Treating mastitis: Balancing cure, money, welfare and resistance Henk Hogeveen With input from Wilma Steeneveld and Claudia Kamphuis
Dec 02, 2014
Treating mastitis: Balancing cure, money, welfare and resistance
Henk Hogeveen
With input from Wilma Steeneveld and Claudia Kamphuis
My paper gives latest results from literature
But ...session on analytics:
The modern dairy farm has at its fingertips an endless array of data. When managed properly, these data can be used to create a competitive advantage. This session will explore the potential of analytical approaches to managing mastitis through the use of on-farm records, decision support for mastitis treatment, and statistical processing of information.
My presentation will have an emphasis on analytics, not on results
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
Important in treatment decisions
Cure
Much knowledge available
Keywords “Cure rate” and “mastitis”
11 scientific papers in 2013
There is more than cure rates: Welfare
Clinical mastitis gives pain (e.g., Kemp et al. 2008 VetRec)
Behaviour is also affected (Medrano-Galarza et al., 2012 JDairySci)
So: Better cure is better welfare
Antibiotic resistance
Heavily in discussion
Resistance of mastitis pathogens
●Self-interest
●No increase seen (Hogan, IDF-factsheet)
Antibiotic resistance in humans
●Externality
●Dairy cattle has very minor contribution (Oliver et al., 2011)
●In the Netherlands (self) regulations
Economics
A farm is a business
Self interest
Costs of antibiotics vs benefits of higher cure rates or better prevention
Difficult task of herd manager
There is:
●Cure rate (welfare)
●Money
●Antibiotic resistance
Should be balanced
Not much knowledge on balancing
11 scientific papers in 2013
One with economics (related to transmission); Down et al., JDairySci
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
Lactational treatment
Much knowledge available on cure, e.g., reviews
● Barkema et al. 2006 JDairySci
● Roberson, 2012 VetClinFoodAnim
● Roy and Keefe, 2012 VetClinFoodAnim
● Suojala et al. 2013 JVetPharmTherap
Some papers on economics
● Steeneveld et al., 2011 JDairySci
● Halasa et al., 2012 JDairySci
● Down et al., 2013 JDairySci
These are averages
Make your own calculations
Straightforward spreadsheet
Two scenarios
Straightforward spreadsheet
Cure rates(input)
Straightforward spreadsheet
Calculation of proportion of cured cows
Straightforward spreadsheet
What if animals do not cure?
We have not really thought that out very
well
Farm specific
Role of advisor
Let’s add economics
Proportion of proportion
Let’s add economics
But we are missing input
Let’s add economics
That depends on cure and
lactation stage
Let’s add economics
Formula for the average nr of days
We are missing some price levels
Let’s add economics
Let’s make final
calculations
Let’s add economics
And finally, what are the costs?
What’s the point
Specific situation -> no papers available
Creation of a tool is not too difficult
Input needed
●Price levels – farmers know these
●Specific situation of the cow – farmers know these
●Cure rates – this is a problem
Cure rates might be available from farm records
●Large farms
●Available data
Evaluate previous assumptions
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
Treatment of clinical mastitis
Causal pathogen●Streptococci (40%), S. aureus (30%), E. coli (30%)
Parity Day in milk Calving interval Most recent SCC-value Repeated CM case yes/no Systematically ill yes/no 305-day milk production
●Wood curve to determine daily milk production at moment CM and remaining milk production during lactation
All stochastic
Steeneveld et al., 2011, JDairySci
Defined treatments
3-day intramammary antibiotic treatment (IMM3)
5-day intramammary antibiotic treatment (IMM5)
3-day intramammary antibiotic treatment + systemic antibiotic treatment (IMM3_S)
5-day intramammary antibiotic treatment
+ systemic antibiotic treatment (IMM5_S)
Immediately culling
Simulating follow-up of treatment
Treatment CM1
Bact. + clin. cure
Bact. + clin. cure
Dry-off quarter
Treatment CM2
No bact. cure,no clin. cure
No bact. cure,no clin. cure
No bact. cure,clin. cure
No bact. cure,clin. cure
End lactation
Dying
Culling
Culling
Culling
Extended treatment
End lactation
End lactation
End lactation
Culling
etc.
Cow-specific cure
Probability of bacteriological cure (%) defined
●for heifers, SCC<200, <60 DIM, no CM before, not systematically ill
IMM3 IMM5 IMM3_S IMM5_S
StreptococciS. aureusE. coli
704080
806085
806085
907095
Defined effects of cow factors• Older cow: 10%• SCC 200-500: 10%• SCC >500: 20%
• >60 DIM:10%
• Repeated case: 20%
• Systematically ill:20%
Average costs ($US)
758
749
789
266
318
310
238
296
284
238
296
284
216
283
253
Causal pathogen
Streptococci
S. aureus
E. coli
766295275270245Overall
CullingIMM5_SIMM3_SIMM5IMM3
478
578
569
577
732
1,026
207
230
255
266
301
351
192
216
230
244
284
332
158
189
200
216
253
307
Daily milk production (kg)
<20
20-25
25-30
30-35
35-40
>40
174
206
223
238
279
326
Original calculations in €; € 1 = $US 1.37
Least cost frontiers
20
30
40
50
60
70
80
90
100
130 150 170 190 210 230 250 270 290
Total costs (€)
Prob
abili
ty o
f cur
e (%
)
IMM3
IMM5
IMM3_S
IMM3_N_SIMM5_S
High cure cowE. coli cow
Average cowS. aureus cowLow cure cow
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
The ongoing debate on dry-cow therapy
Dry-cow therapy has two uses
●Curative
●Preventive
“We” do not want preventive use of antibiotics (anymore)
Which cows to dry-off with antibiotics?
Let’s create another spreadsheet
●Basically the same as the previous one
Probabilities from: Scherpenzeel et al., 2014,
in preparation
Very basic input
Probabilities depend on risk group
Costs with and without
DCT
Average costs drying-off per cow
Costs per cow
We’re also interested in amount of AB
Daily doses
Totals
Let’s optimize (linear programming)
Minimize total costs of mastitis By the numbers
of treated cows
Under constraints:- Nr of cows equal- Less AB than threshold
The constraint
What’s the point
Literature never fits the individual farmer’s situation
Probabilities can be based on farmers on data
Quite straightforward economic modelling
Evaluate previous assumptions
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
Automatic milking
On-line mastitis monitoring (Electrical conductivity, colour, SCC)
Great possibilities for therapy evaluation
But ….. Sensitivity & specificity of mastitis detection
Farmer’s confirmation is needed
●Time consuming
●Mostly negative
●Not the nicest of work
3.5 % of alerts are checked
Study on farmer’s handing of alerts
7 farmers, student checked all alerts
●60 % of alerts false positive
●3.5 % of alerts is checked by farmers
●checked alerts are often clinical cases
●74 % of clinical cases is missed
How bad is this?
Options
1. Maintain the “old” paradigm of treating clinical mastitis cases hold in automatic milking
….. and educate our farmers better to check….. or have better sensors (less false positives)
2. Use the daily sensor measurements differently -> detect acute severe mastitis, for mild mastitis look at chronicity, use on-farm culturing before treatment
Lot’s of questions, no answers (yet)
My presentation
Balancing treatments
Analytics of lactational treatment
Example of published results
Analytics of dry-cow therapy (optimization)
New possibilities with automatic milking
Concluding remarks
I did not present all knowledge
There is more knowledge out there
Mostly economics
Welfare ≈ cure rate
On-farm culture systems
● Lago et al., 2011, JDairySci
● Cameron et al., 2013, PrevVetMed
● Pinzón-Sánchez et al., 2011 JDairySci (Economics)
Use of antibiotics only through dry cow therapy
On-farm analyses
Use straightforward calculation tools
Use farm-specific input
●Price levels
●Incidences
●Cure rates
Use those farm data!!!
Operational use should be automated
There is a future for tailor-made treatment decisions
PS Example models and ppt are available
Thank you for your attention
@henkhogeveen
animal-health-management.blogspot.com
On-line courses on Veterinary Economics on:
www.elevatehealth.eu