February 2006
Volume 6, No. 2
for the West, about the West, from the West
Current mortality rates on U.S. dairies
for a specific illness.
Although producers generally record
their assessment of cause of death in onfarm information systems, there are numerous reasons why these data are unreliable or not useful for analyzing the causes
esults from the
of mortality or directing intervenUSDA:APHIS:VS National Animal
Figure 1. DHI Provo recorded dairy cattle death rates in 8 Westtion strategies.
Health Monitoring System
ern states (1991 to 2002).
First, on-farm assessments of dis(NAHMS) Dairy 2002 survey reease occurrence and cause of death
ported that approximately five per- 1 2 .0
may not be accurate.
cent of U.S. dairy cows die on-farm
1 0 .0
1 0 .0
Second, most record systems on
across the country each year. This
8 .8
8 .2
dairies are focused on reproductive
is a very high death rate compared
7 .9
7 .8
8 .0
and milk production performance.
with that of beef cows or feedlot
6 .8
6 .1
6 .0
Health events are either not monianimals, where annual death rates
5 .7
6 .0
5 .0
tored, are poorly defined (e.g. catewere estimated at just one to 1.5
5 .0
gories such as illness, lame or digespercent.
4 .0
3 .0
In the NAHMS survey, untive are not sufficiently characterknown reasons accounted for the
ized to allow analysis of specific
2 .0
largest proportion (20 percent) of
problems), or are not recorded at all.
producer reported dairy cow
Third, most of the conditions to
0 .0
deaths are attributed are
deaths, followed by calving probwhich
92
93
94
95
96
97
98
99
00
01
02
91
lems (17 percent), mastitis (17 persimilar to the reasons listed for an
Y ear
cent), and lameness or injury (14
animal being sold. Information is
Furthermore, data from the West shows
percent).
not recorded that allows interpretation of
a profound increase trend in death rates,
why the outcomes of disease are different
A liability of the NAHMS survey is that
from 5.0 percent in 1991 to 10.0 percent in
(sold vs. died).
it reported data based on the recollection of
2002 (Figure 1). During this same time peThough herd removal may be recorded,
dairy producers over the preceding year.
riod the culling rate remained fairly steady
typically no reason is provided or it conEvaluation of Dairy Herd Improvement Asat approximately 29 percent. Thus, while
tains no useful information (e.g. the remark
sociation (DHIA) records suggests death
the risk factors for death and culling may
for a DIED event is ‘SHOT’; or a specific
rates in the NAHMS studies are likely unbe similar, the NAHMS and DHIA data indisease such as hemorrhagic bowel disease
derestimated. DHIA death rates also rely
dicate that the risk of death is increasing.
is identified as HBS, BLDGUT, CLOST in
on producer reporting, however they are
However, this data provides no indicathe records).
based on monthly herd removal tallies and
tion as to the causes of death which could
The term ‘herd removal’ is used to dethus are less susceptible to recall bias and
direct management intervention to reduce
scribe the withdrawal of once productive
likely provide a more accurate estimate of
herd losses. Furthermore, there is no indideath loss.
animals from the herd. Mortalities are a
cation whether deaths occurred naturally
DHIA data from the West indicates that
subset of herd removals. Besides being lost
or via elected euthanasia, or if they were
the dairy cow death rate in 1998 (7.8 perdue to death, cows may be culled for
unexpected or occurred during treatment
cent) was similar to that reported across
slaughter or sold for dairy production on
by Frank Garry, MS, DVM
Coordinator of Integrated
Livestock Management Group,
Colorado State University
R
the rest of the country (Figure 1). However,
2004 DHIA data delineated for eight Western states indicated that a wide range of
death rates exists from state to state, and
Colorado (9.1 percent) has one of the highest (Figure 2).
Western Dairy News is a collaborative effort of Dairy Specialists from:
Production effort for Western Dairy News is generously sponsored by Monsanto Dairy Business
Febuary 25, 2006
W-43
of poor clinical disease management in adOther studies have identified clinically
another farm.
dition to any preexisting problem with cow
recognizable health problems that increase
The NAHMS Dairy 2002 survey showed
physiology.
the risk of death or culling in dairy cows,
that approximately 25.5 percent of dairy
Necropsy examination of dead animals to
such as calving difficulty, ketosis/fatty liver
cows left herds permanently during 2001,
assess and monitor cause of death is rarely
disease, coliform mastitis, milk fever, and
and that approximately six percent of these
performed on dairies. This is in sharp conparatuberculosis. The severity of these discows were sold to other dairies, while 94
trast to other intensive livestock manageeases in individual animals is highly influpercent were culled (i.e. sold and not rement systems, including poultry, swine, and
ential on the outcome.
turned to milk production/sent for slaughfeedlot enterprises, where necropsy moniSince most dairy health programs do not
ter). The reasons cows were culled included
toring is routine. Most dairy veterinarians
monitor or analyze the severity or impact of
mastitis and udder problems (27 percent of
focus considerable effort on dairy reproducthese diseases, dairies lack the tools needed
culled cows), lameness or injury (16 pertion, but little time on mortality evaluation.
to associate occurrence with final outcome
cent), other disease (six percent), reproducThis presents a very significant liability
and may fail to manage the problems aptive failure (27 percent), and poor milk proto the dairy industry because efforts to efpropriately. Differences in outcomes for induction not related to these other problems
fectively decrease mortality losses are hamdividual cows may result from failure to
(19 percent). Other miscellaneous reasons
pered by a lack of monitoring and informaapply readily available evaluation and
accounted for about five percent of culling.
tion that provide accurate assessment of
treatment methods appropriate to the speTherefore, on average, the overwhelming
the problem. Dairy workers could be
cific disease and severity of disease.
majority of dairy cows leaving dairies are
trained to more effectively monitor death
An overview of the health challenges
not fit for sale as dairy production animals,
losses, and to perform on-farm necropsy exfaced by dairy cows needs to recognize that
and approximately 50 percent of them are
aminations in consultation with vetleaving because of disease or inFigure 2: DHI Provo recorded dairy cattle death rates by state
erinarians when the veterinarian
jury problems, rather than being
in 2004.
cannot be present to perform the exselectively removed because of subamination on a freshly dead carcass.
optimal productivity.
1 2 .0
1 1 .3
The Integrated Livestock ManThe DHIA data from the West
1 0 .1
agement
Group (ILM) of Colorado
support this idea, reporting a sta1 0 .0
9 .1
State
University
has produced an
ble culling rate but a profound in8 .1
7 .9
7 .8
8 .0
on-line training program for percrease in death rate.
forming field necropsies on our webThe reasons for death have
6 .3
6 .2
5 .9
6 .0
site at http://www.cvmbs.colostate.edu/ilm/
many similar descriptors to the
reasons listed for herd removal.
outreach/necropsy/_notes/INDEX.HTML.
4 .0
Death may be attributed to disorVery few producers or veterinarians
ders stemming from calving, digeshave pursued this approach, attest2 .0
tive or metabolic derangements,
ing to the notion that monitoring acudder health, or accidents, but
tual cause of death has not been
0 .0
there is an inherent subjectivity
seen as a valuable pursuit.
AZ
CA
CO
ID
NM
OR
UT
WA
A ll
within the producer assessment.
Because of the highly complex naSTATE
The limitations of such surveys
ture of dairy management systems,
may lead to the generic classificait is likely that a variety of causes
changes in the modern dairy industry could
tion of a significant percentage of deaths as
are responsible for high cow mortality
lead to systematic problems with animal
due to ‘other reasons’ or ‘unknown’ –
rates, with different rates of occurrence on
care. Specifically, dairies have increased in
nomenclature that does little to delineate
different operations. The wide range of lacsize and focused on business profitability
causality or suggest preventative stratetational incidence risk for common diseases
with great emphasis on low cost. The labor
gies. If the ultimate goal is to decipher the
(milk fever, 0.03 to 22.3 percent; retained
force on most dairies is primarily composed
causes of death and to decrease mortality
placentas, 1.3 to 39.2 percent; metritis, 2.2
of low wage workers without extensive, prerates through management alternatives,
to 37.3 percent; ketosis, 1.3 to 18.3 percent,
existing dairy cow management skills. The
specific diagnoses must be obtained.
LDAs, 0.3 to 6.3 percent; and lameness, 1.8
ability of dairy personnel to adequately
to 30 percent) attests to the complexity of
identify disease in individual animals and
What is causing high mortality rates?
the system.
respond with prompt individual animal atPrevious studies and our preliminary
To adequately address such a complex
tention is limited by the extent of their exdata suggest that there is no single cause of
problem
requires more accurate informaperience and training.
the very high death rates in dairy cows. It
tion
about
current losses, followed by manThe overwhelming majority of sick cows
does not seem plausible that there is some
agement
alterations
that address the unon dairies are identified, diagnosed, and
unseen ‘cow killing disease’ that explains
derlying
problems.
This
will require changtreated by farm workers, rather than vetthese death rates across large numbers of
ing
the
nature
of
information
used in dairy
erinarians. Poor outcomes could be an issue
management systems. An example of masdairies. Rather, it appears that subsequent
culling or death as likely outcomes. The
titis prevalence can illustrate this point:
reasons for removal of cows for slaughter
The specific infectious organism that
Western Dairy News is published as a service to
causes a clinical mastitis episode can have
are closely related to the causes of death,
people interested in the health and welfare of the
a dramatic impact on outcome, and approand most of these are representative of
Western dairy industry. Archives of this publication
priate preventative or therapeutic meahealth issues that can be improved.
may be found at:
sures need to be tailored to the specific
There are some subclinical metabolic or
http://animalscience-extension.tamu.edu/dairy/wdn.html
cause, e.g. gram negative vs. gram positive,
physiologic problems faced by many cows in
For further information contact:
environmental vs. contagious, E. coli vs.
modern dairy systems that could predisDr. Ragan Adams, Editor
Staph. aureus.
pose to poor outcomes in the face of disease
ILM, CSU-VTH
Assessments and record systems that
challenges. Numerous problems have been
300 W. Drake Road
Fort Collins, CO 80523
described and can be identified in some cirtrack “mastitis” without identifying other
970-297-0371
cumstances. These include subclinical
specific details do not provide sufficient inradams@lamar.colostate.edu
hypocalcemia, subacute ruminal acidosis,
formation to promote effective intervenMaterial published in Western Dairy News is not
negative energy balance and metabolic distions. Similarly, monitoring death losses
subject to copyright. Permission is therefore granted
ease in early lactation, trace mineral and
with generic terms such as “lameness” or
to reproduce articles, although acknowledgement
vitamin deficiency, poor immune respon“mastitis” and performing this monitoring
of the source is requested.
siveness in the postpartum period, and feed
on the basis of presumption will not allow
Cooperative Extension programs
quality problems that induce gastrointesticorrection of management problems that
are available to all without discrimination.
nal disturbances or specific toxicoses.
may underlie the death.
W-44
HOARD’S WEST