Australian College of Veterinary Scientists
Write brief notes on the important features of the epidemiology and the control, management or eradication of THREE (3) of the following conditions/agents:
Briefly describe the essential features and application of THREE (3) of the following:
Write brief notes to demonstrate your understanding of TWO (2) of the following:
Using examples, write brief notes on THREE (3) of the following:
Using the data in the table below:
| Diseased | Non-diseased | Total | |
| Exposed | 20 | 5 | 25 |
| Non-exposed | 10 | 10 | 20 |
| Total | 30 | 15 | 45 |
Answer provided by Ian Langstaff:
RR = risk in exposed / risk in unexposed
RR = (20/25) / (10/20)
RR = 1.6
Interpretation: factor positive animals are 1.6 times as likely to develop disease as factor negative animals. Exposure to factor is positively associated with disease when RR > 1.
OR = odds of disease in exposed / odds in unexposed
OR = (20 × 10)/ (5 × 10)
OR = 200 / 50
OR = 4.0
Interpretation: the odds of factor positive in diseased animals are four times the odds of factor positive in non-diseased animals.
Meadures of effect: absolute effect measures are attributable risk (AR) and attributable fraction (AF).
AR = risk in exposed - risk in unexposed
AR = (20/25) - (10/20)
AR = 0.3
For every 100 animals 30 are diseased due to exposure to factor positive. Exposure is protective?
AF = AR/Rexp +
AF = 0.3 / 0.8
AF = 0.37
Interpretation: 37.5% of disease in exposed individuals that are due to exposure.
Measures of total effect: population at risk (PAR) and population attributable fraction (PAF).
Alternative: calculate confidence intervals of the point estimates and see if they overlap. This is the intuitive way for all the four measures above.
Two tests for ovine paratuberculosis (Mycobacterium avium subsp. paratuberculosis) have been developed. One test is an absorbed enzyme-linked immunosorbent assay (ELISA) while the other is an agar-gel immuno-diffusion (AGID) test. You have been asked to evaluate these tests. To assist with the study, sheep have been made available from farms where ovine paratuberculosis is known to be present and from farms in a region thought to be free. Histologic examination of intestinal tissue is regarded as a definitive test for ovine paratuberculosis.
Describe how you would evaluate the performance of these tests (70% of marks). Include in your answer a discussion of how you might compare the performance of the tests (30% of marks).
Answer provided by Paul Bingham:
Evaluation of test performance. We have three types of test:
For both tests to be evaluated:
The quantitative ELISA test:
The ELISA tests will give a continuous outcome of results and the tests performance will depend upon the cut-off selected for the test. Plot all ELISA results in a histogram. From the distribution it may be possible to select a crude-off value. Calculate the Se and Sp at various cut-offs using 2 × 2 tables. This cut-off can be refined by plotting a ROC (receiver operating characteristic) curve (plot of Se against 1 - Sp) for a variety of selected cut-offs. The optimum cut-off can be selected - this is usually the point on the ROC curve in the upper left corner (where both Se and Sp are maximised ).
Alternatives to using a ROC curve are to plot Se against Sp at various cut-offs, where the two lines intersect test performance is maximised.
Use of likelihood ratios. These methods assume that the costs of a false positive and false negative result are equivalent. Formulas are available which take economics into account.
The dichotomous AGID test:
As for the ELISA for a single cut-off. Plot the results of the ADID tests against the results from the gold standard in a two by two table Se and Sp can then be calculated as can other test parameters e.g. Youdens index, PPV NPV, index of efficiency.Comparison of the tests can be done be comparing the test parameters directly (as they have been assessed using the same panel and the same gold standard. For example: direct comparison of Se, Sp, Youdens, PPV, NPV, index of efficiency They can also be compared for agreement by using e.g. a kappa value.
In conclusion, the tests can be compared using these parameters but the test performance for the ELISA can be altered by altering the cut-off which in turn can be selected according to the consequences of a false result.
Canine coronavirus is becoming an agent of concern to small animal practitioners. In particular, there is some controversy regarding the prevalence of seropositive dogs, and the role that this virus may play in canine gastroenteritis. As a consultant epidemiologist, you have been asked to design and carry out a cross-sectional study to investigate these problems. Describe how you would proceed with planning this study, including discussion of important factors affecting study design?
Salmonella enteritidis (SE) is a common cause of food-poisoning in many countries, with the main source of infection being raw or partly cooked eggs and egg products. For example, human illness from SE positive eggs in the United States is now approximately 637,000 cases per year. SE in chickens causes a silent systemic infection that can be detected by both bacteriological and serological techniques. Prevalence of infection in naturally infected commercial layers has been found to be very low.
The Australian and New Zealand egg industries are thought to be free of SE infection, and relatively few human cases occur in either country when compared to overseas. No human cases of SE due to consumption of Australian or NZ eggs have been reported. You have been engaged by the egg industry in your country (Australia or New Zealand) to plan an epidemiological study to demonstrate freedom from SE. Describe how you would proceed with this project and discuss the key issues affecting the study design.
As a government epidemiologist, you have been asked to undertake an import risk assessment for the importation of horse semen from South America. Describe how you would proceed.
Answer provided by Jenny Weston:
An import risk assessment needs to be carried out to describe the risk associated with an event - in this case, the importation of horse semen from South Africa. Potential risks associated with this activity that we would be concerned with are the introduction of diseases that are exotic to New Zealand or are the subject of a control or eradication programme. A variety of exotic diseases could be introduced via this route such as Equine Viral Arteritis, Equine Influenza, a variety of arthropod-borne viruses.
The risk assessment will consist of both the probability of such an event occurring and the consequences should the event occur (usually in a dollar term). This assessment will allow decision makers to determine of a course of action should occur and helps identify those areas that should be closely monitored and could be used to compare a number of possible testing and control options for selection of stallions whose semen may be imported and how this could be managed.
Risk assessment may be quantitative where absolute probabilities of risk and dollar costs of outcome are known or able to be estimated. However, in most cases this can not be known and qualitative or semi-quantitative methods are used. For qualitative analysis, terms such as high, moderate, low and very low are used and for semi-quantitative analysis there is a numerical value ascribed to the descriptive terms used in qualitative analysis.
The steps in the risk analysis process include:
The risk assessment must be able to be justified both domestically (we have responsibilities to protect our borders) but also have obligations to overseas trade organisations and can't just ban all imports and take a "no risk" policy.
Mnemonic:
| I | Identify hazards | I |
| D | Define risks | don't |
| O | Outline model | offer |
| I | Information required | import risk |
| A | Assess risks | assessment |
| S | Sensitivity testing | since I |
| C | Clarity of logic | can't |
| P | Peer review and feedback | provide any |
| S | Stakeholders | security |
Australian College of Veterinary Scientists