The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
LIVESTOCK DISEASE ERADICATION: EVALUATION OF THE COOPERATIVE STATE–FEDERAL BOVINE TUBERCULOSIS ERADICATION PROGRAM
by the detection system, is that most diseases have a complex natural history. Whereas the majority of disease may arise from identifiable sources, some disease occurrences happen for other, often unknown, reasons. This could include aberrant forms of the organism, unusual methods of spread, or the structure or nature of specific animal enterprises. For example, as noted earlier, large herds may have different epidemiologic features than smaller herds, or disease control in herds located in one type of terrain may be more difficult than in similar herds in a different environment. Another reason for using depopulation is that although a test and slaughter program may be capable of achieving an eradication goal, a depopulation program may offer considerable savings in the total number of animals sacrificed, in time, and in money. Finally, given the limitations of the disease detection process at the individual animal level, there may be no reasonable alternative to depopulation to ensure that the last vestiges of disease have been removed.
In principle, determining if and how a disease should be controlled is a straightforward task. In practice, however, there are numerous difficulties, primarily because of uncertainties about the biologic and economic effects of the disease and of the impact of the various control strategies. Depending on the disease and the existing environment, a thorough bioeconomic analysis should help decide whether control efforts can be effectively and efficiently mounted by the affected farms or industries themselves, or if a more widespread government program is needed. If a government program is appropriate, a thorough analysis should help decide on the nature, intensity, and end point of the program(s).
MODELING AS AN AID TO DECISIONS ABOUT CONTROL
Combining economic and epidemiologic methods has proved to be successful in providing guidelines and decision-making criteria for animal health programs (McInerney et al., 1992). One method, modeling of disease processes, will be stressed here. Bioeconomic animal disease simulation models, a combination of epidemiologic and econometric models, have been used primarily to determine the benefits and costs associated with a number of alternative disease control strategies as part of animal health programs directed against diseases such as brucellosis, tuberculosis, screwworm, and foot and mouth disease (Dijkhuizen et al., 1991; Dijkhuizen, 1992).
Paramount to the success of any bioeconomic model is the accuracy with which the epidemiologic coefficients reflect the actual characteristics of the disease within the specific livestock industry. If a valid and representative epidemiologic model has been established, then the effects of the baseline program and of alternative control programs can be simulated. Epidemiologic models are mainly designed to simulate (1) the incidence, prevalence, and spread of infection; (2) the impact of these on the demographics and productivity of the species of concern; and (3) the effects of control program components on the level of infection and related physical losses. Anticipated benefits from disease control include reductions (changes) in physical losses of meat and milk (or other animal products) associated with the alternative programs compared to the baseline program.
Most epidemiologic models consider the probability of an animal (or herd) becoming diseased, the biologic impact of the disease in that herd/animal, and the impact on other components of the industry. The model should take into account the expected variability in these parameters depending on the nature of the enterprise, the local environment and any variation over time (for example, cyclical or seasonal patterns). Some epidemiologic models are deterministic and essentially give the same result every time the model is run with the same start-up parameter values. Other