Computer Simulation Model of Mycobacterium avium subsp. paratuberculosis Control in a Zoo: Can We Improve Our Disease Surveillance and Management Strategies?
American Association of Zoo Veterinarians Conference 2013
Carmel L. Witte1,2,3, MS; Nathaniel Osgood4, MS, PhD; Anthony Gamst5, PhD; Laura L. Hungerford6, DVM, MPH, PhD; Bruce A. Rideout1, DVM, PhD, DACVP
1Wildlife Disease Laboratories, Institute for Conservation Research, San Diego Zoo Global, Escondido, CA, USA; 2Division of Epidemiology, Department of Family and Preventive Medicine, University of California, San Diego, CA, USA; 3Graduate School of Public Health, San Diego State University, San Diego, CA, USA; 4Department of Computer Science, Computational Epidemiology and Public Health Informatics Laboratory, University of Saskatchewan, Saskatoon, SK, Canada; 5Biostatistics and Bioinformatics, Department of Family and Preventive Medicine, University of California, San Diego, CA, USA; 6Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD, USA
Abstract
Mycobacterium avium subsp. paratuberculosis (Map) infection is a significant cause of morbidity and mortality in endangered wildlife and zoo animals.1,2 The goal of this study was to use simulation to explore the impact of different Map surveillance strategies on infection incidence and detection for hoofed mammals housed in a zoo environment. Data from 20 yr of disease surveillance at San Diego Zoo Global facilities and scholarly literature were used to construct an animated, agent-based simulation model using Anylogic software.a The simulation was designed to capture important features of life history and disease biology, including aging, reproduction, infection status, shedding, and testing. Monte Carlo methods were used to predict incidence, proportion of infectious animals detected, and cost of testing over a 10-yr period for seven different scenarios. The model successfully simulated emergent behavior of an endemic disease with low incidence; the baseline strategy of annually testing all animals ≥10 mo of age yielded a mean 10-yr incidence rate of 2.1% (95% CI: 1.9–2.4). None of the scenarios evaluated completely eliminated Map infection. Several management strategies, such as waiting until animals reached 2 yr of age to start testing and testing only 50% of the population, decreased costs and did not significantly alter the incidence or proportion of animals detected. Keeping a test-positive animal for an additional year to breed prior to culling did not significantly affect incidence. Simulation offers an innovative and useful tool for zoos to evaluate different animal management and disease surveillance decision-making strategies.
Endnote
a. The Anylogic Company, St. Petersburg, Russian Federation.
Acknowledgments
This study was funded by the Ellen Browning Scripps Foundation. The authors would like to acknowledge Tristan Burgess (University of California, Davis) for assistance with population summaries and the many veterinarians, laboratory staff, animal managers, and keepers that contributed to data collection and processing during the past two decades.
Literature Cited
1. Manning, E.J.B. 2011. Paratuberculosis in captive and free-ranging wildlife. Vet. Clin. N. Am.- Food A. 27:621–630.
2. Witte C., L.L. Hungerford, and B.A. Rideout. 2009. Association between Mycobacterium avium subsp. paratuberculosis infection among offspring and their dams in nondomestic ruminant species housed in a zoo. J. Vet. Diag. Invest. 21:40–47.