Economic Impact of Mastitis

in Dairy Cows

Christel Nielsen
Faculty of Veterinary Medicine and Animal Science
Department of Animal Breeding and Genetics
Uppsala

Doctoral Thesis
Swedish University of Agricultural Sciences
Uppsala 2009
Acta Universitatis agriculturae Sueciae
2009:29

Cover illustration by Anny Toftkær Nielsen

ISSN 1652-6880
ISBN 978-91-86195-76-2
© 2009 Christel Nielsen, Uppsala
Print: SLU Service/Repro, Uppsala 2009
Economic Impact of Mastitis in Dairy Cows

Abstract
This thesis aims to assess the economic loss associated with clinical (CM) and
subclinical (SCM) mastitis under current Swedish farming conditions.
Stochastic simulation was used to investigate the impact of mastitis on technical
and economic results of a 150-cow dairy herd. The yearly avoidable cost of mastitis,
assuming that the initial incidence (32 and 33 cases of CM and SCM per 100 cow-
years, respectively) could be reduced by 50%, was estimated at €8 095. This figure
corresponded to 5% of the economic net return for the herd given the initial
incidence of mastitis. Expressed as an average per cow/year, the avoidable cost of
mastitis was estimated at €54.
The economic loss associated with mastitis could not be reduced by discarding
milk with high somatic cell count, because this resulted in a substantial decrease of
the volume of sold milk which was not offset by the increase in milk price.
Cases of CM and SCM were on average associated with an average economic
loss of €275 and €60, respectively. Reduced milk production constituted the major
cost component of the economic loss caused by mastitis.
The magnitude of yield loss associated with mastitis occurring in different stages
of lactation was assessed using mixed linear models. The dataset was collected in a
research herd between 1987 and 2004, and consisted of weekly test-day records
sampled in 1200 lactations. The most extensive yield loss was estimated when CM
developed in early lactation and when SCM (modelled by means of increased
somatic cell count) occurred in late lactation. The 305-day yield loss associated with
CM varied between 0 and 705 kg milk in primiparous cows and between 0 and 902
kg milk in multiparous cows, depending on lactation week at onset. Most cases of
CM developed in the first week of lactation and resulted in a yield loss of 578 and
782 kg milk in primiparous and multiparous cows, respectively. Daily yield loss at
an SCC of 500 000 cells/ml ranged from 0.7 to 2.0 kg milk in primiparous cows
and from 1.1 to 3.7 kg milk in multiparous cows. The yield loss in an average 305-
day lactation affected by SCM was 150 and 450 kg milk in primiparous and
multiparous cows, respectively.

Keywords: dairy cow, mastitis, somatic cell count, yield loss, dairy herd, economic
performance, discarding milk

Author’s address: Christel Nielsen, SLU, Department of Animal Breeding and

Genetics, P.O. Box 7023, 750 07 Uppsala, Sweden
E-mail: Christel.Nielsen@hgen.slu.se
Till Birgitta
Jag önskar du hade fått vara med
Contents
List of Publications 9 

Abbreviations 10 

Introduction 11 

Background 13 
What is Mastitis? 13 
Severity and Duration 13 
Somatic Cell Count 14 
Pathogens 15 
Risk Factors 15 
Consequences of Mastitis 16 
Yield Loss 17 
Milk Composition 18 
Veterinary and Treatment Costs 19 
Discarded Milk 19 
Extra Labour 20 
Subsequent Disorders 20 
Culling 21 
Other Effects 22 
Economic Assessment of Mastitis 23 
Methods of Economic Analysis of Animal Disease 25 
Partial Budgeting 25 
Simulation 26 

Aims of Thesis 27 

Summary of Investigations 29 

Material and Methods 29 
Data from Research Herd 29 
Simulated Data 32 
Statistical Approaches 34 
The SimHerd Model 35 
Main Results 36 
Yield Loss caused by Mastitis 36 
Economic Loss associated with Mastitis 38 
Discarding Milk with High SCC 40 

5
General Discussion 41 
Yield Loss caused by Mastitis 41 
Severity 41 
Lactation Stage 41 
Parity 43 
Production Level 43 
Importance of Mastitis to Herd Profit 43 
Implications for Mastitis Control 44 
Prioritization between Cows 44 
Economic Framework 45 
Different Estimates Support Decisions at Different Levels 46 
Economic Loss per Case of Mastitis 46 
Inconsistency of Estimates 46 
Consistency with Previously Published Estimates 48 
Economic Loss caused by Mastitis in Sweden 51 
Discarding Milk with High SCC 52 
Methodological Issues 53 
Reference Level for Yield in Healthy Cows 53 
Recurrent Cases of CM 53 
Estimates of Yield Loss Obtained from One Herd 54 
Definition of SCM 54 
Strategies for Modelling Yield Loss 56 
Method of Economic Analysis 56 

Main Conclusions 61 

Practical Implications 63 

Future Research 65 

Ekonomisk betydelse av mastit hos mjölkkor 67 

Bakgrund 67 
Sammanfattning av avhandlingens delarbeten 68 
Avkastningsförlusternas storlek 68 
Kostnad per fall 68 
Besättningsekonomiska konsekvenser 69 
Sortering av mjölk för att sänka tankcelltalet 69 
Kostnaden för mastit i Sverige 69 
Slutsatser 69 
Praktisk tillämpning av resultaten 70 

6
References 71 

Acknowledgements 79 

7
8
List of Publications
This thesis is based on the work contained in the following papers, referred
to by Roman numerals in the text:

I Hagnestam, C., Emanuelson, U. and Berglund, B. (2007). Yield Losses

Associated with Clinical Mastitis Occurring in Different Weeks of
Lactation. Journal of Dairy Science 90(5), 2260-2270.

II Hagnestam-Nielsen, C. and Østergaard, S. (2009). Economic Impact of

Clinical Mastitis in a Dairy Herd Assessed by Stochastic Simulation using
Different Methods to Model Yield Losses. Animal 3(2), 315-328.

III Hagnestam-Nielsen, C., Emanuelson, U., Berglund, B. and Strandberg,

E. (2009). Relationship Between Somatic Cell Count and Milk Yield in
Different Stages of Lactation. Journal of Dairy Science, accepted,
doi:10.3168/jds.2008-1719.

IV Hagnestam-Nielsen, C., Emanuelson, U., Strandberg, E., Andersson, H.,

Berglund, B. and Østergaard, S. Economic Consequences of Mastitis and
Discarding Milk with High Somatic Cell Count (manuscript).

Papers I-III are reproduced with the permission of the publishers.

9
Abbreviations
BTSCC bulk tank somatic cell count
CM clinical mastitis
SCC somatic cell count
SCM subclinical mastitis
SEK Swedish krona
TD test day

10
Introduction
The dairy sector is subjected to increasing international competition.
Economically effective herds are thus a prerequisite to maintain a sound
Swedish dairy industry. Economic margins of dairy herds are, however,
narrow. Optimization of the economic results, therefore, becomes
important, and the need for cost minimization at every level of production
is accentuated. A means of reducing the costs of production is to decrease
the incidence of production disorders, as such are associated with reduced
production, veterinary costs, and increased replacement rate, and,
consequently, give rise to economically less efficient herds. Mastitis is of
considerable interest because of its high incidence and the extensive costs
associated with the disease. In 2007, the average incidence of veterinary-
treated clinical mastitis (CM) in herds participating in the Swedish milk
recording scheme was 16% (Swedish Dairy Association, 2008). The
incidence of mastitis can however be expected to be even higher, because
there is considerable under-reporting of CM (Mörk et al., 2009). Udder
disease, including udder disorders and high somatic cell count (SCC),
constitutes the most common reason for culling of Swedish dairy cows
(Swedish Dairy Association, 2008). In 2007, 26% of cullings was attributed
to udder disease, and 10% of the total cow population was, consequently,
culled because of udder disorders and high SCC. Indeed, mastitis is the most
costly disease in dairy production (Seegers et al., 2003 (review); Kossaibati &
Esslemont, 1997; Degraves & Fetrow, 1993 (review)).
Clearly, mastitis control is of paramount importance. The incidence of
mastitis can be reduced by implementation of preventive measures. These
are, however, associated with extra costs for the farmer in terms of
investments and labour, and interventions will only be made if the resulting
increase in revenue can be expected to offset the incurred costs. Information
about the economic loss associated with mastitis is, therefore, crucial when

11
evaluating the viability of different preventive measures. The economic loss
incurred by mastitis is also an essential part of other management decisions,
such as treating infected udder quarters, culling mastitic cows, and discarding
milk with high SCC in order to obtain a higher milk price.
The frequency of mastitis in the dairy cow population can also be
decreased by breeding for cows with better ability to resist udder disease.
Udder health is unfavourably genetically correlated with milk yield (Carlén
et al., 2004; Heringstad et al., 2000 (review); Emanuelson et al., 1988), and
selecting only for increased production, which traditionally has been the
focus of dairy cattle breeding in many countries, will therefore result in
deterioration of udder health. This can be counteracted by applying a broad
breeding goal, like the one used in the Nordic countries, which includes not
only production traits, but functional traits such as mastitis resistance. The
genetic progress in a trait is partly determined by the relative weight put on
it in the total merit index of bulls. In order to assign proper economic
weight to mastitis resistance, accurate estimates of the economic loss caused
by mastitis are necessary.
This thesis aims at assessing the economic loss associated with mastitis
under current Swedish farming conditions, in order to provide estimates that
can support decisions regarding mastitis control in individual herds and
facilitate derivation of appropriate economic weight of mastitis resistance in
the breeding goal.

12
Background
What is Mastitis?
Mastitis is defined as an inflammatory reaction of the mammary gland
(International Dairy Federation, 1987). It is induced when pathogenic
microorganisms enter the udder through the teat canal, overcome the cow’s
defence mechanisms, begin to multiply in the udder, and produce toxins
that are harmful to the mammary gland. Mammary tissue is then damaged,
which causes increased vascular permeability. As a result of this, milk
composition is altered: there is leakage of blood constituents, serum
proteins, enzymes, and salts into the milk; decreased synthesis of caseins and
lactose; and decreased fat quality (Østerås, 2000; Harmon, 1994). The extent
of these changes is determined by the severity of the infection (Pyörälä,
2003; Harmon, 1994; International Dairy Federation, 1987).
Mastitis is a multifactorial disease. As such, its incidence depends on
exposure to pathogens, effectiveness of udder defence mechanisms, and
presence of environmental risk factors, as well as interactions between these
factors (Oviedo-Boyso et al., 2007; Suriyasathaporn et al., 2000).

Severity and Duration

Mastitis can be either clinical or subclinical. Clinical cases give rise to visible
symptoms. Mild CM causes flakes or clots in the milk, whereas severe cases
are associated with heat, swelling and discolouration of the udder, as well as
abnormal secretion. Severe CM can also exhibit systemic reactions, such as
fever and loss of appetite. Mastitis can exist in the absence of visible signs of
infection, and is then referred to as subclinical mastitis (SCM). SCM is the
most prevalent form of mastitis (Akers, 2002). In practice, whether a case of
mastitis is classified as clinical or subclinical often depends on how carefully
the cow is observed when diagnosis is made (International Dairy Federation,

13
1987). SCM can be diagnosed by presence of pathogens in bacteriological
cultures of milk, but bacteriological sampling is not practically feasible as a
routine test. The current standard method of detecting SCM is to measure
SCC. Other inflammatory parameters, such as electrical conductivity,
lactose, lactate dehydrogenase, acute phase proteins, etc. (Åkerstedt et al.,
2007; Hamann, 2005; Pyörälä, 2003), have been proposed as indicators of
SCM, and some have the potential of being adapted to in-line use.
The duration of infection further classifies mastitis as acute or chronic
manifestations, where a sudden onset defines acute cases and chronic mastitis
is characterized by an inflammatory process that lasts for months and results
in progressive development of fibrous tissue (International Dairy Federation,
1987; Jain, 1979).

Somatic Cell Count

Milk always contains a certain amount of somatic cells. These consist of
various cell types, and their relative proportions depend on the health status
of the udder. In a healthy lactating mammary gland, the major proportion of
somatic cells is constituted by leukocytes (white blood cells) (Östensson et
al., 1988). These are primarily macrophages and lymphocytes, but a small
fraction consists of neutrophils and epithelial cells (Harmon, 1994; Kehrli &
Shuster, 1994). Microbial infection results in rapid accumulation of large
numbers of somatic cells in the udder, and these are predominantly
neutrophils (Harmon, 1994; Kehrli & Shuster, 1994; Östensson et al., 1988).
The increase in SCC constitutes an important part of the cow’s immune
response, and SCC is, therefore, a widely used indicator of mastitis.
Infection status has been recognized as the main factor affecting SCC
(Schepers et al., 1997; Sheldrake et al., 1983; Dohoo & Meek, 1982), and
SCC often exceeds 1 000 000 cells/ml in milk produced by mastitic cows
(Kehrli & Shuster, 1994). Compositional changes of milk are, however,
significant from 100 000 cells/ml, and are observed already at an SCC as
low as 50 000 cells/ml (Hamann, 2002; Reichmuth, 1975). In
bacteriologically negative milk cultures of udder quarter foremilk, an
average SCC of 68 000 cells/ml has been reported (Djabri et al., 2002).
Laevens et al. (1997) found an SCC of 49 000 cells/ml in composite milk
sampled in lactations from which no pathogens had been isolated. Presently,
an SCC threshold of 100 000 cells/ml in quarter foremilk is internationally
accepted (Hamann, 2005; 2003). In composite milk, an SCC above 50 000
cells/ml is considered indicative of SCM (Hortet & Seegers, 1998a
(review)).

14
Pathogens
The magnitude of increase in SCC partly depends on the causative
pathogen (Djabri et al., 2002; Lam et al., 1997; Schepers et al., 1997). The
primary mastitis causative microorganism is bacteria. These have
traditionally been categorized into major or minor pathogens, depending on
the magnitude of inflammatory response associated with infection. Major
pathogens most often cause CM (Djabri et al., 2002), and give rise to the
most extensive changes of milk composition (Harmon, 1994). These
infections are most often due to Staphylococcus aureus, Streptococci (agalactiae,
dysgalactiae, uberis), Escherichia coli and Klebsiella spp. Minor pathogens,
including Corynebacterium bovis and coagulase-negative staphylococci, cause
only moderate infection and are most often associated with SCM (Djabri et
al., 2002; Harmon, 1994). These pathogens have been reported to have a
protective effect against major pathogens (Lam et al., 1997; Matthews et al.,
1991), and the reason has been suggested to be competitive growth,
antagonism, induced leukocytosis or an increased immunity of the cow
(Black et al., 1972).
Depending on the vector of transmission, bacteria are considered either
contagious or environmental pathogens. Staph. aureus and Strep. agalactiae are
contagious pathogens, for which udders of infected cows serve as the major
reservoir. Contagious pathogens spread from cow to cow, primarily during
milking, and tend to result in chronic subclincal infections with flare-ups of
clinical episodes (Harmon, 1994). Environmental pathogens include E. coli,
Klebsiella spp., Strept. dysgalactiae and Strept. uberis., and have bedding,
manure and soil as their primary sources. The majority of infections caused
by environmental pathogens are clinical and of short duration (Harmon,
1994).
In Sweden, the most common pathogens isolated in cases of CM are
Staph. aureus, Strept. dysgalactiae, and E. coli (Persson Waller et al., 2009).

Risk Factors
Several features of individual cows can be identified, which might indicate
an increased risk of developing mastitis.
Multiparous cows are generally at higher risk of developing CM (Rajala-
Schultz et al., 1999b; Emanuelson et al., 1993; Bendixen et al., 1988), except
in the very early stages of lactation where the relationship is the opposite
(Steeneveld et al., 2008; Barkema et al., 1998). In multiparous cows, the risk
of developing CM increases with increasing parity (Steeneveld et al., 2008).
There is also consistency in the literature as regards higher SCC in older
cows (Harmon, 1994; Reneau, 1986; Dohoo & Meek, 1982).

15
 The risk of developing CM is highest in early lactation (Persson Waller et
al., 2009; Steeneveld et al., 2008; Barkema et al., 1998), whereas the risk of
SCM increases with increasing days in milk (Busato et al., 2000).
Mastitic cows tend to have higher milk yield than non-mastitic cows
before they develop CM (Gröhn et al., 2004; Wilson et al., 2004; Rajala-
Schultz et al., 1999b), indicating that high milk yield is a risk factor for CM.
Previous occurrences of mastitis, CM or high SCC, substantially
increased the risk of a cow developing a new case of CM (Steeneveld et al.,
2008; Bendixen et al., 1988). Other disorders, such as dystocia; milk fever;
retained placenta; metritis; ketosis; and lameness, are also known to increase
the risk of CM (Svensson et al., 2006; Emanuelson et al., 1993; Gröhn et al.,
1990b; Bendixen et al., 1988).
Cows of certain breeds are more prone to mastitis. Among the Swedish
breeds, national statistics (Swedish Dairy Association, 2008), as well as
several studies (Persson Waller et al., 2009; Nyman et al., 2007; Emanuelson
et al., 1993), show that Swedish Red cows have a lower incidence of
mastitis than Swedish Holstein cows.
The incidence of mastitis is influenced by managerial and environmental
factors, such as housing of cows, milking equipment, feeding regime,
hygienic quality of feed and water, cleanliness of cows, implementation of
preventive measures, and general practices related to, for instance, drying-off
(Nyman et al., 2007; Schreiner & Ruegg, 2003; Peeler et al., 2000; Barkema
et al., 1999; Elbers et al., 1998). Season also affects the incidence of mastitis,
and the incidence of CM has been reported to be highest during the winter
months (Steeneveld et al., 2008; Olde Riekerink et al., 2007a; Bendixen et
al., 1988).

Consequences of Mastitis
Mastitis is of great economic importance to milk producers, because the
disease has negative impact on several important aspects of cow and herd
performance. Incurred costs are of both direct and indirect nature
(Kossaibati & Esslemont, 1997). Direct costs include veterinary costs,
increased labour requirement, discarded milk (during the course of
treatment), and reduced milk yield and quality. Indirect costs are those that
are not always obvious to the milk producer, and are therefore referred to as
hidden costs. They include increased risk of subsequent disorders, reduced
fertility (extra services per conception and, as a result of this, an extended
calving interval), increased risk of culling, and, occasionally, mortality. The
total cost of mastitis can, consequently, be much higher than the direct cost

16
(Kossaibati & Esslemont, 1997). The cost associated with each component is
likely to vary between herds; partly because of differences in performance
parameters (yield level, fertility, etc.) and partly because of different
preferences of farmers influencing, for instance, their inclination to contact a
veterinarian when mastitis is detected.

Yield Loss
The extent of yield loss depends on severity, causative pathogen, parity of
cow, and the stage of lactation at which mastitis develops. In primiparous
cows, yield loss is most severe when CM is caused by Staph. aureus, E. coli,
and Klebsiella spp., whereas, in multiparous cows, Streptococcus spp., Staph.
aureus, E. coli, Klebsiella spp., and A. pyogenes are responsible for the largest
yield loss (Gröhn et al., 2004). Multiparous cows suffer more severe yield
loss than primiparous cows (Bennedsgaard et al., 2003; Hortet et al., 1999;
Rajala-Schultz et al., 1999b). CM occurring before peak yield results in the
most extensive yield loss (Rajala-Schultz et al., 1999b; Hortet & Seegers,
1998b (review)), whereas SCM occurring in late lactation is associated with
the highest yield loss (Bennedsgaard et al., 2003; Hortet et al., 1999).
There is substantial variation as regards estimates of the magnitude of
lactational yield loss in the literature. Discrepancies are due to differences in
management, breed and yield level, as well as the analytical method used.
CM is associated with yield loss at the time of diagnosis, and, more
importantly, yield loss often persists throughout lactation (Wilson et al.,
2004; Rajala-Schultz et al., 1999b; Houben et al., 1993). In the latest review
on the subject, Hortet and Seegers (1998b) summarised a lactational yield
loss of 300 to 400 kg (4 to 6%) in multiparous cows and 200 to 300 kg in
primiparous cows. Cases of CM are of different severity, and 40% of CM
cases can be expected to be associated with negligible yield loss, 30% with a
lactational yield loss of 150 ± 250 kg, and 30% with a lactational yield loss
of 950 ± 1050 kg (Hortet & Seegers, 1998b). Before diagnosis, mastitic
cows have a production advantage over their non-mastitic herd mates
(estimated at 2.6 kg by Wilson et al. (2004)). As most studies use the yield
level of non-mastitic cows as reference for yield in healthy cows, the
reported losses probably underestimate the true yield loss associated with
CM.
The reduction in milk yield associated with SCM has been summarised
as 80 kg (1.3%) and 120 kg (1.7%) per 2-fold increase in SCC above 50 000
cells/ml in primiparous and multiparous cows, respectively (Hortet &
Seegers, 1998a).

17
 The economic damage caused by yield loss is to some extent alleviated
by reduced feed cost (Yalcin, 2000). This effect of local CM is small, but
systemic CM has been reported to reduce dry matter intake by 30 kg over a
period of 117 days (Bareille et al., 2003). Still, reduced milk yield is the
major component of the cost associated with both CM and SCM (Huijps et
al., 2008; Hortet & Seegers, 1998b (review); Degraves & Fetrow, 1993
(review)).

Milk Composition
Mastitis decreases the synthetic capacity of the mammary gland, which leads
to decreased concentrations of fat and caseins in the milk (summarized by
Pyörälä, 2003 and Akers, 2002). Indeed, in reviews by Hortet & Seegers
(1998a; 1998b) CM and SCM were found to cause somewhat lower fat
content in the milk, and, on lactational level, estimates of the absolute fat
yield loss due to CM varied between 3 and 22 kg (1.5 to 7.5%). Results
from previous studies indicate slightly increased protein content in milk
produced by mastitic cows (Hortet & Seegers, 1998a; 1998b). This is due to
a higher content of inflammatory, non-coagulating proteins and whey
proteins, but, at the same time, a decreased proportion of casein (Urech et
al., 1999; Auldist et al., 1996; Barbano et al., 1991). In the studies reviewed
by Hortet & Seegers (1998b), the absolute protein yield loss following a case
of CM ranged from 0 to 15 kg (0 to 8.5%). Milk composition changes
caused by mastitis can be neglected in economic calculations (Seegers et al.,
2003), because milk produced in connection with diagnosis is discarded due
to treatment and later losses are proportional to the milk loss.
The bacterial count is increased by mastitis, but the elevation can be
neglected after the withdrawal period (Seegers et al., 2003).
The only changes in milk composition that are of economic importance
to the dairy producer are those that affect the milk price, i.e. components
that are part of the milk payment scheme. The largest dairy association in
Sweden, Arla Foods, pays a premium when bulk tank somatic cell count
(BTSCC) is below 300 000 cells/ml, and incurs a penalty when BTSCC
exceeds 401 000 cells/ml. In a specific herd, the bulk tank SCC and the
amount of milk produced influence whether a single case of mastitis will
affect the milk price or not, because those factors impact on the extent to
which the milk from the mastitic cow is “diluted” in the bulk tank (Østerås,
2005).
In practice, if BTSCC has been too high for a period of time, a cow with
high SCC might be culled in order to bring down BTSCC and avoid
payment penalties. Culling does, however, incur a cost of replacement, and

18
might, in fact, be more costly than accepting milk quality penalty (Dekkers
et al., 1996). An alternative approach to reduce BTSCC and limit the
impact of mastitis on herd profit is to sort out milk with high SCC. Milk-
sorting decisions have traditionally been based on SCC information
obtained from the milk recording scheme. Such information is obtained on
a monthly basis, resulting in crude decisions. Recent technological
advancements now allow SCC to be recorded in-line, and thus provide
means to make decisions on whether or not the milk from a certain cow is
to be delivered to the dairy in connection with each milking. The
profitability of discarding poor quality milk based on the results of an in-line
SCC indicator can, however, be expected to depend on how accurately
SCC can be recorded. If uncertainty is high, there would be a risk of
classifying milk with low SCC as poor quality milk and discard it and vice
versa, resulting in lower economic benefit of discarding milk with high SCC
as compared to if the true SCC could be measured.

Veterinary and Treatment Costs

In Sweden, only veterinarians may prescribe drugs for treatment of mastitis.
Milk producers consulting a veterinarian for treatment of a case of CM are,
1
on average, charged €119 (1 200 SEK, exchange rate of 18 November
2008, Paper IV) for starting fee, travel costs, labour, and drugs. Veterinary
costs are partly subsidised in Sweden.
Veterinary costs are easily identified, and are often perceived by farmers
as the full cost of mastitis. In a study of Norwegian dairy herds, it has,
however, been shown that there is only marginal association between
treatment rate of mastitis and gross margin, indicating that treatment costs
constitute only part of the total cost of mastitis and that most costs are
hidden (Østerås, 2000). Østerås (2000) therefore suggested that treatment
costs should be regarded as an investment to decrease the hidden costs.

Discarded Milk
Milk produced when a cow shows signs of mastitis, or while a cow is
treated with antibiotics, is discarded. The withdrawal period includes the
days when a cow actually receives drugs and a waiting time, usually
consisting of some additional days, when there is a risk of antibiotic residues
in the milk. The length of the withdrawal period depends on the
production system (i.e. conventional or organic), and the drug used.
The cost of discarded milk is comparable to that of milk loss, but with
one important difference: discarded milk is produced by the cow and is
1
Thomas Svensson, Swedish Board of Agriculture, personal communication

19
therefore associated with feed costs. The cost per unit of discarded milk is
thus higher than the corresponding cost of milk not produced (Halasa et al.,
2007; Hogeveen & Østerås, 2005).
Some herds feed discarded milk to calves, and, when this is practised,
discarded milk should be assigned an alternative value corresponding to that
of the amount of milk replacer that would otherwise have been needed.
Feeding milk produced by mastitic cows or cows being treated with
antibiotics to calves is, however, not recommended by Swedish
veterinarians.

Extra Labour
CM is associated with extra labour requirement, for instance in form of
attendance of the visit by the veterinarian and administration of medicine.
Also, CM may affect the order in which cows are milked, and thus gives rise
to less efficient milking routines. The time requirement associated with a
2
case of CM is likely to amount to two hours . The amount of time needed
to treat SCM can be expected to be less than that associated with CM,
because SCM is not always detected, and, when detected, is not always
treated.
Extra labour requirement should be valued based on the opportunity cost
of labour, i.e. the value of the next best alternative foregone as the result of
having to assign time to mastitis. Opportunity cost of labour in agriculture is
often difficult to assess, and is likely to differ between farms (Halasa et al.,
2007; Hogeveen & Østerås, 2005). Opportunity cost is readily calculated if
labour is of external source; because the value of time spent on preventing
mastitis can be estimated as hours times hourly wage. If labour is of internal
source, i.e. the farmers own time, opportunity cost is zero, or the value that
the farmer assigns to his or her free time. If, however, the farmer spends less
time on other tasks consequential upon having to deal with mastitis, then
opportunity cost is the decreased income resulting from spending less time
on these other tasks.

Subsequent Disorders
Cows having experienced one case of CM often develop a subsequent case
of CM later in lactation (Rajala & Gröhn, 1998; Houben et al., 1993). Also,
as contagious pathogens use the udder of infected cows as reservoir, having
mastitic cows in a herd increases the risk of spreading infection to healthy
cows. Mastitis is associated with increased risk of lameness (Peeler et al.,
1994; Dohoo & Martin, 1984a), and CM has been reported to be associated
2
Charlotte Hallén-Sandgren, Swedish Dairy Association, personal communication

20
with concurrent or subsequent diagnosis of ketosis, displaced abomasum,
and non-parturient paresis (Gröhn et al., 1989). CM is not a risk factor for
reproductive disorders (Gröhn et al., 1990a), but both CM and SCM are
known to adversely affect reproductive performance (Petersson et al., 2006;
Maizon et al., 2004; Schrick et al., 2001).
When other disorders or fertility disturbances occur as a consequence of
mastitis, it has been argued that their economic cost should be included in
the total cost of mastitis (Halasa et al., 2007; Hogeveen, 2005). A
complicating factor is, however, that the causative relationships between
mastitis and other disorders are somewhat obscure, and care must, therefore,
be taken in assigning costs of other disorders to mastitis as this might result
in overestimation of the total cost of mastitis.

Culling
CM increases the risk of culling (Schneider et al., 2007; Rajala-Schultz &
Gröhn, 1999a; Gröhn et al., 1998), as well as mortality (Bar et al., 2008a).
The extent to which CM affects the risk of culling depends on lactation
stage at clinical onset (Schneider et al., 2007; Beaudeau et al., 1995; Dohoo
& Martin, 1984b). It is also influenced by reproductive status, and open
cows are at greatest risk of being culled due to CM if they are diagnosed in
early lactation, whereas, pregnant cows are subjected to a relatively similar
risk of being culled because of CM irrespective of when in lactation they are
diagnosed (Schneider et al., 2007). Once cows are pregnant, the risk of
being culled as a consequence of CM drops sharply (Gröhn et al., 1998).
SCC above 300 000 cells/ml has been reported to increase the risk of
culling in primiparous cows (Beaudeau et al., 1995), and, in late lactation,
SCM is the most important disease influencing culling decisions regardless of
parity of the cow (Dohoo & Martin, 1984b).
The cost associated with involuntary culling as a consequence of mastitis
is an important component of the total cost of mastitis. Like milk loss,
increased risk of culling imposes a hidden cost, which is not always obvious
to the farmer. Involuntary cullings are associated with replacement costs,
and hence include costs of rearing or buying a heifer. If a heifer is not
available at the time a cow is culled, capacity utilization is reduced as a stall
will be empty while the fixed costs remain the same. Further economic loss
can be expected as milk yield of primiparous cows is lower than that of
multiparous cows, and because there is a risk that the yield level of a heifer
might be disappointing (Halasa et al., 2007; Hogeveen, 2005; Østerås,
2005). Economic cost also arise as cows culled due to mastitis do not reach
their full production potential (Østerås, 2005). On the other hand,

21
additional returns from meat are obtained if a mastitic cow can be sold to
slaughter.
Economic assessment of the impact of increased risk of culling is not
straightforward. Culling results from a management decision taken by the
farmer, which is based not only on presence or absence of mastitis but also
on milk yield, pregnancy status, parity, stage of lactation, and presence of
other diseases (Gröhn et al., 1998). Cows are culled when replacement is
judged to be the economically optimal alternative. At what point in time it
is optimal to replace a mastitic cow is, indeed, determined by the above
mentioned factors (Houben et al., 1994). Furthermore, it also depends on
mastitis incidence and critical price parameters (Stott & Kennedy, 1993).

Other Effects
Any kind of pathology involves some degree of poor animal welfare
(Broom, 2006). Mastitis is a very painful condition and is one of the major
welfare problems of dairy cows (Broom & Fraser, 2007; Webster, 1999).
Even mild cases of CM cause increased responsiveness to pain and affected
cows become hypersensitized to stimuli normally considered innocuous
(Fitzpatrick et al., 1998).
In European countries, there is a high level of consumer concern for
animal welfare (Harper & Henson, 2001; Moynagh, 2000), which results in
major public demand for improvements in animal welfare. Indeed,
consumers are prepared to pay considerably more for welfare-friendly
production practices (Moynagh, 2000). If milk is produced from cows with
high incidence of mastitis, consumers’ acceptance of dairy production, and
thereby their willingness to buy dairy products, may be adversely affected.
There are no known, direct threats to public health associated with
consuming dairy products made from high SCC milk (Hogan, 2005).
Potential food safety risks do, however, arise from ingestion of human
pathogens, bacterial toxins, and antibiotic residues; factors that are associated
with high SCC in milk (Hogan, 2005).
Mastitis in cattle is the main reason for use of antibiotics in Swedish
animal production (Swedish Board of Agriculture, 2008b). Exposure of
animals to antibiotics is an important factor contributing to development of
resistant bacteria (Mevius et al., 2005), which is considered to be one of the
major public health threats.

22
Economic Assessment of Mastitis
The total economic cost of disease consists of two distinct components;
production loss and control expenditures (McInerney et al., 1992). Losses
include benefits that are taken away and benefits that are not realized. The
former can be exemplified by milk that must be discarded following
treatment with antibiotics and the latter by milk that is never produced as a
result of disease. Expenditures are extra inputs needed to limit losses, either
by reducing the impact of an unplanned event, such as treatment of a
mastitic cow, or by preventing such events from occurring, as in the case of
investments into preventive measures. It has been argued that it is the
relationship between losses and expenditures that is of importance if
estimates of the cost of disease are to be used as input in decision-making
(McInerney et al., 1992), not the economic loss associated with disease per se
(Østerås, 2005; Dijkhuizen et al., 1995; Schepers & Dijkhuizen, 1991).
The primary purpose of economic analyses is to support decisions
regarding mastitis control. The economic losses in situations where nothing
is done to limit the impact of mastitis should, therefore, be compared with
the economic loss in situations where mastitis control is practised
(Hogeveen, 2005). When such assessments reveal that mastitis management
is economically profitable, interventions can be justified. Mastitis control can
be practiced at different levels; udder quarter, cow, herd, or national. The
cost of mastitis can be estimated at all of these levels, and the level of choice
depends on the nature of the decision that is to be supported.
At udder-quarter level, decisions are concerned with whether or not to
dry off an infected udder quarter. Cow-level decisions are directed at
managing occurrences of mastitis, and the options are no treatment,
treatment or culling. Treatment decisions impact also on herd level, as
treatment reduces spread of infection to healthy cows. In the same way,
culling might serve to reduce the overall incidence of mastitis in the herd.
Mastitis control at herd level aims at reducing the incidence of mastitis, and
consists of various proactive and reactive measures. Information on the
national consequences of mastitis is needed to answer whether subsidized
veterinary services and targeted research are necessary in order to reduce the
incidence of mastitis. Other contexts in which estimates of the impact of
mastitis is essential are when milk payment schemes are designed to motivate
producers to produce milk of desired quality, and when breeding programs
are dimensioned with respect to mastitis resistance.
The economic viability of different measures to control mastitis can only
be assessed if reliable estimates of the economic loss brought about by the
disease are available (Seegers et al., 2003). Before management strategies are

23
compared, the magnitude of economic loss must be addressed (Hogeveen,
2005), because mastitis management needs to be based on insight into the
costs associated with CM and SCM (Hogeveen & Østerås, 2005). Huirne
(2003) emphasized that calculations of the economic loss resulting from
mastitis is central, as they aid in providing a better overall view of the
impact of the disease, as well as contribute to better understanding of the
extent to which the loss can be reduced. It is the latter, the avoidable cost,
that is of importance in mastitis control, because complete elimination of
disease is not feasible (McInerney et al., 1992) and calculation of the
economic loss against a situation with zero incidence would thus encourage
overestimation of the economic damage associated with mastitis.
Previous estimates of the cost of mastitis show large variation (reviewed
by Halasa et al., 2007; Degraves & Fetrow, 1993; and Schepers &
Dijkhuizen, 1991). Some reasons for this variation seem to be origin of data,
definition of mastitis, differences in sources of loss included, and analytical
approach applied. Furthermore, studies have been conducted in different
spatiotemporal contexts, which can be assumed to influence the results as
circumstances of production and price levels vary between countries and
over time. Differences as regards the economic consequences of mastitis can
also be expected between farms, because of differences in incidence of
mastitis, pathogen frequency, severity of mastitis cases, number of cases per
affected cow and management routines. Additionally, the impact of mastitis
upon the economic performance of individual cows will be influenced by
their production level, age and reproductive status, and this will, in turn,
affect decisions regarding mastitis management on cow level (Hogeveen &
Østerås, 2005). The external validity of results might, therefore, be
questioned, and any generalizations must be made with caution. In order to
support decisions regarding mastitis management in individual herds,
evaluations of the economic loss associated with mastitis must be as specific
as possible (Hogeveen, 2005; Hogeveen & Østerås, 2005). Preferably, they
should be conducted for a specific herd and in a specific economic context
(Seegers et al., 2003).
The level at which the impact of mastitis has been estimated obviously
affects the results. Even though it has been suggested that the herd-level cost
of mastitis can be obtained by aggregating the costs at cow level (Østerås,
2005), this might impose bias on the results. Several of the direct costs of
mastitis, such as treatment, discarded milk, increased labour and decreased
milk production, can readily be assessed in individual cows. Indirect
consequences, however, frequently arise through herd dynamics, and often
reflect management decisions taken by the farmer. Increased risk of culling

24
and thus increased replacement costs, as well as penalties or loss of premiums
connected with increased SCC in the bulk tank milk, are good examples of
dynamic factors. The decision to cull a mastitic cow, in order to increase the
average milk yield per cow and decrease BTSCC, incurs a replacement cost,
and illustrates the way in which dynamics within the herd impact on herd-
level economy. Some herd-level effects can therefore not be assessed simply
by summing up the cow-level effects, and this type of dynamics require the
economic impact of mastitis to be addressed at herd level (Seegers et al.,
2003).

Methods of Economic Analysis of Animal Disease

There are several analytical approaches that can be applied in the assessment
of economic effects of disease and disease control. Which one that is most
suitable for a certain analysis depends on the nature of the decision problem;
the complexity of the disease and its effects; the data available; the intended
use of the model and the preferences and capabilities of the model builder
and/or decision maker; and the resources available (Bennett, 1992). A brief
description of the methods most frequently applied in economic analyses of
the economic loss associated with mastitis is given below.

Partial Budgeting
Partial budgeting can be used for rather simplistic economic comparisons of
different situations, such as high or low incidence of mastitis or changes
associated with implementation of a new control measure. It is a marginal
approach, i.e. it is concerned with changes in costs and returns due to an
actual decision or a disease event. Consequently, the analysis considers
variable costs (such as veterinary costs, labour, milk yield), but typically
ignore fixed costs (for instance maintenance costs and interest). Establishing
a partial budget requires information on:

1. Additional returns (returns that will not be received unless the

change is undertaken)
2. Reduced costs (costs present in the initial situation that will be
avoided if the change is made)
3. Returns foregone (returns received in the initial situation that will
not be received if the change is made)
4. Extra costs (costs associated with the change that are not present in
the initial situation)

25
If the sum of additional returns and reduced costs is greater than that of
returns foregone and extra costs, the change can be economically justified.
Advantages of using partial budgeting are that the method is a fairly
uncomplicated and time effective. Drawbacks are that it cannot account for
stochastic events nor involve development over time (Dijkhuizen et al.,
1995).

Simulation
Simulation is a more sophisticated method to analyse the impact of disease.
A simulation model is a simplified mathematical model of the unit of
concern (for instance a dairy herd), which can be manipulated by
modification of input parameters and, thus, adjusted to various real-life
situations. Simulation is well suited to investigate the impact of strategies
before they are applied, or as an alternative to intervention studies where
such would be time consuming or associated with great costs. Simulation
models can be either static or dynamic. Static models do not include time as
a variable, whereas dynamic models do. Dynamic models are, thus, capable
of modelling the development of the system over time. Models are further
classified as deterministic or stochastic. Deterministic models make definite
predictions of quantities, whereas stochastic models account for uncertainty
in the behaviour of the system (i.e. the same conditions can give different
outcomes), and, thus, reflect biological variation.
Dynamic and stochastic simulation models can account for complicated
interactions, and are thus capable of mimicking the dynamics taking place in
a dairy herd. Indeed, stochastic simulation has been suggested to be the most
relevant method to use when studying the effects of disease in a system
(Allore & Erb, 1999; Dijkhuizen et al., 1995). Disadvantages of stochastic
simulations are that the methodology requires an extensive amount of input,
lots of computational power, and is rather time consuming.

26
Aims of Thesis
The overall aim was to assess the economic loss associated with mastitis
under current Swedish farming conditions. More specifically, the objectives
were to:

 Estimate the yield loss associated with CM and SCM occurring in

different stages of lactation
 Assess the consequences of mastitis on the results of a dairy herd,
and to estimate the economic gain following a reduction of the
current incidence of mastitis
 Investigate whether recognition that yield loss varies depending on
when in lactation CM occurs results in an evaluation of the
economic loss caused by CM that differs from that derived when
the lactational timing of CM is ignored
 Study the impact of discarding milk with high SCC on herd net
return

27
28
Summary of Investigations
Material and Methods
Detailed descriptions of studied materials and applied methods are given in
Papers I to IV. Here, a condensed version is presented.

Data from Research Herd

Papers I and III were based on test-day records collected at weekly intervals
in the research herd of the Department of Animal Breeding and Genetics,
Swedish University of Agricultural Sciences. Data sampled between
September 1987 and April 2004 were available. Cows were of the Swedish
Red and Swedish Holstein breeds, and had an average yearly production of
8 900 and 10 600 kg milk, respectively, in 2004. The research herd was kept
at two locations during the study period. Before 1992, the herd was kept on
a farm with tie-stall housing. The current farm has a free-stall barn (n = 50)
and a tie-stall barn (n = 50). All cows were milked twice daily. In 2004, the
median BTSCC was 150 000 cells/ml.
Cases of CM were detected by the milkers by presence of abnormal milk
in the first milk streams or by signs of inflammation in one or more udder
quarters. All cases were diagnosed by a veterinarian. Not all cases were
necessarily treated. Treatment decisions were made according to a Standard
Operating Protocol based on stage of lactation as well as possible designation
for culling. Affected udder quarters were sampled and the milk cultured to
determine the pathogens present. Additional samples were taken two and
five weeks after a completed treatment. Milk samples for bacteriological
culture were also taken from cows with composite milk SCC > 180 000
cells/ml on two subsequent test days (TD). Furthermore, milk samples were
routinely cultured from each udder quarter in the fourth week of lactation,
as well as two weeks before a cow was dried off. From the autumn of 1997
until the autumn of 2000, milk samples were also taken in the first week of

29
lactation. From 1993 to 2004, milk samples were taken from 1 560 udder
quarters and 40% of the samples had positive cultures. Table 1 shows the
pathogens present in the culture-positive milk samples.

Table 1. Pathogens present in culture-positive milk samples (n = 624) taken from mastitic udder
quarters as well as by routine (Paper I)
Pathogen Frequency (%)
Mixed culture 25
1
Staphylococcus aureus 10
Staphylococcus aureus2 1
1
Coagulase-negative Staphylococci 12
Coagulase-negative Staphylococci2 6
Streptococcus agalactiae 0
Streptococcus dysgalactiae 11
Streptococcus uberis 7
Other Streptococcus spp. 1
Escherichia coli 16
Klebsiella 5
Arcanobacterium pyogenes 3
Other pathogens 3
1
Sensitive to penicillin
2
Resistant to penicillin

The lactational incidence risk (lactations with at least one case of CM

divided by the total number of lactations at risk) of CM was 19.9 and 28.9%
in primiparous and multiparous cows, respectively (Paper I). Cases of CM
most often developed in the first week of lactation (Figure 1). The overall
recurrence rate of CM was 23%, and the average number of cases per
lactation was 1.3.
The geometric mean SCC on TD free of CM was 55 000 and 95 000
cells/ml in primiparous and multiparous cows, respectively (Paper III).
Median-values of SCC on TD free of CM, sampled in different stages of
lactation, are given in Table 2.

30
 Primiparous cows

Cases of clinical mastitis (n)

40

30

20

10

0
1 6 11 16 21 26 31 36 41
Week in lactation

60 Multiparous cows
Cases of clinical mastitis (n)

50

40

30

20

10

0
1 6 11 16 21 26 31 36 41
Week in lactation
Figure 1. Number of cases of clinical mastitis in each week of lactation in
primiparous and multiparous cows (Paper I).

31
Table 2. Median-values of somatic cell count (x 103 cells/ml) in different stages of lactation on test days
free of clinical mastitis (Paper III)
Lactation weeks
1-2 3-8 9-16 17-24 25-32 33-44
Primiparous cows 133 46 35 37 43 59
Multiparous cows 114 48 51 77 109 164

Paper I was based on 38 535 test-day records sampled from 307 Swedish
Red and 199 Swedish Holstein cows. Out of 1 192 lactations, 298 were
affected by at least one case of CM. The first lactational incidence of CM,
irrespective of causative pathogen, was studied.
Test-day records without information on SCC were discarded in Paper
III, leaving data sampled between November 1989 and April 2004 from 303
Swedish Red and 194 Swedish Holstein cows. Two datasets were created.
Dataset A excluded test-day records sampled on days where cows were
affected by CM, and a case of CM was assumed to last for eight days
following diagnosis. It comprised 36 117 test-day records collected in 1 155
lactations. A subset of dataset A (dataset B) was created by excluding all
lactations in which CM occurred. Dataset B contained 27 753 test-day
records sampled in 863 lactations.

Simulated Data
The impact of mastitis on technical and economic results in a Swedish 150-
cow dairy herd was studied by means of simulation (Papers II and IV).
Scenarios were simulated over ten (Paper II) and twenty (Paper IV) years,
but only the average annual results from the latter half of the period was
used in the analyses. For each scenario, 250 replicates were performed.
The effect of mastitis on milk yield was not modelled in the same
manner in the two studies. In Paper II, the effect of CM on milk yield was
modelled directly (based on results from Paper I), and no effect of SCC on
milk yield was included in the model. The estimated average economic loss
per case of CM, therefore, included possible correlated effects of CM on
SCM, because the drop in production some weeks prior to diagnosis as well
as part of the lactation-long impairment of milk yield subsequent to CM,
may be due to SCM. It, thus, expressed the total economic loss caused by
mastitis in lactations where CM occurred. In Paper IV, we were interested
in the entire effect of mastitis on herd economy, and thus modelled both
CM and SCM. Because of programming considerations (mastitis is modelled
as one trait with different severities in SimHerd), the effect of mastitis on
milk yield was mediated primarily through its effect on SCC, i.e. as an

32
indirect effect. Mastitis started to affect SCC on the day of diagnosis. Cases
of CM were supplemented with a direct effect on milk yield, because it has
been reported that CM gives rise to a larger yield loss than what can be
explained simply by the increase in SCC (Bennedsgaard et al., 2003). In
Paper IV, recurrent cases of mastitis were allowed to develop within the
same lactation.
In Paper II, results given the initial incidence of CM (26 cases per 100
cow-years, first lactational incidence) were studied, together with the
consequences of reducing the incidence of CM by 50% and 90%
throughout lactation and the consequences of reducing the incidence by
50% and 90% only before peak yield. Different approaches to model yield
loss subsequent to CM were compared; a conventional modelling strategy -
i.e. one employing a single yield-loss pattern irrespective of when, during
the lactation, the cow developed CM - and a new modelling strategy in
which CM was assumed to affect production differently depending on its
lactational timing (based on the results obtained in Paper I). The effect of
choice of reference level when estimating yield loss was investigated by
combining the modelling strategies with two different reference levels; the
potential yield of mastitic cows, had they not developed CM, and the yield
of non-mastitic cows.
The full effect of mastitis, including both CM and SCM, was estimated
in Paper IV. In the initial scenario, a CM incidence of 32 cases per 100
cow-years (multiple cases could occur within the same lactation), based on
the incidence of CM in the research herd studied in Papers I and III, was
modelled. According to the distribution between clinical and subclinical
cases of mastitis in a study by De Haas et al. (2002), the probability of a case
becoming clinical was assumed to be 47%. In the initial scenario, an
incidence of SCM of 33 cases per 100 cow-years was, consequently,
modelled. Herd results given the initial incidence of mastitis were studied,
together with the consequences of reducing and increasing the incidence of
mastitis by 50% and the consequences of modelling no CM while the
incidence of SCM was kept constant, and vice versa.
Paper IV also included an assessment of the economic benefit of
discarding milk with high SCC in order to obtain a higher price for the
delivered milk. This was done by comparing results obtained when no milk
with high SCC was discarded with results given different strategies for
deciding when sorting of milk was to be initiated. When the decision of
whether to discard milk with high SCC was based on herd-level
information, sorting of milk was initiated when bulk tank SCC exceeded
220 000, 200 000 and 180 000 cells/ml, respectively, whereas when the

33
decision was based on cow-level information, milk was discarded when
SCC in individual cow’s milk exceeded 1 000 000, 750 000, and 500 000
cells/ml, respectively. The impact of uncertainty in the measurement of
SCC on the consequences of discarding milk with high SCC was
investigated by simulating different levels of uncertainty; high, low, or none.

Statistical Approaches
The effects of CM and SCC on test-day yield were estimated using mixed
linear models in SAS 8.2 and 9.1, respectively (PROC MIXED, SAS
Institute Inc., Cary, USA). Due to different shapes of their lactation curves,
primiparous and multiparous cows were analyzed separately. Dependent
variables in the models were test-day milk (Papers I and III), and fat and
protein yield (Paper I). The general model used in Papers I and III included
fixed effects of parity, breed, pregnancy status, year-season of calving, and
various disorders. Additionally, fixed effects of season of test-day and
housing system were included in Paper III. Variables with P-value ≤ 0.05
were considered statistically significant and were kept in the final models.
Model validation was conducted by visual examination of normal
probability plots of residuals against standardized residuals (q-q plots).
Yield loss associated with CM occurring in different stages of lactation
(Paper I) was investigated by including an interaction term between a
mastitis index and lactation stage in the general model. The mastitis index
was used to distinguish between cows with and without CM, as well as to
indicate time (test day) with respect to day of diagnosis. The clustered
nature of test-days, and declining correlation between test-days as the time
interval between them increased, were accounted for by specifying an auto-
regressive residual correlation structure within lactations (Paper I). Least-
squares means of the interaction term were used as estimates of daily yield in
a certain week of lactation at a certain time with respect to diagnosis, and
305-day yields were extrapolated from the daily estimates. Yield loss was
expressed relative to the yield of non-mastitic cows.
The association between SCC and daily milk yield in different stages of
lactation in cows free of CM was investigated in Paper III. Fixed linear,
quadratic and cubic regressions of log2-transformed and centered SCC were
nested within lactation stage. The shape of the lactation curve was described
by two fixed effects; lactation stage and weeks in milk. Moreover, a random
regression, which modelled the deviation of individual lactations from the
general lactation curve, was fitted. Daily milk yield at different SCC, in
different stages of lactation, was calculated based on the estimated regression
coefficients. Daily yield loss was expressed relative to milk yield on test-days

34
free of SCM, defined as having SCC below 50 000 cells/ml in lactation
weeks 2 to 44 and below 175 000 and 200 000 cells/ml, respectively, in
primiparous and multiparous cows in the first week of lactation. Lactational
(305-day) milk loss caused by SCM was calculated based on the regression
coefficients. Affected TD were assigned SCC corresponding to the week-
specific geometric mean of TD with SCC above the thresholds, whereas
healthy TD were assigned SCC corresponding to the week-specific
geometric mean of TD with SCC below the thresholds. Lactational milk
loss in an average lactation affected by SCM was obtained by comparing the
sum of weekly yields in affected cows, weighted by the prevalence of SCM
in each week of lactation, with the 305-day yield of healthy cows.

The SimHerd Model

SimHerd is a dynamic bio-economic model with stochastic elements. It
simulates production and associated events in a dairy herd over time through
weekly time-increments. The simulation unit in the model is the individual
animal. Herd-level production is simulated through the changes in state and
production of individual animals. The state of an animal in each week is
defined by age, parity, lactation stage, milk yield, body weight, reproductive
status (oestrus and pregnancy) and disease status. Discrete events, such as
oestrus detection, conception, sex and viability of the calf, disease
occurrences, non-voluntary culling and mortality, are triggered
stochastically. The modelling of mastitis within a lactation in Paper IV is
summarized in Figure 2.
The impact of different scenarios on herd net return was evaluated by
applying Swedish market prices to the results. A treatment cost of €119
(1 200 SEK, exchange rate of 18 November 2008, Paper IV) per incidence
of CM, including veterinary fees and antibiotics, was assumed in Paper IV.
Results were expressed as averages of 250 replicates, and were analyzed by
univariate ANOVA. Simultaneous pair-wise comparisons of scenarios were
conducted using t-tests (P < 0.05).

35
 Completed
lactation
Cure
Risk factors Culling
CM
Parity
Week in lactation Culling
Yield level Completed
Previous mastitis case lactation
Cure
Cure SCM* Culling
CM Culling
Culling
Completed
No lactation
mastitis
Culling
Mastitis
Culling
No
mastitis
Completed
lactation
Culling
SCM Culling
Cure CM Culling
Cure
Completed
Production effects lactation
Increased SCC Culling
Reduced milk yield SCM Culling
Reduced feed intake
Cure
Milk withdrawal (CM) Completed
lactation

Figure 2. A simplified description of the modelling of mastitis in SimHerd (Paper IV). Mastitis
only affected culling indirectly (....). Subclinical mastitis (SCM*) could only occur after
clinical mastitis (CM) if the yield loss incurred by SCM was higher than the current yield loss
caused by the earlier occurrence of CM. There was no upper limit as regards the number of
mastitis cases in a lactation (indicated with --- in the figure).

Main Results
Yield Loss caused by Mastitis
The magnitude of the yield loss was affected by the stage of lactation in
which the cow developed mastitis. CM gave rise to the most extensive yield
loss when cows were diagnosed in early lactation (Paper I), whereas
increased SCC caused greatest loss when it occurred in late lactation (Paper
III). Multiparous cows generally suffered more severe yield loss than
primiparous cows (Papers I and III). Cows developing CM generally had a
higher initial milk yield than non-mastitic cows (Paper I).
Daily milk yield tended to decline 2 to 4 weeks prior to CM, and, after a
case of CM, milk yield was suppressed throughout lactation (Paper I). At the
time of diagnosis, daily milk loss in primiparous cows was close to 5 kg in
Paper I, whereas it ranged from 1 to 8 kg in multiparous cows. On test-days
free of CM, daily milk loss at an SCC of 500 000 cells/ml was estimated at
0.7 to 2.0 and 1.1 to 3.7 kg in primiparous cows and multiparous cows,
respectively (Paper III). The higher figures applied to TD sampled in
lactation weeks 33 to 44. In Paper III, an increase in SCC had a certain
relationship with daily milk yield in primiparous cows irrespective of
whether the cow developed CM in the lactation or not. In multiparous

36
cows, on the other hand, an increase in SCC was associated with a higher
milk loss in lactations where the cow did not develop CM.
The relative yield loss associated with CM occurring in different weeks
of lactation is illustrated in Figure 3. On 305-day basis, primiparous cows
affected by CM suffered a yield loss in the range of 0 to 705 kg (0 to 9%),
depending on the week of lactation in which the cow was diseased (Paper
I). The most severe yield loss occurred when primiparous cows developed
CM in lactation week six. Most cases of CM occurred in the first week of
lactation, and yield loss in primiparous cows diagnosed at this point in time
amounted to 578 kg milk. In multiparous cows, lactational yield loss varied
from 0 to 902 kg (0 to 11%). The highest yield loss applied to multiparous
cows developing CM in lactation week three. When CM occurred in the
first week of lactation, yield loss in multiparous cows was estimated at 782
kg milk.
In an average lactation affected by SCM, primiparous cows suffered a
yield loss of 155 kg milk, which corresponded to 2% of their 305-day yield
(Paper III). Milk yield of multiparous cows was substantially more affected,
and, in an average lactation affected with SCM, yield loss in multiparous
cows amounted to 445 kg milk (5%).

37
 15 Primiparous cows

10

Relative yield loss (%)

5

-5

-10

-15
1 6 11 16 21 26 31 36 41
Lactation week of CM diagnosis

15 Multiparous cows
Relative yield loss (%)

10

-5

-10

-15
1 6 11 16 21 26 31 36 41
Lactation week of CM diagnosis
Figure 3. Proportional change in 305-day milk yield in primiparous and multiparous cows
diagnosed with clinical mastitis in different weeks of lactation, expressed relative to milk yield
of non-mastitic cows (Paper I).

Economic Loss associated with Mastitis

The maximum avoidable cost of CM (i.e. the increase in net return if the
initial incidence could be reduced by 90%) in a Swedish 150-cow dairy herd
was estimated at €14 504 per year (Paper II). In Paper IV, the yearly
avoidable cost of mastitis (CM and SCM) in a herd of the same size was
estimated at €8 095, under the assumption that the initial incidence of
mastitis could be reduced by 50%. Assuming that the relationship between

38
mastitis incidence and herd net return was linear, which appeared to be the
case in Paper II, this would correspond to a maximum avoidable cost of
€14 571. The avoidable cost of mastitis estimated in Paper IV corresponded
to 5% of the herd net return given the initial incidence of mastitis. Figure 4
shows the changes in various cost items that resulted from the 50%
reduction of the initial incidence of mastitis in Paper IV.

6000

4000
Change (€)

2000

0
Milk sales

Culled cows

Sold heifers

Sold bull calves

Feed, cows

Feed, heifers

Vet and drugs

Inseminations

Other costs, cows

Other costs, heifers

Replacement heifers

Interest
-2000

-4000
Figure 4. Changes in various cost components as a consequence of a 50% reduction of the
initial incidence of mastitis (Paper IV).

When the economic consequences of CM were investigated, an average

economic loss of €428 per case was arrived at (Paper II). In Paper IV, where
the economic impact of CM and SCM were assessed simultaneously, an
estimate of €275 was obtained. These figures are not comparable, because
the estimate from Paper II accounts for correlated effects of CM on SCM,
whereas the estimate from Paper IV does not. The average economic loss
per case of SCM was estimated at €60 (Paper IV).
The average economic loss per case of CM was only slightly affected by
the strategy for modelling yield loss when mastitic cows’ own yield level,
had they not developed CM, was used as the reference for production in
healthy cows when the yield loss was estimated. When the same reference

39
level for yield in healthy cows was used, applying specific yield-loss patterns
for different periods in lactation did not have any substantial impact on herd
results, as compared to using just one yield-loss pattern irrespective of the
lactational timing of CM. Neither did the modelling strategy affect the
response to a reduction of the risk of CM. Differences between modelling
strategies were more pronounced when yield of non-mastitic cows was used
as the reference for production in healthy cows.

Discarding Milk with High SCC

Discarding milk with high SCC was never profitable, because sorting of
milk generated a substantial amount of milk withdrawal that was not offset
by a sufficient increase in milk price. The most negative impact of
discarding milk with high SCC on net return per cow-year was,
consequently, observed when high incidence of mastitis was simulated. Milk
sorting based on SCC measured with low uncertainty reduced the amount
of withdrawn milk, and thus had less negative effect on net return per cow-
year as compared with when SCC was measured with high uncertainty.

40
General Discussion

It is generally accepted that mastitis is the most costly disease in dairy

production. The results of this thesis clearly emphasize the economic
importance of mastitis, and stress the need for strategies to limit the impact
of the disease. In the following, the most important findings will be
discussed in detail.

Yield Loss caused by Mastitis

Severity
Clinical mastitis was associated with higher milk loss than SCM (Papers I
and III). In primiparous cows, CM occurring in early lactation gave rise to a
yield loss that was three times the size of the milk loss in an average lactation
affected by SCM. In multiparous cows, the reduction in milk yield
following CM in early lactation was twice as large as that estimated in an
average lactation affected by SCM.
CM is the more severe form of mastitis, and a higher yield loss resulting
from CM as compared with SCM is, thus, a logic finding. CM is associated
with a larger amount of affected udder tissue, which results in more severe
udder damage with decreased synthetic capacity of the mammary gland as a
consequence.

Lactation Stage
The timing of mastitis in lactation had profound impact on the magnitude
of yield loss (Papers I and III). Milk yield in cows developing CM before
peak yield was most severely affected (Paper I). Partly, this can be expected
to be due to the effects of CM being experienced over a longer period of
time. Udder-tissue damage caused by CM in early lactation will, thus,

41
suppress milk yield throughout lactation (in accordance with the findings in
Paper I). Also, it may be speculated that CM in early lactation obstructs the
differentiation of secretory cells, which takes place in this period and is
responsible for the increase in milk yield until peak lactation (Capuco et al.,
2001), thus resulting in more severe reduction of milk yield. Another
explanation of the high yield loss associated with early occurrences of CM is
related to that, in this stage of lactation, milk yield is increasing. High
production can cause impairment of the immune system due to metabolic
stress (Knegsel et al., 2007). When cows are in negative energy balance,
body fat is converted to ketone bodies, and hyperketonemia has been
suggested to be one of the most important factors causing impairment of the
udder defence mechanisms (Janosi et al., 2003; Suriyasathaporn et al., 2000).
It is likely that the impaired immune system in cows in early lactation results
in reduced ability to battle infection, which contributes to the more severe
depression of milk yield associated with CM. Cows are in negative energy
balance until approximately weeks 7 to 9 of lactation (Suriyasathaporn et al.,
2000), which coincides with the stage of lactation in which CM gives rise to
the most extensive yield loss.
Increased SCC had the most detrimental effect on milk yield in late
lactation. It has been suggested (Hortet et al., 1999), that poorer udder
health status, caused by increased exposure to pathogens, increased
prevalence of infection, and subsequent permanent glandular damage from
previous infections, is responsible for the higher milk loss associated with an
increase in SCC in late lactation. In Paper III, this explanation was tested
using a sub-dataset containing only test-day records sampled in lactations in
which CM did not occur and that were not immediately preceded by a
lactation affected by CM. The obtained regression coefficients were similar
to those estimated in the original analyses, and it was, therefore, concluded
that milk loss related to an increase in SCC was highest towards the end of
lactation, irrespective of whether cows had a history of CM or not. A
different explanation for the higher yield loss associated with high SCC in
late lactation was proposed: the fact that the udder is in a catabolic state.
The degenerative process taking place might influence both the udder’s
ability to repair itself after infection and the compensatory ability of
uninfected quarters. The compensatory ability of uninfected quarters has,
indeed, been found to be lower in late than in early lactation (Hamann &
Reichmuth, 1990). Hamann and Reichmuth (1990) argued that the
potential for yield compensation is related to the number and activity of
secretory cells, supporting the above reasoning.

42
Parity
Parity of the cow was another important determinant of the extent of yield
loss caused by mastitis (Papers I and III). Multiparous cows consistently
suffered more severe yield loss than primiparous cows. This might be
explained by multiparous cows having a higher yield level, and that they,
simply, have more milk to lose. The relative yield loss was, however, also
larger in multiparous cows, which indicates that this is not the full
explanation. It may be hypothesized that the immune system becomes less
efficient as cows get older, which might result in less complete cure of
mastitis, with more extensive tissue damage and more severe yield loss as a
consequence.

Production Level
Cows that developed CM tended to have higher initial milk yield than cows
without CM (Paper I). The fact that high-yielding cows are more likely to
develop CM further adds to the economic damage of the disease, and might
suggest that the estimates reported in Paper I possibly underestimated the
actual yield loss caused by CM. In Paper I, yield loss was expressed relative
to the production level of cows without CM. Cows that developed CM in
lactation weeks 1 to 8 and 9+ produced 2.5 and 0.8 kg ECM per day,
respectively, more milk than their non-mastitic herd mates prior to three
weeks before diagnosis (Paper II). Considering that most cases of CM
develop in early lactation, these figures correspond well with the production
advantage of mastitic cows of 2.6 kg milk reported by Wilson et al. (2004).
In view of this, it might actually be reasonable to assume that another 2.5 kg
milk could be added to the daily yield loss estimated in Paper I without
overestimating the yield loss caused by CM.
Most cows in a herd are in second or later lactations. Considering that
most CM cases occurred in multiparous cows (Paper I), and that the median
SCC in later parts of lactation in multiparous cows was more than twice the
size of that in primiparous cows (Paper III), it can be concluded that the
majority of milk loss caused by mastitis in a herd is suffered by multiparous
cows. Furthermore, most of this yield loss is likely to occur in early and late
stages of lactation.

Importance of Mastitis to Herd Profit

The yearly avoidable cost of mastitis in a Swedish 150-cow herd, assuming
that the initial incidence of mastitis could be reduced by 50%, corresponded
to 5% of the herd net return in the initial situation (Paper IV). Mastitis was

43
thus of considerable importance to the economic performance of the herd,
confirming findings of Hansson (2007) who reported that average milk
production per cow-year and incidence of mastitis were the only significant
indicators of economic efficiency in Swedish dairy herds among several
investigated predictors (incidence of mastitis, average milk production per
cow-year, average protein content of milk, average herd fertility, and
involuntary culling rate).
Main components of the increase in net return when the incidence of
mastitis was reduced were increased income from milk sales, because of
higher production per cow-year and higher milk price as a result of an
increased proportion of delivered milk with an SCC below 200 000
cells/ml, and reduced veterinary costs (Paper IV, Figure 4). When all cases
of CM were eliminated, income from milk sales increased substantially more
than when no SCM cases were modelled. Also, CM was always associated
with veterinary costs, and, when all cases where eliminated, these costs were
avoided. It is thus clear from Paper IV that CM is more influential on herd
profit than SCM.

Implications for Mastitis Control

The impact of mastitis on herd-level economy can be reduced by various
interventions, which are of proactive or reactive nature. Measures aiming at
preventing new cases of mastitis from occurring include breeding,
improvement of milking hygiene, implementation of post-milking teat
disinfection, regular controls of the milking equipment, implementation of
milking order, and improvement of bedding material. Antibiotic treatment
protocols, dry cow therapy, and culling regimes are examples of actions that
are taken in order to limit the impact of mastitis once it has occurred.
Discarding milk with high SCC, in order to increase the price of the
delivered milk, can also be regarded as a means to reduce the economic
damage caused by mastitis.

Prioritization between Cows

The most extensive yield loss occurred in multiparous cows diagnosed with
CM in early lactation. To make matters worse from an economic point of
view, most cases of CM developed in this category of cows (Paper I).
Preventive measures should, therefore, focus on reducing the incidence of
mastitis in multiparous cows at early stages of lactation. This can be
attempted by enhanced mastitis control in primiparous cows, because cows

44
that have suffered a previous case are at substantially increased risk of
developing subsequent cases (Steeneveld et al., 2008; Bendixen et al., 1988).

Economic Framework
Mastitis control incurs additional costs for the milk producer in terms of
investments and extra labour requirement. Estimates of the avoidable cost of
mastitis, which expresses the maximum viable expenditure on preventive
measures per year, are thus necessary to determine the amount of money
that can be invested into mastitis control. In a Swedish 150-cow herd with
initial incidences of CM and SCM of 32 and 33 per 100 cow-years,
respectively, € 8 095 could be spent on mastitis control on a yearly basis, if
that investment results in a 50% reduction of the incidence of mastitis (Paper
IV). This amount of money equates to 1.2 hours of external labour per day
(at a cost of €17.8 per hour (Agriwise, 2008), which can be spent on
improving management routines. Alternatively, €57 000 could be invested
in improved milking equipment which might contribute to better udder
health status, or be invested in technical tools, such as an in-line SCC
indicator, which would facilitate monitoring of udder health in the herd.
The maximal investment is calculated by the annuity method, with an
interest rate of 7% per year and a depreciation period of ten years.
The costs associated with implementation of different preventive
measures differ. Moreover, different preventive measures are not equally
effective in reducing the incidence of mastitis. The profitability of a certain
control strategy is thus determined by the difference between its cost and
the value of the reduction in mastitis incidence that it can achieve. To
further complicate matters, the efficiency of different approaches to control
mastitis varies depending on milking system. For example, Yalcin et al.
(1999) showed that dry-cow therapy reduced SCC in herds milking in
parlour, but not in herds with tie-stalls. Further studies are therefore needed
to answer which preventive measures are most appropriate in the context of
a specific herd.
Control expenditures arise as a consequence of a mastitis problem and
should therefore be attributed to mastitis (McInerney et al., 1992). Thus, in
order to obtain an estimate of the total economic cost of mastitis, costs of
prevention should be added to the estimates of economic loss proposed in
this thesis.

45
Different Estimates Support Decisions at Different Levels
Mastitis control is practiced at different levels. The nature of the decision
that is to be supported determines which estimate of the economic loss
associated with mastitis that is of interest in a certain situation.
Cow-level decisions most often concern which actions that should be
taken after a case of mastitis has been confirmed. In this case, the economic
loss per case is the most relevant estimate. The expected extent of yield loss
must then be taken into consideration, because it constitutes the major
component of economic loss caused by mastitis.
The avoidable cost of mastitis at herd level is pertinent when the
economic viability of mastitis control programs is to be assessed.
When economic weights of traits in the breeding goal are calculated, it is
the avoidable economic cost associated with mastitis per cow-year that is the
most appropriate input. It can be debated whether yield loss should be
included in the economic loss caused by mastitis in this context, because if
yield loss is already part of the genetic evaluation for milk yield, i.e. the milk
yield of an affected cow is considered her production potential, then
including yield loss also when deriving the economic weight of mastitis
would double-count this effect. Milk loss is, however, a consequence of
mastitis, and neglecting it in the genetic evaluation might underestimate the
economic importance of mastitis.

Economic Loss per Case of Mastitis

Inconsistency of Estimates
Considerably different estimates of the average economic loss per case of
CM were obtained in Papers II and IV. These figures are, however, not
directly comparable. The model in Paper IV was a development of that in
Paper II and there were, consequently, some differences in the underlying
assumptions. All of these contributed to the lower estimate of average
economic loss per case of CM obtained in Paper IV.
The cost of extra labour requirement associated with treatment of CM
was included in the fixed economic loss per case of CM (veterinary and
treatment costs) in Paper II, but not in Paper IV. Which approach that is
most correct depends on the assumed opportunity cost of labour, which, in
turn, depends on the source of labour. Whether or not extra labour has
been included in the economic loss caused by mastitis in previous studies
has, indeed, varied (Table 4). SimHerd does not consider labour as a variable
cost (Østergaard et al., 2005), and we, therefore, found it most suitable not

46
to include extra labour requirement in the economic loss associated with
CM when the economic parameters were updated for the analyses
conducted in Paper IV.
Milk sorted out due to high SCC was produced by cows without CM,
and such milk was modelled to be fed to calves, thus, reducing the need for
milk replacer. It was, consequently, assigned alternative value. SimHerd did
not distinguish between this milk and milk not sold due to treatment of
CM. As a result of this, milk produced in the withdrawal period was
assigned alternative value in Paper IV, but not in Paper II. In this context, it
needs to be stressed that feeding milk produced by cows receiving treatment
to calves is not a recommended practice, because it increases antibiotic
resistant bacteria in the lower gut of calves (Langford et al., 2003).
Another, and probably the most important, explanation for the
differences is that Paper II addressed the impact of the first lactational
incidence of CM whereas Paper IV considered recurrent cases within the
same lactation. When multiple cases developed, yield loss was defined based
only on the impact of the most recent case. For instance, if a cow developed
CM in lactation week two, and then again in lactation week four, the yield
loss in lactation week five would be related only to the case occurring in
lactation week four. The case developing in lactation week two would,
thus, only incur yield loss in lactation weeks two and three. As a
consequence, the average yield loss per case of CM was lower in Paper IV,
551 kg ECM, as compared with 797 kg ECM in Paper II. Considering that
the average price per delivered kg ECM was €0.288 in Paper IV, the value
of a 246 kg ECM lower yield reduction equates to €71. If the average yield
loss would have been the same in Paper IV as in Paper II, €71 should be
added to the estimated average economic loss per case of CM, and a new
estimate of €346 (compared with €428 in Paper II) would be arrived at.
Also, yield loss in the weeks before clinical onset was attributed to CM in
Paper II. This effect was assumed to be caused by SCM in Paper IV and
was, thus, not included in the yield loss caused by CM.
Based on these arguments, it is relatively safe to assume that the actual
average economic loss per case of CM lies somewhere between €275 and
€428. The former estimate was, however, derived from a more complete
model, where recurrent cases of CM as well as SCM were adjusted for. The
importance of representing effects of both CM and SCM when assessing the
impact of mastitis in a dairy herd is stressed by the fact that estimates of net
return per cow-year are highly sensitive to the severity of mastitis cases
(Østergaard et al., 2005), and by subclinical cases being responsible for a
large proportion of the total economic loss associated with mastitis (Huijps et

47
al., 2008; Yalcin, 2000). The estimate obtained in Paper IV is, consequently,
likely to be more correct and the average economic loss of a case of CM
will, in the remaining sections, be assumed to be €275.

Consistency with Previously Published Estimates

Estimates of the cost per case of mastitis, published since 1990, are
summarized in Table 3. The average estimate is presented when studies
reported several estimates. Studies based on old data (e.g. McInerney et al.,
1992) or concerning non-lactating cows (Hillerton et al., 1992) are not
included. Estimates have been adjusted for the effect of inflation by the use
of consumer price indices (Organization for Economic Co-operation and
Development, 2009) and are expressed in the price level of 2007, which
agrees with the price level applied in Paper IV. Estimates originally
expressed in other currencies have been converted to Euro (exchange rate of
18 November 2008, Paper IV) to facilitate comparison of results.

Table 3. Estimates, published since 1990, of the cost per case of clinical (CM) and subclinical mastitis
(SCM) in lactating cows
Reference Country Severity Method Cost per case (€)
Paper IV Sweden CM SS1 275
Paper II Sweden CM SS 413
Bar et al. (2008b) US CM DP2 146
Huijps et al. (2008) The Netherlands CM PB3 205
Wolfová et al. (2006) Czech Republic CM PB 71
Østergaard et al. (2005) Denmark CM4 SS 360
Kossaibati & Esslemont (1997) UK CM5 PB 519
Sandgren & Emanuelson (1994) Sweden CM PB 350
Miller et al. (1993) US CM PB 142
Belotti (1991) Sweden CM4 PB 420

Paper IV Sweden SCM SS 60

6
Steeneveld et al. (2007) The Netherlands SCM SS 1157
8
Swinkels et al. (2005a) The Netherlands SCM PB 116
6,9
Swinkels et al. (2005b) The Netherlands SCM PB 130
1
Stochastic simulation; 2dynamic programming; 3partial budget; 4assuming that SCM is a natural
part of CM; 5severe case; 6Streptococcus uberis; 7without treatment; 8Staphylococcus aureus; 9Streptococcus
dysgalactiae.

The cost per case of mastitis differed between studies: estimates of the cost
of a case of CM ranged between €71 and €519. Thus, compared to findings

48
in the literature, the average economic loss of €275 per case of CM
obtained in Paper IV is a relatively intermediate value.
Cost components included in the total cost of mastitis differed widely
between studies (Table 4), which probably explains a substantial part of the
large variation in estimates. Veterinary costs, drugs, discarded milk, yield
loss, and culling were frequently included, whereas other cost components,
such as effects on milk composition, mortality, and spread of infection to
healthy cows, were more infrequently accounted for. None of the reviewed
studies considered increased risk of subsequent disorders (other than mastitis)
or reduced fertility.

Table 4. Cost components considered in studies, published since 1990, estimating the cost of mastitis in
lactating cows
Considered consequences of mastitis

Discarded milk

Contagiousness
Extra labour

Prevention
New case
Yield loss

Mortality

Culling
Drugs

SCC
Feed
Vet

Reference
Paper IV1 X X X X X X X X X
Paper II X X X X X X X X X
Bar et al. (2008b) X X X X X (X)2 (X)2
Huijps et al. (2008) X X X X X X X
Wolfová et al. (2006) X X X X X3
Østergaard et al. (2005) X X X X X X X X X
Kossaibati & Esslemont (1997) X X X X X
Sandgren & Emanuelson (1994) X X X X X
Miller et al. (1993) X X X X X X X
Belotti (1991) X X X X X X X

4
Paper IV X X X X X X
5 6
Steeneveld et al. (2007) X X6 X6 X X X
6 6
Swinkels et al. (2005a) X X X X
Swinkels et al. (2005b) X X X
1
Clinical mastitis; 2not included in cost, but accounted for in model; 3dry-cow therapy; 4subclinical
mastitis; 5without treatment; 6applicable to clinical flare-up.

Earlier Swedish studies have arrived at higher estimates of the economic loss
per case of CM than Paper IV. These investigations used a simpler analytical
approach, partial budgeting, and herd dynamics were, consequently,
neglected. Differences are likely to arise as a consequence of dissimilar

49
analytical approaches applied. It has been pointed out (Bar et al., 2008b),
that static models, such as partial budgets, often overestimate the economic
loss associated with an average case of mastitis. The cost of mastitis is not the
same in all cows, even if they are of the same parity and develop mastitis in
the same stage of lactation. Mastitis affecting a cow that should be replaced
anyway is, in practice, associated with a lower cost, whereas mastitis in cows
of high future expected income is more costly (Bar et al., 2008b). The
expected value of a mastitic cow will influence whether she receives
treatment, which, in turn, will affect the magnitude of yield loss suffered.
Dynamic models, therefore, often include a lower average yield loss per case
of mastitis. Culling of mastitic cows is, to a large extent, a voluntary decision
made by the farmer, and depends on the lactational timing of CM and
pregnancy status of the cow (Schneider et al., 2007). Such interactions are
not easily accounted for in static models. Static models attributing treatment
costs, yield loss, and costs associated with increased risk of culling to every
case of mastitis might, thus, overestimate the economic loss.
It is difficult to make unbiased inferences about the impact of
methodology on estimated cost per case of mastitis based on the summarized
studies, because most of them obtained their estimates by means of partial
budgeting. Studies applying stochastic simulation or dynamic programming
arrived at estimates that fell within the range of estimates in studies using
partial budgeting, and variability was thus larger within a specific analytical
approach than between analytical approaches. Other factors, such as the
economic value assigned to each component of loss included in the cost per
case of mastitis, therefore seemed to be more important than the analytical
approach applied.
Estimates of the cost per case of CM in Western Europe were higher
than those in the US, which was somewhat unexpected due to the absence
of a milk-quota system in the US. Mastitis is more costly in a non-quota
situation, because the reduction in milk yield subsequent to mastitis directly
affects herd net returns by means of reduced milk sales. Where a quota
system is in place, farmers usually compensate the lower average production
brought about by mastitis with keeping cows in the herd beyond their
intended culling date (Swinkels et al., 2005a). In Papers II and IV, no milk
quota was modelled even though it is well known that the economic loss
associated with mastitis depends on the presence or absence of a milk quota,
and despite the fact that milk is produced under a quota system in Sweden.
The amount of milk produced in Sweden is, however, not enough to fill
the quota (Swedish Dairy Association, 2009), and Swedish milk production,
thus, functions as if no milk quota was present.

50
 Differences between studies conducted in different spatiotemporal
contexts can, at least to some extent, also be explained by different pathogen
frequencies. For instance, CM caused by environmental pathogens
(primarily E. coli) have been reported to be more costly than CM caused by
contagious pathogens (Miller et al., 1993). Swinkels et al. (2005a; 2005b)
did, however, estimate relatively similar cost per case of SCM, irrespectively
of whether Staph. aureus, Strept. uberis or Strept. dysgalactiae was the causative
agent.
Different circumstances of production, management routines, and
attitudes towards the use of antibiotics are other factors likely to contribute
to differences between estimates obtained in different countries. Likewise,
prices and costs can be expected to differ. Indeed, the cost per case of CM
has been found to depend on exogenous factors, primarily the milk price
(Bar et al., 2008b): when a 20% higher milk price was modelled an 18%
higher cost per case of CM was reported.
All reviewed studies investigating the cost per case of SCM originated
from the Netherlands and were conducted closely in time (Table 3). They
also arrived at very similar estimates. The estimate of €60 per average case of
SCM obtained in Paper IV was about half the size compared with those
calculated by Steeneveld et al. (2007) and Swinkels et al. (2005a; 2005b).
These studies investigated the cost of SCM caused by specific pathogens,
whereas Paper IV did not take the causative pathogen into account. More
importantly, different definitions of SCM were applied. In the three Dutch
studies, SCM was defined as two out of three consecutive milk samples,
taken at 3 or 4-week intervals, with SCC > 250 000 cells/ml. This
definition requires that SCC is elevated for at least seven weeks, and, if it
were to be applied in Paper IV, no cases of SCM would have occurred in
primiparous cows because of the insufficient response in SCC. Modelled
cases of SCM in Paper IV were, consequently, milder than those
investigated by Steeneveld et al. (2007) and Swinkels et al. (2005a; 2005b),
and a lower estimate of average economic loss per case is, thus, reasonable.

Economic Loss caused by Mastitis in Sweden

There are 350 000 dairy cows in Sweden (Swedish Board of Agriculture,
2008a). Considering the estimated avoidable cost of mastitis of €54 per
cow-year (Paper IV), the national avoidable cost can roughly be estimated at
€19 millions per year. This figure does not include costs of prevention and
costs that mastitis incurs to the processing industry. It is thus not an estimate
of the total cost of mastitis, and, as such, it must be interpreted with care.

51
The proposed estimate does, however, give a hint of the extent of the
economic damage caused by mastitis each year.
No comparison of the national economic costs due to different
production diseases in cattle has been conducted under Swedish
circumstances of production. A British study (Bennett et al., 1999) has,
however, investigated the economic impact of numerous diseases
simultaneously, and came to the conclusion that mastitis, because of its high
incidence and high financial impact per affected cow, had the largest
economic impact on cattle production in the UK.

Discarding Milk with High SCC

The possibility that discarding milk with high SCC could reduce the
economic loss associated with mastitis in dairy herds was refuted in Paper
IV. Discarding milk with high SCC was never profitable in herds with
average BTSCC between 158 000 and 236 000 cells/ml, because substantial
amounts of milk had to be discarded. That generated an economic loss that
was not offset by the increase in the milk price. It should be pointed out
that discarding milk with high SCC was not economically justified, even
though the discarded milk was assigned alternative value and extra labour
costs were neglected. If discarded milk were not fed to calves, or if sorting
of milk would require some monitoring, then the impact of discarding milk
with high SCC would have been even more negative.
The results of Paper IV suggest that the quality of delivered milk has less
impact on herd net return than the amount of delivered milk, supporting
the findings of Hansson (2007) who concluded that milk quality was not a
significant indicator of economic performance in Swedish dairy herds
whereas milk yield per cow was. The current Swedish milk-pricing system
does, consequently, not serve as a strong motivation to discard milk with
high SCC. Premiums and penalties are, however, an important motivation
for farmers to improve udder health management (Valeeva et al., 2007), and
the current milk-pricing system is, thus, likely to stimulate efforts to reduce
BTSCC. Some concerns regarding the current milk-pricing system are,
nevertheless, raised by the fact that the impact of discarding milk with high
SCC on net return per cow-year was most negative when the incidence of
mastitis was high. The extra premiums received when delivering milk with
lower SCC did not compensate for the economic loss incurred by
discarding milk. If the milk-pricing system is supposed to serve as an
incentive to farmers to improve the quality of delivered milk, then
premiums and penalties need to be revised. This, of course, invokes the

52
issue of designing a new milk-payment scheme that would make it
profitable to discard milk with high SCC. It needs to be stressed that it is
not recommended to proceed with making changes in the milk-pricing
system until scientifically supported regulatory limits have been proposed.

Methodological Issues
Reference Level for Yield in Healthy Cows
Cows that develop CM tend to have a production advantage over non-
mastitic cows before diagnosis (Gröhn et al., 2004; Wilson et al., 2004;
Rajala-Schultz et al., 1999b). In Paper I, the seemingly higher 305-day
yields of cows developing CM in mid or late lactation were most certainly a
consequence of mastitic cows producing above the level of non-mastitic
cows to such an extent that their cumulative 305-day yields were higher,
even though they had suffered from CM towards the end of lactation.
Therefore, using the yield of non-mastitic cows as the reference for
production in healthy cows will underestimate the true yield loss caused by
CM.
In future studies aiming at assessing the impact of mastitis on milk yield,
it is recommended to account for the higher initial milk yields of mastitic
cows. One means to do this is to express yield loss relative to cows’ pre-
mastitic milk yield. This approach does, however, impose problems because
most cases of CM develop in early lactation where no information on pre-
mastitic yield is available. This can, to some extent, be overcome by using
the yield level in the previous lactation(s) for multiparous cows, and
accounting for the expected yield change to this lactation. An appealing
possibility would be to correct for a cow’s genetic and permanent
environmental effect for milk-yield in the models, and, thus, obtain
estimates that are less biased due to the relation between milk yield and
mastitis.

Recurrent Cases of CM
Cows that have suffered from a case of mastitis are at increased risk of
developing subsequent cases (Rajala & Gröhn, 1998; Houben et al., 1993).
Still, the average number of CM cases per lactation was only 1.3 in the
dataset studied in Paper I. Only the impact of the first lactational incidence
of CM was, however, studied. By only including the first incidence, the
entire lactational yield loss was attributed to just one case. Thus, in lactations
with recurrent cases, the loss assigned to the first case will be overestimated.
To obtain less biased estimates of the yield loss caused by a single case of

53
CM, lactations with multiple cases could have been excluded from the
analyses.

Estimates of Yield Loss Obtained from One Herd

Yield losses estimated in Papers I and III were based on a dataset sampled in
only one herd. In comparison with average Swedish figures, the herd has
low BTSCC but high incidence of CM. The studied herd was, however, a
research herd, and the surveillance of animals was probably more extensive
than what can be expected in commercial dairy herds. As a consequence of
this, more CM cases than normally were likely detected. Also, some of the
reported cases were presumably milder than those reported by practicing
veterinarians, which might have influenced the size of the estimated yield
loss. The estimates obtained in Paper I can, therefore, be utilized in
decision-support without any risk of overestimating the yield loss caused by
CM.
Among the pathogens present in culture-positive milk samples taken in
the research herd, a relatively large proportion was of environmental origin.
As different pathogens give rise to different extent of yield loss, it might be
argued that the estimates are not applicable in the average Swedish herd,
where environmental pathogens, generally, are less frequent (Persson Waller
et al., 2009). The sampling procedure in the research herd can, however,
not be compared with that applied in commercial herds. Milk samples for
bacteriological culture were collected by routine at pre-determined weeks in
lactation, and not only when clinical symptoms were observed or when
cows were suspected to be subclinically infected. It may be speculated that
the bacteriological findings do, indeed, represent a pattern of infection that
is representative also in other herds, but that is currently undetected due to
the way in which sampling normally is conducted.

Definition of SCM
From a management perspective, estimates of the lactational yield loss
caused by SCM occurring in different stages of lactation would have been of
great interest, because such information could provide insight into the
expected economic benefit that could be achieved if proper actions were
taken. In order to obtain such estimates, however, a definition of a case of
SCM would have been required. Subclinical mastitis does, by definition, not
give rise to visible symptoms, and cannot be diagnosed by the milkers.
Instead, SCM is defined based on SCC or other indicators. SCC is measured
on a continuous scale, and, even though SCM is a dynamic condition,

54
different threshold values are usually used to distinguish between healthy
and affected cows.
In Paper III, a threshold of 50 000 cells/ml was used to distinguish
between healthy TD and TD affected by SCM in lactation weeks 2 to 44.
This threshold was suggested by Hortet and Seegers (1998a), and was based
on a review of studies on SCC in bacteriologically negative cultures; i.e.
Laevens et al. (1997) who reported an SCC of 49 000 cells/ml in composite
milk samples and Schepers et al. (1997) who found SCC around 14 000
cells/ml in udder quarter foremilk samples. Higher thresholds of 175 000
and 200 000 cells/ml in primiparous and multiparous cows, respectively,
were applied in the first week of lactation (Paper III), because SCC is
elevated in the colostrum period irrespective of whether the cow is affected
with mastitis or not (Dohoo & Meek, 1982). These thresholds were defined
based on bacteriological cultures of milk samples, taken by routine in the
research herd in the first week of lactation from the fall of 1997.
A threshold of 200 000 cells/ml is commonly applied to distinguish
between healthy cows and cows affected by SCM. With respect to the
findings of the previously mentioned studies, such a threshold is far too
high. Also, the results obtained in Paper III indicate that milk yield is
reduced on TD with an SCC of 200 000 cells/ml, as compared to TD with
an SCC of 50 000 cells/ml. Applying a threshold of 200 000 cells/ml will,
consequently, result in that the importance of SCM is underestimated, and
that proper measures are not taken to avoid unnecessary milk loss in affected
cows.
A clear drawback with a definition of SCM based only on SCC at a
single TD is that SCC is affected by a number of factors. Infectious status of
the mammary gland is clearly the most important one, but season, stress and
management mishaps also assert an effect on SCC (Reneau, 1986; Dohoo &
Meek, 1982). Individual TD results are therefore rather inconclusive.
In order to increase the accuracy of diagnosis, it has been suggested (De
Haas et al., 2004; Reneau, 1986; Dohoo & Meek, 1982) that SCM should
be defined based on measurements of SCC taken over a series of
consecutive TD. In Sweden, “udder disease status”, based on SCC
measured at three monthly test milkings, is used as a prognostic tool to
identify individual cows that are likely to be affected by infectious mastitis
(Brolund, 1985). De Haas et al. (2004) derived the SCC-patterns of
clinically infected cows based on 3 to 5 TD results, recorded with intervals
of three or four weeks. With such long sampling intervals, actually detecting
an increased SCC will depend on when, with respect to sampling, infection
occurred, but also on the duration of the increase in SCC, as pointed out by

55
De Haas et al. (2004). If this method were to be applied to SCM, subclinical
infections could cure spontaneously or turn into a clinical infection before
diagnosis based on a completed SCC-pattern could be made. A shorter time
interval between TD is, thus, preferable if diagnosis of SCM is to be based
on SCC-patterns. Patterns should preferably be based on daily observations,
because SCC fluctuates from day to day (Reneau, 1986), and even between
between milkings on the same day (Olde Riekerink et al., 2007b).

Strategies for Modelling Yield Loss

Using specific yield-loss patterns for different periods in lactation or just one
yield-loss pattern, irrespective of the lactational timing of CM, only
marginally affected the estimated economic impact of CM (Paper II). This
was particularly evident when mastitic cows’ own yield level, had they not
developed CM, was used as reference for production of healthy cows when
yield losses were estimated. The lack of effect of modelling strategy on
average economic loss per case of CM was probably due to that the average
yield loss did not change much as most CM cases were modelled to occur in
early and late lactation. In these periods, yield loss modelled by specific
yield-loss patterns in different periods of lactation was higher and lower,
respectively, compared with the yield loss modelled when one yield-loss
pattern was used throughout lactation. This is likely to have evened out any
difference between the modelling approaches.
The results of Paper II, consequently, suggest that using one yield-loss
pattern, irrespective of when in lactation CM occurs, is good enough, and
that this approach to assess yield loss is adequate for application into
decision-support systems.

Method of Economic Analysis

The analytical approach is often suggested to be one of the reasons for the
variation in estimates of cost of mastitis obtained in different studies. The
aim of the studies reviewed earlier was to estimate the cost of mastitis, and
none of them, therefore, investigated a possible impact of methodology on
the result. Stochastic modelling is suggested to be the most relevant method
for assessment of economic impact of disease in dairy cattle because of its
capability to account for herd dynamics. Simulation models are, however,
not transparent, and are therefore not a very appealing tool for advisors.
Partial budgets, on the other hand, are transparent, but are sometimes
argued not to provide reliable estimates because of their inability to account
for stochastic or dynamic elements.

56
 Paper II provided estimates of the average economic loss per case of CM
estimated by stochastic simulation. To provide some insight into the impact
of methodology on economic loss per case of CM, the economic loss of a
case of CM will now be calculated using a simple partial budgeting
approach. As far as possible, cost components will be assigned the same
values as in Paper II. The estimate obtained in Paper II applied to an average
case of CM. The extent of yield loss caused by CM shows large variation
depending on in which week of lactation the cow is diseased (Paper I). In
the partial budget analysis it is therefore assumed that CM occurred in the
first lactation week, because this was the week in which most cows
developed CM in the dataset used in Paper I. Furthermore, as most cows in
Paper II (65%) were simulated to be in their second or later parity, the
economic loss of a case of CM in a multiparous cow is estimated.

Extra Costs
When a cow is diagnosed with CM and treatment is necessary, a
veterinarian is consulted (required by Swedish legislation). An average fixed
cost of €133 per case, including extra labour requirement, veterinary costs
and drugs, was assumed. Clinical mastitis increases the risk of culling, and,
thus, incurs replacement costs. In previous studies, it has been assumed that
15 to 20% of CM cases result in culling (Huijps et al., 2008; Sandgren &
Emanuelson, 1994). The cost of culling can, roughly, be estimated as the
cost of a replacement heifer minus the slaughter value of the culled cow.
The value of a heifer is €952, and the meat from a slaughtered cow
generates an income of €0.86 per kg live weight. Assuming that 15% of CM
cases leads to culling and that live weight is 630 kg, extra costs associated
with increased culling can be calculated as 0.15 x (€952 - €0.86 x 630) =
€62. Extra costs associated with a case of CM in a multiparous cow in
lactation week one are, consequently, €133 + €62 = €195.
There are a couple of other aspects as regards culling due to mastitis that
ought to be considered. Replacement animals are almost exclusively heifers,
and primiparous cows have lower lactational milk yields than multiparous
cows. By replacing an older cow with a younger, the production advantage
of the older cow is lost. The value of the extra volume of milk that would
have been produced if the older cow had not been culled should be
included in the economic loss caused by CM. On the other hand, a heifer is
on average genetically superior to the culled cow. Finally, a heifer is not
always available when a cow must be replaced. If so, a stall will be empty
and the production potential of the herd will not be fully utilized. It might
be argued that the loss in income as a consequence of reduced capacity

57
utilization should be attributed to CM. These considerations will, however,
be ignored for reasons of simplicity.

Returns Foregone
A multiparous cow developing CM in the first week of lactation suffers a
yield loss of 782 kg milk (Paper I). Considering an average price per kg milk
of €0.29, the economic loss caused by reduced milk yield amounts to €227.
When the cow is diagnosed with mastitis, she receives antibiotics. Neither
the milk produced during the treatment period nor the milk produced
during the waiting period following treatment can be sold. The total
withdrawal period subsequent to treatment usually lasts for eight days. The
average daily milk yield in a multiparous cow diagnosed with CM in the
first week of lactation is about 17 kg (Paper I), and the cost of the discarded
milk is, consequently, €39. This is actually an underestimation of its
economic value, because discarded milk is produced by the cow and is,
consequently, associated with feed costs. The value of discarded milk is,
thus, higher than that of lost milk yield. In spite of this, all milk not
delivered will now be assigned the same value, resulting in an economic loss
caused by returns foregone of €227 + €39 = €266.

Additional Returns
Milk produced in the withdrawal period is sometimes fed to calves. In this
case, the discarded milk has an alternative value corresponding to the value
of the milk replacer saved. The economic loss associated with discarded milk
would then, obviously, be less. Feeding milk produced by cows treated with
antibiotics is, however, not recommended and discarded milk will not be
assumed to have an economic value.

Reduced Costs
When cows are affected by CM, feed intake is reduced (Bareille et al.,
2003). Østergaard et al. (2005), for instance, modelled a reduction of feed
intake of 7, 5, and 2%, respectively, in the three weeks following CM
diagnosis. The amount of money saved from reduced feed intake is,
however, difficult to assess, because it will depend on the feedstuffs used,
and will be neglected here.

Consequences Not Considered

Extra labour requirement and the increased risk of mortality caused by CM
are not accounted for, and these factors were also not included in Paper II.
Effects of CM on milk composition were not included, because it has been
argued (Seegers et al., 2003) that changes in fat and protein content can be

58
neglected in economic calculations as milk produced in the withdrawal
period is discarded and losses occurring later in lactation are proportional to
the milk loss. Also, increased SCC is not considered as BTSCC and the
amount of milk produced in the herd will influence whether or not a case
of CM affects milk price (Østerås, 2005).
In this simple partial budgeting approach, an economic loss of €461 per
case of CM was estimated. This figure is similar to the estimate of €428
obtained in Paper II, supporting the notion that the value assigned to each
cost component is more influential on the estimate than the analytical
approach. A well-parameterized partial budget is, therefore, able to provide
reliable estimates of the economic loss associated with production disorders.

59
60
Main Conclusions
The yearly avoidable cost of mastitis in a Swedish 150-cow dairy herd,
assuming that the initial incidence (32 and 33 cases of CM and SCM per
100 cow-years, respectively) can be reduced by 50%, is slightly more than
€8 000. This figure corresponds to 5% of the net return given the initial
incidence of mastitis. Expressed per cow/year, the avoidable cost of mastitis
is €50.
The economic loss associated with mastitis cannot be reduced by
discarding milk with high SCC, because this resulted in a substantial
decrease of the volume of sold milk, which was not offset by the increase in
milk price. Under the current milk-pricing system, it is, consequently, more
profitable for farmers to sell a larger volume of milk with higher SCC than
to discard high SCC milk in order to obtain a higher average milk price.
The average economic loss per case of CM and SCM are €275 and €60,
respectively. Reduced milk production constitutes the major cost
component of the total economic loss caused by mastitis. The magnitude of
yield loss is determined by the stage of lactation in which the cow develops
mastitis: milk yield is most severely affected when CM occurs in early and
when SCM occurs in late lactation. The lactational yield loss associated with
CM varies between 0 and 705 kg in primiparous cows and between 0 and
902 kg in multiparous cows, depending on lactation week at clinical onset.
Most cases of CM develop in the first week of lactation and results in a yield
loss of 578 and 782 kg milk in primiparous and multiparous cows,
respectively. This yield loss is, most likely, underestimated, because it is
expressed relative to the production level of non-mastitic cows, and cows
developing CM tends to have a production advantage over non-mastitic
herd mates before becoming diseased. Daily milk loss at an SCC of 500 000
cells/ml amounts to 0.7 to 2.0 kg in primiparous cows and 1.1 to 3.7 kg in
multiparous cows in different stages of lactation. SCM is difficult to define

61
accurately, and the impact of SCM on milk yield was, therefore, assessed in
terms of increased SCC. No estimates of the lactational yield loss caused by
a case of SCM could, consequently, be obtained. The yield loss in an average
305-day lactation affected by SCM is, however, 150 and 450 kg milk in
primiparous and multiparous cows, respectively.

62
Practical Implications
This thesis clearly demonstrates that the economic performance of dairy
herds can be improved by reducing the incidence of mastitis. A lower
incidence of mastitis can be achieved by implementation of mastitis control
programs. In order to make scientifically supported decisions regarding
mastitis control in dairy herds, however, accurate estimates of the avoidable
cost of mastitis are a necessary prerequisite. These are provided in this thesis.
The negative economic consequences of mastitis can be reduced either
by preventing new cases from occurring or by strategies aiming to limit the
impact of mastitis once it has occurred. It is demonstrated that prevention is
preferred over discarding milk with high SCC, because a reduction of the
incidence of mastitis resulted in significantly increased net return per cow-
year, whereas the effects of discarding milk with high SCC seemed to be
poor. Farmers are, therefore, recommended to invest in preventive measures
rather than in milk-sorting equipment.
The largest proportion of economic loss associated with mastitis arises
from reduced milk production. Yield loss is, however, a hidden cost and it is
not always obvious to farmers. It is, therefore, the task of advisors to
communicate the relative economic importance of lost milk production, as
compared to veterinary costs, and to convey that veterinary costs are to be
regarded as a means to limit the potentially more severe economic loss
caused by reduced milk production.
Considering that most cows in a herd are of second or later parities and
that the incidence of CM, as well as the proportion of TD with increased
SCC, is highest in multiparous cows, the majority of the production loss
caused by mastitis is likely to be experienced by multiparous cows. Special
emphasis should, therefore, be put on reducing the incidence of mastitis in
multiparous cows. This can be attempted by enhanced mastitis control in
primiparous cows, because cows that have suffered from mastitis are more

63
likely to develop subsequent cases. Attention should be paid to cows in early
and late stages of lactation. During these periods, the incidence of mastitis is
highest and yield losses are most severe, but, by taking proper actions,
unnecessary milk loss can be avoided. High-yielding cows should be given
priority, because they are at increased risk of developing CM.
Partial budgets are relatively easy to construct and are able to provide
accurate estimates of the economic impact of disease. If reliable information
on the effects of disease and the value of cost components are available, they
can be used by advisors and farmers to support decisions regarding animal
health management.
The incidence of mastitis can be reduced by breeding for increased
mastitis resistance. Genetically, CM is not the same trait over time and this
should be considered in the genetic evaluation for mastitis resistance
(Carlén, 2008). Based on the findings of this thesis, it might be economically
wise to use a selection criterion which attaches more weight to CM in early
lactation.
The substantial economic loss associated with mastitis is a powerful
incentive to improve the udder health of dairy cows, encouraging farmers to
put effort into mastitis prevention. In a long-term perspective, this thesis
can, therefore, result in improved welfare of dairy cows as it sheds light on
the economic importance of mastitis.

64
Future Research
The presented estimates of yield loss caused by mastitis were based on data
sampled in a single herd, where production records sampled at weekly
intervals were available. The validity of the results in herds with other
production characteristics should be verified by repeating the analyses on
data sampled in herds with different incidences of mastitis, yield levels, and
pathogen frequencies.
In Papers I and III, production records sampled at weekly intervals were
used to estimate the magnitude of yield loss associated with mastitis. The
obtained estimates were, consequently, more fine-tuned as compared with
those presented in the literature, where similar studies usually have been
conducted based on datasets collected at monthly samplings. As milking
systems are becoming more and more technologically advanced, data
collected even more frequent is becoming readily available. Based on data
recorded on a daily basis, or even at individual milkings, the impact of
mastitis on milk yield could be further dissected. Estimates obtained in such
analyses would improve mastitis management, because farmers could make
more well-informed decisions.
An interesting topic to examine is a possible breed difference in the
extent of yield loss suffered due to mastitis. Swedish Red cows have lower
incidences of diseases than Swedish Holstein cows, which might indicate
differences in innate immune function between the breeds. It may be
speculated that a better immune function in Swedish Red cows can
contribute also to less yield loss due to faster, and more complete, cure. The
economic loss associated with mastitis might, consequently, not be the same
in both breeds. Therefore, it is worth studying whether economic
considerations actually favour the somewhat lower producing, but more
robust, Swedish Red cow.

65
 All strategies of mastitis control are not equally effective as regards their
ability to increase the economic performance of dairy herds. The
profitability of a certain preventive measure is determined by the cost of its
implementation and the value of the reduction in mastitis incidence that it
can achieve. In order to improve decision support concerning whether
individual herds ought to invest in preventive measures and to facilitate
prioritization between different strategies, the expected economic viability
of different preventive measures should be investigated.
®
There are technological tools, such as Herd Navigator , available on the
market that enable detection of SCM at an early stage of infection and, thus,
allow for early intervention. With early treatment of mastitis, cure rate can
be expected to be higher and the economic loss can thereby be reduced. In
theory, such equipment has the potential to revolutionize udder-health
management. It is, however, associated with an investment of considerable
magnitude, and research is required to assess the economic viability of this
kind of technology. Preferably, economic calculations should consider the
consequences of a, possibly, increased usage of antibiotics resulting from
more cases of mastitis being detected.
Discarding milk with high SCC was not an effective strategy to increase
herd net return under the current milk-pricing system. The simulated herds
did, however, have a relatively low BTSCC, and the impact of sorting of
milk in herds with higher BTSCC and higher incidence of CM must also be
looked into. If such studies confirm that discarding milk with high SCC is
not profitable, an important task for research is to demonstrate the reduced
product value of high-SCC milk. Such information can support decisions
regarding whether the premium and penalty system needs to be revised, or
if new regulatory limits with respect to SCC should be proposed, so that
farmers are motivated not to deliver milk of poor quality.
Because mastitis is considered the most costly disease affecting dairy
cows, it is often prioritized in herd-health management. Other production-
related disorders, such as lameness, might as a result receive less attention,
although they are responsible for substantial economic loss. In order to
achieve correct on-farm prioritizations regarding disease control, effects of
all disorders must be considered simultaneously. To take a comprehensive
approach to herd-health management, the economic impact of lameness,
metabolic disorders, reproductive disorders, and calving-related disorders,
therefore, need to be assessed. In order to facilitate comparison between the
obtained results and our estimates of economic loss caused by mastitis, such
studies should be performed under Swedish production circumstances.

66
Ekonomisk betydelse av mastit hos mjölkkor
Bakgrund
Mjölksektorn är utsatt för hårdnande internationell konkurrens. Ekonomiskt
effektiva besättningar är därför en förutsättning för att en livskraftig svensk
mjölkproduktion ska kunna upprätthållas. Detta medför att det blir allt
viktigare att optimera produktionens alla delar och behovet av
kostnadsminimering accentueras i takt med att produktvärdet minskar. Ett
sätt att minska produktionskostnaderna är att reducera förekomsten av
produktionssjukdomar, då dessa leder till reducerad avkastning, kostnader för
veterinär och behandling samt ökad utslagning. I detta sammanhang är
mastit (juverinflammation) av stor betydelse, eftersom sjukdomen är vanligt
förekommande och ger upphov till stora ekonomiska förluster. I
besättningar anslutna till kokontrollen förekommer mastit i 16 % av alla
laktationer. Sjukdomen kan dock antas vara betydligt vanligare än så
eftersom antalet mastiter som upptäcks av djurägare visat sig vara 33 % högre
än antalet mastiter i djursjukdata. Juversjukdom, inklusive höga celltal (ett
mått på den immunologiska aktiviteten i juvret), utgör den vanligaste
utslagsorsaken bland svenska mjölkkor; i drygt en fjärdedel av fallen då kor
gallras ut anges juversjukdom som orsak. Detta innebär att 10 % av den
svenska mjölkkopopulationen årligen slås ut på grund av juversjukdom.
Förutom rent ekonomiska förluster ger mastit upphov till ett flertal andra
negativa påföljder. Sjukdomen medför en ökning av mjölkens celltal, vilket
leder till sämre processegenskaper. Vidare förknippas mastit med försämrad
djurvälfärd och ökad användning av antibiotika, vilket påverkar
konsumenternas attityd gentemot mjölkproduktionen negativt.
De ekonomiska konsekvenserna av mastit kan minskas genom åtgärder
som syftar till att förebygga nya fall av mastit eller begränsa effekterna av
mastiter då de inträffat. Sådana åtgärder innebär dock extra kostnader för

67
mjölkproducenten i form av produkter och arbete. En förutsättning för att
investeringar ska göras är därför att dessa kan förväntas generera ökade
intäkter som överstiger kostnaderna. Kunskap om den ekonomiska
betydelsen av mastit är därför avgörande när lönsamheten av att investera i
mastitprevention ska utvärderas. Vidare påverkar sådan information andra
skötselbeslut i en besättning, t.ex. om en ko med mastit ska behandlas,
sinläggas eller slås ut, samt om mjölk med högt celltal ska sorteras bort för att
på så vis få ett högre pris för den levererade mjölken. På nationell nivå är
uppgifter om kostnaden avgörande för t.ex. investeringar i bättre
registreringar av mastit i kokontrollen och för att ge mastit rätt ekonomisk
vikt i avelsmålet.
Syftet med denna avhandling är att skatta kostnaden för mastit under
svenska produktionsförhållanden.

Sammanfattning av avhandlingens delarbeten

Avkastningsförlusternas storlek
Avkastningsförlustens storlek beror på när i laktationen kon får mastit.
Klinisk mastit orsakar störst produktionsbortfall då korna insjuknar i början
av laktationen, medan förhöjt celltal ger upphov till störst avkastningsförlust
när det inträffar i slutet av laktationen. Avkastningsförlusterna är konsekvent
högre hos äldre kor än hos förstakalvare.
Klinisk mastit uppträder oftast i första laktationsveckan. Förstakalvare och
äldre kor som diagnosticeras med klinisk mastit under den perioden drabbas
av ett produktionsbortfall på 578 respektive 782 kg mjölk. Mjölkförlustens
storlek varierar mellan 0 och 705 kg (0 till 9 %) hos förstakalvare och mellan
0 och 902 kg (0 till 11 %) hos äldre kor, beroende på när mastiten
uppträder. Den dagliga avkastningsförlusten vid ett celltal på 500 000
celler/ml uppgår till 1 till 2 kg mjölk hos förstakalvare och 1 till 4 kg mjölk
hos äldre kor, beroende på i vilket laktationsstadium som kon befinner sig. I
en genomsnittlig laktation där subklinisk mastit förekommer reduceras
avkastningen med 150 och 450 kg mjölk hos förstakalvare respektive äldre
kor.

Kostnad per fall

Varje fall av klinisk mastit är i genomsnitt förknippat med en ekonomisk
förlust motsvarande 2 800 kr, vilken i huvudsak utgörs av reducerad mjölk-
produktion. Ett fall av subklinisk mastit värderas i medeltal till 600 kr.

68
Besättningsekonomiska konsekvenser
Om den nuvarande förekomsten av mastit i en genomsnittlig svensk 150-
kors besättning halverades skulle täckningsbidraget öka med 80 000 kr per
år, vilket motsvarar 5 % av täckningsbidraget. Detta belopp motsvarar den
maximala årliga investeringen i mastitförebyggande åtgärder, och kan
innebära t.ex. ytterligare 1,2 timmars lönekostnad per dag eller en
investering om 580 000 kr i ny teknik.

Sortering av mjölk för att sänka tankcelltalet

Ett tänkbart sätt att reducera de besättningsekonomiska effekterna av mastit
skulle kunna vara att sortera bort mjölk med högt celltal för att på så vis
undvika betalningsavdrag. Flera olika tröskelvärden för sortering har
undersökts, men samtliga hade negativ inverkan på täckningsbidraget
eftersom en avsevärd mängd mjölk sorterades bort. Under det nuvarande
mjölkbetalningssystemet är den levererade mjölkens kvantitet av större
betydelse för besättningens ekonomiska resultat än dess kvalitet.

Kostnaden för mastit i Sverige

I Sverige finns ungefär 350 000 mjölkkor. Täckningsbidraget per ko och år
är 550 kr lägre än vad det skulle ha varit om mastitfrekvensen hade varit
hälften så hög som idag. På nationell nivå uppgår alltså den motsvarande
ekonomiska förlusten på grund av mastit till 192 miljoner kr. Denna siffra
innefattar inte kostnader för förebyggande åtgärder mot mastit. Om dessa
hade inkluderats, vilket skulle varit mer korrekt, skulle den totala kostnaden
för mastit varit högre.

Slutsatser
De viktigaste resultaten från avhandlingen kan summeras enligt följande:

 Mastit kostar årligen Sveriges mjölkproducenter 192 miljoner kr

 Kostnaden för mastit uppgår i genomsnitt till ca 550 kr per ko och
år
 En klinisk mastit kostar i genomsnitt ca 2 800 kr medan en
subklinisk mastit i medeltal kostar ca 600 kr
 De ekonomiska förlusterna p.g.a. mastit kan inte reduceras genom
att mjölk med högt celltal sorteras bort
 Den största kostnadsposten i samband med mastit är reducerad
avkastning

69
  Klinisk mastit ger upphov till större avkastningsförlust än subklinisk
mastit
 Klinisk mastit orsakar högst avkastningsförluster hos kor som
insjuknar tidigt i laktationen
 Subklinisk mastit reducerar dygnsavkastningen mest då kor drabbas
sent i laktationen
 Mastit orsakar större avkastningsförlust hos äldre kor än hos
förstakalvare
 Kor drabbas oftast av klinisk mastit i första laktationsveckan
 Hög mjölkproduktion är en riskfaktor för klinisk mastit

Praktisk tillämpning av resultaten

Det är tydligt att besättningars ekonomiska resultat kan förbättras genom att
förekomsten av mastit reduceras. I en genomsnittlig svensk besättning kan
550 kr per ko och år investeras i mastitprevention, under förutsättning att
dagens mastitfrekvens därigenom halveras. Det inte lönsamt för lantbrukare
att sortera bort mjölk med högt celltal. Lantbrukare rekommenderas därför
istället att investera i mastitförebyggande åtgärder. Särskild vikt ska läggas vid
att förhindra mastiter i tidig och sen laktation, eftersom flest kor insjuknar i
dessa perioder och avkastningsförlusterna är som störst då.

70
References
Agriwise. (2008). Databoken 2008. [online] Available from: http://www.agriwise.org/
databoken/ [2009-03-15].
Akers, R. M. (2002). Lactation and the Mammary Gland. Iowa: Iowa State Press.
Allore, H.G. & Erb, H.N. (1999). Approaches to Modeling Intramammary Infections in
Dairy Cattle. Preventive Veterinary Medicine 39(4), 279-293.
Auldist, M.J., Coats, S., Sutherland, B.J., Mayes, J.J., McDowell, G.H. & Rogers, G.L.
(1996). Effects of Somatic Cell Count and Stage of Lactation on Raw Milk
Composition and the Yield and Quality of Cheddar Cheese. Journal of Dairy
Research 63(2), 269-280.
Bar, D., Gröhn, Y.T., Bennett, G., Gonzalez, R.N., Hertl, J.A., Schulte, H.F., Tauer, L.W.,
Welcome, F.L. & Schukken, Y.H. (2008a). Effects of Repeated Episodes of
Generic Clinical Mastitis on Mortality and Culling in Dairy Cows. Journal of Dairy
Science 91(6), 2196-2204.
Bar, D., Tauer, L.W., Bennett, G., Gonzalez, R.N., Hertl, J.A., Schukken, Y.H., Schulte,
H.F., Welcome, F.L. & Gröhn, Y.T. (2008b). The Cost of Generic Clinical
Mastitis in Dairy Cows as Estimated by Using Dynamic Programming. Journal of
Dairy Science 91(6), 2205-2214.
Barbano, D.M., Rasmussen, R.R. & Lynch, J.M. (1991). Influence of Milk Somatic Cell
Count and Milk Age on Cheese Yield. Journal of Dairy Science 74(2), 369-388.
Bareille, N., Beaudeau, F., Billon, S., Robert, A. & Faverdin, P. (2003). Effects of Health
Disorders on Feed Intake and Milk Production in Dairy Cows. Livestock Production
Science 83(1), 53-62.
Barkema, H.W., Schukken, Y.H., Lam, T.J., Beiboer, M.L., Benedictus, G. & Brand, A.
(1999). Management Practices Associated with the Incidence Rate of Clinical
Mastitis. Journal of Dairy Science 82(8), 1643-1654.
Barkema, H.W., Schukken, Y.H., Lam, T.J., Beiboer, M.L., Wilmink, H., Benedictus, G. &
Brand, A. (1998). Incidence of Clinical Mastitis in Dairy Herds Grouped in Three
Categories by Bulk-Milk Somatic Cell Counts. Journal of Dairy Science 81(2), 411-
419.
Beaudeau, F., Ducrocq, V., Fourichon, C. & Seegers, H. (1995). Effect of Disease on Length
of Productive Life of French Holstein Dairy Cows Assessed by Survival Analysis.
Journal of Dairy Science 78(1), 103-117.
Belotti, C. (1991). Vad betyder mjölkkornas hälsa för ekonomin? SLU/info ekonomi 5.
Swedish Univerisity of Agricultural Sciences, Uppsala.
Bendixen, P.H., Vilson, B., Ekesbo, I. & Astrand, D.B. (1988). Disease Frequencies in Dairy
cows in Sweden. 5. Mastitis. Preventive Veterinary Medicine 5(4), 263-274.

71
Bennedsgaard, T.W., Enevoldsen, C., Thamsborg, S.M. & Vaarst, M. (2003). Effect of
Mastitis Treatment and Somatic Cell Counts on Milk Yield in Danish Organic
Dairy Cows. Journal of Dairy Science 86(10), 3174-3183.
Bennett, R., Christiansen, K. & Clifton-Hadley, R. (1999). Preliminary Estimates of the
Direct Costs Associated with Endemic Diseases of Livestock in Great Britain.
Preventive Veterinary Medicine 39(3), 155-171.
Bennett, R.M. (1992). The Use of Economic Quantitative Modelling Techniques in
Livestock Health and Disease-Control Decision-Making - a Review. Preventive
Veterinary Medicine 13(1), 63-76.
Black, R.T., Bourland, C.T. & Marshall, R.T. (1972). California Mastitis Test Reactivity
and Bacterial Invasions in Quarters Infected with Corynebacterium Bovis. Journal of
Dairy Science 55(7), 1016-1017.
Brolund, L. (1985). Cell Counts in Bovine Milk. Causes of Variation and Applicability for
Diagnosis of Subclinical Mastitis. Acta Veterinaria Scandinavica Supplementum 80, 1-
123.
Broom, D.M. (2006). Behaviour and Welfare in Relation to Pathology. Applied Animal
Behaviour Science 97(1), 73-83.
Broom, D.M. & Fraser, A.F. (2007). Domestic Animal Behaviour and Welfare. Fourth Edition.
Wallingford, UK: CABI Publishing.
Busato, A., Trachsel, P., Schallibaum, M. & Blum, J.W. (2000). Udder Health and Risk
Factors for Subclinical Mastitis in Organic Dairy Farms in Switzerland. Preventive
Veterinary Medicine 44(3-4), 205-220.
Capuco, A.V., Wood, D.L., Baldwin, R., Mcleod, K. & Paape, M.J. (2001). Mammary Cell
Number, Proliferation, and Apoptosis During a Bovine Lactation: Relation to
Milk Production and Effect of bST. Journal of Dairy Science 84(10), 2177-2187.
Carlén, E., Strandberg, E. & Roth, A. (2004). Genetic Parameters for Clinical Mastitis,
Somatic Cell Score, and Production in the First Three Lactations of Swedish
Holstein Cows. Journal of Dairy Science 87(9), 3062-3070.
Carlén, E. (2008). Genetic Evaluation of Clinical Mastitis in Dairy Cattle. Doctoral Thesis.
Swedish University of Agricultural Sciences. Uppsala, Sweden.
De Haas, Y., Barkema, H.W. & Veerkamp, R.F. (2002). The Effect of Pathogen-Specific
Clinical Mastitis on the Lactation Curve for Somatic Cell Count. Journal of Dairy
Science 85(5), 1314-1323.
De Haas, Y., Veerkamp, R.F., Barkema, H.W., Gröhn, Y.T. & Schukken, Y.H. (2004).
Associations Between Pathogen-Specific Cases of Clinical Mastitis and Somatic
Cell Count Patterns. Journal of Dairy Science 87(1), 95-105.
Degraves, F.J. & Fetrow, J. (1993). Economics of Mastitis and Mastitis Control. Veterinary
Clinics of North America-Food Animal Practice 9(3), 421-434.
Dekkers, J.C.M., VanErp, T. & Schukken, Y.H. (1996). Economic Benefits of Reducing
Somatic Cell Count Under the Milk Quality Program of Ontario. Journal of Dairy
Science 79(3), 396-401.
Dijkhuizen, A.A., Huirne, R.B.M. & Jalvingh, A.W. (1995). Economic Analysis of Animal
Diseases and Their Control. Preventive Veterinary Medicine 25(2), 135-149.
Djabri, B., Bareille, N., Beaudeau, F. & Seegers, H. (2002). Quarter Milk Somatic Cell
Count in Infected Dairy Cows: a Meta-Analysis. Veterinary Research 33(4), 335-357.
Dohoo, I.R. & Martin, S.W. (1984a). Disease, Production and Culling in Holstein-Friesian
Cows. III. Disease and Production as Determinants of Disease. Preventive Veterinary
Medicine 2(5), 671-690.
Dohoo, I.R. & Martin, S.W. (1984b). Disease, Production and Culling in Holstein-Friesian
Cows. V. Survivorship. Preventive Veterinary Medicine 2(6), 771-784.
Dohoo, I.R. & Meek, A.H. (1982). Somatic Cell Counts in Bovine Milk. Canadian
Veterinary Journal 23(4), 119-125.
Elbers, A.R., Miltenburg, J.D., De Lange, D., Crauwels, A.P., Barkema, H.W. & Schukken,
Y.H. (1998). Risk Factors for Clinical Mastitis in a Random Sample of Dairy

72
 Herds from the Southern Part of The Netherlands. Journal of Dairy Science 81(2),
420-426.
Emanuelson, U., Danell, B. & Philipsson, J. (1988). Genetic Parameters for Clinical Mastitis,
Somatic Cell Counts, and Milk Production Estimated by Multiple-Trait
Restricted-Maximum Likelihood. Journal of Dairy Science 71(2), 467-476.
Emanuelson, U., Oltenacu, P.A. & Gröhn, Y.T. (1993). Nonlinear Mixed-Model Analyses
of 5 Production Disorders of Dairy-Cattle. Journal of Dairy Science 76(9), 2765-
2772.
Fitzpatrick, J.L., Young, F.J., Eckersall, D., Logue, D.N., Knight, C.J. & Nolan, A. (1998).
Recognising and Controlling Pain and Inflammation in Mastitis. In: Proceedings of
the British Mastitis Conference 1998, 36-44.
Gröhn, Y., Erb, H.N., Mcculloch, C.E. & Saloniemi, H.S. (1990a). Epidemiology of
Reproductive Disorders in Dairy Cattle: Associations Among Host Characteristics,
Disease and Production. Preventive Veterinary Medicine 8(1), 25-39.
Gröhn, Y.T., Eicker, S.W., Ducrocq, V. & Hertl, J.A. (1998). Effect of Diseases on the
Culling of Holstein Dairy Cows in New York State. Journal of Dairy Science 81(4),
966-978.
Gröhn, Y.T., Erb, H.N., McCulloch, C.E. & Saloniemi, H.S. (1989). Epidemiology of
Metabolic Disorders in Dairy Cattle: Association Among Host Characteristics,
Disease, and Production. Journal of Dairy Science 72(7), 1876-1885.
Gröhn, Y.T., Erb, H.N., McCulloch, C.E. & Saloniemi, H.S. (1990b). Epidemiology of
Mammary Gland Disorders in Multiparous Finnish Ayrshire Cows. Preventive
Veterinary Medicine 8(4), 241-252.
Gröhn, Y.T., Wilson, D.J., Gonzalez, R.N., Hertl, J.A., Schulte, H., Bennett, G. &
Schukken, Y.H. (2004). Effect of Pathogen-Specific Clinical Mastitis on Milk
Yield in Dairy Cows. Journal of Dairy Science 87(10), 3358-3374.
Halasa, T., Huijps, K., Osteras, O. & Hogeveen, H. (2007). Economic Effects of Bovine
Mastitis and Mastitis Management: a Review. Veterinary Quarterly 29(1), 18-31.
Hamann, J. (2002). Relationships Between Somatic Cell Count and Milk Composition.
Bulletin of the International Dairy Federation 372, 56-59.
Hamann, J. (2003). Definition of the Physiological Cell Count Threshold Based on Changes
in Milk Composition. Bulletin of the International Dairy Federation 381, 9-12.
Hamann, J. (2005). Diagnosis of Mastitis and Indicators of Milk Quality. In: Hogeveen, H.
(Ed.) Proceedings of 4th IDF International Dairy Conference: Mastitis in Dairy Production -
Current Knowledge and Future Solutions. Wageningen: Wageningen Academic
Publishers. pp. 82-90.
Hamann, J. & Reichmuth, J. (1990). Compensatory Milk Production Within the Bovine
Udder: Effects of Short-Term Non-Milking of Single Quarters. Journal of Dairy
Research 57(1), 17-22.
Hansson, H. (2007). The Links Between Management's Critical Success Factors and Farm
Level Economic Performance on Dairy Farms in Sweden. Acta Agriculturae
Scandinavica. Section C, Food Economics 4(2), 77-88.
Harmon, R.J. (1994). Symposium - Mastitis and Genetic Evaluation for Somatic Cell Count
- Physiology of Mastitis and Factors Affecting Somatic Cell Counts. Journal of Dairy
Science 77(7), 2103-2112.
Harper, G.C. & Henson, S.J. (2001). Consumer Concerns About Animal Welfare and the Impact
on Food Choice - Final Report EU FAIR CT98-3678 The University of Reading,
U.K. ISBN
Heringstad, B., Klemetsdal, G. & Ruane, J. (2000). Selection for Mastitis Resistance in Dairy
Cattle: a Review with Focus on the Situation in the Nordic Countries. Livestock
Production Science 64(2-3), 95-106.
Hillerton, J.E., West, J.G.H. & Shearn, M.F.H. (1992). The Cost of Summer Mastitis.
Veterinary Record 131(14), 315-317.

73
Hogan, J. (2005). Human Health Risks Associated with High SCC Milk. In: Proceedings of the
British Mastitis Conference 2005, Warwickshire, U.K.
Hogeveen, H. (2005). Mastitis is an Economic Problem. In: Proceedings of the British Mastitis
Conference 2005. Warwickshire, U.K.
Hogeveen, H. & Østerås, O. (2005). Mastitis Management in an Economic Framework. In:
Hogeveen, H. (Ed.) Proceedings of 4th IDF International Dairy Conference: Mastitis in
Dairy Production - Current Knowledge and Future Solutions. Wageningen: Wageningen
Academic Publishers. pp. 41-52.
Hortet, P., Beaudeau, F., Seegers, H. & Fourichon, C. (1999). Reduction in Milk Yield
Associated with Somatic Cell Counts up to 600 000 cells/ml in French Holstein
Cows Without Clinical Mastitis. Livestock Production Science 61(1), 33-42.
Hortet, P. & Seegers, H. (1998a). Calculated Milk Production Losses Associated with
Elevated Somatic Cell Counts in Dairy Cows: Review and Critical Discussion.
Veterinary Research 29(6), 497-510.
Hortet, P. & Seegers, H. (1998b). Loss in Milk Yield and Related Composition Changes
Resulting from Clinical Mastitis in Dairy Cows. Preventive Veterinary Medicine 37(1-
4), 1-20.
Houben, E.H.P., Dijkhuizen, A.A., Vanarendonk, J.A.M. & Huirne, R.B.M. (1993). Short-
Term and Long-Term Production Losses and Repeatability of Clinical Mastitis in
Dairy Cattle. Journal of Dairy Science 76(9), 2561-2578.
Houben, E.H.P., Huirne, R.B.M., Dijkhuizen, A.A. & Kristensen, A.R. (1994). Optimal
Replacement of Mastitic Cows Determined by a Hierarchical Markov Process.
Journal of Dairy Science 77(10), 2975-2993.
Huijps, K., Lam, T.J. & Hogeveen, H. (2008). Costs of Mastitis: Facts and Perception. Journal
of Dairy Research 75(1), 113-120.
Huirne, R.B.M. (2003). Economic Analysis of Farm-Level Health Problems in Dairy Cattle.
Cattle Practice 11, 227-235.
International Dairy Federation (1987). Bovine Mastitis: Definition and Guidelines for
Diagnosis. Bulletin of the International Dairy Federation 211, 24 pp.
Jain, N.C. (1979). Common Mammary Pathogens and Factors in Infection and Mastitis.
Journal of Dairy Science 62(1), 128-134.
Janosi, S., Kulcsar, M., Korodi, P., Katai, L., Reiczigel, J., Dieleman, S.J., Nikolic, J.A.,
Salyi, G., Ribiczey-Szabo, P. & Huszenicza, G. (2003). Energy Imbalance Related
Predisposition to Mastitis in Group-Fed High-Producing Postpartum Dairy Cows.
Acta Veterinaria Hungarica 51(3), 409-424.
Kehrli, M.E., Jr. & Shuster, D.E. (1994). Factors Affecting Milk Somatic Cells and Their
Role in Health of the Bovine Mammary Gland. Journal of Dairy Science 77(2), 619-
627.
Knegsel, A.T.M.v., Vries Reilingh, G.d., Meulenberg, S., Brand, H.v.d., Dijkstra, J., Kemp,
B. & Parmentier, H.K. (2007). Natural Antibodies Related to Energy Balance in
Early Lactation Dairy Cows. Journal of Dairy Science 90(12), 5490-5498.
Kossaibati, M.A. & Esslemont, R.J. (1997). The Costs of Production Diseases in Dairy Herds
in England. Veterinary Journal 154(1), 41-51.
Laevens, H., Deluyker, H., Schukken, Y.H., De Meulemeester, L., Vandermeersch, R., De
Muecaronlenaere, E. & De Kruif, A. (1997). Influence of Parity and Stage of
Lactation on the Somatic Cell Count in Bacteriologically Negative Dairy Cows.
Journal of Dairy Science 80(12), 3219-3226.
Lam, T.J., Schukken, Y.H., van Vliet, J.H., Grommers, F.J., Tielen, M.J. & Brand, A.
(1997). Effect of Natural Infection with Minor Pathogens on Susceptibility to
Natural Infection with Major Pathogens in the Bovine Mammary Gland. American
Journal of Veterinary Research 58(1), 17-22.
Langford, F.M., Weary, D.M. & Fisher, L. (2003). Antibiotic Resistance in Gut Bacteria
from Dairy Calves: a Dose Response to the Level of Antibiotics Fed in Milk.
Journal of Dairy Science 86(12), 3963-3966.

74
Maizon, D.O., Oltenacu, P.A., Gröhn, Y.T., Strawderman, R.L. & Emanuelson, U. (2004).
Effects of Diseases on Reproductive Performance in Swedish Red and White Dairy
Cattle. Preventive Veterinary Medicine 66(1-4), 113-126.
Matthews, K.R., Harmon, R.J. & Langlois, B.E. (1991). Effect of Naturally Occurring
Coagulase-Negative Staphylococci Infections on New Infections by Mastitis
Pathogens in the Bovine. Journal of Dairy Science 74(6), 1855-1859.
McInerney, J.P., Howe, K.S. & Schepers, J.A. (1992). A Framework for the Economic
Analysis of Disease in Farm Livestock. Preventive Veterinary Medicine 13(2), 137-154.
Mevius, D., Sampimon, O. & Sol, J. (2005). Antimicrobial Resistance in Mastitis Organisms
th
as a Public Health Threat. In: Hogeveen, H. (Ed.) Proceedings of 4 IDF International
Dairy Conference: Mastitis in Dairy Production - Current Knowledge and Future Solutions.
Wageningen: Wageningen Academic Publishers. pp. 102-108.
Miller, G.Y., Bartlett, P.C., Lance, S.E., Anderson, J. & Heider, L.E. (1993). Costs of
Clinical Mastitis and Mastitis Prevention in Dairy Herds. Journal of the American
Veterinary Medical Association 202(8), 1230-1236.
Moynagh, J. (2000). EU Regulation and Consumer Demand for Animal Welfare.
AgBioForum 3(2&3), 107-114.
Mörk, M., Lindberg, A., Alenius, S., Vågsholm, I. & Egenvall, A. (2009). Comparison
Between Dairy Cow Disease Incidence in Data Registered by Farmers and in Data
from a Disease-Recording System Based on Veterinary Reporting. Preventive
Veterinary Medicine 88(4), 298-307.
Nyman, A.K., Ekman, T., Emanuelson, U., Gustafsson, A.H., Holtenius, K., Waller, K.P. &
Sandgren, C.H. (2007). Risk Factors Associated with the Incidence of Veterinary-
Treated Clinical Mastitis in Swedish Dairy Herds with a High Milk Yield and a
Low Prevalence of Subclinical Mastitis. Preventive Veterinary Medicine 78(2), 142-
160.
Olde Riekerink, R.G., Barkema, H.W. & Stryhn, H. (2007a). The Effect of Season on
Somatic Cell Count and the Incidence of Clinical Mastitis. Journal of Dairy Science
90(4), 1704-1715.
Olde Riekerink, R.G., Barkema, H.W., Veenstra, W., Berg, F.E., Stryhn, H. & Zadoks,
R.N. (2007b). Somatic Cell Count During and Between Milkings. Journal of Dairy
Science 90(8), 3733-41.
Organization for Economic Co-operation and Development. (2008). Consumer Price Indices.
[online] Available from: http://stats.oecd.org/wbos/ [2008-12-15].
Oviedo-Boyso, J., Valdez-Alarcon, J.J., Cajero-Juarez, M., Ochoa-Zarzosa, A., Lopez-Meza,
J.E., Bravo-Patino, A. & Baizabal-Aguirre, V.M. (2007). Innate Immune Response
of Bovine Mammary Gland to Pathogenic Bacteria Responsible for Mastitis. Journal
of Infection 54(4), 399-409.
Peeler, E.J., Green, M.J., Fitzpatrick, J.L., Morgan, K.L. & Green, L.E. (2000). Risk Factors
Associated with Clinical Mastitis in Low Somatic Cell Count British Dairy Herds.
Journal of Dairy Science 83(11), 2464-2472.
Peeler, E.J., Otte, M.J. & Esslemont, R.J. (1994). Interrelationships of Periparturient Diseases
in Dairy Cows. Veterinary Record 134(6), 129-132.
Persson Waller, K., Bengtsson, B., Lindberg, A., Nyman, A. & Ericsson Unnerstad, H.
(2009). Incidence of Mastitis and Bacterial Findings at Clinical Mastitis in Swedish
Primiparous Cows - Influence of Breed and Stage of Lactation. Veterinary
Microbiology 134(1-2), 89-94.
Petersson, K.J., Strandberg, E., Gustafsson, H. & Berglund, B. (2006). Environmental Effects
on Progesterone Profile Measures of Dairy Cow Fertility. Animal Reproduction
Science 91(3-4), 201-214.
Pyörälä, S. (2003). Indicators of Inflammation in the Diagnosis of Mastitis. Veterinary Research
34(5), 565-578.

75
Rajala-Schultz, P.J. & Gröhn, Y.T. (1999a). Culling of Dairy Cows. Part I. Effects of
Diseases on Culling in Finnish Ayrshire Cows. Preventive Veterinary Medicine 41(2-
3), 195-208.
Rajala-Schultz, P.J., Gröhn, Y.T., McCulloch, C.E. & Guard, C.L. (1999b). Effects of
Clinical Mastitis on Milk Yield in Dairy Cows. Journal of Dairy Science 82(6), 1213-
1220.
Rajala, P.J. & Gröhn, Y.T. (1998). Disease Occurrence and Risk Factor Analysis in Finnish
Ayrshire Cows. Acta Veterinaria Scandinavica 39(1), 1-13.
Reichmuth, J. (1975). Somatic Cell Counting - Interpretation of Results. Annual Bulletin,
International Dairy Federation 85, 93-109.
Reneau, J.K. (1986). Effective Use of Dairy Herd Improvement Somatic Cell Counts in
Mastitis Control. Journal of Dairy Science 69(6), 1708-1720.
Sandgren, C. & Emanuelson, U. (1994). Ekonomiskt effektiv mjölkproduktion, hög
mjölkmängd och/eller god hälsa? Memo, Swedish Dairy Association.
Schepers, A.J., Lam, T.J.G.M., Schukken, Y.H., Wilmink, J.B.M. & Hanekamp, W.J.A.
(1997). Estimation of Variance Components for Somatic Cell Counts to Determine
Thresholds for Uninfected Quarters. Journal of Dairy Science 80(8), 1833-1840.
Schepers, J.A. & Dijkhuizen, A.A. (1991). The Economics of Mastitis and Mastitis Control in
Dairy Cattle - a Critical Analysis of Estimates Published Since 1970. Preventive
Veterinary Medicine 10(3), 213-224.
Schneider, M.d.P., Strandberg, E., Emanuelson, U., Grandinson, K. & Roth, A. (2007). The
Effect of Veterinary-Treated Clinical Mastitis and Pregnancy Status on Culling in
Swedish Dairy Cows. Preventive Veterinary Medicine 80(2-3), 179-192.
Schreiner, D.A. & Ruegg, P.L. (2003). Relationship Between Udder and Leg Hygiene
Scores and Subclinical Mastitis. Journal of Dairy Science 86(11), 3460-3465.
Schrick, F.N., Hockett, M.E., Saxton, A.M., Lewis, M.J., Dowlen, H.H. & Oliver, S.P.
(2001). Influence of Subclinical Mastitis During Early Lactation on Reproductive
Parameters. Journal of Dairy Science 84(6), 1407-1412.
Seegers, H., Fourichon, C. & Beaudeau, F. (2003). Production Effects Related to Mastitis
and Mastitis Economics in Dairy Cattle Herds. Veterinary Research 34(5), 475-491.
Sheldrake, R.F., Hoare, R.J.T. & McGregor, G.D. (1983). Lactation Stage, Parity, and
Infection Affecting Somatic Cells, Electrical Conductivity, and Serum Albumin in
Milk. Journal of Dairy Science 66(3), 542-547.
Steeneveld, W., Hogeveen, H., Barkema, H.W., van den Broek, J. & Huirne, R.B.M.
(2008). The Influence of Cow Factors on the Incidence of Clinical Mastitis in
Dairy Cows. Journal of Dairy Science 91(4), 1391-1402.
Steeneveld, W., Swinkels, J. & Hogeveen, H. (2007). Stochastic Modelling to Assess
Economic Effects of Treatment of Chronic Subclinical Mastitis Caused by
Streptococcus uberis. Journal of Dairy Research 74(4), 459-467.
Stott, A.W. & Kennedy, J.O.S. (1993). The Economics of Culling Dairy Cows with Clinical
Mastitis. Veterinary Record 133(20), 495-499.
Suriyasathaporn, W., Heuer, C., Noordhuizen-Stassen, E.N. & Schukken, Y.H. (2000).
Hyperketonemia and the Impairment of Udder Defense: a Review. Veterinary
Research 31(4), 397-412.
Svensson, C., Nyman, A.K., Waller, K.P. & Emanuelson, U. (2006). Effects of Housing,
Management, and Health of Dairy Heifers on First-Lactation Udder Health in
Southwest Sweden. Journal of Dairy Science 89(6), 1990-1999.
Swedish Board of Agriculture. (2008a). Husdjur i juni 2008. [online] Available from:
http://www.sjv.se/ [2009-03-15].
Swedish Board of Agriculture. (2008b). Djurhälsa år 2007. [online]. Available from:
http://www.sjv.se/ [Accessed 2009-02-15].
Swedish Dairy Association. (2009). Marknadsinformation -kvotutnyttjande.
[online] Available from: http://www.husdjur.se/html/marknad/
kvotutnyttjande.asp [2009-03-11].

76
Swedish Dairy Association. (2008). Husdjursstatistik, Cattle Statistics, 2008. Svensk Mjölk,
Stockholm, Sweden.
Swinkels, J.M., Hogeveen, H. & Zadoks, R.N. (2005a). A Partial Budget Model to Estimate
Economic Benefits of Lactational Treatment of Subclinical Staphylococcus aureus
mastitis. Journal of Dairy Science 88(12), 4273-4287.
Swinkels, J.M., Rooijendijk, J.G., Zadoks, R.N. & Hogeveen, H. (2005b). Use of Partial
Budgeting to Determine the Economic Benefits of Antibiotic Treatment of
Chronic Subclinical Mastitis caused by Streptococcus uberis or Streptococcus dysgalactiae.
Journal of Dairy Research 72(1), 75-85.
Urech, E., Puhan, Z. & Schallibaum, M. (1999). Changes in Milk Protein Fraction as
Affected by Subclinical Mastitis. Journal of Dairy Science 82(11), 2402-2411.
Valeeva, N.I., Lam, T.J.G.M. & Hogeveen, H. (2007). Motivation of Dairy Farmers to
Improve Mastitis Management. Journal of Dairy Science 90(9), 4466-4477.
Webster, J. (1999). Animal Welfare. A Cool Eye Towards Eden. First Edition. Oxford, UK:
Blackwell Science Ltd.
Wilson, D.J., Gonzalez, R.N., Hertl, J.A., Schulte, H.F., Bennett, G.J., Schukken, Y.H. &
Gröhn, Y.T. (2004). Effect of Clinical Mastitis on the Lactation Curve: A Mixed
Model Estimation Using Daily Milk Weights. Journal of Dairy Science 87(7), 2073-
2084.
Wolfová, M., Stípková,M. & Wolf, J. (2006). Incidence and Economics of Clinical Mastitis
in Five Holstein Herds in the Czech Republic. Preventive Veterinary Medicine 77(1-
2), 48-64.
Yalcin, C. (2000). Cost of Mastitis in Scottish Dairy Herds with Low and High Subclinical
Mastitis Problems. Turkish Journal of Veterinary and Animal Sciences 24(5), 465-472.
Yalcin, C., Stott, A.W., Logue, D.N. & Gunn, J. (1999). The Economic Impact of Mastitis-
Control Procedures Used in Scottish Dairy Herds with High Bulk-Tank Somatic
Cell Counts. Preventive Veterinary Medicine 41(2/3), 135-149.
Åkerstedt, M., Persson Waller, K. & Sternesjö, A. (2007). Haptoglobin and Serum Amyloid
A in Relation to the Somatic Cell Count in Quarter, Cow-Composite and Bulk-
Tank Milk Samples. Journal of Dairy Research 74(2), 198-203.
Östensson, K., Hageltorn, M. & Åström, G. (1988). Differential Cell Counting in Fraction-
Collected Milk from Dairy Cows. Acta Veterinaria Scandinavica 29(3-4), 493-500.
Østergaard, S., Chagunda, M.G.G., Friggens, N.C., Bennedsgaard, T.W. & Klaas, I.C.
(2005). A Stochastic Model Simulating Pathogen-Specific Mastitis Control in a
Dairy Herd. Journal of Dairy Science 88(12), 4243-4257.
Østerås, O. (2000). The Cost of Mastitis - an Opportunity to Gain More Money. In:
Proceedings of the British Mastitis Conference - 2000, Shepton Mallet, U.K. pp. 67-77.
Østerås, O. (2005). Economic Consequences of Mastitis. Bulletin of the International Dairy
Federation 394, 1-25.

77
78
 Acknowledgements
The majority of the work underlying this thesis was performed at the
Department of Animal Breeding and Genetics, Swedish University of
Agricultural Sciences, Uppsala. The analyses in Papers II and IV were
conducted within the group of Herd Health and Production Management,
Faculty of Agricultural Sciences, Aarhus University. The Swedish Farmers’
Foundation for Agricultural Research is gratefully acknowledged for funding
the project. Financial support was also provided by the department. The
SLU fund for internationalization of postgraduate studies funded my stay in
Denmark and provided travel grants for participation in a conference.

A lot of people has supported me during my time as a PhD-student. I

would like to express my sincere thanks to all of you, and give special credit
to some of you:

My supervisors, Erling Strandberg, Ulf Emanuelson, Britt Berglund,

Jan Philipsson, Hans Andersson, and Søren Østergaard, it has been a pleasure
to work within a group with such a diverse field of knowledge. Thanks a lot
for fruitful discussions and all your support!
Erling, min huvudhandledare, för det stora intresse du visat för
projektet! Tack för att du på ett så ypperligt sätt bidragit till att lyfta
diskussionen på handledarmötena och för dina finurliga tankegångar som har
förbättrat mina manuskript.
Uffe, du har varit en ovärderlig support under hela min doktorandtid!
Tack för engagemang, snabb feedback och kloka åsikter. Du är min idol i
forskarvärlden!
Britt, för att du alltid trodde på min förmåga och aldrig tvivlade på att
jag skulle ro projektet i land. Tack för att du hjälpte mig att utvecklas till en
självständig forskare!

79
 Jan, det var du som fick mig att hoppa på projektet från allra första
början. Det har jag inte ångrat! Tack för all uppmuntran och alla samtal om
smått och stort.
Hans, du lyckades med konststycket att övertyga mig om att ekonomi
inte är så tokigt ändå. Det är jag väldigt tacksam över nu! Tack också för alla
tänkbara (och otänkbara!) anekdoter som kryddade handledarmötena.
Søren, for at du uden tøven involverede dig i mit projekt! Tak for at du
introducerede mig for simuleringens forvirrende verden og et endnu større
tak for din tålmodighed, når jeg ganske tydeligt ikke havde nogen ide om
hvad du snakkede om! Jeg har virkeligt sat stor pris på at arbejde sammen
med dig, og dit input i projektet har været uvurderlig!

Tack alla ni som arbetar på Hgen för att ni bidragit till att skapa en trivsam
arbetsmiljö som gjort att det har varit roligt att gå till jobbet!
Monica, Bijan, Harriet, Helena, Siw och Jörgen, för att ni håller
ordning på institutionen och alltid har hjälpt till när det behövts!
Dan och Stickan, för all hjälp med min krånglande dator (men mest
tack för fredagsmejlen och dörröppnandet!).
Gudrun och Lena, för att ni samlat in allt data i Jällabesättningen och
tålmodigt försett mig med allt som jag efterfrågat.
Jörgen (igen!), min livscoach och extrapappa! Tack för alla råd vad gäller
jobb, relationer och kärleksliv! Och tack för att din dörr alltid är öppen!
Alla doktorander på Hgen, nuvarande och före detta, tack för att ni
gjort vardagen på institutionen roligare! Jag kommer att sakna att vara en av
er (fast kanske inte så mycket att vara doktorand…).
Doktorandmaffian: Jessica, Sofia, Jennie, Therese, Anna O,
Emma C och Emelie, tack för allt som inte hade med jobbet att göra!
Tiden på institutionen hade inte varit densamma utan er!
Johanna, jag uppskattar våra små pratstunder mycket och ser fram emot
att ni snart (!) flyttar hem till Sverige så att vi kan ses lite oftare.
Emma TH, tack för all peppning under slutet av min doktorandtid och
för att du lånar ut Ellin till mig. Stunderna i stallet har hjälpt mig att hålla
mig över ytan under de sista stressiga månaderna.
Karl-Johan, för givande diskussioner om Jälladata, SAS (du är ett SAS-
geni!) och allt annat som hört doktorerandet till.

Gruppen på Foulum, tak for jeres varme modtagelse og for at I fik mig
at føle mig som en del af gruppen, på trods af at mine ophold var korte!
Johanna, Bart, Tina, Jehan, Elise og Kirstine, fordi I sørgede for at
jeg havde et socialt liv under mine ophold på Foulum!

80
 Personalen på Wapnö, för att ni tog er an ”den där hästtjejen” och på
riktigt väckte mitt intresse för kor! Ett särskilt tack till Jörgen för att du var
fast besluten att lära mig allt som har med en mjölkbesättning att göra!

Anette, dig kan jag lita på! Tack för att du tog hand om Myran i vått
och torrt och för att du alltid ställt upp för mig oavsett vad det gällt!
Birgitta, min inspirationskälla framför alla andra! Du väckte mitt intresse
för god djurhållning redan i unga år och hejade ivrigt på mig i mitt yrkesval,
övertygad om att jag skulle lyckas förändra världen. Jag önskar att du hade
fått vara med på min stora dag!
Mamma och pappa, för att ni alltid uppmuntrade mitt djurintresse när
jag var liten och lät allt från hamstrar till hästar bli fullvärdiga
familjemedlemmar. Ofantligt tack för att ni har varit barnvakter åt Buster
och Linus i sex (!) år!
Per, för att du finns i mitt liv och för att du alltid stöttar mig! Du är det
bästa som hänt mig! Jag ser fram emot vårt fortsatta liv tillsammans, utan
avhandlingar som ska skrivas…

81