Animal Health & Welfare

Optimising a diagnostic test for use in the field

Frequently, diagnostic tests make a continuous measurement on a single individual with the intention of establishing whether that animal is ill or healthy. Determining the status of the animal is usually done by setting a cut-off threshold above which an animal is considered ill, and below which it is considered healthy or vice versa. The choice of cut-off value depends on the preferred balance between the mistakes of wrongly diagnosing healthy animals as ill, and ill animals as healthy.

In collaboration with scientists from MRI, BioSS has developed a model to conduct a cost-benefit analysis of diagnostic tests, making allowance for the properties of disease epidemiology, test characteristics and the consequences of diagnostic mistakes. With this model we can determine the most cost-effective cut-off value for a test, given how the test is to be used, or we can determine the most cost effective number of tests to make on a group of animals, given a specified cut-off. Although this work was motivated by research on sheep scab, this approach can inform the wider development of protocols for the use of new tests in the field, helping to optimise the benefits of diagnostic tests in livestock.

prevalence mapDemonstration of the effect of cut-off and disease prevalence on the cost of applying a diagnostic test. Blue colours represent cheaper tests, red more costly tests. Note how the cut-off corresponding to the minimum cost changes as the prevalence varies.

sheep being given a drenchOur work will help to optimise the diagnosis and treatment regimes for economically important diseases of sheep and cattle.

Further details from: Giles Innocent

Article date 2013

Consultancy Advice & Collaboration

Plant Science

Animal Health & Welfare

Ecology & Environmental Science

Human Health & Nutrition