Cost-effectiveness of selective dry cow therapy

Similar documents
Proper Dry-Off Procedures to Prevent New Infections and Cure Existing Cases of Mastitis. Stephen C. Nickerson University of Georgia

Mastitis MANAGING SOMATIC CELLS COUNTS IN. Somatic Cell Count Are Affected by. Somatic Cells are NOT Affected by:

Decision tree analysis of treatment strategies for mild and moderate cases of clinical mastitis occurring in early lactation

MILK COMPOSITIONAL CHANGES DURING MASTITIS

Mastitis: Background, Management and Control

DRY YOUR BEST. The incidence and importance of milk leakage in the dry cow. Ana I. de Prado Corporate Technical Manager Ceva Santé Animale (France)

The Bimeda Guide to Selective Dry Cow Therapy

MASTITIS CASE MANAGEMENT

Last 2-3 months of lactation

Milk Quality Management Protocol: Fresh Cows

A Partial Budget Model to Estimate Economic Benefits of Lactational Treatment of Subclinical Staphylococcus aureus Mastitis

Evaluation of the use of dry cow antibiotics in low somatic cell count cows

Lactation. Macroscopic Anatomy of the Mammary Gland. Anatomy AS 1124

Evaluation of intervention strategies for subclinical and clinical mastitis

, Pamela L. Ruegg

Mastitis cows and immunization

Mastitis in ewes: towards development of a prevention and treatment plan

Selective dry cow therapy and the influence of milk yield at drying-off under Dutch practical circumstances

Understanding the Basics of Mastitis

Premiums, Production and Pails of Discarded Milk How Much Money Does Mastitis Cost You? Pamela Ruegg, DVM, MPVM University of Wisconsin, Madison

Options for Handling Mastitis during Lactation in Modern Dairy Farms

MASTITIS. Therefore, mastitis is an inflammation of the mammary gland.

Milk quality & mastitis - troubleshooting, control program

The mastitis situation in Canada where do you stand?

Economics of mastitis. Kirsten Huijps and Henk Hogeveen

TECHNOTE 14. Decide dry cow management strategy DRYING OFF

DISSERTATION. Paige Nicole Gott, M.S. Graduate Program in Comparative and Veterinary Medicine. The Ohio State University. Dissertation Committee:

Milk Secretion and Quality Standards

Responsible Use of Antimicrobials in Dry Cow Strategies

Minna Koivula & Esa Mäntysaari, MTT Agrifood Research Finland, Animal Production Research, Jokioinen, Finland

TREATMENT DECISIONS FOR MILD AND MODERATE CASES OF CLINICAL MASTITIS. Carolina Pinzón-Sánchez

AUTOMATIC MILKING SYSTEMS AND MASTITIS

Strep. ag.-infected Dairy Cows

Statistical Indicators E-27 Breeding Value Udder Health

Controlling Contagious Mastitis

Quality Milk on Pasture Based Dairy Farms. Scott E. Poock, DVM University of Missouri Clinical Assistant Professor DABVP Beef and Dairy Cattle

Presented at Central Veterinary Conference, Kansas City, MO, August 2013; Copyright 2013, P.L Ruegg, all rights reserved

New Developments in Mastitis Research

Differential Somatic Cell Count with the Fossomatic 7 DC - a novel parameter

Selective Dry Cow Therapy

Walter M. Guterbock, DVM, MS Veterinary Medicine Teaching and Research Center University of California, Davis

LOOKING FOR PROFITS IN MILK QUALITY

Using SCC to Evaluate Subclinical Mastitis Cows

Case Study: Dairy farm reaps benefits from milk analysis technology

Milk Quality Evaluation Tools for Dairy Farmers

Outline MILK QUALITY AND MASTITIS TREATMENTS ON ORGANIC 2/6/12

Influence of Experimentally- induced clinical mastitis on Reproductive Performance of Dairy Cattle

Your clients need you to help them reach their full dairy potential.

Economic effects of bovine mastitis and mastitis management: A review

MATERIALS AND METHODS

Mastitis Management and SCC Control in Once a Day Herds. Don Crowley- Teagasc

2012 Indiana Regional Dairy Meetings. Purdue University College of Veterinary Medicine Dr. Jon Townsend Dairy Production Medicine

Incidence of subclinical mastitis in Dutch dairy heifers in the first 100 days in lactation and associated risk factors

29/11/2017. Best Milking Practices. Greg Strait- Fulton County Extension Amber Yutzy- Huntingdon County Extension

Somatic Cell Count as an Indicator of Subclinical Mastitis. Genetic Parameters and Correlations with Clinical Mastitis

Effect of omitting post-milking teat disinfection on the mastitis infection rate of dairy cows over a full lactation

Decide dry cow strategy - drying off dates, antibiotic Dry Cow and Internal Teat Sealant treatments?

Mastitis Reminders and Resources LAURA SIEGLE EXTENSION AGENT VIRGINIA COOPERATIVE EXTENSION AMELIA COUNTY

Practical Strategies for Treating Mastitis Pamela L. Ruegg, DVM, MPVM University of Wisconsin, Madison

Mastitis Prevention and Cure Rates in Heifers Treated with Spectramast Dry Cow Therapy and/or Orbeseal Dry Cow Teat Sealant

Caused by microorganisms (usually bacteria) that invade the udder, multiply, and produce toxins that are harmful to the mammary gland

Guideline on the conduct of efficacy studies for intramammary products for use in cattle

Genetic parameters for pathogen specific clinical mastitis in Norwegian Red cows

Comparison of different methods to validate a dataset with producer-recorded health events

How to Decrease the Use of Antibiotics in Udder Health Management

DeLaval Cell Counter ICC User Strategies Guide

8. MILK Udder health and milk quality is important to farm productivity and profitability. Every farm stands to gain from less mastitis.

Update on Staphylococcus aureus Mastitis. John R. Middleton College of Veterinary Medicine, University of Missouri, Columbia

Field Efficacy of J-VAC Vaccines in the Prevention of Clinical Coliform Mastitis in Dairy Cattle

ENVIRACOR J-5 aids in the control of clinical signs associated with Escherichia coli (E. coli) mastitis

Best practice guide for on-farm mastitis control

Check that milk is suitable to go in the vat

MILK QUALITY PROGRAMS FOR TRANSITION COWS AND HEIFERS. Leo Timms Iowa State University, Ames IA

Trouble-Shooting a Mastitis Problem Herd 1

The organism Infection process Tissue reaction SCC response Prevention Treatment

Cepravin and Combination Dry Cow Therapy Trial Work

MASTITIS DNA SCREENING

Mastitis and On-Farm Milk Cultures - A Field Study - Part 1

Emerging Mastitis Threats on the Dairy Pamela Ruegg, DVM, MPVM Dept. of Dairy Science

Prototheca Mastitis in Dairy Cows

Actions and Outcomes of Wisconsin Dairy Farms Completing Milk Quality Teams

Management and treatment of summer cattle mastitis

Ubroseal Dry Cow 2.6 g intramammary suspension for cattle

TEAT DIP- POST DIP- PRE DIP- STRIPING

Prevention of clinical and subclinical mastitis

International Journal of Science, Environment and Technology, Vol. 6, No 2, 2017,

Herd Navigator and mastitis management

Dr. Michelle Arnold, DVM DABVP (Food Animal) Ruminant Extension Veterinarian University of Kentucky Veterinary Diagnostic Laboratory

Rate of transmission: A major determinant of the cost of clinical mastitis

MALLA HOVI & STEVE RODERICK, Veterinary Epidemiology and Economics Unit, University of Reading, PO Box 236, READING RG6 6AT

Management Practices and Intramammary Infections: New Ideas for an Old Problem

Association between teat skin colonization and intramammary infections with Staphylococcus aureus and Streptococcus agalactiae

A Review on Mastitis Control through Dry Cow Therapy

REPORT FROM THE. 3 rd GLOBAL MILK QUALITY EXPERT FORUM

1 st EMP-meeting: European boom in AMS and new tools in mastitis prevention

Dairy/Milk Testing Report Detecting Elevated Levels of Bacteria in Milk-On-Site Direct- From-The-Cow Within Minutes as Indicator of Mastitis

Experts on mastitis evaluate the efficiency of management measures to prevent the disease

Mathematical and Molecular Epidemiology of Subclinical Mastitis Treatment in Lactating Dairy Cows

Using DHIA and bacteriology to investigate herd milk quality problems.

Milking behaviour in dairy cows naturally infected with clinical mastitis

Transcription:

Wageningen University - Department of Social Sciences MSc Thesis Chair Group: Business Economics (BEC) Cost-effectiveness of selective dry cow therapy Optimization of selecting animals for selective dry cow therapy based on cow-level characteristics September 2017: Master Animal Sciences Ilonka van der Wagt Name of Supervisor: dr. M (Mariska) van der Voort Thesis code: BEC-80424

Cost-effectiveness of selective dry cow therapy Optimization of selecting animals for selective dry cow therapy based on cow-level characteristics Name: Ilonka van der Wagt Registration number: 940703924030 Animal Sciences Minor thesis Business Economics Course code: BEC-80424 Supervisor: dr. M (Mariska) van der Voort Date: 8 September, 2017 Place of publication: Wageningen University, Wageningen

Preface This minor thesis report is part of my master program Animal Science and contains work done from October 2016 to July 2017. As I am an Animal Science student with limited economic background, I combined my knowledge about production animals with economy. This thesis is mostly based on literature, that was found by consulting the (digital) library of Wageningen University and has been used to acquire knowledge about subjects to write the body of the thesis. I would like to express my thanks to my supervisor Mariska van der Voort for the good guidance and support during the past months. Ilonka van der Wagt Spijkenisse, 08-09-2017

Abstract A lot of research has already been done on the mastitis management on a dairy farm. To control mastitis, blanket dry cow therapy (BDCT) was long successfully in many parts of the world. However, due to a growing concern about antibiotic use in animal husbandry, BDCT is no longer allowed in several European countries, including the Netherlands. Selective dry cow therapy (SDCT) had been proposed as an alternative of BDCT. In the Netherlands, the selection of cows to receive dry cow therapy (DCT) is based on somatic cell count (SCC). However, many other cow characteristics (i.e. parity, infection history, breed, number of quarters infected, teat morphology and lactation number) can play a role in the development of an intramammary infection (IMI). In this study, we did an optimization of the selection criteria (parity, SCC and infection history of mastitis) to select cows for SDCT. This optimization was done for different farm Bulk Tank Somatic Cell Count (BTSCC) categories (low, average and high) in different scenarios with different selection criteria for the cows to receive SDCT. The goal was to minimize the total costs related to SDCT. This was done using Linear Programming (LP). A fictive herd of 100 dairy cows was created to determine which selection criteria is most cost-effective for the dairy farmer. Antibiotic use and costs related to SDCT were evaluated from the start of dry period until the first 100 days in milk (DIM). The total costs of SDCT varied from 4,295 to 9,488. The occurrence of subclinical mastitis in the fictive herd varied from 8.50% to 28.89% and the occurrence of clinical mastitis varied from 12.63% to 28.77%. The total antimicrobial usage for SDCT and clinical mastitis treatment varied from 3.0 to 4.0 animal daily dosages. Farms with a low BTSCC, no infection history and different thresholds SCC for drying off has the lowest total costs, costs per animal and antibiotic use related to SDCT. Total costs, costs per animal, antibiotic use and percentage of (sub)clinical cases were higher in farms where cows had an infection history of mastitis compared to farms where cows had no mastitis infection history. A sensitivity analysis was performed to assess the effect of varying the input values (antibiotic use and risk on mastitis after a previous infection) on the outcome (total costs and antibiotic use). The total costs, costs per animal and the percentage of (sub)clinical cases increases with a stricter antibiotic constraint. This was also the case with a higher risk on (sub)clinical mastitis after a previous mastitis infection. It can be concluded that drying off cows with antibiotics influences the costs related to SDCT and antibiotic use. The costs of mastitis around dry period are sensitive to a change in risk on mastitis and antibiotic restriction. However, the optimal decision to dry off cows with antibiotics depends on the farmer. Keywords: (sub)clinical mastitis, antibiotics, dry cow therapy, economics, dry period, linear programming

Table of Contents Abstract... iv 1. Introduction... 1 1.1 Background information... 1 1.2 Research problem... 2 1.3 Objective... 2 1.4 Outline... 3 2. Literature review... 5 2.1 Dry period and intramammary infections... 5 2.1.1 Dry period... 5 2.2.1 Intramammary infections... 5 2.2 Mastitis in the Netherlands... 7 2.2.1 National mastitis control program... 7 2.2.2 Economic impact... 7 2.2.3 Antibiotic use... 8 2.3 Preventive measurements of subclinical and clinical mastitis... 9 2.3.1 Blanket dry cow therapy... 9 2.3.2 Selective dry cow therapy... 10 2.3.3 Teat sealant... 10 2.4 Individual cow characteristics... 11 2.4.1 Somatic Cell Count (SCC) and parity... 11 2.4.2 Milk production and immunosuppression... 11 2.4.3 Infection history... 12 3. Materials and methods... 15 3.1 Model description... 15 3.1.1 Linear Programming... 15

3.1.2 Age groups... 15 3.2 Model parameters... 17 3.2.1 Economic costs... 17 3.2.2 Antibiotic use... 18 3.2.3 Risk of mastitis... 19 3.2 Scenarios... 21 3.3 Sensitivity analysis... 21 4. Results... 23 4.1 General results... 23 4.2 Scenarios... 25 4.3 Sensitivity analysis... 25 4.3.1 Antibiotic restriction... 25 4.3.2 (Sub)clinical mastitis risk... 28 5. Discussion... 31 6. Conclusion... 35 References... 37

1. Introduction 1.1 Background information Farmers use dry cow management from the two weeks prior until the two weeks following drying-off and includes attention to proper procedures for drying off dairy cows and concern about the cow s environment (Dingwell et al., 2004). The goal of dry cow management is to have healthy cows starting a new lactation with an uninfected healthy mammary gland (Dingwell et al., 2003). Although it is known that the dry period is an important time for optimizing udder health, the mammary gland is more susceptible to intramammary infections (IMI) during this period (Henderson et al., 2016). This increased susceptibility to IMI during the dry period may be affected by (1) the rate of exposure to potential pathogens, (2) the susceptibility of an individual cow to infection and (3) the effectiveness of protection from medical interventions (Green et al., 2007). IMI are a multifactorial problem and contributing factors are related to cow characteristics, management and environmental conditions, as well as interactions between these factors (Barkema et al., 1999; Huijps, 2009). IMI, such as mastitis, are one of the most frequently occurring health disorders in dairy cows and it has a negative economic as well as a negative welfare impact. Mastitis causes a reduction in animal welfare (Peters et al., 2015; Kemp et al., 2008; Schukken et al., 2003) since cows are suffering from pain and discomfort (Valeeva et al., 2007). Reduced animal welfare can also result in a reduction in the job satisfaction of the dairy farmer (Valeeva et al., 2007), since cows that suffer from mastitis are less easy to approach (Ivemeyer et al., 2011). This can affect the human-animal relationship and the working routine of the farmer (Jansen, 2010). Besides the negative impact on animal welfare, mastitis is considered to be the costliest health issue in dairy cows (Halasa et al., 2007). Economic losses related to clinical or subclinical mastitis are increased veterinary costs, increased replacement costs due to premature culling, product quality changes, increased labour costs, milk production losses and indirectly a higher susceptibility for other diseases (Halasa et al., 2007). Hogeveen et al. (2011) estimated the losses due to clinical and subclinical mastitis in the Netherlands between 17 and 198 per cow per year, with an average of 78 per cow per year. Heikkilä et al. (2012) estimated the cost of clinical mastitis in Finland at 458 per CM case or 147 per cow per year. Although the estimations of the costs differ between studies and countries, it is clear that mastitis has a great negative economic impact for the dairy farms. There are several measures that can be used to prevent a mastitis infection during the dry period (Godden et al., 2003). Those measures are blanket dry cow therapy (BDCT), selective dry cow therapy (SDCT) and an internal/external teat sealant. In many countries, BDCT is the standard way to dry off cows (Huijps and Hogeveen, 2007). During BDCT, all four quarters of the udder are treated with a long-acting antibiotic at drying off. BDCT has two important functions: (1) to eliminate IMI present at drying off and (2) to prevent new infections by protecting the healthy quarters against IMI caused by (environmental) pathogens (Huijps and Hogeveen, 2007). According to St. Rose et al. (2003), the dry period is the best period to use antibiotics because of the long duration of antibiotic treatment, which makes it more effective. Another reason to use antibiotics during the dry period is the fact that they have no effect on the milk production of the cow, because the cow is in a non-lactating phase. 1

However, there is a growing public concern about the use of antibiotics in animal husbandry. An overuse of antibiotics results in antibiotic resistance and residues of antibiotics in the food chain (Bhutto et al., 2011). The Dutch animal production industry signed a covenant in 2008 ( Convenant antibioticaresistentie dierhouderij ) to reduce the resistance of antibiotics by promoting a responsible use of veterinary antibiotics by means of a selective and restrictive use of antibiotics in food-producing animals. This has led to the introduction of SDCT in the Dutch dairy sector in 2013 as an alternative for BDCT. During SDCT only cows with an IMI are selected to receive antibiotics at drying off. In the Netherlands, this selection is based on the somatic cell count (SCC) in the milk. It is only allowed to use antibiotics at drying off in primiparous cows with a SCC >150,000 cells/ml and in multiparous cows with a SCC >50,000 cells/ml according to the last milk recording test day up to six weeks before drying off (KNMvD, 2013). In this way, unnecessary antibiotic use can be avoided and this has produced positive results already in the Netherlands (Berry and Hillerton, 2002; Speksnijder et al., 2015). Other preventive measures are the use of an intramammary teat sealant or external teat sealant. An internal teat sealant, made from bismuth subnitrate in a paraffin base, has been available since 1978. This inert viscous paste is infused into the quarter at the time of drying off forming an immediate physical barrier in the distal portion of the teat cistern. This prevents bacteria from ascending through the teat canal (Berry and Hillerton, 2002; Godden et al., 2003). There is also an external teat seal available. The materials in the external teat sealant dry and generate a latex, acrylic or other polymer-based film over the teat. This prevents the entry of pathogens into the teat canal (Godden et al., 2003). Several studies found that application of an internal teat seal to all quarters at drying off results in significantly fewer new IMI at calving, less (sub)clinical mastitis during the dry period and fewer cases of clinical mastitis in the 100 days of the subsequent lactation compared to no treatment (Berry and Hillerton, 2002). 1.2 Research problem Cost of mastitis can be looked upon at different levels as quarter level, cow level, herd level, regional level, country level and world level (Østerås, 2000). At each level an impression of the cost will be important for decision making. Already many studies investigated different forms of dry cow management and their economic consequences (Hogeveen et al., 2011; Halasa et al. 2007; Scherpenzeel et al., 2014; 2016). However, many of these studies mainly focused on farm level and not on individual cow level. Individual cow characteristics such as breed, parity, infection history, SCC, number of quarters infected, teat morphology and lactation number play a role in the development of new IMI (Oviedo-Boyso et al., 2007; Bakken, 1981). These individual cow characteristics can therefore be used as a selection criteria for SDCT to determine which cows receive antibiotics at drying off. These characteristics influences the risk on IMI and thus the treatment and economic consequence. It is therefore important to take (some of) these characteristics into account in the calculation of the expected economic losses. Cost-effective SDCT need to be estimated based on the risk to infection of the individual cow. In this way, farmers can offer a cow-specific treatment that is tailor made for the individual animal and the costs related to SDCT can be estimated more accurate. 1.3 Objective This study will focus on SDCT, as it is the most used method in the Netherlands. Given the variety in possible approaches for selecting cows for SDCT when implementing SDCT in practice, and the consequences of that for economics, selection criteria need further attention. 2

Therefore, in this study, we optimized selecting animals for SDCT, taking into account variation in parity, cow-level SCC and the presence of an infection history at drying off. The objective of this study was to evaluate antibiotic use and economics related to SDCT when using different scenarios (with different selection criteria) for selecting cows for SDCT. The goal is to minimize the total costs related to SDCT. To reach this objective, the following research questions will be answered: 1. How do individual cow characteristics contribute to the development of an IMI? 2. What are the possible selection criteria for drying off with antibiotics? 3. What are the expected costs of selective dry cow therapy? 4. Which selection criteria for SDCT is the most cost-effective strategy? 1.4 Outline This report is divided into different chapters. Chapter 2 presents a literature review. This literature review describes the dry period, intramammary infections in dairy cows in the Netherlands, possible preventive measures for IMI and several individual cow characteristics that can contribute to the development of an IMI are discussed. Chapter 3 presents the materials and methods, which gives information about the model and the parameters that will be used. Chapter 4 outlines the results and in chapter 5 these results are discussed. Chapter 6 gives an overall conclusion. 3

4

2. Literature review 2.1 Dry period and intramammary infections 2.1.1 Dry period The dry period is the non-lactating period between two active lactating phases to maximize milk production in the subsequent lactation (Jánosi and Huszenicza, 2001). It is a period of anatomical, metabolic and physiological changes for many body systems, including the mammary gland (Petzer et al., 2009). The mammary gland undergoes biochemical, cellular, and immunologic changes (Erskine, 2016). Generally speaking, the mammary gland changes consist of three phases (Arnold and Bewley, 2012). Phase 1: Involution This phase prepares the mammary gland for stopping milk production and starts immediately following drying off. Milk accumulates in the udder causing changes in the structures and secretions in the gland, decreased secretory activity and increased pressure. During this phase, there is an increased risk of IMI, because there is no flushing of bacteria from the streak canal, no teat dip protection and milk leakage. Normally, a natural keratin plug is formed to prevent the entry of bacteria into the gland, but studies have shown that 3-5% of the quarters never form a functional keratin plug (Williamson et al., 1995). This process of active involution is most likely completed within 21 days after drying off (Smith and Todhunter, 1982). Phase 2: Steady state involution Once fully involuted, the mammary gland is protected to infection. During this phase, protective factors inhibit bacterial growth in the udder. The duration of this steady state involution depends on the length of the dry period (Jánosi and Huszenicza, 2001). Smith and Todhunter (1982) indicated that a dry period shorter than 40 days have a decreased hormonally lactogenic response which may result in a suboptimal milk production in the next lactation. Phase 3: Colostrogenesis This phase starts when calving approaches and prepares the udder for milk synthesis. There is growth of mammary tissue and increased secretory activity in the last two weeks of gestation. There is an increased susceptibility to infection as the keratin plug breaks down, leukocyte function is impaired (the protective white blood cells do not work as well), and leakage of colostrum often occurs. 2.2.1 Intramammary infections Especially in the dry period, the mammary gland is more susceptible to new IMI (Jánosi and Huszenicza, 2001). The flushing effect of milking on bacteria colonizing in the teat canal is terminated and there is an increased intramammary pressure that may cause leakage of milk and facilitate bacterial penetration (Jánosi and Huszenicza, 2001). Also, the defence mechanism of the mammary gland is low during the dry period: low numbers of lymphocytes, polymorphonuclear neutrophils and a low concentration of immunoglobulins and lactoferrin (Oliver and Sordillo, 1989). In literature, mastitis is often discussed when it comes to IMI. Mastitis is an inflammation of the mammary gland that develops as a response to pathogenic microorganism. These microorganisms enter through the teat canal and multiply inside the gland. During this 5

inflammatory response, the immune system of the mammary gland is activated to eliminate the pathogen (Oviedo-Boyso et al., 2007). This defence mechanism consists of anatomical, cellular and soluble factors that coordinate. Many cells are involved in this defence mechanism. For example, the white blood cells (leukocytes) fight the infection and assist in the repair of damaged tissue in the udder (Sharma et al., 2011). These leukocytes in the mammary gland release chemo-attractive products for especially neutrophils. This causes an increase in somatic cells (neutrophils) in the milk. This is also known as an increase in SCC in the milk (Sharma et al., 2011; Oviedo-Boyso et al., 2007). Somatic cells are therefore indicators of resistance and susceptibility of cows to intramammary infections such as mastitis. Figure 2.1 shows a schematic overview of the process of infection and the cow s defence to the infection. Figure 2.1: Schematic overview of the development of mastitis and the cow's defence against the infection. Bacteria invaded the teat canal and if they are not entirely destroyed, they continue to multiply and begin to invade the smaller ducts and alveolar areas (A). Milk-secreting cells are damaged by toxins and other irritants release substances that lead to increased permeability of the blood vessels (B). Additional leukocytes move to the site of infection, enter the alveolar tissue by squeezing between damaged milksecreting cells (C). As the infection persists and ducts remain clogged, the entrapped milk causes the secretory cells to revert to a resting (non-producing) state and the alveoli begin to shrink (D). Substances released by leukocytes lead to the complete destruction of alveolar structures, which are replaces by connective and scar tissue (E and F) (Wattiaux, 1995). There are two forms of mastitis: subclinical mastitis (SCM) and clinical mastitis (CM). Subclinical mastitis is the presence of a mammary infection without clear signs of local inflammation or systemic involvement. No visible changes occur in the milk, but milk production decreases, bacteria are present in the milk and the composition is altered (Harmon, 1994; Wattiaux, 1995). Clinical mastitis is an inflammatory response to the infection causing abnormal milk and swelling or pain in the udder. It may be accompanied by systemic signs such as an elevated rectal temperature, or anorexia (Erskine, 2016; Harmon, 1994). 6

Studies have shown that over 60% of the new IMI occur during the dry period (Todhunter et al., 1991) (Figure 2.2). Figure 2.2: Illustration of the incidence of new intramammary infection during the lactation cycle (Bradley and Green, 2004) 2.2 Mastitis in the Netherlands 2.2.1 National mastitis control program Mastitis is a worldwide problem and is an important disease because of its common occurrence and its significant economic effect (Huijps et al., 2008; Lam et al., 2013). In the Netherlands, the last representative estimate of CM was conducted as a part of the national udder health program in 2009 (Lam et al., 2013). They calculated the costs of CM based on the estimated incidence rate of clinical mastitis (IRCM) during the years 2004 and 2009. In 2004, the Dutch dairy industry decided to invest in a 5-year national mastitis control program after field reports indicated that a substantial number of dairy herds were experiencing problems with CM (Lam et al., 2013). The goal of this control program was to decrease the IRCM by 10 percentage points (van der Zwaag et al., 2005). The control program was run by the Dutch Udder Health Centre (UGCN). The UCGN served as an independent information source on udder health for farmers and stakeholders from the dairy sector (Lam et al., 2013). 2.2.2 Economic impact In the Netherlands, the estimated total costs of mastitis varies from 17 to 198 per cow per year (Huijps et al., 2008). The study of Huijps et al. (2008) also found that most farmers (72%) expected lower costs. Underestimating the economic losses of mastitis is a general problem in the dairy sector. Most economic losses are related to decreased milk production, premature culling and (antibiotic) treatment (van Soest et al., 2016). In both subclinical and clinical mastitis, there is a loss in milk production (Grohn et al., 2004). It is a hidden cost, because the loss in milk production is not visible for the farmer as it is never produced (Halasa et al., 2007; Østerås, 2000). There is also a loss of milk due to 7

discarded milk. Milk from cows treated with antibiotics must be discarded for 3 or 4 days, because of antibiotic residues in the milk. This discarded milk can be fed to the calves on the farm, although it is not advisable from a veterinary point of view (Halasa et al., 2007). This economic damage is comparable with the decrease in milk production, but differs at one point; the discarded milk is actually produced by the cows, which means that feeding costs for the amount of milk produced should be taken into account in the economic calculations. Mastitis also influences the quality of the milk (Hoblet and Miller, 1991; Østerås, 2000). Some of these quality changes cause a less efficient processing of the milk and might result in products with less favourable properties (Santos et al., 2003). Examples are a lower cheese yield, longer renneting time in cheese production, less stability of the cheese structure and taste and a longer whipping time for cream (Hortet and Seegers, 1998; Østerås, 2000). Another cost are the veterinary services. A veterinarian visit is needed to diagnose a mastitis case or for veterinary advice related to (sub)clinical mastitis (Halasa et al., 2007). Some cows need drugs to treat the mastitis infection. The duration of the treatment varies between cows and may vary from days to weeks to months (Halasa et al., 2007). Cows with mastitis have a higher risk of being culled (Cha et al., 2013). Premature replacement costs of cows due to mastitis is probably one of the largest economic cost (Halasa et al., 2007). These costs involve rearing or buying a replacement animal and a possible decreased efficiency of milk production by the replacement animal (Halasa et al., 2007). Next to economic losses, there are also economic benefits from (sub)clinical mastitis. Seegers et al. (2003) performed a partial budget simulation to calculate the economic impact of mastitis and found several economic benefits. For example, there is an increase in revenues due to higher meat sales (extra culls) and higher calf sales due to extra replacement heifers. Mastitis also causes a decrease in certain costs, such as lower use of concentrates, since cows suffering from (sub)clinical mastitis eat less concentrates (Seegers et al., 2003). If discarded milk is given to the calves, there is a lower use of milk replacer. 2.2.3 Antibiotic use Agriculture accounts for the highest volume of antibiotic use (Cogliani et al., 2011). Globally, an estimated 50% of all antimicrobials serve veterinary purposes (Teuber, 2001). Repeated exposure of microorganisms to antimicrobials increase the selection pressure for antibiotic resistance (Cully, 2014). The Dutch government began to impose fines for overuse of antibiotics in 2009; veterinary consumption of antibiotics subsequently dropped by more than 50% within three years (CLO, 2017). Between 2009 and 2014, the agricultural antibiotic use in the Netherlands decreased with 58%. However, the goal is to have 70% reduction in antibiotic use (CLO, 2017). Therefore, action is still needed. Since 2011, the Netherlands Veterinary Medicines Authority (Stichting Diergeneesmiddelenautoriteit, SDa) has been monitoring antibiotic use at Dutch livestock farms by means of benchmark indicators (SDa, 2014). Specific benchmark indicators have been defined for the various livestock sectors and types of livestock farms. Collected data from livestock farms facilitated detailed reporting by the SDa expert panel on developments regarding antibiotic use at Dutch livestock farms. The 'Defined Daily Dose Animal' (DDDA) is used to calculate the antibiotic use of a farm. It is determined by first calculating the total number of treatable kilograms at the farm for a specific year, and then dividing this number by the average number of kilograms of animals present at the farm. It reflects the amount of 8

antibiotics used at dairy farm level. The DDDA is used to set two benchmark values: a signalling threshold and an action threshold, defining by three zones (Table 2.1). 1. The target zone: the amount of antibiotics used is equal to or below the signaling threshold. 2. The signaling zone: the amount of antibiotics used exceeds the signaling threshold but is below the action threshold. If a livestock farm's antibiotic use falls within this zone, further attention is required and action may have to be taken to reduce the amount of antibiotics used. 3. The action zone: the amount of antibiotics used exceeds the action threshold. The livestock farmer should take action immediately to quickly reduce the amount of antibiotics used. Table 2.1: Benchmark values for dairy cattle (SDa, 2014) Threshold values for individual livestock farms (DDDA F) Target zone Signaling zone Action zone Dairy cattle 0-4 >4-6 >6 2.3 Preventive measurements of subclinical and clinical mastitis There are several measures that can be used to prevent IMI during the dry period (Godden et al., 2003). Those measures are BDCT, SDCT and an internal/external teat sealant. 2.3.1 Blanket dry cow therapy In many countries, BDCT is the standard way to dry off cows (Huijps and Hogeveen, 2007; Higgings et al., 2017). During BDCT, all four quarters of the udder are treated with a long-acting antibiotic immediately after the last milking of the lactation (Godden et al., 2003). The goal is to reduce the prevalence of IMI by eliminating existing IMI at drying off and to prevent new IMI from occurring during the dry period. Studies have shown that cows treated with antibiotics at drying off have less CM infections compared with untreated cows (Huijps and Hogeveen, 2007; Scherpenzeel et al., 2016). However, according to Berry and Hillerton (2002), many cows have not adequate antibiotic protection for the whole period and can thus be susceptible to a mastitis infection later in the dry period. In the Netherlands, approximately 90% of all dairy cows were treated with dry-cow antimicrobials in the period from 2005 to 2010. In the United Kingdom, this is estimated to be even higher, with 99% of the dairy cows treated at drying off (Scherpenzeel et al., 2016). However, due to public concerns about antimicrobial resistance and food safety, preventive use of antimicrobials has become questionable (Scherpenzeel et al., 2016). For many years, approximately 60% of the antimicrobial use in dairy cattle was related to mastitis, of which 67% was related to DCT (Kuipers et al., 2016). The Dutch animal production industry signed a covenant in 2008 ( Convenant antibioticaresistentie dierhouderij ) to reduce the resistance of antibiotics and to promote a responsible use of veterinary antibiotics by means of a selective and restrictive use of antibiotics in food-producing animals (Halasa et al., 2010). Therefore, in the Netherlands, preventive use of antimicrobials, like with BDCT, is no longer allowed. As result SDCT was proposed as an alternative in 2013 (Scherpenzeel et al., 2016). 9

2.3.2 Selective dry cow therapy SDCT has been proposed as an alternative to BDCT to avoid unnecessary antibiotic use in animal husbandry (Huijps and Hogeveen, 2007). With SDCT, only cows with an IMI are selected to receive antibiotics at drying off (Halasa et al., 2010). To select cow for SDCT, IMI at drying off need to be identified. This identification can be based on different criteria, such as SCC, bacterial culture and CM history (Scherpenzeel et al., 2016). In the Netherlands, antibiotic treatment is only allowed in primiparous cows with a SCC higher than 150,000 cells/ml and in multiparous cows with a SCC higher than 50,000 cells/ml according to the last milk recording (KNMvD, 2013). However, there can be differences between the average SCC before drying off and SCC at drying off (Den Uijl et al., 2012). The longer the period between last milk recording and drying off, the greater the difference in SCC. It is therefore strongly advised to dry the cow off as soon as possible after the last milk recording. 2.3.3 Teat sealant With an increasing interest in organic production, there is also interest in a reduction in the use of antibiotics as prophylactics (Berry and Hillerton, 2002). Some of the prophylactic antibiotics are in categories that contain antibiotics, or are in a class that contains antibiotics of value in human medicine. This overuse of antibiotics in animal husbandry promote the selection of antibiotic resistance in bacterial populations. The resistant bacteria from agricultural environments may be transmitted to humans, in whom they cause disease that cannot be treated by conventional antibiotics (Khachatourians, 1998). Normally, the cow will develop a natural keratin plug in the teat canal at dry-off, but research has shown that about 50% of all cows do not form a keratin plug within two weeks after drying off and some cows never develop this plug. This leaves the teat canal open for infection (Hutjens and Aalseth, 2005). A measure that can be used to prevent new IMI during the dry period is an external teat sealant. No antibiotics are used with this measure. Once applied, the products in this sealant dries and generates a latex, acrylic or other polymer-based film over the teat that prevents entry of pathogenic bacteria into the teat canal (Godden et al., 2003). However, Berry and Hillerton (2002) have had only limited success in reducing IMI during the dry period with teat sealers, due to poor persistence on the teat. An alternative of an external teat sealant is the use of an internal teat sealant (Godden et al., 2003). An internal teat sealant made from bismuth subnitrate in a paraffin base has been available since 1978 (Godden et al., 2003). This inert viscous paste is infused into the quarter at the time of drying off forming an immediate physical barrier in the distal portion of the teat cistern to prevent bacteria from ascending through the teat canal (Berry and Hillerton, 2002; Godden et al., 2003). However, successful implementation of this treatment approach may be a challenge, because it will require careful attention to infusion techniques to avoid the introduction of pathogens into an unmedicated quarter. In North-America, internal teat sealants are most commonly used in combination with intramammary antibiotics (NMC, 2006). In the Netherlands, teat sealant as therapy is rare, probably because it is time-consuming, challenging to implement and not 100% reliable (Huijps and Hogeveen, 2007). 10

2.4 Individual cow characteristics On an increasing number of dairy farms, a lot of information on individual cows is automatically available via management programs (Steeneveld et al., 2011). Individual cow characteristics such as breed, parity (age), infection history, SCC, number of quarters infected, teat morphology and lactation number play a role in the development of new IMI (Oviedo-Boyso et al., 2007; Bakken, 1981). As they play a role in the development of an IMI, they also influence the treatment. Some of these cow characteristics will be discussed below. 2.4.1 Somatic Cell Count (SCC) and parity The normal composition of somatic cells in the milk varies and depends on the stage of lactation, age, parity and lactation cycle (Sharma et al., 2011). According to Bytyqi et al. (2010), the milk of a healthy mammary gland has a SCC of lower than 100,000 cells/ml, while during a bacterial infection this can increase to above 1,000,000 cells/ml. The SCC is used to identify cows that are subclinically infected with a certain mastitis pathogen. The use of SCC to measure the inflammatory response to an IMI is nowadays a practical tool to measure udder health (Scherpenzeel et al., 2014). Farms often use bulk tank somatic cell count (BTSCC) as a well-established parameter to describe milk quality and to diagnose multiple problems that might be present in a dairy herd related to milk quality and the presence of mastitis pathogens (Jayarao et al., 2004). As mentioned earlier, the SCC level depends on parity. Multiparous cows have an increased prevalence of infection compared to primiparous cows (Zadoks et al., 2001; Leelahapongsathon et al., 2014). This may be because the anatomical changes in the teat over time cause disruption of the natural defence mechanism and there is a decrease in the integrity of the streak canal (Cousins et al., 1980). This might cause a systemic reduction in immune capability, associated with aging that increases susceptibility to infection (Green et al., 2007). In the Netherlands, a SCC threshold of >250,000 cells/ml is commonly used to indicate IMI in multiparous cows, whereas >150,000 cells/ ml is used in primiparous cows (Santman-Berends et al., 2012; Sampimon et al., 2010). 2.4.2 Milk production and immunosuppression Several studies found a link between high milk yield and the incidence of CM. Milk yield is genetically unfavourably correlated with disease resistance (Hooijer et al., 2001; Carlén et al., 2004; König et al., 2008). The average genetic correlation between mastitis and milk yield is 0.43 (Sander Nielsen et al., 1996; Luttinen and Juga, 1997; Lund et al., 1999). Rajala- Schultz et al. (2005) found that the odds ratio for a cow being infected with environmental pathogens increase with 77% with every time milk yield at drying off is increased with 5 kg. Even tough milk is no longer removed from the mammary gland during the dry period, the mammary gland continues to synthesize milk during the involution phase, causing milk accumulation and leakage via the teats and thereby facilitating the entry of microorganisms into the teat canal and colonize the mammary gland (Cousins et al., 1980). Schukken et al. (1993) found that cows that leaked milk following drying off were four-times more likely to develop CM. The natural keratin plug in the teats forms faster in cows with a milk production of <21 kg at drying off (Dingwell et al., 2004). Also, the increased volume of milk in the mammary gland contains lower concentrations of natural protective factors, such as lactoferrin, phagocytotic cells and immunoglobulins 11

(Bushe and Oliver, 1987; Paape et al., 1992). Thus, the risk of acquiring a new infection during the dry period increases rapidly with the level of milk production (Rajala-Schultz et al., 2005). Drying off includes reducing the concentrate intake of a cow (Drackley, 1999). However, a drastic reduction of concentrates for high yielding cows may provide severe metabolic stress and may induce health problems like mastitis (Odensteen et al., 2007). Metabolic stress compromises the host s immune defences, meaning that high producing dairy cows have a reduced immune competence. This can lead to a higher incidence of CM in high producing cows compared to low producing cows (Walsh et al., 2011). 2.4.3 Infection history The major causative agents of mastitis in modern Dutch dairy herds are Escherichia coli, Staphylococcus aureus, Streptococcus uberis and Streptococcus dysgalactiae (Barkema et al., 1999; Miltenburg et al., 1996). Several studies did study the risk of CM after a previous case of CM (Zadoks et al., 2001; Cha et al., 2016). Zadoks et al. (2001) found that previous infection of one (or more) quarter(s) may lead to changes at systemic level, e.g. development of immunity. This could affect the susceptibility to future occurrences of mastitis in all quarters of the cow. Bradley and Green (2000) did research on the incidence and significance of intramammary enterobacterial infections acquired during the dry period. They found that of all enterobacterial mastitis occurring in the first 100 d of lactation, 53% arose in quarters previously infected. Houben et al. (1993) reported finding an increased risk of mastitis in the current lactation, due to mastitis in the previous lactation, ranging from a factor of 2.0 to 2.9. Cha et al. (2016) also found that the incidence of clinical mastitis doubled when a previous mastitis case was experienced. This means that there is twice as much risk of mastitis in the current lactation, due to mastitis in the previous lactation. Swinkels et al. (2013) studied the recurrence of clinical mastitis at both cow and quarter level and found that the recurrence was higher at cow level (58%) than quarter level (43%). This also shows that infection history can both be defined at quarter level and cow level. Cow-level infection history is strongly correlated with quarter-level infection history (Zadoks et al., 2001). 2.4.4 Breed As already mentioned before, breeding is an important way to reduce the incidence of CM. Several studies have shown that selection for cows with a higher milk production causes negative effects on udder health (Heringstad et al., 2003). Therefore, it is suggested that dual purpose breeds such as Maas-Rijn-IJssel (MRY) cows are supposed to be more resistant to mastitis due to their lower milk yield (Neuenschwander et al., 2005). Barth and Aulrich (2008) compared the dual-purpose breed Red-and-White Holsteins (RW) and the milk oriented breed Holstein Friesian (HF) regarding their disease occurrence. Their findings did not support the popular assumption that it is better to use dual purpose breeds, as they found no significant differences between both breeds. Also, Schukken et al. (1990) found an increased clinical mastitis rate in dual purpose MRY breed versus the Holstein Friesians. This may be associated with udder conformation or a genetic trait. Another Dutch study also found a higher clinical mastitis incidence in the MRY breed (Grootenhuis, 1981). This suggests that immune response to IMI differs between breeds. The difference in susceptibility to mastitis between breeds could be the result of difference in the prevalence of specific genes (Compton et al., 2007). No conclusions can be drawn regarding the breed of an individual cow and its risk for mastitis (Schukken et al., 1990). 12

It is clear that mastitis has a major economic impact on a dairy farm (Halasa et al., 2007; Hogeveen et al., 2011). The last years, a lot of research has been done on mastitis management. A 5-year national mastitis control program was introduced after field reports indicated that a substantial number of herds were experiencing problems with CM (Lam et al., 2013). The goal of this control program was to decrease the IRCM by 10 percentage points (van der Zwaag et al., 2005). Several preventive measurements are available to prevent mastitis during the dry period. Those measures are BDCT, SCDT and an external/internal teat seal. Due to antimicrobial resistance and public concern about the antibiotic use in livestock, BDCT was no longer allowed and SDCT was proposed as alternative to avoid the unnecessarily use of antibiotics. As not all cows can be treated, cows must be selected for SDCT. This selection is often based on SCC, but many studies found other cow characteristics that could affect the development of IMI and mastitis (Oviedo-Boyso et al., 2007; Bakken, 1981). Breed, parity, infection history, number of quarters infected, teat morphology and lactation number can play a role in the development of mastitis. 13

14

3. Materials and methods 3.1 Model description 3.1.1 Linear Programming In this study, linear programming (LP) was used to distribute the antibiotic use at drying off over different groups of cows (primiparous and multiparous), with the objective to minimize the total costs related to SDCT and (sub)clinical mastitis until 100 days in milk (DIM). LP is a method which is used to optimize profit or minimize losses by finding the optimal combination of different parameters with respect to a set of fixed constraints (Norton and Hazell, 1986). The mathematical formulation for the linear programming model is: MIN Z = Ci Xi Aij Xi Bj Xi 0 Z = Total cost related to SDCT Ci = Economic costs per unit of activity Xi = Activity Aij = Technical coefficient per unit of activity Bj = Resources To model the use of antibiotics and put a restriction on the use it needs to be determined whether and how much antibiotics should be used at drying off. Therefore, the animals were categorized in several groups based on the SCC at last milk recording before drying off and the presence of an infection history (IH). Animals with high SCC (primiparous cows >150,000 cells/ml and multiparous >250,000 cells/ml) are already associated with an IMI and should therefore be treated during the dry period (Schukken et al., 2003). The LP model was based on a fictive dairy farm with 100 dairy cows with an average age distribution of Dutch dairy herds with 33% primiparous cows and 67% multiparous cows (Scherpenzeel et al., 2014; 2016). Therefore, of the 100 cows, 33 were cows in their first dry period and 67 were cows in their multi dry period. 3.1.2 Age groups Because different age groups of cows have different characteristic regarding udder health and dry period, both primiparous and multiparous cows were used. In this study, a primiparous cow was defined as a first-lactation cow until the moment she calved for the second time. A multiparous cow was defined as a cow that has calved at least twice. 16 groups were formed based on the SCC before drying off and infection history (table 3.1). A cow was considered to have a history of infection when she was infected with a mastitis pathogen in a previous lactation. Each group consists of cows that were either dried off with antibiotics or without antibiotics. Therefore, there were in total 32 activity levels (Xi) in the LP model (table 3.1). 15

Table 3.1: Groups with activities used in the linear programming model based on parity, SCC on the last milk recording before drying off and infection history 1 (0= no infection history, 1= infection history) Group Activity level (Xi) Parity SCC IH 1 (x1000 cells/ml) Dried off with antibiotics Not dried off with antibiotics 1 1 2 Primiparous 0 50 0 2 3 4 Multiparous 0 50 0 3 5 6 Primiparous 50 100 0 4 7 8 Multiparous 50 100 0 5 9 10 Primiparous 100 150 0 6 11 12 Multiparous 100 150 0 7 13 14 Multiparous 150 250 0 8 15 16 Multiparous >250 0 9 17 18 Primiparous 0 50 1 10 19 20 Multiparous 0 50 1 11 21 22 Primiparous 50 100 1 12 23 24 Multiparous 50 100 1 13 25 26 Primiparous 100 150 1 14 27 28 Multiparous 100 150 1 15 29 30 Multiparous 150 250 1 16 31 32 Multiparous >250 1 1 IH = infection history. A cow was considered to have a history of infection when she was infected with a mastitis pathogen in a previous lactation Dairy farms differ in the average SCC of the herd and can therefore be categorised into three categories regarding their bulk tank somatic cell count (BTSCC); a low BTSCC (<100,000 cells/ml), an average BTSCC (100,000 250,000 cells/ml) and a high BTSCC (>250,000-400,000 cells/ml) (Den Uijl et al., 2012). This study will simulate all three farm categories. The distribution of the animals over SCC groups for different BTSCC was based on the study of Scherpenzeel et al. (2016) (table 3.2). For example, they found that 80% of the cows at a low BTSCC farm have a SCC between 0 and 50,000 cell/ml. 16

Table 3.2: Distribution of the animals over the different BTSCC farm categories to the model based on last SCC before drying off. Each category has a fictive herd of in total 100 dairy cows. Individual SCC (x1,000 cells/ml) Parity BTSCC Low Average High 0 50 Primiparous 26 11 8 0 50 Multiparous 53 21 12 50 100 Primiparous 5 11 8 50 100 Multiparous 11 21 13 100 150 Primiparous 1 5 5 100-150 Multiparous 2 10 8 150-200 Multiparous 2 10 12 >250 Multiparous 0 11 34 Total 100 100 100 3.2 Model parameters 3.2.1 Economic costs To calculate the economic consequence for the farm, the total costs related to SDCT, CM and SCM were calculated. The disease costs were estimated by determining the milk production losses, discarded milk, treatments, veterinarian, farmers labour and death and culling. The data is based on literature and expert opinions, collected from the Netherlands. The total costs of mastitis (TCM) consists of the total costs of clinical mastitis (TCCM) and the total costs of subclinical mastitis (TCSM). Both TCCM and TCSM are calculated by multiplying the probability for clinical mastitis and subclinical mastitis with the costs of clinical and subclinical mastitis. TCMi = Total Costs of Clinical mastitis (TCCM)i + Total Costs of Subclinical Mastitis (TCSM)i TCCMi = Probability of Clinical Mastitis (PCM)i Costs of Clinical Mastitis (CCM) TCSMi = Probability of Subclinical Mastitis (PSM)i Costs of Subclinical Mastitis (CSM) Table 3.3 gives an overview of the input values of the LP model. Table 3.3: Input values for the LP model Parameter Value Source Drying off ( ) 12.40 Halasa et al., 2010; Scherpenzeel et al., 2016 Costs of clinical mastitis ( ) 235 Huijps et al., 2008; Hogeveen et al., 2011 Costs of subclinical mastitis primiparous cow 31.76 Halasa et al., 2009 ( ) Costs of subclinical mastitis multiparous cow 59.70 Halasa et al., 2009 ( ) Daily doses antibiotic treatment for drying off 4 Halasa et al., 2009 Daily doses antibiotic treatment during lactation 1.5 Sol et al., 2000 17

The costs of clinical mastitis were set at 235, - per case of clinical mastitis (Huijps et al., 2008; Hogeveen et al., 2011). This cost is based on milk production losses related to CM, drugs, discarded milk, veterinarian visits, culling, labour and penalties (Huijps et al., 2008). The costs of subclinical mastitis (CSM) consist of milk production losses (Halasa et al., 2007). It was found that milk production losses due to SCM were set at 0.5 L/day for primiparous cows and 0.94 L/day for multiparous cows (Halasa et al., 2009). In a non-quota system, the milk production losses consist of the milk price minus the saved feed costs (concentrates). The duration of milk production loss was estimated to be 219 days (Halasa et al., 2009). The milk price in the Netherlands is 0.36/kg of milk (Agrimatie, 2017). CSM = Milk production losses due to subclinical mastitis (L/day) duration of milk production losses (days) (Milk price ( /kg) saved concentrate costs ( /L milk) For primiparous cows the costs of SCM are 31.76 per case and 59.70 per case for multiparous cows. The costs of drying off a cow consist of labour costs and antibiotics. The costs for antibiotics were estimated to be 10 and an hourly wage of labour of 18/hour (Scherpenzeel et al., 2016). It was estimated that it takes 8 minutes to apply SDCT (Halasa et al., 2010). The costs for drying off were set at 12.40 per cow. 3.2.2 Antibiotic use Costs of drying off minutes to apply DCT = costs for antibiotics + ( ) 60 hourly wage of labour ( /hour) In this LP model two restrictions were incorporated; restriction on the number of animals and restriction on the antibiotics use. The number of animals dried off with or without antibiotics together should be equal to the number of animals per group. This is to make sure that all animals from all groups are included in the model. Without this restriction, the model could decide to exclude the animals that are not treated with antibiotics. This influences the calculation. Antibiotic usage for SDCT and the treatment of CM was expressed as the number of daily dosages. Antibiotic usage for SDCT was calculated based on the number of cows in the different groups depending on the activity (dried off with antibiotics or not). Antibiotic usage for CM was calculated by multiplying the probability of CM with the daily dosage needed for treatment. The total antibiotic usage for the fictive dairy farm was calculated by summing up the antibiotic usage per group. At drying off four antibiotic injectors were used, one per quarter. All the cows that are dried off with antibiotics (AB) had 4 daily dosages for the dry period up to 100 days in milk (DIM). Besides the preventive use of antibiotics, antibiotics are also used to treat a possible CM case in lactation. It was assumed that infected quarters were treated three times, with a 12-h interval between the treatments (Sol et al., 2000). This sums up to 1.5 daily dosages. 18