic and conservation concern (WDNR 2010; Texas Parks and Wildlife and Texas Animal Health Commission 2015).

Despite traditionally concerning wildlife professionals, new emphasis on CWD's spillover potential has garnered attention from the CDC (2019b) and the European Food Safety Authority (EFSA Panel on Biological Hazards 2018). Experimental challenge via intracranial prion inoculation demonstrates other wildlife, such as raccoons (Procyon lotor; Moore et al. 2019), common livestock species including sheep (Ovis aries) and cattle (Bos taurus; Hamir et al. 2006, 2007), and nonhuman primates (Marsh et al. 2005) can be susceptible to prion infection. Similarly, swine (Sus scrofa) have shown susceptibility to PrPCWD after consuming prion-infected tissues (Moore et al. 2017). Although there are no reports of PrPCWD infecting humans, public health experts discourage consuming CWD-positive cervids (CDC 2019b) and posit that CWD's ever-increasing spread and exposure warrant action (Osterholm et al. 2019).

To determine which measures of surveillance, control, and prevention for infectious diseases are appropriate, epidemiologists work to comprehend the phenomena and mechanisms that trigger and facilitate disease spread. Since the mid-1700s, epidemiologists have used statistical and mathematical models for describing epidemiologic data and complex infection processes (Heesterbeek et al. 2015). Models are often valued for their abilities to simplify complex biologic systems, describing and forecasting infectious disease events, and evaluating control methods under diverse, what-if scenarios (Garner and Hamilton 2011). A plethora of studies have examined diverse epidemiologic modeling approaches for infectious diseases caused by viruses (Gambhir et al. 2015; Herzog et al. 2017), protozoa (Wallace et al. 2014), fungi (Van Maanen and Xu 2003), and bacteria and nematodes (Hollingsworth et al. 2015; Lou and Wu 2017).

Like other wildlife diseases, research on the ecology of CWD has relied on modeling of epidemics. For example, models have found associations between deer demographics and CWD infection, typically with highest CWD prevalences in older males followed by older females and then yearling males (Grear et al. 2006; Heisey et al. 2010), leading to the development and testing of demographically weighted harvesting systems for disease control (Walsh and Miller 2010; Jennelle et al. 2018). Other studies have expanded knowledge on CWD transmission mechanisms, prion dynamics, and host population dynamics (Mejía-Salazar et al. 2016; Samuel and Storm 2016). Recent landscape epidemiologic studies have identified factors that explain roles of the landscape in CWD prevalence (Walter et al. 2011a; O'Hara Ruiz et al. 2013; Edmunds et al. 2018). Similarly, physical landscape features such as geomorphology (Mateus-Pinilla et al. 2013) and rivers and roads (Robinson et al. 2013) have shown linkages with CWD transmission (Rees et al. 2012) and modify the shape of CWD epidemics (Robinson et al. 2013).

Identifying trends in research is commonly used to identify promising approaches and gaps of knowledge and to guide additional efforts in epidemiology (Allen et al. 2012; Heesterbeek et al. 2015; McCallum 2016; Herzog et al. 2017). Chronic wasting disease is of interest for wildlife and veterinary professionals, livestock industries, and public health agencies and has therefore been the subject of previous assessment of its trends (Schauber and Woolf 2003; Conner et al. 2008). A recent assessment of 16 articles found that models support the role of management interventions (i.e., selective and nonselective culling, seasonal hunting, and vaccination) and identify uncertainty in models (Uehlinger et al. 2016). We used a broad definition of modeling in the epidemiologic sense (statistical or mathematical). Our aim was to explore past trends in CWD epidemiology. We identified analytical approaches, diagnostic methods applied, collaboration structure among researchers, model parameterizations, and gaps of information.

Search and screening approach

We collected articles from Web of Science (Clarivate Analytics, Philadelphia, Pennsylvania, USA) in January 2019. Keywords included chronic wasting disease, prion, model*, landscape, and spatial, combined to capture articles published from January 1980 to December 2018, encompassing >35 yr of CWD research since its formal recognition in Colorado (Williams and Young 1980). We conducted an initial screening of titles and abstracts to retain online peer-reviewed research manuscripts (i.e., nonliterature reviews) related to statistical or mathematical modeling. We applied the following selection criteria: 1) articles written in English language, 2) models applied were statistical or mathematical (i.e., not animal models), 3) articles that were not clinical or pathogenetic, 4) CWD infection accounted for in models (i.e., not loosely implied), and 5) model approaches accounted for cervid ecology (i.e., CWD reservoirs explicitly considered in models). Next, the bibliographies of articles were inspected manually to identify articles not detected in our initial search and falling within our inclusion criteria.

Data collection

Articles were reviewed, and data were extracted and assembled in four major groups: article title, publication year, journal name, and authors, to address what, when, where, and by whom, respectively, articles were published. In addition, we extracted epidemiologic data from articles by using a content analysis (Hsieh and Shannon 2005) considering different research approaches. Specifically, research approaches were defined based on biologic organizational levels as 1) individual-level studies (i.e., focused on individual or cohort pathogenesis/survival); 2) population-level studies (i.e., one or more cervid populations defined by the article were the dependent variables of the modeling application); 3) community-level studies (i.e., models integrating multiple species from diverse taxa); 4) ecosystem-level studies (i.e., models integrating environmental features and epidemiologic data); and 5) biogeographic-level studies (i.e., coarse-scale, broad-extent studies). We also collected the geographic location, host species involved in the modeling, data source (i.e., primary or secondary), sample size and CWD prevalences, and diagnostic methods used to detect CWD infection. We also categorized between studies based on the use of empirical or simulated data (i.e., virtually created populations and/or environments). Finally, we identified the modeling algorithms used, model evaluation methods, type of modeling (predictive vs. descriptive), and variables assessed. Variables included in the modeling were characterized in 11 categories: 1) control/management method (i.e., exploration of methods for management control, such as harvest); 2) demographic (i.e., population-centric variables); 3) epidemiologic (i.e., characteristics of pathogens or hosts); 4) landscape (e.g., land cover types); 5) life cycle (e.g., functions of population viability); 6) location; 7) sampling method (e.g., route of data collection); 8) time; 9) sampling effort; 10) trophic-related variables; and 11) spatial, following Auchincloss et al. (2012).

Data analyses

We organized, summarized, and visualized data with R software (R Core Team 2019) by using ggplot2 and dplyr packages in the tidyverse platform (Wickham 2018). In addition, we used ArcMap 10.5 (Environmental Systems Research Institute, Redlands, California, USA) for choropleth map generation to show geographic distribution of studies by state. We used a social network analysis to describe the structure of the CWD modeling community (Newman 2004). We compiled an adjacency matrix containing the number of selected publications written by and between authors. We extracted each authors' affiliations listed in the articles and categorized affiliations as 1) state wildlife agencies, 2) academia, 3) federal science agencies, and 4) other governmental agencies (e.g., city government). Finally, we used Gephi 0.9.2 network analysis software (Bastian et al. 2009) to quantify author influence in connectivity and affiliation-based structures of the research community.

The search strings yielded a total of 679 articles. After removing duplicates, 589 unique articles remained. Following our selection criteria, 79 research articles were found, including 8 additional articles not captured by the search on Web of Science but recovered from articles' bibliography (Joly et al. 2003; Grear et al. 2006; Johns and Mehl 2006; Miller et al. 2008; Al-Arydah et al. 2012; Edmunds et al. 2016; Galloway et al. 2017; Schuler et al. 2018). The 79 articles and their corresponding metadata can be found in the Supplementary Material (Table S1).

The number of articles on CWD modeling has steadily increased since 2000 (Fig. 1), with a mean of 4.2 (SD=2.2) publications being published per year. Miller et al. (2000) and Conner et al. (2000) published the first research articles applying analytical modeling to CWD. The first article integrating spatial statistics across scales was published in 2003 (Joly et al. 2003). Ecologic modeling including environmental covariates started in 2005 (Farnsworth et al. 2005; Krumm et al. 2005). Integrating genetics in CWD epidemic modeling started in 2008 (Miller et al. 2008).

Figure 1

Selected chronic wasting disease modeling studies published from 2000 to 2018 (=4.2 publications/yr, SD=2.2). Modeling studies started in 2000. Bar plots (gray) show number of publications (left) by year. The solid line represents cumulative number of articles (right) across years.

Figure 1

Selected chronic wasting disease modeling studies published from 2000 to 2018 (=4.2 publications/yr, SD=2.2). Modeling studies started in 2000. Bar plots (gray) show number of publications (left) by year. The solid line represents cumulative number of articles (right) across years.

Close modal

In total, 37 journals contained the articles collected. Of these, the Journal of Wildlife Diseases, PLoS One, Journal of Wildlife Management, and Ecological Applications contained about half of the articles (see the Supplementary Figure). Articles were published mainly in journals related to ecology (e.g., Ecology, Journal of Applied Ecology, Ecosphere) and biomathematics (e.g., Journal of Mathematical Biology, Bulletin of Mathematical Biology, ISRN Biomathematics), with a limited presence in veterinary journals (e.g., Preventative Veterinary Medicine, Veterinaria Italiana). In total, 180 individual authors participated in the 79 research articles. The social network of the CWD community revealed that a few specific researchers (nodes) occurred in most modeling studies (Fig. 2). In addition, individual authors' level of influence in overall network connectivity (eigenvector centrality) was unevenly distributed among the CWD-modeling community, with a small number being the most influential in connectivity. State and federal agencies (e.g., natural resource departments and US Geological Survey) comprised a considerable number of connections with other researchers (Fig. 3A) and with academic institutions (Fig. 3B). Approximately 18% articles were generated by isolated groups of authors (i.e., mathematicians), whereas authors from other disciplines were generally well connected to the major network. Only one researcher was not affiliated with a “science, technology, engineering, and mathematics” department.

Figure 2

Collaboration network of chronic wasting disease modeling research. Nodes (circles) represent authors having between 1 (small circle) and 19 publications (largest circle). Edge (connecting lines) thickness represents magnitude of collaboration between authors in terms of shared publications and vary from low (thin lines) to high (thick lines). The influence of authors for connectivity (eigenvector centrality) is denoted as showing in decreasing densities of shades from high (dark gray) to low influence (light gray). Note that a few authors have been central to connect chronic wasting disease research in the community (large circles) and a few isolated clusters reflect research conducted independently. Inset text: High-influence authors in the network.

Figure 2

Collaboration network of chronic wasting disease modeling research. Nodes (circles) represent authors having between 1 (small circle) and 19 publications (largest circle). Edge (connecting lines) thickness represents magnitude of collaboration between authors in terms of shared publications and vary from low (thin lines) to high (thick lines). The influence of authors for connectivity (eigenvector centrality) is denoted as showing in decreasing densities of shades from high (dark gray) to low influence (light gray). Note that a few authors have been central to connect chronic wasting disease research in the community (large circles) and a few isolated clusters reflect research conducted independently. Inset text: High-influence authors in the network.

Close modal
Figure 3

Network of chronic wasting disease researchers' affiliations. Network models showing individual authors and their number of publications (node size), whereas edge (connecting line) thickness denotes strength of collaboration via number of papers written between authors. Author affiliations recorded from authors' publications. A) Collaboration among state wildlife agencies (yellow), federal science agencies (blue), other government (gray), and academia (red). Note the apparent role of state and federal science agencies in connecting academia. B) Academia-affiliated authors categorized as biologically and ecologically based departments (green), veterinary and animal science (orange), epidemiology and health (pink), mathematics and mathematical biology (yellow), statistics and biostatistics (blue), biophysics (turquoise), soils (brown), and business (red). Note the strong connection (large nodes) between ecology (green) and veterinary fields (orange).

Figure 3

Network of chronic wasting disease researchers' affiliations. Network models showing individual authors and their number of publications (node size), whereas edge (connecting line) thickness denotes strength of collaboration via number of papers written between authors. Author affiliations recorded from authors' publications. A) Collaboration among state wildlife agencies (yellow), federal science agencies (blue), other government (gray), and academia (red). Note the apparent role of state and federal science agencies in connecting academia. B) Academia-affiliated authors categorized as biologically and ecologically based departments (green), veterinary and animal science (orange), epidemiology and health (pink), mathematics and mathematical biology (yellow), statistics and biostatistics (blue), biophysics (turquoise), soils (brown), and business (red). Note the strong connection (large nodes) between ecology (green) and veterinary fields (orange).

Close modal

The research approaches used were not evenly distributed across the pool of articles. For example, two articles were performed at the individual level, in studies of cohort survival and transmission (Monello et al. 2017; Davenport et al. 2018). Population-level studies were the most common scale studied (n=48), followed by ecosystem-level studies (n=20). The number of population-level articles remained relatively constant over time, but ecosystem-level analyses became more frequent during the 2010s. Six community-level articles explored predator populations directly or indirectly through predation-associated mortality on cervids and infection status (Miller et al. 2008; Walsh and Miller 2010; Wild et al. 2011; Monello et al. 2014; DeVivo et al. 2017; Maji et al. 2018). Nine articles mixed methods by including cervid genetics, spatial analyses, and landscape-level variables (Blanchong et al. 2008; Miller et al. 2008; Grear et al. 2010; Cullingham et al. 2011a, b; Rogers et al. 2011; Robinson et al. 2013; Kelly et al. 2014; Mejía-Salazar et al. 2017). We did not identify studies using biogeographic level approaches or studies across large study areas.

Thirteen studies used artificial, simulated environments for modeling (Gross and Miller 2001; Diefenbach et al. 2004; Johns and Mehl 2006; Nusser et al. 2008; Almberg et al. 2011; Wild et al. 2011; Al-Arydah et al. 2012; Potapov et al. 2012; Cortez and Weitz 2013; Oraby et al. 2014; Sun et al. 2015; Vasilyeva et al. 2015; Maji et al. 2018). Excluding aspatial simulations, all modeling studies were conducted in the US and Canada (Fig. 4). Wisconsin (n=25) and Colorado (n=21) were the most represented states in the literature, followed by neighboring states Wyoming (n=10) and Illinois (n=7). In Canada, provinces with most articles were Saskatchewan (n=7) and Alberta (n=6). Similarly, cervid species used in the pool of research articles were unevenly represented. Articles modeled white-tailed deer (n=39) and mule deer (n=37) most frequently (88%), followed by elk, which were rarely the target species (n=5). Five studies did not report the host species studied (Wild et al. 2011; Oraby et al. 2014; Sun et al. 2015; Vasilyeva et al. 2015; Maji et al. 2018). We did not find studies focused on other known CWD-susceptible host species (i.e., moose, sika deer, red deer, or caribou).

Figure 4

Spatial distribution of chronic wasting disease modeling studies by administrative area. In the US, studies were conducted in Wisconsin, Illinois, Colorado, Wyoming, Utah, Maryland, Virginia, Pennsylvania, West Virginia, North Dakota, and South Dakota. In Canada, studies occurred in the provinces of Saskatchewan, Alberta, and British Columbia. No modeling studies were conducted in other states, provinces, or countries (white). Research articles based on artificial, simulated data were excluded (n=13), and we retained the remaining articles (n=66).

Figure 4

Spatial distribution of chronic wasting disease modeling studies by administrative area. In the US, studies were conducted in Wisconsin, Illinois, Colorado, Wyoming, Utah, Maryland, Virginia, Pennsylvania, West Virginia, North Dakota, and South Dakota. In Canada, studies occurred in the provinces of Saskatchewan, Alberta, and British Columbia. No modeling studies were conducted in other states, provinces, or countries (white). Research articles based on artificial, simulated data were excluded (n=13), and we retained the remaining articles (n=66).

Close modal

Most articles (n=54) relied on secondary data sources for modeling (i.e., published data collected from surveillance or literature), whereas primary-sourced data (original field or experimental data) studies were less common (n=25). Excluding simulation-based studies with artificial populations, articles using secondary data sources (e.g., statewide surveillance programs) had sample sizes that accounted for >96% of cervids used in models. These secondary sources typically revealed higher CWD prevalences (µ=11.7%; range=0–94.7%) and larger sample sizes (µ=12,405; range=39–152,133) across articles. Reported prevalences and number of cervids sampled in articles at the year of their publication was highly variable (Fig. 5).

Figure 5

Reported prevalences and sampling sizes of chronic wasting disease modeling studies by year for published articles. The cumulative sample sizes (×103; black line) and reported prevalence in percent (gray boxes). Note the variability of prevalence values among studies. No published articles reported prevalences in 2001 and 2003. Studies using simulated data were omitted.

Figure 5

Reported prevalences and sampling sizes of chronic wasting disease modeling studies by year for published articles. The cumulative sample sizes (×103; black line) and reported prevalence in percent (gray boxes). Note the variability of prevalence values among studies. No published articles reported prevalences in 2001 and 2003. Studies using simulated data were omitted.

Close modal

Regarding CWD prion diagnostic or detection, nearly all studies (72/79) relied on immunohistochemical (IHC) and/or enzyme-linked immunosorbent assay (ELISA). These methods were used directly or indirectly for determining CWD infection status in cervids. Five studies did not report diagnostic methods used or assumed for CWD detection (Diefenbach et al. 2004; Almberg et al. 2011; Cullingham et al. 2011a; Galloway et al. 2017; Maji et al. 2018). Only one study (Davenport et al. 2018) incorporated real-time quaking induced conversion (RT-QuIC), and no studies reported use of protein-misfolding cyclic amplification (PMCA).

Analytical methods

Phenomenologic models, such as regression analyses, and mechanistic models, such as compartmental models using differential equations (i.e., susceptible, infected, recovered), were the most common analytical approaches, followed by hierarchical Bayesian models, population matrix models, and descriptive statistics (Fig. 6). Less common methods included diffusion models, machine learning (i.e., boosted regression trees and Maxent), and network models, among others (Table S2). In addition, 25 of 79 (32%) studies were descriptive, aiming to reconstruct past epidemics, whereas 46 of 79 (58%) relied on modeling algorithms that were predictive in nature, aiming to forecast unknown CWD scenarios (Table S1).

Figure 6

Modeling methods used in chronic wasting disease research. Regressions (e.g., logistic, linear, generalized linear mixed-effects model, negative binomial) were the most commonly used models, followed by compartment models using differential equations (e.g., susceptible, infected, recovered models), hierarchical Bayesian, matrix population models, and descriptive statistics, respectively. Less common methods were omitted from the figure and are described in Supplementary Material Table S2.

Figure 6

Modeling methods used in chronic wasting disease research. Regressions (e.g., logistic, linear, generalized linear mixed-effects model, negative binomial) were the most commonly used models, followed by compartment models using differential equations (e.g., susceptible, infected, recovered models), hierarchical Bayesian, matrix population models, and descriptive statistics, respectively. Less common methods were omitted from the figure and are described in Supplementary Material Table S2.

Close modal

From the 79 modeling articles, 49 articles relied on time explicitly (e.g., prevalence over time) or implicitly (e.g., in compartment model timesteps), and 40 articles incorporated spatial information (e.g., Game Management Unit, geographic coordinates, or Township-Region-Section; Conner and Miller 2004; Krumm et al. 2005; Kelly et al. 2014, respectively). Four studies explored sampling effort in CWD research (Joly et al. 2009; Walsh and Miller 2010; Rees et al. 2012; Mateus-Pinilla et al. 2013), and six studies examined trophic variables as top-down effects of predation on cervids and infection status (Miller et al. 2008; Walsh and Miller 2010; Wild et al. 2011; Monello et al. 2014; DeVivo et al. 2017; Maji et al. 2018). No studies investigated bottom-up effects, and a few explored cervid body condition (Edmunds et al. 2016; DeVivo et al. 2017) and consumption rates (Potapov et al. 2013) as functions of movement habits and prion deposition, respectively. Common landscape variables of ecosystem-level studies included characteristics of forest composition, agricultural habitat, and soils, whereas wetlands and riparian habitats were rarely incorporated (Table 1). Demographic variables including sex and age of cervids, and epidemiologic variables such as prevalence were the most frequent in CWD modeling (Fig. 7).

Table 1

Landscape variables used in chronic wasting diseases research published in journals from 1980 to 2018. From ecosystem-level studies in meta-analysis, columns show categorized landscape variables, how often they appeared in articles, an example of the variable or its description, and their respective sources. Variables related to forests, urbanization, and agriculture were more common, whereas riparian and broad-scale variables (e.g., ecoregions) were seldom examined. Note that climactic variables are absent.

Landscape variables used in chronic wasting diseases research published in journals from 1980 to 2018. From ecosystem-level studies in meta-analysis, columns show categorized landscape variables, how often they appeared in articles, an example of the variable or its description, and their respective sources. Variables related to forests, urbanization, and agriculture were more common, whereas riparian and broad-scale variables (e.g., ecoregions) were seldom examined. Note that climactic variables are absent.
Landscape variables used in chronic wasting diseases research published in journals from 1980 to 2018. From ecosystem-level studies in meta-analysis, columns show categorized landscape variables, how often they appeared in articles, an example of the variable or its description, and their respective sources. Variables related to forests, urbanization, and agriculture were more common, whereas riparian and broad-scale variables (e.g., ecoregions) were seldom examined. Note that climactic variables are absent.
Figure 7

Variables used for model parameterization in chronic wasting disease modeling research. Demographic variables (e.g., sex and age) and epidemiologic variables (e.g., prevalence and transmission rate) were frequently included in the models. Variables relating to sampling methods, effort, or control measures were less common. Landscape variables are presented in Table 1. SRS=simple random sampling.

Figure 7

Variables used for model parameterization in chronic wasting disease modeling research. Demographic variables (e.g., sex and age) and epidemiologic variables (e.g., prevalence and transmission rate) were frequently included in the models. Variables relating to sampling methods, effort, or control measures were less common. Landscape variables are presented in Table 1. SRS=simple random sampling.

Close modal

Using 79 research articles, our research revealed trends in epidemiologic modeling of CWD. We collected and standardized metadata by using methods comprehensible and accessible for both epidemiologists and wildlife professionals. We offer a synthesis of analytical modeling of CWD, the prion disease with the highest spillover potential (Escobar et al. 2019). The articles we outlined have generated valuable findings to guide current and future management actions and efforts from the research community remain critical in understanding this emerging infectious disease. We note the following patterns: 1) population-level studies were predominant; 2) models relied on diagnostic tests of limited sensitivity; 3) the research community is collaborative among professions and institutions; and 4) the data collected are geographically clustered, representing portions of CWD's distribution.

Our review indicated that the first modeling applications began >30 yr after the initial detection of CWD. This lag is likely a result of the timeline of the diagnosis of CWD as a TSE and identification in wild herds in the early 1980s (Spraker et al. 1997). Also, delays in publication are subject to lengths of surveillance periods, data cleaning and analysis, and other practical limitations. Still, in less than 2 decades, the magnitude of CWD modeling research has increased consistently along with the number of researchers.

The CWD modeling community is tightly linked with major researchers at federal and state agencies. In academia, veterinary scientists and ecologists possess strong ties expressed as co-authorship across academic realms. Supportively, publications in ecology and veterinary journals featured authors with larger collaboration networks, whereas isolated clusters of authors often published in specialized journals. The fact that the community of researchers was well connected, with a few isolated clusters, suggests strong collaboration among disciplines and agencies and multidisciplinarity in the study of CWD. A deficiency of social science researchers suggests little input from experts in human-dimension, economics, and policy, which could limit the use of CWD models for management.

Research approaches focused on population-level modeling that has revealed demographic patterns of CWD infection (Miller and Conner 2005; Grear et al. 2006) useful for testing control scenarios at the local level, including culling practices to reduce prevalence (Wasserberg et al. 2009; Potapov et al. 2012). Whether CWD transmission is frequency dependent or density dependent (or a hybrid of the two) is still in debate, and models show mixed results (Grear et al. 2010; Cortez and Weitz 2013; Storm et al. 2013; Jennelle et al. 2014; Oraby et al. 2014). We found that population-level models that incorporate spatial analysis revealed that CWD is not randomly distributed; rather, it is observed in geographic clusters (Joly et al. 2003, 2006). These clusters result in hotspots of infection that could serve as a source for posterior CWD spread to other regions (Nusser et al. 2008; Heisey et al. 2010). Population-level studies incorporating genetics identified heterogeneous transmission risk (Matsumoto et al. 2013), allelic selection (Monello et al. 2017), impacts of genetic relatedness on probability of transmission (Grear et al. 2010; Mejía-Salazar et al. 2017), and patterns of geographic spread (Blanchong et al. 2008; Cullingham et al. 2011a, b; Rogers et al. 2011; Robinson et al. 2013; Kelly et al. 2014).

Community-level studies often determined the roles of cervid predation in CWD spread. These studies were generally empirically based (Miller et al. 2008; Walsh and Miller 2010; Monello et al. 2014; DeVivo et al. 2017), and theoretical applications were less common (Wild et al. 2011; Maji et al. 2018). Ecosystem-level studies revealed the importance of landscape features for CWD spread (Garlick et al. 2011; Nobert et al. 2016; Hefley et al. 2017b). For example, landscape epidemiology identified positive associations between CWD prevalence and specific variables such as urban (Farnsworth et al. 2005), forested (O'Hara Ruiz et al. 2013; Storm et al. 2013), and riverine landscapes (Edmunds et al. 2018), whereas soil composition had limited effects (O'Hara Ruiz et al. 2013; Storm et al. 2013; Manjerovic et al. 2014) in contrast to Walter et al. (2011b). One individual-level study determined cohort pathogenesis and transmission (Davenport et al. 2018), whereas the other investigated survival in relation to genotype (Williams et al. 2014).

Most studies were conducted in regions severely afflicted by CWD (USGS 2019). For example, modeling studies were conducted in 11 of the 26 CWD-affected states and in two of the five countries in which CWD has been detected in wild cervids. This is probably a result of available funding in CWD endemic areas, low detection of prions outside foci of infection, and limited data available for modeling other regions. Most articles in our review used study areas selected pragmatically or based on political boundaries, a selection process that has been cautioned against, considering the lack of biologic support and artifactual results of such study designs (Barve et al. 2011). Interestingly, some study areas were defined through Game Management Units, delineations of which can vary in their accommodation to wildlife-population (Miller and Conner 2005), landscape, and human-associated variables, such as land cover types and infrastructure (WDNR 1998; Joly et al. 2009). Fortunately, despite traditional wildlife management operating within political boundaries, recent recommendations from wildlife managers are now promoting CWD management and monitoring across state and provincial boundaries (Western Association of Fish and Wildlife Agencies 2018).

Many susceptible cervid species (e.g., moose, caribou, red deer) were neglected in the revised modeling studies. The three species generally included in models (i.e., white-tailed deer, mule deer, and elk) are highly economically valued as game species (Koontz and Loomis 2005; WDNR 2010; Mule Deer Working Group 2015), suggesting an implicit economic bias in the species selection. Alternatively, the lack of modeling studies for moose, caribou, and red deer could reflect either lower prevalences detected in these species, or a deficiency of journal publications of the research conducted on these species.

Sample sizes of studies showed a tendency to increase, but CWD prevalences were considerably variable across time. This may suggest an increased sampling effort, but inconsistent sampling designs. Also, it is unclear whether prevalence estimations represented different stages of CWD epidemics. In addition, it is possible that variability of prevalence across time and studies is linked to inclusion (or not) of samples from passive surveillance (e.g., road-killed deer) that may have different prevalences than samples from active epidemiologic surveillance (e.g., hunting for sampling). The high variability in prevalences reveals the increase in research teams studying CWD prevalence in wildlife using nonstandardized methods for estimations, making studies hard to compare. For example, some studies rely on decade-long surveillance data and others on data across years. Unfortunately, efforts to elucidate correction of prevalence values by sampling bias as well as controlling for expanding geographic coverage has been limited. Indeed, geographic areas and temporal periods, which are critical units to estimate CWD prevalence, varied considerably among studies. We call for more informed, standardized reports of CWD prevalence that include the sample size (e.g., number of animals tested), sampling period (e.g., annual basis, duration since local discovery, or stage of epidemic), and study area extent (e.g., hectares). This will make metadata more reliable as a baseline to facilitate prevalence comparison across regions, periods, and populations.

Most CWD data originated from two diagnostic tests: IHC and ELISA. These tests are considered the gold standard for CWD diagnosis (Haley and Richt 2017). Routine procedure uses ELISA for initial screening, followed by IHC for confirmation. Strikingly, recent studies showed that IHC has questionable sensitivity when detecting low concentrations of prions in the asymptomatic phase of infection. For example, IHC could fail to detect PrPCWD-infected individuals in controlled, transgenic mice (McNulty et al. 2019). Similarly, ELISA fails to detect low concentrations of CWD prions in brain homogenate (McNulty et al. 2019), which may be consistent with earlier stages of the disease. Therefore, considering that diagnostic methods used in CWD modeling produce false negatives, previous epidemiologic models could be underestimating CWD prevalences. In the US, IHC and ELISA are the only accepted methods for CWD diagnosis officially (US Department of Agriculture 2014). Thus, state agencies in the US are restricted from using other non–US Department of Agriculture–validated tests (i.e., ultrasensitive methods such as PMCA or RT-QuIC) in routine CWD surveillance. This could influence our understanding of CWD in wildlife, including increased temporal lags in detecting the effects of management intervention. Future research should consider developing corrector parameters to account for uncertainties in CWD detection. Such parameters could be estimated by quantifying CWD detection error using other diagnostic methods with elevated sensitivity (e.g., PMCA, RT-QuIC) on subsets of samples tested using conventional diagnostic tests. We note that this approach may be impractical for large sample sizes but could help detect the circulation of lower prion concentrations otherwise missed, which may be specifically crucial in early stages of CWD invasion when outbreak control could be more effective.

Analytical methods to analyze CWD were dominated by regression and compartment-based models. These models have generated a baseline to guide management efforts and build new research hypotheses. Nevertheless, the quality of compartment models is subject to the robustness and biologic realism of specific parameters and is context, population, and time specific (Uehlinger et al. 2016). Regression models chiefly identified linkages between CWD infection and environmental features (Mateus-Pinilla et al. 2013). This empirical work can help guide experimental work to assess CWD transmission under specific landscape features (e.g., plant species) to build upon preliminary work with scrapie prions (Pritzkow et al. 2015, 2018). This will elucidate mechanisms of causation of CWD environmental transmission and will guide evidence-based interventions based on landscape modification for CWD control (Goñi et al. 2015). Additional research could include exploring the effects of more sensitive CWD detection on parameter estimation for compartmental, hierarchical, and matrix models as it pertains to prion transmission dynamics and demographic trends in wild cervids. For example, environmental prion loads may be better quantified with accurate detection of cervids in the preclinical phase of infection (Henderson et al. 2017). Finally, hunter-sourced data has been invaluable in CWD surveillance programs; however, these data can be biased (Conner et al. 2000) or incomplete compared with active epidemiologic sampling. Future research could include assessing bias mitigation strategies for the epidemiologic surveillance derived from samples obtained opportunistically from hunters.

Model evaluation and field validation remain barely explored; current models only focus on calibration. That is, CWD descriptive models prioritize model-fit to the data available (e.g., Akaike information criteria). For predictive models, evaluation approaches should split data to ensure temporal and spatial independence among datasets. We found that current evaluation methods do not ensure statistical independence between calibration and evaluation datasets (e.g., cross-validation) and require accurate detection of positive and negative individuals (e.g., evaluations based on sensitivity and specificity), which is not a robust approach considering the CWD detection capacities of major diagnostic methods used (i.e., IHC and ELISA).

Few modeling studies incorporated sampling effort, and no studies integrated biogeographic approaches in models (e.g., continental geomorphology), variables that are often starting points in which infectious diseases are generally explored (Emmanuel et al. 2011; Peterson 2014). Instead, models were restricted to finer scale population-level variables and localized geomorphology (i.e., soil composition). Considering resistance and environmental transmissibility of prions (Zabel and Ortega 2017) and recent experiments showing potential of prion deactivation from natural weathering processes (Yuan et al. 2018), research on CWD based on landscape epidemiology warrants further investigation.

A new frontier in CWD modeling research includes addressing the detection, quantification, and mitigation of sampling biases in surveillance. Biases could be linked to the geographic areas studied, species inspected, and diagnostic approaches used. Specifically, broad-scale biogeographic models are encouraged to account for the roles of environmental variation on past CWD epidemics (Evans et al. 2016). Such coarse-scale studies will allow researchers to reconstruct patterns in CWD spread, establishment, and maintenance in novel areas and populations as has been proven useful for other infectious disease agents including worms, bacteria, and viruses (Reisen 2010; Cadavid Restrepo et al. 2016). Continental-level assessment of CWD surveillance will elucidate whether current CWD distributions are driven by sampling effort, specifically in more probabilistic areas for CWD infection (e.g., CWD-free counties neighboring endemic CWD areas).

Chronic wasting disease and other prion diseases remain a challenge in wildlife disease modeling. Prion diseases may not follow traditional rules and assumptions derived from other pathogens for which more information is available. For example, unlike most diseases, hosts fail to demonstrate immune responses to prion infection, and prions lack genetic identity necessary for coevolution (Zabel and Avery 2015). In addition, prions remain resilient in extreme conditions otherwise fatal to other pathogens (Jung et al. 2003), and their unclear origins further complicate CWD tracking. More importantly, CWD is the only prion disease affecting free-ranging wildlife, it has no treatment, and its zoonotic potential cannot be discredited, thereby limiting our abilities to develop experimental work in basic laboratory settings. Our overview of CWD modeling could serve as a baseline for future CWD research.

We thank the Editor in Chief, Assistant Editor, David R. Edmunds, Emily A. Almberg, Emmanuel A. Fringpong, Megan S. Kirchgessner, and an anonymous reviewer for helpful comments and editions. We also thank Fran Astorga and Spencer L. Waddle for assistance in data analysis. The study was supported by a 2019 grant from the Rural Health Initiative of the Destination Areas/Strategic Growth Areas, Virginia Tech.

Supplementary material for this article is online at http://dx.doi.org/10.7589/2019-08-213.

Al-Arydah
M
,
Smith
RJ
,
Lutscher
F
.
2012
.
Modeling gender-structured wildlife diseases with harvesting: Chronic wasting disease as an example.
ISRN Biomath
2012
:
1
18
.
Allen
LJS
,
Brown
VL
,
Jonsson
CB
,
Klein
SL
,
Laverty
SM
,
Magwedere
K
,
Owen
JC
,
Van Den Driessche
P
,
Feinstone
WH
.
2012
.
Mathematical modeling of viral zoonoses in wildlife.
Nat Resour Model
25
:
5
51
.
Almberg
ES
,
Cross
PC
,
Johnson
CJ
,
Heisey
DM
,
Richards
BJ
.
2011
.
Modeling routes of chronic wasting disease transmission: Environmental prion persistence promotes deer population decline and extinction.
PLoS One
6
:
e19896
.
Auchincloss
AH
,
Gebreab
SY
,
Mair
C
,
Diez Roux
AV
.
2012
.
A review of spatial methods in epidemiology, 2000–2010.
Annu Rev Public Health
33
:
107
122
.
Baeten
LA
,
Powers
BE
,
Jewell
JE
,
Spraker
TR
,
Miller
MW
.
2007
.
A natural case of chronic wasting disease in a free-ranging moose (Alces alces shirasi).
J Wildl Dis
43
:
309
314
.
Barve
N
,
Barve
V
,
Jiménez-Valverde
A
,
Lira-Noriega
A
,
Maher
SP
,
Peterson
AT
,
Soberón
J
,
Villalobos
F
.
2011
.
The crucial role of the accessible area in ecological niche modeling and species distribution modeling.
Ecol Model
222
:
1810
1819
.
Bastian
M
,
Heymann
S
,
Jacomy
M
.
2009
.
Gephi: An open source software for exploring and manipulating networks.
In
:
Proceedings of the third international AAAI conference on weblogs and social media, Association for the Advancement of Artificial Intelligence, San Jose, California, 17–20 May
, pp.
361
362
. .
Benestad
SL
,
Mitchell
G
,
Simmons
M
,
Ytrehus
B
,
Vikøren
T
.
2016
.
First case of chronic wasting disease in Europe in a Norwegian free-ranging reindeer.
Vet Res
47
:
88
.
Blanchong
JA
,
Samuel
MD
,
Scribner
KT
,
Weckworth
BV
,
Langenberg
JA
,
Filcek
KB
.
2008
.
Landscape genetics and the spatial distribution of chronic wasting disease.
Biol Lett
4
:
130
133
.
Cadavid Restrepo
AM
,
Yang
YR
,
McManus
DP
,
Gray
DJ
,
Giraudoux
P
,
Barnes
TS
,
Williams
GM
,
Soares Magalhães
RJ
,
Hamm
NAS
,
Clements
ACA
.
2016
.
The landscape epidemiology of echinococcoses.
Infect Dis Poverty
5
:
13
.
Carlson
CM
,
Hopkins
MC
,
Nguyen
NT
,
Richards
BJ
,
Walsh
DP
,
Walter
WD
.
2018
.
Chronic wasting disease: Status, science, and management support by the US Geological Survey.
US Geological Survey Open-File Report 2017-1138.
US Geological Survey
,
Reston, Virginia
,
8
pp.
Centers for Disease Control and Prevention (CDC)
.
2019a
.
CWD in animals.
.
CDC
.
2019b
.
Prevention.
.
Conner
MM
,
Ebinger
MR
,
Blanchong
JA
,
Cross
PC
.
2008
.
Infectious disease in cervids of North America: Data, models, and management challenges.
Ann N Y Acad Sci
1134
:
146
172
.
Conner
MM
,
McCarty
CW
,
Miller
MW
.
2000
.
Detection of bias in harvest-based estimates of chronic wasting disease prevalence in mule deer.
J Wildl Dis
36
:
691
699
.
Conner
MM
,
Miller
MW
.
2004
.
Movement patterns and spatial epidemiology of a prion disease in mule deer population units.
Ecol Appl
14
:
1870
1881
.
Cortez
MH
,
Weitz
JS
.
2013
.
Distinguishing between indirect and direct modes of transmission using epidemiological time series.
Am Nat
181
:
E43
E52
.
Cullingham
CI
,
Merrill
EH
,
Pybus
MJ
,
Bollinger
TK
,
Wilson
GA
,
Coltman
DW
.
2011a
.
Broad and fine-scale genetic analysis of white-tailed deer populations: Estimating the relative risk of chronic wasting disease spread.
Evol Appl
4
:
116
131
.
Cullingham
CI
,
Nakada
SM
,
Merrill
EH
,
Bollinger
TK
,
Pybus
MJ
,
Coltman
DW
.
2011b
.
Multiscale population genetic analysis of mule deer (Odocoileus hemionus hemionus) in western Canada sheds new light on the spread of chronic wasting disease.
Can J Zool
89
:
134
147
.
Davenport
KA
,
Mosher
BA
,
Brost
BM
,
Henderson
DM
,
Denkers
ND
,
Nalls
AV
,
McNulty
E
,
Mathiason
CK
,
Hoover
EA
.
2018
.
Assessment of chronic wasting disease prion shedding in deer saliva with occupancy modeling.
J Clin Microbiol
56
:
e01243
17
.
DeVivo
MT
,
Edmunds
DR
,
Kauffman
MJ
,
Schumaker
BA
,
Binfet
J
,
Kreeger
TJ
,
Richards
BJ
,
Schätzl
HM
,
Cornish
TE
.
2017
.
Endemic chronic wasting disease causes mule deer population decline in Wyoming.
PLoS One
12
:
e0186512
.
Diefenbach
DR
,
Rosenberry
CS
,
Boyd
RC
.
2004
.
From the field: Efficacy of detecting chronic wasting disease via sampling hunter-killed white-tailed deer.
Wildl Soc Bull
32
:
267
272
.
Dubé
C
,
Mehren
KG
,
Barker
IK
,
Peart
BL
,
Balachandran
A
.
2006
.
Retrospective investigation of chronic wasting disease of cervids at the Toronto Zoo, 1973–2003.
Can Vet J
47
:
1185
1193
.
Edmunds
DR
,
Albeke
SE
,
Grogan
RG
,
Lindzey
FG
,
Legg
DE
,
Cook
WE
,
Schumaker
BA
,
Kreeger
TJ
,
Cornish
TE
.
2018
.
Chronic wasting disease influences activity and behavior in white-tailed deer.
J Wildl Manag
82
:
138
154
.
Edmunds
DR
,
Kauffman
MJ
,
Schumaker
BA
,
Lindzey
FG
,
Cook
WE
,
Kreeger
TJ
,
Grogan
RG
,
Cornish
TE
.
2016
.
Chronic wasting disease drives population decline of white-tailed deer.
PLoS One
11
:
e0161127
.
EFSA Panel on Biological Hazards
,
Ricci
A
,
Allende
A
,
Bolton
D
,
Chemaly
M
,
Davies
R
,
Escámez
PSF
,
Gironés
R
,
Herman
L
,
Koutsoumanis
K
, et al.
2018
.
Scientific opinion on chronic wasting disease (II).
EFSA J
16
:
e05132
.
Emmanuel
NN
,
Loha
N
,
Okolo
MO
,
Ikenna
OK
.
2011
.
Landscape epidemiology: An emerging perspective in the mapping and modelling of disease and disease risk factors.
Asian Pac J Trop Dis
1
:
247
250
.
Escobar
LE
,
Pritzkow
S
,
Winter
S
,
Grear
DA
,
Kirchgessner
MS
,
Dominguez-Villegas
E
,
Machado
G
,
Peterson
AT
,
Soto
C
.
2020
.
The ecology of chronic wasting disease in wildlife.
Biol Rev Camb Philos Soc
95
:
393
408
.
Evans
TS
,
Kirchgessner
MS
,
Eyler
B
,
Ryan
CW
,
Walter
WD
.
2016
.
Habitat influences distribution of chronic wasting disease in white-tailed deer.
J Wildl Manag
80
:
284
291
.
Farnsworth
ML
,
Hoeting
JA
,
Hobbs
NT
,
Conner
MM
,
Burnham
KP
,
Wolfe
LL
,
Williams
ES
,
Theobald
DM
,
Miller
MW
.
2007
.
The role of geographic information systems in wildlife landscape epidemiology: Models of chronic wasting disease in Colorado mule deer.
Vet Ital
43
:
581
593
.
Farnsworth
ML
,
Hoeting
JA
,
Hobbs
NT
,
Miller
MW
.
2006
.
Linking chronic wasting disease to mule deer movement scales: A hierarchical Bayesian approach.
Ecol Appl
16
:
1026
1036
.
Farnsworth
ML
,
Wolfe
LL
,
Hobbs
NT
,
Burnham
KP
,
Williams
ES
,
Theobald
DM
,
Conner
MM
,
Miller
MW
.
2005
.
Human land use influences chronic wasting disease prevalence in mule deer.
Ecol Appl
15
:
119
126
.
Galloway
NL
,
Monello
RJ
,
Brimeyer
D
,
Cole
E
,
Thompson Hobbs
N
.
2017
.
Model forecasting of the impacts of chronic wasting disease on the Jackson elk herd.
Biological Research Division, National Park Service
,
Washington, DC
,
32
pp.
Gambhir
M
,
Bozio
C
,
O'Hagan
JJ
,
Uzicanin
A
,
Johnson
LE
,
Biggerstaff
M
,
Swerdlow
DL
.
2015
.
Infectious disease modeling methods as tools for informing response to novel influenza viruses of unknown pandemic potential.
Clin Infect Dis
60
(
Suppl
1
):
S11
S19
.
Garlick
MJ
,
Powell
JA
,
Hooten
MB
,
McFarlane
LR
.
2011
.
Homogenization of large-scale movement models in ecology.
Bull Math Biol
73
:
2088
2108
.
Garlick
MJ
,
Powell
JA
,
Hooten
MB
,
MacFarlane
LR
.
2014
.
Homogenization, sex, and differential motility predict spread of chronic wasting disease in mule deer in southern Utah.
J Math Biol
69
:
369
399
.
Garner
M
,
Hamilton
S
.
2011
.
Principles of epidemiological modelling.
Rev Sci Tech
30
:
407
416
.
Goñi
F
,
Mathiason
CK
,
Yim
L
,
Wong
K
,
Hayes-Klug
J
,
Nalls
A
,
Peyser
D
,
Estevez
V
,
Denkers
N
,
Xu
J
, et al.
2015
.
Mucosal immunization with an attenuated Salmonella vaccine partially protects white-tailed deer from chronic wasting disease.
Vaccine
33
:
726
733
.
Grear
DA
,
Samuel
MD
,
Langenberg
JA
,
Keane
D
.
2006
.
Demographic patterns and harvest vulnerability of chronic wasting disease infected white-tailed deer in Wisconsin.
J Wildl Manag
70
:
546
553
.
Grear
DA
,
Samuel
MD
,
Scribner
KT
,
Weckworth
BV
,
Langenberg
JA
.
2010
.
Influence of genetic relatedness and spatial proximity on chronic wasting disease infection among female white-tailed deer.
J Appl Ecol
47
:
532
540
.
Gross
JE
,
Miller
MW
.
2001
.
Chronic wasting disease in mule deer: Disease dynamics and control.
J Wildl Manag
65
:
205
215
.
Haley
NJ
,
Richt
J
.
2017
.
Evolution of diagnostic tests for chronic wasting disease, a naturally occurring prion disease of cervids.
Pathogens
6
:
35
.
Hamir
AN
,
Kunkle
RA
,
Cutlip
RC
,
Miller
JM
,
Williams
ES
,
Richt
JA
.
2006
.
Transmission of chronic wasting disease of mule deer to Suffolk sheep following intracerebral inoculation.
J Vet Diagn Invest
18
:
558
565
.
Hamir
AN
,
Miller
JM
,
Kunkle
RA
,
Hall
SM
,
Richt
JA
.
2007
.
Susceptibility of cattle to first-passage intracerebral inoculation with chronic wasting disease agent from white-tailed deer.
Vet Pathol
44
:
487
493
.
Heesterbeek
H
,
Anderson
RM
,
Andreasen
V
,
Bansal
S
,
De Angelis
D
,
Dye
C
,
Eames
KTD
,
Edmunds
WJ
,
Frost
SDW
,
Funk
S
, et al.
2015
.
Modeling infectious disease dynamics in the complex landscape of global health.
Science
347
:
aaa4339
.
Hefley
TJ
,
Hooten
MB
,
Hanks
EM
,
Russell
RE
,
Walsh
DP
.
2017a
.
Dynamic spatio-temporal models for spatial data.
Spat Stat
20
:
206
220
.
Hefley
TJ
,
Hooten
MB
,
Russell
RE
,
Walsh
DP
,
Powell
JA
.
2017b
.
When mechanism matters: Bayesian forecasting using models of ecological diffusion.
Ecol Lett
20
:
640
650
.
Heisey
DM
,
Osnas
EE
,
Cross
PC
,
Joly
DO
,
Langenberg
JA
,
Miller
MW
.
2010
.
Linking process to pattern: Estimating spatiotemporal dynamics of a wildlife epidemic from cross-sectional data.
Ecol Monogr
80
:
221
240
.
Henderson
DM
,
Tennant
JM
,
Haley
NJ
,
Denkers
ND
,
Mathiason
CK
,
Hoover
EA
.
2017
.
Detection of chronic wasting disease prion seeding activity in deer and elk feces by real-time quaking-induced conversion.
J Gen Virol
98
:
1953
1962
.
Herzog
SA
,
Blaizot
S
,
Hens
N
.
2017
.
Mathematical models used to inform study design or surveillance systems in infectious diseases: A systematic review.
BMC Infect Dis
17
:
775
.
Hollingsworth
TD
,
Adams
ER
,
Anderson
RM
,
Atkins
K
,
Bartsch
S
,
Basáñez
MG
,
Behrend
M
,
Blok
DJ
,
Chapman
LAC
,
Coffeng
L
, et al.
2015
.
Quantitative analyses and modelling to support achievement of the 2020 goals for nine neglected tropical diseases.
Parasit Vectors
8
:
630
.
Hsieh
HF
,
Shannon
SE
.
2005
.
Three approaches to qualitative content analysis.
Qual Health Res
15
:
1277
1288
.
Jennelle
CS
,
Henaux
V
,
Wasserberg
G
,
Thiagarajan
B
,
Rolley
RE
,
Samuel
MD
.
2014
.
Transmission of chronic wasting disease in Wisconsin white-tailed deer: Implications for disease spread and management.
PLoS One
9
:
e91043
.
Jennelle
CS
,
Walsh
DP
,
Samuel
MD
,
Osnas
EE
,
Rolley
R
,
Langenberg
J
,
Powers
JG
,
Monello
RJ
,
Demarest
ED
,
Gubler
R
, et al.
2018
.
Applying a Bayesian weighted surveillance approach to detect chronic wasting disease in white-tailed deer.
J Appl Ecol
55
:
2944
2953
.
Johns
CJ
,
Mehl
CH
.
2006
.
A dynamic spatial model for chronic wasting disease in Colorado.
J Data Sci
4
:
21
37
.
Joly
DO
,
Ribic
CA
,
Langenberg
JA
,
Beheler
K
,
Batha
CA
,
Dhuey
BJ
,
Rolley
RE
,
Bartelt
G
,
Van Deelen
TR
,
Samuel
MD
.
2003
.
Chronic wasting disease in free-ranging Wisconsin white-tailed deer.
Emerg Infect Dis
9
:
599
601
.
Joly
DO
,
Samuel
MD
,
Langenberg
JA
,
Blanchong
JA
,
Batha
CA
,
Rolley
RE
,
Keane
DP
,
Ribic
CA
.
2006
.
Spatial epidemiology of chronic wasting disease in Wisconsin white-tailed deer.
J Wildl Dis
42
:
578
588
.
Joly
DO
,
Samuel
MD
,
Langenberg
JA
,
Rolley
RE
,
Keane
DP
.
2009
.
Surveillance to detect chronic wasting disease in white-tailed deer in Wisconsin.
J Wildl Dis
45
:
989
997
.
Jung
M
,
Pistolesi
D
,
Panà
A
.
2003
.
Prions, prion diseases and decontamination.
Ig Sanità Pubbl
59
:
331
344
.
Keane
DP
,
Barr
DJ
,
Bochsler
PN
,
Hall
SM
,
Gidlewski
T
,
O'Rourke
KI
,
Spraker
TR
,
Samuel
MD
.
2008
.
Chronic wasting disease in a Wisconsin white-tailed deer farm.
J Vet Diagn Invest
20
:
698
703
.
Kelly
AC
,
Mateus-Pinilla
NE
,
Brown
W
,
Ruiz
MO
,
Douglas
MR
,
Douglas
ME
,
Shelton
P
,
Beissel
T
,
Novakofski
J
.
2014
.
Genetic assessment of environmental features that influence deer dispersal: Implications for prion-infected populations.
Popul Ecol
56
:
327
340
.
Kim
TY
,
Shon
HJ
,
Joo
YS
,
Mun
UK
,
Kang
KS
,
Lee
YS
.
2005
.
Additional cases of chronic wasting disease in imported deer in Korea.
J Vet Med Sci
67
:
753
759
.
Koontz
BL
,
Loomis
JB
.
2005
.
Economic importance of elk hunting in Jackson Hole, Wyoming.
US Geological Survey Open-File Report 2005-1183.
US Geological Survey
,
Reston, Virginia
,
24
pp.
Krumm
CE
,
Conner
MM
,
Miller
MW
.
2005
.
Relative vulnerability of chronic wasting disease infected mule deer to vehicle collisions.
J Wildl Dis
41
:
503
511
.
Lou
Y
,
Wu
J
.
2017
.
Modeling lyme disease transmission.
Infect Dis Model
2
:
229
243
.
Maji
C
,
Mukherjee
D
,
Kesh
D
.
2018
.
Deterministic and stochastic analysis of an eco-epidemiological model.
J Biol Phys
44
:
17
36
.
Manjerovic
MB
,
Green
ML
,
Mateus-Pinilla
N
,
Novakofski
J
.
2014
.
The importance of localized culling in stabilizing chronic wasting disease prevalence in white-tailed deer populations.
Prev Vet Med
113
:
139
145
.
Marsh
RF
,
Kincaid
AE
,
Bessen
RA
,
Bartz
JC
.
2005
.
Interspecies transmission of chronic wasting disease prions to squirrel monkeys (Saimiri sciureus).
J Virol
79
:
13794
13796
.
Mateus-Pinilla
N
,
Weng
HY
,
Ruiz
MO
,
Shelton
P
,
Novakofski
J
.
2013
.
Evaluation of a wild white-tailed deer population management program for controlling chronic wasting disease in Illinois, 2003-2008.
Prev Vet Med
110
:
541
548
.
Matsumoto
T
,
Samuel
MD
,
Bollinger
T
,
Pybus
M
,
Coltman
DW
.
2013
.
Association mapping of genetic risk factors for chronic wasting disease in wild deer.
Evol Appl
6
:
340
352
.
McCallum
H.
2016
.
Models for managing wildlife disease.
Parasitology
143
:
805
820
.
McNulty
E
,
Nalls
AV
,
Mellentine
S
,
Hughes
E
,
Pulscher
L
,
Hoover
EA
,
Mathiason
CK
.
2019
.
Comparison of conventional, amplification and bio-assay detection methods for a chronic wasting disease inoculum pool.
PLoS One
14
:
e0216621
.
Mejía-Salazar
MF
,
Goldizen
AW
,
Menz
CS
,
Dwyer
RG
,
Blomberg
SP
,
Waldner
CL
,
Cullingham
CI
,
Bollinger
TK
.
2017
.
Mule deer spatial association patterns and potential implications for transmission of an epizootic disease.
PLoS One
12
:
e0175385
.
Mejía-Salazar
MF
,
Waldner
C
,
Stookey
J
,
Bollinger
TK
.
2016
.
Infectious disease and grouping patterns in mule deer.
PLoS One
11
:
e0150830
.
Miller
MW
,
Conner
MM
.
2005
.
Epidemiology of chronic wasting disease in free-ranging mule deer: Spatial, temporal, and demographic influences on observed prevalence patterns.
J Wildl Dis
41
:
275
290
.
Miller
MW
,
Swanson
HM
,
Wolfe
LL
,
Quartarone
FG
,
Huwer
SL
,
Southwick
CH
,
Lukacs
PM
.
2008
.
Lions and prions and deer demise.
PLoS One
3
:
e4019
.
Miller
MW
,
Williams
ES
,
McCarty
CW
,
Spraker
TR
,
Kreeger
TJ
,
Larsen
CT
,
Thorne
ET
.
2000
.
Epizootiology of chronic wasting disease in free-ranging cervids in Colorado and Wyoming.
J Wildl Dis
36
:
676
690
.
Monello
RJ
,
Galloway
NL
,
Powers
JG
,
Madsen-Bouterse
SA
,
Edwards
WH
,
Wood
ME
,
O'Rourke
KI
,
Wild
MA
.
2017
.
Pathogen-mediated selection in free-ranging elk populations infected by chronic wasting disease.
Proc Natl Acad Sci U S A
114
:
12208
12212
.
Monello
RJ
,
Powers
JG
,
Hobbs
NT
,
Spraker
TR
,
Watry
MK
,
Wild
MA
.
2014
.
Survival and population growth of a free-ranging elk population with a long history of exposure to chronic wasting disease.
J Wildl Manag
78
:
214
223
.
Moore
SJ
,
Smith
JD
,
Richt
JA
,
Greenlee
JJ
.
2019
.
Raccoons accumulate PrPSc after intracranial inoculation of the agents of chronic wasting disease or transmissible mink encephalopathy but not atypical scrapie.
J Vet Diagn Invest
31
:
200
209
.
Moore
SJ
,
West Greenlee
MH
,
Kondru
N
,
Manne
S
,
Smith
JD
,
Kunkle
RA
,
Kanthasamy
A
,
Greenlee
JJ
.
2017
.
Experimental transmission of the chronic wasting disease agent to swine after oral or intracranial inoculation.
J Virol
91
:
e00926
17
.
Mule Deer Working Group
.
2015
.
The Wyoming mule deer initiative.
Wyoming Game and Fish Department
,
Cheyenne, Wyoming
,
83
pp.
Newman
MEJ.
2004
.
Coauthorship networks and patterns of scientific collaboration.
Proc Natl Acad Sci U S A
101
(
Suppl
1
):
5200
5205
.
Nobert
BR
,
Merrill
EH
,
Pybus
MJ
,
Bollinger
TK
,
Hwang
YT
.
2016
.
Landscape connectivity predicts chronic wasting disease risk in Canada.
J Appl Ecol
53
:
1450
1459
.
Nusser
SM
,
Clark
WR
,
Otis
DL
,
Huang
L
.
2008
.
Sampling considerations for disease surveillance in wildlife populations.
J Wildl Manag
72
:
52
60
.
O'Hara Ruiz
M
,
Kelly
AC
,
Brown
WM
,
Novakofski
JE
,
Mateus-Pinilla
NE
.
2013
.
Influence of landscape factors and management decisions on spatial and temporal patterns of the transmission of chronic wasting disease in white-tailed deer.
Geospat Health
8
:
215
227
.
Oraby
T
,
Vasilyeva
O
,
Krewski
D
,
Lutscher
F
.
2014
.
Modeling seasonal behavior changes and disease transmission with application to chronic wasting disease.
J Theor Biol
340
:
50
59
.
Osterholm
MT
,
Anderson
CJ
,
Zabel
MD
,
Scheftel
JM
,
Moore
KA
,
Appleby
BS
.
2019
.
Chronic wasting disease in cervids: Implications for prion transmission to humans and other animal species.
MBio
10
:
e01091
19
.
Peterson
AT.
2014
.
Mapping disease transmission risk: Enriching models using biogeography and ecology.
Johns Hopkins University Press
,
Baltimore, Maryland
,
224
pp.
Potapov
A
,
Merrill
E
,
Lewis
MA
.
2012
.
Wildlife disease elimination and density dependence.
Proc Biol Sci
279
:
3139
3145
.
Potapov
A
,
Merrill
E
,
Pybus
M
,
Coltman
D
,
Lewis
MA
.
2013
.
Chronic wasting disease: Possible transmission mechanisms in deer.
Ecol Model
250
:
244
257
.
Pritzkow
S
,
Morales
R
,
Lyon
A
,
Concha-Marambio
L
,
Urayama
A
,
Soto
C
.
2018
.
Efficient prion disease transmission through common environmental materials.
J Biol Chem
293
:
3363
3373
.
Pritzkow
S
,
Morales
R
,
Moda
F
,
Khan
U
,
Telling
GC
,
Hoover
E
,
Soto
C
.
2015
.
Grass plants bind, retain, uptake, and transport infectious prions.
Cell Rep
11
:
1168
1175
.
Prusiner
SB.
1982
.
Novel proteinaceous infectious particles cause scrapie.
Science
216
:
136
144
.
R Core Team
.
2019
.
R: A language and environment for statistical computing.
R Foundation for Statistical Computing
,
Vienna, Austria
.
http://www.R-project.org. Accessed October 2019
.
Rees
EE
,
Merrill
EH
,
Bollinger
TK
,
Hwang
YT
,
Pybus
MJ
,
Coltman
DW
.
2012
.
Targeting the detection of chronic wasting disease using the hunter harvest during early phases of an outbreak in Saskatchewan, Canada.
Prev Vet Med
104
:
149
159
.
Reisen
WK.
2010
.
Landscape epidemiology of vector-borne diseases.
Annu Rev Entomol
55
:
461
483
.
Robinson
SJ
,
Samuel
MD
,
Rolley
RE
,
Shelton
P
.
2013
.
Using landscape epidemiological models to understand the distribution of chronic wasting disease in the midwestern USA.
Landsc Ecol
28
:
1923
1935
.
Rogers
KG
,
Robinson
SJ
,
Samuel
MD
,
Grear
DA
.
2011
.
Diversity and distribution of white-tailed deer mtdna lineages in chronic wasting disease (CWD) outbreak areas in southern wisconsin, USA.
J Toxicol Environ Health A
74
:
1521
1535
.
Samuel
MD
,
Storm
DJ
.
2016
.
Chronic wasting disease in white-tailed deer: Infection, mortality, and implications for heterogeneous transmission.
Ecology
97
:
3195
3205
.
Schauber
EM
,
Woolf
A
.
2003
.
Chronic wasting disease in deer and elk: A critique of current models and their application.
Wildl Soc Bull
31
:
610
616
.
Schuler
KL
,
Jenks
JA
,
Klaver
RW
,
Jennelle
CS
,
Bowyer
RT
.
2018
.
Chronic wasting disease detection and mortality sources in semi-protected deer population.
Wildl Biol
2018
:
wlb.00437
.
Skuldt
LH
,
Mathews
NE
,
Oyer
AM
.
2008
.
White-tailed deer movements in a chronic wasting disease area in south-central Wisconsin.
J Wildl Manag
72
:
1156
1160
.
Spraker
TR
,
Miller
MW
,
Williams
ES
,
Getzy
DM
,
Adrian
WJ
,
Schoonveld
GG
,
Spowart
RA
,
O'Rourke
KI
,
Miller
JM
,
Merz
PA
.
1997
.
Spongiform encephalopathy in free-ranging mule deer (Odocoileus hemionus), white-tailed deer (Odocoileus virginianus) and Rocky Mountain elk (Cervus elaphus nelsoni) in northcentral Colorado.
J Wildl Dis
33
:
1
6
.
Statens Veterinärmedicinska Anstalt
.
2019
.
Wasting disease (CWD) discovered on moose in Norrbotten County.
.
Storm
DJ
,
Samuel
MD
,
Rolley
RE
,
Shelton
P
,
Keuler
NS
,
Richards
BJ
,
Van Deelen
TR
.
2013
.
Deer density and disease prevalence influence transmission of chronic wasting disease in white-tailed deer.
Ecosphere
4
:
1
14
.
Sun
L
,
Lee
C
,
Hoeting
JA
.
2015
.
Parameter inference and model selection in deterministic and stochastic dynamical models via approximate Bayesian computation: Modeling a wildlife epidemic.
Environmetrics
26
:
451
462
.
Texas Parks and Wildlife, Texas Animal Health Commission
.
2015
.
Chronic wasting disease management plan.
Texas Parks and Wildlife and Texas Animal Health Commission
,
Houston, Texas
,
17
pp.
Uehlinger
FD
,
Johnston
AC
,
Bollinger
TK
,
Waldner
CL
.
2016
.
Systematic review of management strategies to control chronic wasting disease in wild deer populations in North America.
BMC Vet Res
12
:
173
.
US Department of Agriculture
.
2014
.
Chronic wasting disease program standards.
US Department of Agriculture, Animal and Plant Health Inspection Service
,
Washington, DC
,
17
pp.
USGS (US Geological Survey)
.
2019
.
Distribution of chronic wasting disease in North America (updated).
.
Van Maanen
A
,
Xu
XM
.
2003
.
Modelling plant disease epidemics.
Eur J Plant Pathol
109
:
669
682
.
Vasilyeva
O
,
Oraby
T
,
Lutscher
F
.
2015
.
Aggregation and environmental transmission in chronic wasting disease.
Math Biosci Eng
12
:
209
231
.
Wallace
DI
,
Southworth
BS
,
Shi
X
,
Chipman
JW
,
Githeko
AK
.
2014
.
A comparison of five malaria transmission models: Benchmark tests and implications for disease control.
Malar J
13
:
268
.
Walsh
DP
,
Miller
MW
.
2010
.
A weighted surveillance approach for detecting chronic wasting disease foci.
J Wildl Dis
46
:
118
135
.
Walter
WD
,
Baasch
DM
,
Hygnstrom
SE
,
Trindle
BD
,
Tyre
AJ
,
Millspaugh
JJ
,
Frost
CJ
,
Boner
JR
,
VerCauteren
KC
.
2011a
.
Space use of sympatric deer in a riparian ecosystem in an area where chronic wasting disease is endemic.
Wildlife Biol
17
:
191
209
.
Walter
WD
,
Walsh
DP
,
Farnsworth
ML
,
Winkelman
DL
,
Miller
MW
.
2011b
.
Soil clay content underlies prion infection odds.
Nat Commun
2
:
200
.
Wasserberg
G
,
Osnas
EE
,
Rolley
RE
,
Samuel
MD
.
2009
.
Host culling as an adaptive management tool for chronic wasting disease in white-tailed deer: A modelling study.
J Appl Ecol
46
:
457
466
.
Western Association of Fish and Wildlife Agencies
.
2018
.
Recommendations for adaptive management of chronic wasting disease in the West.
Western Association of Fish and Wildlife Agencies
,
Edmonton, Alberta, Canada and Fort Collins, Colorado
,
20
pp.
Wickham
H.
2018
.
Tidyverse: Easily install and load the “tidyverse”.
RStudio.
.
Wild
MA
,
Hobbs
NT
,
Graham
MS
,
Miller
MW
.
2011
.
The role of predation in disease control: A comparison of selective and nonselective removal on prion disease dynamics in deer.
J Wildl Dis
47
:
78
93
.
Williams
AL
,
Kreeger
TJ
,
Schumaker
BA
.
2014
.
Chronic wasting disease model of genetic selection favoring prolonged survival in Rocky Mountain elk (Cervus elaphus).
Ecosphere
5
:
60
.
Williams
ES
,
Young
S
.
1980
.
Chronic wasting disease of captive mule deer: A spongiform encephalopathy.
J Wildl Dis
16
:
89
98
.
WDNR (Wisconsin Department of Natural Resources)
.
1998
.
Wisconsin's deer management program. The issues involved in decision-making.
2nd Ed.
Wisconsin Department of Natural Resources
,
Madison, Wisconsin
,
46
pp.
WDNR
.
2010
.
Wisconsin's chronic wasting disease response plan: 2010–2025.
Wisconsin Department of Natural Resources
,
Madison, Wisconsin
,
42
pp.
Yuan
Q
,
Telling
G
,
Bartelt-Hunt
SL
,
Bartz
JC
.
2018
.
Dehydration of prions on environmentally relevant surfaces protects them from inactivation by freezing and thawing.
J Virol
92
:
e02191
17
.
Zabel
MD
,
Avery
AC
.
2015
.
Prions—Not your immunologist's pathogen.
PLoS Pathog
11
:
e1004624
.
Zabel
MD
,
Ortega
A
.
2017
.
The ecology of prions.
Microbiol Mol Biol Rev
81
:
e00001
17
.

Supplementary data