Section of Epidemiology and Public Health, University of Pennsylvania, School of Veterinary Medicine, New Bolton Center, 382 W. Street Road, Kennett Square, PA 19348, USA
A mathematical model for infection with bovine viral diarrhea virus (BVDV) was created comprising a series of coupled differential equations. The model architecture is a development of the traditional model framework using susceptible, infectious and removed animals (the SIR model). The model predicts 1.2% persistent infection (within the range of field estimates) and is fairly insensitive to alterations of structure or parameter values. This model allows us to draw important conclusions regarding the control of BVD, particularly with respect to the importance of persistently infected (PI) animals in maintaining BVD as an endemic entity in the herd. Herds without PI animals are likely to experience episodic reproductive losses at intervals of two to three years, unlike herds with PI animals which will not see such marked episodic manifestations of infection. Instead, these herds will experience an initial peak of disease which will settle to low-level chronic reproductive losses. The model indicates that vaccine coverage for herd immunity (to avoid episodic manifestations of disease) need be only 57% without PI animals, although 97% coverage is required when PI animals are present. Analysis of model behavior suggests, a program of detection and removal of PI animals may enhance the effectiveness of a vaccine program provided these animals are in the herd for 10 days or less. The best results would be seen with PI animals in the herd for 5 or fewer days.
This article was published in Preventive Veterinary Medicine, 33, B. R. Cherry, M. J. Reeves and G. Smith, Evaluation of bovine viral diarrhea virus control using a mathematical model of infection dynamics, 91-108, Copyright Elsevier 1998.