Материал: Bovine Viral Diarrhea Virus Diagnosis, Management, and Control

mean heart girth at 80 days of age among 8 male PI calves was 96 cm as compared to 100.5 cm among 13 male non-PI calves (P<0.05) (Larsson et al., 1994). The same figures at 180 days were 123.3 vs. 130.2, respectively (P<0.05).

The incidence risk of dying or being slaughtered due to unthriftiness within 1 year among 34 PI animals in 10 Danish dairy herds was 0.28 and 0.31, respectively (Houe, 1993). The attributable risk for PI animals of leaving the herd for any reason within 1 year was calculated as 0.35. In eight Danish dairy herds with outbreaks of mucosal disease, 31 of 52 PI animals died that were identified as PI either at outbreak or following whole herd testing (Houe et al., 1994). However, as these were selected outbreaks, this mortality might have been higher than what would occur among PI animals in the whole cattle population. On the other hand, had whole herd testing not been performed more PI animals would probably have died before they were identified as PI animals and slaughtered. Another way of judging the mortality among PI animals is to look at age specific prevalence. Thus, the prevalence of PI animals among all cattle in 19 Danish dairy herds was 1.4% (Houe and Meyling, 1991). However, among cattle younger than 1 year of age the prevalence was 2.9% indicating a high mortality (or at least culling) also among younger PI animals. In a prevalence study by Frey et al. (1996), 66% of the PI animals were younger than 1 year, whereas only 10% were older than 2 years.

ECONOMIC IMPACT OF BVDV

Economic calculations and models depend on the underlying data. The types of data necessary for calculation of economic impact have been outlined in previous sections of this chapter. The significant variation in occurrence, frequency, and damaging effects following infection will affect the uncertainty in economic estimations. Also, there are indications of a changing disease pattern of BVDV over time (Evermann and Ridpath, 2002) and hence incorporation of appropriate variables in the economic models should be a dynamic process.

There are two major functions of performing economic analyses. First, calculation of the pure costs of disease will give an indication of the relevance of disease as compared to other diseases in animal husbandry. In addition to economics, there are other criteria for ranking the relevance and importance of diseases, such as animal welfare and human safety. These criteria will not be discussed further although animal welfare is an important aspect of BVDV in-

fections due to the severe suffering from mucosal disease. The second purpose of performing an economic analysis is to guide decision makers in determining whether it is cost-effective to control the infection and also to compare the cost effectiveness of different control options. The different control options are discussed in Chapter 14.

The direct economic cost (C) of a disease is the sum of the production losses (L) and the expenditures (E) for treatment and prevention (McInerney et al., 1992; Bennett et al., 1999a):

C = L + E

The expenditures for treatment and prevention are often relatively easy to obtain as they follow specific actions in the herd. The production losses on the other hand are often associated with much higher uncertainty due to biological variation. The production losses can be obtained if there is general information about the population size, the incidence of infection, damaging effect (including both the probability and magnitude) on disease and production, and the value of each damaging effect:

L = ∑ population size × incidence × effect × value

The calculations need to be done for each damaging effect and the associated value (drop in milk production, abortion, death, etc.).

The economic losses at the herd level are often determined from what actually happened in a particular herd or group of herds. In larger populations and at the national level, it is more common to first establish the relevant variables (population size, incidence, effect, and values) and then include them in economic models. The following section gives some examples of economic estimations that have been published. The size of the economic estimations is cited as they appeared in the original publications, sometimes together with recalculations in other publications. In addition, if standardization has been made (e.g., measuring economic values per cow), this is also included.

Present values (PV) can be calculated based on earlier values (EV) and the real rate of interest (r), where the real rate of interest is the difference between the market rate of interest and the inflation rate (Huirne and Dijkhuizen, 1997). The formula for calculating the present value based on a given earlier value n years ago is

PV = EV × (1 + r)n

In addition, the exchange rate between currencies should also be considered.

ECONOMIC LOSSES AT THE HERD LEVEL

There are many ways to estimate the economic impact at the herd level in any given outbreak situation. The outbreak situation can consist of accumulated clinical data of both reproductive disorders and losses among PI animals over a longer time period, or they may only estimate the sudden losses of an outbreak of mucosal disease over a few months. In any case they usually reflect observable problems. Often only outright losses are included.

In an outbreak of abortion, neonatal death and subsequent mucosal disease in a 67-cow dairy herd the estimated economic losses varied between £1720 and £4115, depending on whether dead or culled animals were replaced (Duffel et al., 1986). In U.S. dollars this corresponds to approximately $40–95 per dairy cow. A study of 14 outbreaks in the Netherlands including losses due to abortions, stillbirths, various clinical lesions, and mucosal disease had losses in the range of 42Dfl. to 285Dfl. ($24–161) per dairy cow (Wentink and Dijkhuizen, 1990). In outbreaks of mucosal disease in eight Danish dairy herds, the losses due to mucosal disease only were estimated to be $13–39 per animal corresponding to $33–98 per dairy cow (Houe et al., 1994).

Case descriptions of severe outbreaks of acute BVD have resulted in losses several times higher. An outbreak with a combined infection with BVDV,

Leptospira Hardjo and Coxiella burnetii was, associated with death in adult cows, abortion, and neonatal mortality. This resulted in a total cost of more than £50,000 (Pritchard et al., 1989) corresponding to $410 per cow. Seven outbreaks of severe acute BVD in Canada with high mortality and abortion rate caused losses in the range of $40,000–100,000 per herd (Carman et al., 1998). These herds had between 40 and 191 animals, so the losses per cow would be higher than $400 per cow.

This means that whether we are dealing with a more “classic” outbreak with losses in the range of $50–100 per cow (note that these figures are per all cows being in the herd and not only diseased cows) or we are dealing with severe outbreaks of acute BVD with losses of more than $400 per cow, the losses make up a significant percentage of the total value of the livestock. Due to the variation of these estimations and the uncertainty of how representative these cases are, we can only have rough indications of the true losses or costs. Therefore BVDV is among the diseases where an outbreak can be devastating for the economy of the individual farmer.

ECONOMIC LOSSES IN LARGER

POPULATIONS AND AT THE NATIONAL LEVEL

The losses or costs for the cattle industry may be better calculated using average values for variables obtained from epidemiological studies rather than summing up the losses described in case studies. There are significant differences in the virulence of BVDV strains and both genotype 1 and genotype 2 have high virulent strains. We only have sparse information on the occurrence of strains with different virulence, so this information cannot be directly incorporated in the models for calculating the economic impact. Therefore, many calculations have been performed anticipating an average effect of infections within the area, or the estimations have been made anticipating either a high-virulent or a low-virulent strain being present.

Using the formula for production loss, the economic losses in Denmark before initiation of the eradication campaign have been estimated as $20 million per million calvings (Houe et al., 1993b; Houe, 1999). Including the variables on damaging effects from virulent strains the losses were estimated as $57 million per million calvings (Houe, 1999). These calculations were based on an annual incidence risk of 34% corresponding to very high prevalence of infection, and therefore the magnitude of the losses should be adjusted when considering areas with a lower prevalence (for details see Houe, 1999). The losses in Norway (at the start of the eradication program) were estimated as 26 million NKr per year (Valle et al., 2000b) corresponding roughly to $10 million per million calvings. This corresponds well with the lower prevalence seen in Norway as compared to the prevalence in Denmark. In Great Britain, the direct costs associated with BVDV have been estimated as between 5 and 31 million£ (1996 values) (Bennett et al., 1999b), which with an estimated 4.5 million calvings roughly corresponds to $6 million per million calvings. In this case, the population at risk was considerably lower compared to the one used in the Danish model. In Canada, the total annual cost for an average 50-cow dairy herd have recently been estimated as $48 per cow (Chi et al., 2002).

It is noteworthy that in a high-prevalence area the average loss due to infection is almost half the amount of losses calculated in the outbreak situation. But in many herds the losses were not very obvious because most herds were partly immune and abortions and deaths were more often seen as occasional cases than actual outbreaks. Therefore calcu-

lation of the economic losses has been an important motivator for considering control programs both when farmers have experienced outbreaks and when there was a more continuous occurrence of various losses.

OPTIMIZING DECISIONS BASED ON

ECONOMIC CALCULATIONS

From an economical point of view the optimal situation is when the total costs are minimized—i.e., the sum of losses and expenditures are minimized, which is the basic economical principle (McInerney et al., 1992). Naturally, the more money we spend on treatment and prophylaxis the more the production losses will be reduced. However, we should only continue to increase expenses as long as $1 spent on treatment and prophylaxis will cause a reduction in production loss of more than $1. In practice, it is not always possible to change the size of the expenditures on a continuous scale because a control strategy (e.g., vaccination) will only make sense if it is performed to a certain extent. But still calculating losses and expenses for different control strategies makes them directly comparable from an economic point of view.

ECONOMIC EVALUATION OF CONTROL

STRATEGIES AT THE HERD LEVEL

Decision tree analyses have been used to support decision of selecting a blood testing strategy following outbreaks of mucosal disease (Houe et al., 1994). The calculations showed that testing a risk group of animals being no more than 3 months younger or older than the age of the index case of mucosal disease was most beneficial to the greatest number of farms. However, the most beneficial strategy was always dependent on the individual farm. Furthermore, decision tree models are not suitable for handling the long-term effect of control decisions. The effect of different control strategies has been evaluated in simulation models. A model simulating Dutch conditions showed that culling of PI animals was unattractive (Pasman et al., 1994). However, the conclusion was highly dependent on the risk of reinfection. For comparison, a Markov Chain model assuming that reinfection could be avoided was in favor of eradication. Bennett (1992) described a decision support system showing that depending on the farm-specific situation, the recommendation could be any of the following: a do nothing strategy, a culling strategy, or a vaccination strategy. In conclusion, a general optimal control strategy cannot be given because it will depend on an evaluation of the

farmer’s capability to avoid reinfection and certainly also of the general infection status of the area the farm is located in.

ECONOMIC EVALUATION OF NATIONAL

ERADICATION PROGRAMS

In Norway, a cost-benefit evaluation of the control and eradication program from 1993 to 1997 was made by subtracting the program costs from the benefits (Valle et al., 2000b). The 1993 net present value for the 5 years of the program was calculated as $6 million, and the program was cost-effective in the second year. For comparison, it was estimated that an eradication program in France would take 15 years to be cost-effective (Dufour et al., 1999). Before the eradication program was initiated, the annual losses in Denmark were estimated as approximately $20 million. Thus, this amount is a possible benefit that can be obtained after total eradication. In the initial phases the cost of the program was approximately $9 million per year (Bitsch and Rønsholt, 1995), thereafter being reduced to $3.5 million per year. The costs for continuous monitoring have not been established but will be significantly lower. Because the country is close to total eradication, an annual benefit of almost $20 million is obtained and the program can therefore be considered highly cost-effective. As for the individual herd level, the calculations will be highly dependent on the capability to stay free of infection. Another important factor for making the Scandinavian program very cost-effective is the low cost of using bulk tank milk testing for antibody for the continued monitoring of dairy herds.

CONCLUDING REMARKS AND PERSPECTIVES

This chapter summarizes results of epidemiological studies on the occurrence of bovine virus diarrhea virus (BVDV) infections, identification of risk factors, and effects of the infection on disease and production. The occurrence is described as prevalence and incidence of antibodies and virus, both at the animal level and at the herd level. It is concluded that laboratory methods are sufficiently valid at the animal level for their use in epidemiological studies. At the herd level the sensitivity and specificity show larger variation. Studies show that in many areas the prevalence of PI animals is 0.75–2% and that 60–70% of animals are antibody-positive. However, there may be regional differences in prevalence. For example, many studies in the U.S. have revealed a lower prevalence of PI animals than those reported

in European studies. Other differences in prevalence may be related to cattle density and management style. For example, some areas with low cattle population density and small herds seem to have relatively low prevalence.

Susceptible animals housed together with PI animals will have a very high incidence rate of infection, and almost 100% of the animals will be infected within a few months. In production systems in which subgroups of the herd are segregated from PI animals, these subgroups of animals can stay uninfected for long periods.

Risk factors for BVDV infections are often a reflection of the risk of direct or indirect contact with PI animals. Risk factors include livestock trade, pasturing of animals, use of common pasture, cattle density in the area, herd size, number of infected neighbors, fence breakout, other animal contact between herds (e.g., exhibitions), insufficient hygienic procedure, and occurrence of other species (sheep, wild life). The documentation in literature on the importance of risk factors is often surprisingly low as compared to the obvious effect one would anticipate. Often the proven risk factors explain only a relatively low percentage of infections. Some of the reasons for the difficulties of establishing the importance of risk factors may be due to uncertainties of the time of infection of herds. The improved surveillance systems for herds make larger risk factor studies possible in the future with more precise identification of the herd infection time. Furthermore many risk factors are confounded among each other increasing the need for studies on a larger scale. With the high number of prevalence studies being performed, it would be recommended to gather information on risk factors in a uniform way between studies in order to combine them in formal metaanalyses. These studies should also include information on cattle demographics (e.g., cattle density, herd density, herd size, trade patterns, use of vaccination).

Postnatal infection of immunocompetent animals is often subclinical in most animals. Some may show the typical clinical signs of acute BVD. Transient infection in cows may be followed by reduced milk production, higher incidence of other diseases—such as mastitis—ketosis, and retained placenta. Transient infection in calves may be followed by increased frequency of respiratory diseases and diarrhea. The infection can have significant effect on reproductive disorders, such as conception rate, abortions, congenital effects, and weak-born calves. PI calves are often weak and un-

dersized and have a considerable increased risk of either being culled early due to unthriftiness or dying throughout their lifetime. Different observational studies on the effect of infection show remarkable differences in the clinical consequences. Although some of the differences obviously seem to be due to differences in virulence, a closer understanding of the host-agent-environment triad seems relevant for the understanding of the variation. There is a high variation in the general disease level in herds without BVDV infection, and the importance of infection to herds with a general high health status as compared to herds with low health status needs to be clarified. A lot of the variation seen in the observational studies can also simply be due to chance. If for example the virus is spread at a time when there are many cattle in first trimester means different consequences compared to the situation where most cattle would be in the last trimester. Such differences would be most obvious in herds with seasonal calvings. Furthermore many observational studies are based on the fact that some infection took place, but without knowing exactly how many seroconversions occurred.

Estimations of economic impact are often attempted by combining all the information presented in this chapter, including the occurrence of the infection (prevalence and incidence) and quantifications on the different clinical and production consequences following infection. In order to improve the economic models, a closer description of distribution of the different genotypes and especially the differences in virulence both within and between genotypes would be needed.

ACKNOWLEDGMENTS

Thank you to associate professor Annette Kjær Ersbøll and assistant professors Søren Saxmose Nielsen and Nils Toft for critical reading of the manuscript.

REFERENCES

Alban L, Stryhn H, Kjeldsen AM, et al.: 2001, Estimating transfer of bovine virus-diarrhoea virus in Danish cattle by use of register data. Prev Vet Med 52:133–146.

Alenius S, Jacobsen SO, Cafaro E: 1986, Frequency of bovine viral diarrhea virus infections in Sweden among heifers selected for artificial insemination.

Proc World Congr Diseases of Cattle 14:204–207. Alenius S, Lindberg A, Larsson B: 1997, A national

approach to the control of bovine viral diarrhoea virus. In: Proceedings of the Third ESVV

Symposium on Pestivirus Infections. Eds. Edwards S, Paton DJ, Wensvoort G, 19–20 September 1996, pp.162–169. Lelystad, The Netherlands.

Ames TR: 1986, The causative agent of BVD: Its epidemiology and pathogenesis. Vet Med 81:848–869.

Baker JC: 1995, The clinical manifestations of bovine viral diarrhea infection. Vet Clin North Am Food Anim Pract 11:425–445.

Barber DML, Nettleton PF: 1993, Investigations into bovine viral diarrhoea virus in a dairy herd. Vet Rec 133:549–550.

Beaudeau F, Belloc C, Seegers H, et al.: 2001, Informative value of an indirect enzyme-linked immunosorbent assay (ELISA) for the detection of bovine viral diarrhoea virus (BVDV) antibodies in milk. J Vet Med B 48:705–712.

Bennett RM: 1992, Case-study of a simple decision support system to aid livestock disease control decisions. Agric Sys 38:111–129.

Bennett RM, Christiansen K, Clifton-Hadley RS: 1999a, Estimating the costs associated with endemic diseases of dairy cattle. J Dairy Res 66:455–459.

Bennett RM, Christiansen K, Clifton-Hadley RS: 1999b, Modelling the impact of livestock disease on production: Case studies of non-notifiable disease of farm animals in Great Britain. Anim Sci 68:681–689.

Bitsch V, Hansen K-EL, Rønsholt L: 2000, Experiences from the Danish programme for eradication of bovine virus diarrhoea (BVD) 1994–1998 with special reference to legislation and causes of infection. Vet Microbiol 77:137–143.

Bitsch V, Rønsholt L: 1995, Control of bovine viral diarrhea virus infection without application of vaccines. Vet Clin North Am Food Anim Pract

11:627–640.

Bolin SR, McClurkin AW, Coria MF: 1985, Frequency of persistent bovine viral diarrhea virus infection in selected cattle herds. Am J Vet Res 46:2385–2387.

Bolin SR, Ridpath J: 1998, Prevalence of bovine viral diarrhea virus genotypes and antibody against those viral genotypes in fetal bovine serum. J Vet Diagn Invest 10:135–139.

Bolin SR, Ridpath JF: 1996, The clinical significance of genetic variation among bovine viral diarrhea viruses. Vet Med 91:958–961.

Braun U, Landolt G, Brunner D, Giger T: 1997, Epidemiologishe Untersuchungen über das Vorkommen von BVD/MD bei 2892 Rindern in 95 Milchviehbetriben. Schweizer Archiv für Tierheilkunde 139:172–176.

Braun U, Schönmann M, Ehrensperger F, et al.: 1998, Epidemiology of bovine virus diarrhoea in cattle on

communal alpine pastures in Switzerland. J Vet Med A 45:445–452.

Canal CW, Strasser M, Hertig C, et al.: 1998, Detection of antibodies to bovine viral diarrhoea virus (BVDV) and characterization of genomes of BVDV from Brazil. Vet Microbiol 63:85–97.

Carman S, van Dreumel T, Ridpath J, et al.: 1998, Severe acute bovine viral diarrhea in Ontario 1993–1995. J Vet Diagn Invest 10:27–35.

Chi J, VanLeeuwen JA, Weersink A, Keefe GP: 2002, Direct production losses and treatment costs from bovine viral diarrhoea virus, bovine leukosis virus,

Mycobacterium avium subspecies paratuberculosis and Neospora caninum. Prev Vet Med 55:137–153.

Cho HJ, Masri SA, Deregt D, et al.: 1991, Sensitivity and specificity of an enzyme-linked immunosorbent assay for the detection of bovine viral diarrhea virus antibody in cattle. Can J Vet Res 55:56–59.

Couvreur B, Letellier C, Collard A, et al.: 2002, Genetic and antigenic variability in bovine viral diarrhea virus (BVDV) isolates from Belgium. Virus Res 85:17–28.

David GP, Crawshaw TR, Gunning RF, et al.: 1994, Severe disease in adult dairy cattle in three UK dairy herds associated with BVD virus infection. Vet Rec 134:468–472.

Done JT, Terlecki S, Richardson C, et al.: 1980, Bovine virus diarrhoea-mucosal disease virus: Pathogenicity for the fetal calf following maternal infection. Vet Rec 106:473–479.

Drew TW, Sandvik T, Wakeley P, et al.: 2002, BVD virus genotype 2 detected in British cattle. Vet Rec 151:551–551.

Drew TW, Yapp F, Paton DJ: 1999, The detection of bovine viral diarrhoea virus in bulk milk samples by the use of a single-tube RT-PCR. Vet Microbiol 64:145–154.

Duffel SJ, Sharp MW, Bates D: 1986, Financial loss resulting from BVD-MD virus infection in a dairy herd. Vet Rec 118:38–39.

Dufour B, Repiquet D, Touratier A: 1999, Economic studies in animal health decision-making: The costbenefit ratio of eradicating bovine virus diarrhoea in France. Rev Sci Tech Off Int Epiz 18:520–532.

Durham PJK, Hassard LE: 1990, Prevalence of antibodies to infectious bovine rhinotracheitis, parainfluenza 3, bovine respiratory syncytial, and bovine viral diarrhea viruses in cattle in Saskatchewan and Alberta. Can Vet J 31:815–820.

Edwards S, Drew TW, Bushnell SE: 1987, Prevalence of bovine virus diarrhoea virus viraemia. Vet Rec 120:71.

Ersbøll AK, Stryhn H: 2000, Epidemiological modelling of infectious disease in animals: Bovine virus diarrhoea in Danish dairy herds. In: Proceedings of