Modelling the economic impact of three lameness causing diseases - - PowerPoint PPT Presentation

Modelling the economic impact of three lameness causing diseases using herd cow level evidence.

Jehan Ettema, Søren Østergaard, Anders Ringgaard Kristensen.

Claudia Bono

Introduction

• Lameness causing diseases, have large

economic impact in modern dairy farming.

• In order to support farmer’s decisions on

preventing and treating lameness, decision support models can be used to predict the economic profitability of such actions.

Introduction

• Reported incidences of clinical lameness

vary from 21% up to 70% per lactation in a 1.5 year study period.

• This means that there is a strong impact
• This means that there is a strong impact
• n economic situation.
• 192€ have been estimated per case of

clinical lameness per cow-year in a typical Danish dairy herd (Ettema & Østergaard 2006).

Task

The objective of this study was to simulate the economic feasibility of reducing the risk of lameness reducing the risk of lameness causing diseases in a herd where disease risk is described by hyper- distributions.

Material and methods

Model

A Danish herd simulation model was used : SimHerd IV. As mentioned during last SimHerd IV. As mentioned during last lecture, this is a dynamic, stochastic and mechanistic Monte Carlo simulation model. SimHerd IV simulates the production and state changes in a dairy herd with additional young stock.

Material and methods - Model

State of an animal defined by:

• Age,
• Parity,
• Lactation stage,

Production and development within the herd

• Lactation stage,
• Milk yield,
• Body weight,
• Culling status,
• Reproductive status,
• SCC (Somatic cell count).,
• Disease status.

within the herd are determined indirectly by simulation of production and change in state

• f the individual

cow and heifer.

Material and methods - Model

Model behaviour could be controlled by a set of decision variables, which define certain production systems and management strategies. However, the state-of-nature of a livestock model is never known with certainty. Particularly, risk in decision making is underestimated when using point estimates in the state-of-nature.

Material and methods

Three underlying diseases causing lameness were modelled:

• Digital Dermatitis (DD);
• Interdigital Hyperplasia (IH);
• Interdigital Hyperplasia (IH);
• Claw horn disease (CHD).

IH is considered as a separate disease because of the chronic nature and the incurability. Heel horn erosion (HHE) and Interdigital Dermatitis (ID) are no took in consideration in this study.

Material and methods

Rapresenting uncertenty:

To simulate the economic feasibility of reducing the risk of lameness causing diseases in a herd where the risk is described by hyper- distributions.

Traditionally… Traditionally…

State-of-nature parameters are described by fixed estimates.

In this study:

A joint posterior distribution described the nine state-of-nature parameters representing three disease risks for three categories of parity.

Material and methods

– Formulated given observations (y) made in the specific herd. – Based on prior knowledge of disease prevalence in the entire population,

Joint posterior distribution:

prevalence in the entire population, combined with herd and cow evidence (y). – For every replicate of the simulation model, a state-of-nature was randomly drawn from the joint probability distribution. – Ran 1000 replicates with SimHerd uncertainty around herd level risk represented.

Material and methods

Calibrations of disease prevalence

– Assume a value for disease duration; – Simulation model run with random draws from the hyper-distributions; – Model output for disease prevalence compared to the prevalence described by the hyper-distributions; – Simulation model run again where all random draws multiplied with the same factor until desirable model output for prevalence was produced.

Material and methods

Goal: to simulate certain management strategies in a specific (fictitious) herd using different hyper-distributions which represent different levels of knowledge. Nine different sets of marginal Nine different sets of marginal probability distributions used in this study for the probability’s logit value for having three lameness causing diseases:

Material and methods

The effect of lameness on production parameters, are summarized in table 2:

36 Scenarios were run with the simulation model.

Material and methods

Scenarios:

– First 8 scenarios: fixed estimates for disease risk; – 9-15: hyper-distributions used to describe disease risk; – 9-15: hyper-distributions used to describe disease risk; – 1000 samples drawn from different sets of joint posterior distributions (sets 1-7 of table 1); – 16-22: weekly disease risk halved (average reproductive performance); – 23-36: as above, but for poor reproductive performance.

Material and methods

Simulation Procedure:

A scenario, described with point estimates, was simulated over 10 years. estimates, was simulated over 10 years. Then, all scenarios simulated over 20 years with 1000 replications. Average results over last 15 years of the 20 years simulation was studied.

Results

Tables 4 &5: Technical and economic consequences of a scenario where all three disease risks are at a high level, in contrast to a scenario where the risk of all three diseases was halved. For herds with average and poor reproductive performance. and poor reproductive performance.

Results

Results

Low risk herd, with average reproduction

• When halving the risk of all disease in a low risk

herd:

Margin per cow-year increased by €29.3.

• In a herd with poor reproductive performance

and low risk for all diseases:

Herd size was maintainded, Margin per cow-year increased by € 28.1.

Results

Change in margin per cow- year and the SD when using 7 hyper-distributions in a herd with average and poor reproduction . 7: Case in which only herd reproductive efficiency was reproductive efficiency was known to be average, herd size >125 cows 6, 4, 2: with information on prevalence of lame cows, prevalence of hoof lesions among 45 and 180 cows 7, 5, 3: by adding more proof

• f lameness and the

underlying disease being present in the herd

Discussion

In the current study, the prior distribuitions were created systematically using Bayesian statistics. This study also demonstrated the concept This study also demonstrated the concept

• f using field data on diseae prevalence for

specification of cow level disease risk in a consistent way.

Discussion

• Need for better knowledge of the effects of

production diseases

– Especially DD and IH, where it was necessary to make assumptions of reoccurrence rates, milk loss etc.

• Economic analysis

– Cost of gathering information not included

• Trimming cows is costly, and should be considered.
• Representing uncertainty
• Nine of the simulation model’s parameters were

described by a hyper distribution; over 1000 parameters remain fixed.

Discussion

Dynamic aspects of the simulation model

• Value of drawn sample did not change over time,
• Another approach can be, changing disease risk in

time due to herd effects/preventve measures. time due to herd effects/preventve measures.

Not an optimization model! Opportunities for application:

• Both lameness prevalence and hoof trimming

registration are combined,

• Preventive strategies.

Conclusion

• Novel approach of using hyper-distributions

to describe the risk of three different types of lameness in dairy cows, by an existing dynamic, stochastic and mechanistic simulation model. simulation model.

• Uncertainty in input parameters is reflected

in the uncertainty of the simulation model

• utput.
• However, uncertainty in model outcomes is

still underestimated.

2

?

Thank you