Chapter 24 Outbreak Response Model (in development)

24.1 What is the Outbreak Response Model?

The Outbreak Response Model is used after at least one case of CWD is confirmed to draw a spatial containment zone around the outbreak and to determine how best to proceed with managerial actions within the zone.

The containment zone is calculated as an integrated risk surface that combines (1) contact probability between deer with (2) predicted space use by deer (modeled as a resource selection function) to calculate probability of CWD transmission from the infected individual.

Two types of CWD transmission are modeled:

  1. Direct: transmission following contact between one individual and another.

  2. Indirect: transmission when a susceptible host encounters the disease in the environment. Indirect contact has a slightly wider area of effect than direct.

Six total resource selection functions were developed for this model based on the tested animal’s sex and season when it was active (winter, spring & summer, and fall). The specific function(s) used to calculate the final integrated risk surfaces are chosen based on user-selected parameters (described below).

24.2 What Questions Does it Answer?

This model can be used to answer two distinct questions related to CWD containment and response. Though each question can be answered using the same data, they cannot answered in the same run, so multiple runs are needed to answer both questions.

Question 1. Where has the diseased host come into direct or indirect contact with other susceptible hosts? The model allows the user to locate areas around the diseased host where contact with susceptible hosts has likely occurred. When this question is selected, the model uses the resource selection function from the season when the sample was collected to identify recent areas of likely transmission. Direct and indirect integrated risk surfaces are calculated separately for each submitted sample.

Question 2. Where can targeted culling within the containment zone be conducted to best stem or thwart transmission of the disease to new hosts? The model allows the user to pinpoint areas most likely to be visited by infected host(s). Target culling in these high risk areas is mostly likely to thwart transmission to new hosts. When this question is selected, the model calculates integrated risk surfaces for each sample for each season selected by the user, and combines them into one direct and one indirect probability surface.

24.3 Abbreviated Tutorial

  1. Choose your model parameters (see below).
  2. Run model with those desired parameters to draw the containment zone around the index case(s).
  3. In the results, the containment zone depicts the area in which management should target outbreak response.
  4. Also in results, pay particular attention to areas within the containment zone with the highest relative risk, as these are areas where targeted culling might best thwart continued disease transmission.
  5. Rerun model with different parameter values to see if and to what extent use of indirect or direct contact rates alter the containment radius or high risk zones.

24.4 Parameters Needed to Execute the Model

  • Model type: Select ‘Outbreak Response Model’ from the drop-down list.

  • Reference name: Label the run.

  • (Optional) Applicable season year: Label the season-year. This label is not used in model execution and is intended to assist the provider in documenting the model execution.

  • (Optional) Notes: Enter any additional remarks about the run.

  • Model: Select New York (Quinn et al. 2009).

  • Probability: Select the proportion of contacts that you would like to model within your containment zone. This parameter defines the maximum size of your final probability surfaces. The higher the number, the larger the zone. Must be greater than 0 and less than 1.

  • Summary type: Select the question (described above) to address with this run. Options are either “season harvested” for question 1 (where has the diseased host come into contact with other hosts?) or “across samples & seasons” for question 2 (where should culling be targeted to thwart disease transmission?).

  • Sample source: Select how to read in CWD test data. To read in samples directly from the warehouse, select “warehouse.” To manually enter samples into the model with coordinates and test details, select “user-submitted”.

Parameters only needed for samples read in from warehouse:

  • Season-year: Select a season-year to read those tests directly into the model. Must contain at least one positive CWD test for model to run.

  • Modeled season: Select season(s) across which you want to combine integrated risk surface calculations. Available seasons are winter (January 1 - April 30), spring & summer (May 1 - September 30), and fall (October 1 - December 31).

    • Choose seasons in which the infected deer was active and transmitting CWD to other deer or to the environment. For most samples, we recommend selecting all three available seasons. Fewer seasons may be selected if for example the individual is known to be less than a year old and thus not shedding CWD during one or more seasons.

Parameters only needed for samples entered manually:

  • Date harvested: Date sample was collected. Informs which season-specific resource selection function(s) used to calculate integrated risk surfaces.

  • Sex: Sex of tested deer. Informs which sex-specific resource selection function(s) used to calculate integrated risk surfaces.

  • Latitude and Longitude: Location of sample collection.

24.5 Output Details

Both questions output Integrated Risk Surfaces that show probability of a deer encountering CWD from the diseased host, centered about the index case(s). The number and calculation steps of the output surfaces differ based on the selected question. Those outputs are:

Question 1. Where has the diseased host(s) come into direct or indirect contact with other susceptible hosts?

  • A number of integrated risk surfaces modelling direct and indirect transmission each equal to the number of samples provided.
    • For example, if three samples were provided to the model, the output contains three direct and three indirect surfaces for a total of six.
    • These surfaces are calculated using direct and indirect contact probabilities for each sample and the resource selection function corresponding to sex and harvest date of the sample.
  • A number of quantile polygon summaries of direct and indirect transmission surfaces each equal to the number of samples provided. The surface is binned based on quantiles calculated within the maximum extent of the surface as defined by the user.

Question 2. Where can targeted culling within the containment zone be conducted to best stem or thwart transmission of the disease to new hosts?

  • Two integrated risk surfaces, one modelling direct transmission and one modelling indirect transmission across all provided samples and user-selected seasons. This surface is useful for determining where CWD transmission is more likely on the landscape from all diseased hosts.

  • Two quantile polygon summaries of direct and indirect risk surfaces. Each surface is binned based on quantiles calculated within the maximum extent of the surface as defined by the user.

24.6 Details on the Theory

Dechen Quinn A, Williams D, Porter W, Smith M, DeSantis F, McNulty S. 2009. Risk Assessment of a Chronic Wasting Disease Outbreak in New York: Final Report. State University of New York College of Environmental Science and Forestry.

Williams D, Dechen Quinn A, Porter W. 2014. Informing Disease Models with Temporal and Spatial Contact Structure among GPS-Collared Individuals in Wild Populations. PLoS ONE 9(1): e84368. https://doi.org/10.1371/journal.pone.0084368