Chapter 22 Simple Undetected Prevalence Estimator

22.1 What is the Simple Undetected Prevalence Estimator?

The Simple Undetected Prevalence Estimator estimates the maximum possible underlying (and unobserved) prevalence of CWD after sampling has been completed and no positive cases were found.

Note: This tool does not consider age, sex, source of samples, which are known to influence prevalence computations. If the sample data of interest contains information on age, sex, and source, consider using the Prevalence Estimator Data Export.

Note: This tool does not consider clustering behavior of hosts. If you can estimate clustering behavior in hosts, and want to know the probability that an area is disease-free, consider using the Probability of Disease Freedom Using Clustering Model (https://pages.github.coecis.cornell.edu/CWHL/CWD-Data-Warehouse/DiseaseFreedomUsingClustering.html)

22.2 What Questions Does it Answer?

Question 1. What is the maximum possible unobserved prevalence of CWD in each sub-administrative area given our sampling effort? The Simple Undetected Prevalence Estimator is used after sampling when no positive cases have been found to estimate the maximum possible prevalence of CWD in each sub-administrative area.

Question 2. What is the maximum possible unseen prevalence of CWD in each sub-administrative area if we adopt sampling quotas from a non-statistical quotas model? Non-statistical surveillance quota models include the Risk-weighted Surveillance Quotas Model and the Sample Allocation Model (SAM). The Simple Undetected Prevalence Estimator Model can be used in conjunction with these non-statistical models to explore the level of statistical assurances in CWD status that arise when non-statistical sample quotas are followed.

22.3 Output Details

  • A map containing the maximum underlying unseen prevalence based on Bayesian approach.

  • A map containing the maximum underlying unseen prevalence based on a Frequentist approach.

22.4 Abbreviated Tutorial

  1. Run the model in the CWD Data Warehouse.
  2. Click through the maps to explore the maximum underlying prevalence in each area given sample effort.
  3. Explore the model logs, input file, and output files used in the run.
  4. If the model did not run, check the model logs to understand required data that was missing. ## Parameters Needed to Execute the Model

  • Model type: Select ‘Simple Undetected Prevalence Estimator Model’ from the drop-down list.

  • Reference name: Label the run.

  • (Optional) Applicable season-year: Label the season-year of the run to assist in documentation.

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

  • Season-year: Select the season-year to be used in the estimation.

Note: Underlying prevalence can be estimated for only one season at a time.

  • Demography: Select one or more demography data sets to be used in the estimation.

Note: Only demography data sets that have corresponding sample data for the species are allowed as input options.

  • False Positive Rate: Select the false positive rate. A false positive rate of 0.05 (default) leads to statistical confidence of 95%.

-Diagnostic Assurance: Select the desired diagnostic test sensitivity. Default is 0.95 (95% of the time a test will declare a true positive).

22.5 Details on the Theory

This code was developed by Dr. James Booth, Professor of Statistics and Data Science at Cornell University.