1398
INTRODUCTION
Bovine respiratory disease (BRD) is the most
common and economically signifcant disease af-
fecting cattle in the United States (Galyean et al.,
1999; Lechtenberg et al., 2011). Diagnosis of BRD
is commonly performed by observing clinical signs
including depression, anorexia, coughing, nasal dis-
charge, and lack of rumen fll (Smith et al., 2001).
Current methods used to diagnose BRD based on
visual inspection have poor sensitivity and specifc-
ity (Amrine et al., 2013; Leruste, 2012; White and
Renter, 2009). Poor sensitivity and specifcity cause
frequent erroneous diagnoses by not identifying and
treating truly diseased cattle (low sensitivity) and un-
necessarily treating disease-free animals (low speci-
fcity). Treatment effcacy is evaluated and wellness
assessments are commonly performed using these
low sensitivity and specifcity measures demonstrat-
ing economic losses and misrepresentation of the ef-
fect BRD has on the industry.
Performance differences between calves di-
agnosed with BRD compared to clinically healthy
calves vary in published literature. The objective
of this study is to determine the economic value of
A stochastic model to determine the economic value of changing diagnostic test
characteristics for identi fcation of cattle for treatment of bovine respiratory disease
1
M. E. Theurer,* B. J. White,†
2
R. L. Larson,† and T. C. Schroeder‡
*Department of Diagnostic Medicine and Pathobiology, College of Veterinary Medicine, Kansas State University, Manhattan
66506; †Department of Clinical Sciences, College of Veterinary Medicine, Kansas State University, Manhattan 66506; and
‡Department of Agricultural Economics, College of Agriculture, Kansas State University, Manhattan 66506
ABSTRACT: Bovine respiratory disease is an eco-
nomically important syndrome in the beef industry,
and diagnostic accuracy is important for optimal dis-
ease management. The objective of this study was to
determine whether improving diagnostic sensitivity
or specifcity was of greater economic value at var-
ied levels of respiratory disease prevalence by using
Monte Carlo simulation. Existing literature was used
to populate model distributions of published sensi-
tivity, specifcity, and performance (ADG, carcass
weight, yield grade, quality grade, and mortality risk)
differences among calves based on clinical respira-
tory disease status. Data from multiple cattle feeding
operations were used to generate true ranges of respi-
ratory disease prevalence and associated mortality.
Input variables were combined into a single model
that calculated estimated net returns for animals by
diagnostic category (true positive, false positive,
false negative, and true negative) based on the preva-
lence, sensitivity, and specifcity for each iteration.
Net returns for each diagnostic category were multi-
plied by the proportion of animals in each diagnostic
category to determine group proftability. Apparent
prevalence was categorized into low (<15%) and
high (≥15%) groups. For both apparent prevalence
categories, increasing specifcity created more rapid,
positive change in net returns than increasing sensi-
tivity. Improvement of diagnostic specifcity, perhaps
through a confrmatory test interpreted in series or
pen-level diagnostics, can increase diagnostic val-
ue more than improving sensitivity. Mortality risk
was the primary driver for net returns. The results
from this study are important for determining future
research priorities to analyze diagnostic techniques
for bovine respiratory disease and provide a novel
way for modeling diagnostic tests.
Key words: bovine respiratory disease, diagnostic tools, economic modeling, Monte Carlo simulation
© 2015 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2015.93:1398–1410
doi:10.2527/jas2014-8487
1
There was no extrainstitutional funding or support for this project.
2
Corresponding author: bwhite@vet.k-state.edu
Received September 5, 2014.
Accepted December 10, 2014.
Published March 31, 2015