| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
).1 With such a low income, costs associated with echinococcosis can become a great burden not only for the affected individual and his or her family, but also for the community as a whole. Shiqu County, with an estimated prevalence rate of 4.6% and 4.9% for human AE and human CE, respectively, has one of the highest levels of human echinococcosis ever recorded.2 With a population of 63,000, Shiqu County has been shown to loose approximately 50,000 lifetime prevalence disability-adjusted life years (DALYs) or 1,100 DALYs per year due to human echinococcosis, resulting in a loss of 0.81 DALY per individual.2
|
Economic losses due to lost income during illness, treatment, and the convalescent period must be taken into account, as should mortality related loss of income. In addition, economic and social losses associated with undiagnosed and, therefore, untreated cases need to be considered. Animal production losses must also be evaluated in the case of CE. These include losses from infected sheep, goat, and yak livers, as well as decreased hide value, carcass weight, and reproduction. Currently, the vast majority of expenses attributable to both human echinococcosis and production losses due to livestock infected with E. granulosus are being absorbed by the local community. This includes infected individuals who have to pay for treatment and lose money due to lost work, as well as local herdsmen who must absorb the costs of decreased livestock production. Due to the public health threat from the local infected dog population, as well as the impact on the local economy, it is suggested that a publicly funded control program be implemented for this region, with the cost of this program shared between the public health and agricultural sectors. The most economically and logistically feasible way to decrease the incidence of human echinococcosis in the Shiqu County region is the practice of deworming local dogs combined with a sheep and goat vaccination program.3 Praziquantel is inexpensive, if purchased in bulk for a control program, and requires limited effort and technical skill to distribute. A livestock vaccination program would help to decrease E. granulosus prevalence, even though it would not effect the transmission of E. multilocularis. The addition of an education program would also be beneficial through decreasing the amount of raw offal fed to dogs.4
Mathematical models have suggested that a combined dog deworming and sheep and goat vaccination program would be most effective in substantially reducing the prevalence of cystic echinococcosis in animals.4 This information is further supported by large-scale vaccination field trials in Xinjiang (Hutubi County) and Sichuan (Ganzi County), People’s Republic of China.3 In addition, a study performed in Shiqu County examining the transmission dynamics of E. granulosus and E. multilocularis in owned dogs indicated a mean infection pressure of one infectious insult every 4.8 years for E. granulosus and one infectious insult every 1.9 years for E. multilocularis, assuming a 5-month E. multilocularis life span or once every 1.2 years, assuming a 3-month E. multilocularis life span, which has recently been suggested by experimental infection in dogs (Kapel CMO and others, unpublished data).5 This indicates that deworming once every six months should help control both E. granulosus and E. multilocuaris if sufficient coverage is obtained and the number of susceptible individuals in the population is somewhat greater than the current number of cases. A more intensive deworming program (e.g., monthly or every six weeks) was considered for the Shiqu County area, but was judged impractical due to the pastoral lifestyle of the local inhabitants. However, we propose the addition of a stray dog baiting program at the same time as deworming of owned dogs. Cost-benefit analysis was used, along with DALYs lost, to determine costs per DALY saved if the proposed control program was implemented. Findings were then evaluated to see whether the proposed control plan was within the World Health Organization criteria for a cost-effective strategy.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Income losses. Based on findings from the short form 12 version 2 quality of life survey (QualityMetric, Inc., Lincoln, RI), a generic measure of general health and well being, subjects diagnosed with echinococcosis had a significantly decreased quality of life for all areas tested, including physical functioning, compared with an age and sex cross-matched control population.2 Such a decrease in quality of life is likely to affect the ability to work and thus generate income. Therefore, a case-control study of individuals with echinococcosis was undertaken on the same population as previously described. Ethical approval for all work carried out within China was obtained from the Medical Sciences Expert Consultant Committee, Sichuan Provincial Health Bureau (Sichuan Province, People’s Republic of China). Questions relating to income were used to categorize adult subjects into four income brackets: < U.S.$120 per year, U.S.$121–U.S.$241 per year, U.S.$242–U.S.$362 per year, and > U.S.$362 per year.
Analysis of human-associated losses. The proportions of adult subjects in each income bracket with and without echinococcosis were compared using the chi-square test. Significant differences were used to estimate income losses for affected individuals and this data was used for further analysis. A spreadsheet model was then constructed in Excel® (Microsoft, Redmond, WA). Variables affecting human economic losses due to treatment costs and income losses were randomly varied along their distributions and summed using Monte-Carlo techniques to model outcome uncertainty. Overall, 10,000 simulations were performed.
Monte-Carlo resampling techniques were again used to assign clinical severity and income loss to human echinococcosis cases. The AE and CE cases were assigned a clinical severity outcome according to a multinomial distribution based on literature values of cases treated solely with albendazole.2 A multinomial distribution is an extension of the binomial distribution when there are more than two possible outcomes for each iteration. Individuals in each category were then assigned a reduction of income based on projected severity of disease. The various categories were assigned a reduction in per capita GDP at a level of 2% for 5 years, 5% for 46 years (average estimated life span at time of diagnosis), 10% for 46 years, 25% for 46 years, 50% for 5 years, followed by 100% for 41 years, or 100% for 46 years indicating death. Undiagnosed cases, based on extrapolation of the ultrasound positive cases to the entire population of Shiqu County and corrected for age and sex bias, were also allocated a loss of income. A uniform distribution of 0–5% loss of income was applied to undiagnosed CE cases, which is comparable to losses assumed in past studies.6,7 Due to the increased clinical severity associated with AE, a 0–10% loss of income was assigned to undiagnosed AE cases. Loss of income for undiagnosed cases was applied until the end of the expected life span based on the West Level 26 life table used to calculate DALYs lost for this population.2 The West Level 26 life table is based on the Japanese estimated life span, which is one of the longest known, and was used to standardize DALYs lost in accordance with the Global Burden of Disease Study. A 3% annual discount rate was applied to all income calculations.8 All distributions were sampled across 10,000 times and mean and 95% confidence intervals (CIs) were obtained for losses.
Analysis of livestock-associated losses.
Livestock numbers for Shiqu County were derived from 1997 published statistics, with population size assumed to be normally distributed.9 Prevalence values were taken from the most comprehensive E. granulosus intermediate host study, which was performed during the 1980s. During this study, 7,874 animals (3,645 yaks, 4,104 sheep, and 125 goats) were examined in Shiqu County with an infection rate of 49.9% for yaks, 81.7% for sheep, and 40.8% for goats.10 Beta distributions were then used to model uncertainty in the prevalence estimates. Values for livestock-related products were determined for the region based on local market conditions or extrapolated from other studies (Tables 2–4).1,7,10–12 Log-normal distributions were applied to losses associated with liver condemnation, decreased carcass weight, and decreased number of young born to infected sheep and goats. A log-normal distribution was chosen since the values for each are most likely skewed towards the lower end of the distribution and cannot be less than 0. This is due to the fact that most infected animals will be lightly to moderately infected, with only a few in the highly infected range. A log-normal distribution was also applied to decreased hide value in sheep and liver condemnation and decreased carcass weight and hide value in yaks. A uniform distribution of 1–5% decrease in calves born to infected female yaks was applied. A uniform distribution allows for an equal probability of occurrence over the entire distribution. This type of distribution was chosen since there is no prior information on how Echinococcus infection affects reproduction in yaks, yet some decrease is being presumed since the phenomena has been suggested in sheep and goats.11 Overall annual losses were estimated for liver associated losses alone, as well for livestock associated losses with and without decreased reproduction, carcass weight, and hide value in yaks due to the lack of data on the effect of E. granulosus infection in yaks. Monte-Carlo resampling techniques were used across the distributions 10,000 times and a mean and 95% CI was determined. All computations were performed using an Excel® spreadsheet along with the statistical add-in PopTools (Commonwealth Scientific and Industrial Research Organisation, North Ryde, New South Wales, Australia).
|
|
|
Costs/benefits. Complete eradication of both E. granulosus and E. multilocularis is extremely unlikely due the near impossibility of attaining 100% compliance, as well as the continued maintenance of E. multilocularis in a wildlife cycle. In light of past control studies, a mean compliance rate of 75% is a more attainable goal for a rural area such as Shiqu County.13 Due to the variation in life cycles and parasite life expectancy, a single control program will effect amount of control obtained for each parasite differently. Mathematical models for E. granulosus have indicated the possibility of a more than 90% decrease in intermediate host prevalence within 10 years and near complete eradication of the parasite in 15–20 years, assuming a biannual dog deworming scheme together with a sheep and goat vaccination program, with an average compliance of 75%.4 A conservative estimate of a long-term (e.g., 20 years) E. granulosus control program, based on two times per year owned dog deworming and stray dog baiting plus sheep and goat vaccination, was estimated at a mean of approximately 80% reduction in intermediate host prevalence, with a minimum level of 65% and a maximum estimate of approximately 95% reduction. These values were then used in a cost-benefit analysis.
The proposed intervention strategy will have less effect on E. multilocularis because the parasite is only being controlled at one point in the life cycle, i.e., the definitive host. In addition, the parasite has a shorter life expectancy and is maintained in a wildlife cycle.14 More probable is the establishment of a new equilibrium, with a lower prevalence in the dog population and human cases reported at a lower incidence level. Pre-control abundance 
in the dog definitive host in the absence of parasite-induced immunity can be modeled as
 |
where h is the prevailing infection pressure in number of parasites per year, t is the dog’s age, and µ is the rate of loss of infection (1/µ = parasite life span).15 In this case, time t is modeled as the average age of the dogs from this population (4.5 years).16 The same equation can be used to estimate the average abundance of E. multilocularis in dogs that become infected over the six-month period between anthelmintic treatments. Upon treatment with praziquantel, all parasites will be removed from the dogs. By making the conservative assumption that all infections in dogs are transmitted as a spillover from the fox-small mammal life cycle, the same equation can be used as the infection pressure remains unchanged. The prevalence six months after treatment can be calculated by equation 1 using the steady state infection pressure and making t = 0.5. Thus ', the new mean abundance of E. multilocularis in the dog population, can be estimated by finding the solution of
 |
which is the mean value of equation 1 between zero and six months. To model uncertainty in the estimates of parameters µ and h, the parameters were assigned distributions based on abundance models applied to data from dogs of this region.5 Therefore, µ is modeled as a normal distribution with a mean of 2.4 and a standard deviation of 0.5, with a corresponding normally distributed h with a mean of 334 and standard deviation of 60. The parameter µ was also modeled as a normal distribution with a mean of 4 and a standard deviation of 0.8, with a corresponding normally distributed h with a mean of 533 and standard deviation of 100.5 Both sets of values were used to model E. multilocularis life span in the dog at five months and at three months, with values for h determined by abundance data for the area.5 This function was then weighted according to a uniformly distributed compliance rate with a mean of 75% and lower and upper limits of 60% and 90%, respectively. Post-control abundance was estimated as post-control abundance, assuming 100% compliance, multiplied by the compliance distribution with a mean of 75% and added to pre-control abundance multiplied by one minus the compliance rate.
Change in human incidence of AE was then modeled taking into account the number of susceptible individuals in the population and the number of current cases. A simple model can be derived to model the numbers of human cases N as a function of the abundance in dogs
 |
where S is the number of susceptible individuals and 
is a transmission parameter that encompasses the contact rate between parasite and humans, as well as factors affecting the viability of the free living eggs in the environment. The parameter is the mean abundance of parasites in the dog population. From this it can be shown that
 |
The number of potentially susceptible people in a population of 63,000 was modeled as a log normal distribution with a mean of 3,000 and a standard deviation of 5,000, which was then shifted to the right by 3,000 and maximum number of susceptible individuals truncated at the population size of 63,000. This was to model possible numbers of susceptible individuals from an estimated minimum of approximately 5% of the population (5% were found to be positive by abdominal ultrasound screening) up to a maximum of the total population.2 However, this distribution will skew the numbers of cases towards the lower limit because it is possible that a high proportion of susceptible individuals are already infected due to the local conditions of severe and widespread poverty combined with the population living in conditions of poor hygiene and in intimate association with the dog population. The number of current cases in the population was modeled as a beta distribution based on 180 AE-positive individuals of 3,135 abdominal ultrasound–screened individuals and multiplied by a correction factor for age and sex structure of the Shiqu County population.2 The transmission parameter can be calculated for each value of N, S, and drawn from the prior distributions. This value is used on each occasion to estimate the numbers of new cases N' assuming that
 |
The prior distributions were sampled 10,000 times, with the model recalculating the posterior value of N' on each occasion. The upper 97.5% and lower 2.5% values of N' were used to calculate the 95% CI for the number of new cases. These values were then used in a cost-benefit analysis for the reduction in human disease.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Total cost for the current population is estimated at U.S.$1,507,224 (95% CI = U.S.$525,737–2,496,698), with a per capita lifetime cost of U.S.$23.94 (95% CI = U.S.$8.30–39.38) and an annual cost of U.S.$32,788 (95% CI = U.S.$11,120–54,215), equating to a loss of approximately 0.2% of per capita GDP each year.
|
. Losses due to infected livers alone equate to U.S.$185,635 (95% CI = U.S.$167,793–205,389) per year. Total losses associated with discarded livers, decreased reproduction, and decreased carcass weight in sheep, goats, and yaks in addition to decreased value of sheep and yak hides are U.S.$903,649 (95% CI = U.S.$717,158–1,113,354) annually. When losses due to decreased calf production, losses in carcass weight of yaks, and decreases in the value of yak hides were subtracted from this value, annual losses were U.S.$449,189 (95% CI = U.S.$444,275–559,131). Annual combined human and animal losses due to echinococcosis were U.S.$218,676 (95% CI = U.S.$189,850–247,871), when only liver-associated losses in livestock are assumed. This equates to approximately U.S.$3.47 per person annually or 1.4% of per capita GDP. A loss of U.S.$532,249 (95% CI = U.S.$472,112–595,561) was estimated assuming the only loss in yaks is due to the discarded liver, resulting in a loss of U.S.$8.44 per individual annually or a 3.5% loss of per capita GDP. The worse case scenario would include losses associated with discarded livers, decreased reproduction, and decreased carcass weight in sheep, goats, and yaks, in addition to decreased value of sheep and yak hides, which when added to human losses, would result in losses of U.S.$936,408 (95% CI = U.S.$746,807–1,148,242) annually.
|
Costs/benefits.
An 80% long-term post-control reduction in human and livestock CE incidence, with a minimum level of 65% and a maximum estimate of approximately 95%, is assumed based on mathematical models of control intervention. A post-control decrease in human incidence of AE, with a mean of 31% and a 95% CI of 13–50%, was estimated based on a five-month life span for E. multilocularis in the dog definitive host. A 21% decrease in human incidence, with a 95% CI of 9–38%, was predicted using a three-month E. multilocularis life span in the dog definitive host. Potential economic benefits assuming prevention of AE (assuming a five-month life span) and CE-associated human losses in addition to liver-associated losses due to E. granulosus, livestock losses with only liver associated losses in yaks, and all livestock losses due to discarded livers, decreased reproduction, and decreased carcass weight in sheep, goats, and yaks, in addition to decreased value of sheep and yak hides, are shown in Table 7. Economic benefits assuming a three-month E. multilocularis life span in the dog definitive host, are not shown since the difference between the use of a five-month versus a three-month life span results in less than a U.S.$3,000 per year difference for any category.
|
Cost per DALY averted. Assuming that an average of 80% of DALYs due to CE are averted and 31% of DALYs due to AE are averted (based on an average E. multilocularis life span of five months), with the proposed dog deworming and sheep and goat vaccination program, the cost per DALY saved is U.S.$106.88 (95% CI = U.S.$88.63–127.99). If an average of 80% of CE DALYs and 21% of AE DALYS (based on an average E. multilocularis life span of three months) are averted, the cost per DALY saved is U.S.$123.46 (95% CI = U.S.$102.29–148.15). Lower and upper limits, respectively, of the number of potential DALYs saved, assuming a three-month E. multilocularis life span, result in 65% of CE DALYs being averted and 9% of AE DALYS being averted, resulting in a cost of U.S.$179.75 (95% CI = U.S.$147.50–217.76) per DALY saved, and 95% of CE DALYs and 38% of AE DALYS being averted, resulting in a cost per DALY saved of U.S.$88.81 (95% CI = U.S.$73.63–106.42). Lower and upper limits of potential DALYs averted, assuming a five-month E. multilocularis life span, results in a lower limit of 65% of CE DALYS being averted and 13% of AE DALYs being averted, resulting in a cost of U.S.$164.77 (95% CI = U.S.$136–198.22) per DALY saved, and an upper limit of 95% of CE DALYS being averted, and 50% of AE DALYs being averted, resulting in a cost of U.S.$78.35 (95% CI = U.S.$65.02–93.93) per DALY saved.
However, if cost-sharing was implemented between the public health and agricultural sectors proportional to the overall benefit of each sector, assuming the suggested control program and livestock losses due to the most conservative estimate of liver-associated losses only, the cost per DALY averted attributable to the public health sector, assuming a three-month E. multilocularis life span, would be U.S.$11.73 (95% CI = U.S.$9.72–14.07). If a five-month E. multilocularis life span is assumed, the cost to the public health sector is estimated to be U.S.$10.15 (95% CI = U.S.$8.42–12.15) per DALY averted.
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In regards to the economic impact of echinococcosis due to losses in domesticated livestock, very little is known of the impact of E. granulosus infection in yaks. Therefore, calculations were performed with and without losses due to decreased reproduction, carcass weight, and hide value. Studies on the production impact of echinococcosis in yaks will need to be performed to narrow the estimated economic losses in these animal populations. In addition to sensitivity analysis on the contribution of infected yaks to overall livestock losses, Monte-Carlo techniques were used by varying parameter values across distributions to show variability in economic losses. The overall economic impact of echinococcosis in the population of Shiqu County is severe, especially in light of the low economic status of the region. An annual loss of at least 1.4% of per capita GDP due to echinococcosis compares unfavorably to studies performed in Jordan and Uruguay, which indicated losses of 0.074% and 0.058%, respectively, of annual per capita GDP due to E. granulosus infection.6,7 However, this comparison is not optimal since whole country losses for Jordan and Uruguay, which include substantial urbanized areas, are being compared with a remote rural area of the Tibetan plateau. A better comparison would be between rural areas of Jordan and Uruguay, which are more likely to be poorer than the overall country averages, and the Shiqu County study area. In this case, the proportion of annual GDP lost due to echinococcosis would be higher for Jordan and Uruguay than the original studies suggested.
Due to the severity of echinococcosis infection, a control program aimed at anthelmintic prophylaxis of dogs and vaccination of sheep and goats would have a beneficial result even if a relatively small percentage of human and livestock cases could be avoided. In addition, because of the large impact that E. granulosus has on the domestic livestock industry, which results in the majority of economic losses, the control program suggested here would be beneficial even without taking into account benefits due to the prevention of human AE. Therefore, in essence all savings in terms of AE can be considered an added benefit of the control program. In Shiqu County, the definitive host assumed to have the most impact on Echinococcus spp. transmission to humans is the domestic dog. Previous studies have emphasized the need for the destruction of stray dogs to truly effect the transmission of Echinococcus spp.19 In the case of the Tibetan plateau, this control method was considered. However, it has not been implemented due to the strong religious beliefs of the people of this region. In addition, expanding the proposed vaccination program to include yaks was considered. However, studies have shown that the yak (Bos grunniens) most likely is not an adequate host for E. granulosus due to arrested metacestode development in this species.20 Therefore, vaccination of yaks is not necessary to control echinococcosis in this district.
Deworming of the wildlife definitive hosts of E. multilocularis, in this case the red fox (Vulpes vulpes) and Tibetan fox (Vulpes ferrilata), through the use of baits is not a viable option due to the large geographic area and the substantial funding necessary to implement such a program. Some baits distributed for stray dogs will most likely be consumed by foxes. However, the vast majority of the wildlife hosts will not be reached. Controlling the infection in domestic dogs will alleviate the pressure on humans in the area, but will not eliminate the principle cycle of E. multilocularis since small mammals will continue to be infected by wild canids. Another concern when dealing with a cycle maintained in a wildlife population is that any cessation of the control program would likely result in an eventual return to previous prevalence levels. Therefore, a control program would have to be a permanent commitment for the community and eventual dismantling would not be an option. Echinococcosis multilocularis life spans of both five months and three months were used in analysis based on the five-month life span reported in the fox and newly reported information citing a three-month life span in experimentally infected dogs (Kapel CMO and others, unpublished data).14
The role of resistance to E. multilocularis in the human host is currently being studied. Preliminary research has shown possible genetically based mechanisms of susceptibility/resistance including the influence of the HLA B8, DR3, DQ2 haplotype.21,22 If genetic susceptibility plays a large role in the acquisition of human AE, control programs could have very different outcomes for different human populations. For example, if a population has low numbers of susceptible individuals, a control program may have a more limited effect on decreasing local human incidence. This can be explained by the fact that the infection pressure and transmission rate is high enough to result in most of the susceptible individuals being infected. Thus, a decrease in infection pressure could have a relatively limited effect on the number of cases because of the non-linear relationship between infection pressure and cases described by equation 4. This hypothesis has been suggested to explain increasing prevalence in the fox definitive host in Switzerland, which has yet to lead to a significant increase in human incidence.23 However, the genetic susceptibility of a population is currently not a measurable variable. Therefore, a log normal distribution has been used in the analysis for Shiqu County data to explore various degrees of potential susceptibility in the population.
The most well known E. multilocularis control program, in which domestic dogs played an important role in the cycle, occurred on St. Lawrence Island, Alaska in the 1970s and 1980s. In this region, prevalence rates of E. multilocularis in dogs ranged from 0% to 25% depending on location.24 This can be compared with preliminary findings for Shiqu County indicating an overall E. multilocularis prevalence between 13% and 33%.16 On St. Lawrence Island, the control program consisted of monthly dosing of village dogs with the anthelmintic praziquantel, with a capture rate of approximately 90%. Examination of the northern vole (Microtus oeconomus), which acted as the primary intermediate host, was used as an index for the parasite in the environment. Over the 10-year control program, the prevalence of E. multilocularis in the village vole population was reduced from approximately 29% to 3%. Since overall prevalence in Shiqu County dogs is comparable to that of St. Lawrence Island, a biannual deworming scheme, versus the monthly deworming program on St. Lawrence Island, can be predicted to have a significantly more limited impact assuming all other variables are similar. This is confirmed in the model predictions. In Shiqu County, deworming would also have to be carefully timed with the movement of the herdsmen and their dogs between summer and winter pastures, and the impact of a nomadic lifestyle on Echinococcus transmission would need to be further investigated. In addition, it is not known to what extent dogs in Shiqu County are actively involved in the transmission cycle and thus infecting small mammals, or if they are primarily acting as a dead-end host from spillover from the fox-small mammal cycle. If the former situation is true, then the regular deworming of dogs would have a greater impact on the reduction of AE compared with the latter situation.
Previous studies have also evaluated possible control programs geared towards the eradication or decreased prevalence of E. granulosus.4,25 However, since the use of vaccination against E. granulosus in livestock is still in its infancy, there are few case studies looking at the long-term impact of a control plan incorporating vaccination. One example of an E. granulosus control program was carried out in La Rioja, Spain and consisted of deworming herding dogs with praziquantel every six weeks and non-herding dogs with praziquantel every four months, with questionable compliance, in addition to an education program and culling of stray dogs. At the end of the 14-year program, the prevalence in dogs had decreased from 7.0% to 0.2% (reduction of 97.2%) the prevalence in sheep had decreased from 82.3% to 20.3% (reduction of 75.4%), and the rate of diagnosis of new human cases had decreased by 78.9%.26 Preliminary E. granulosus prevalence rates for Shiqu County of 8–19% in dogs and 81.9% in sheep make it a comparable initial situation.16 However, complete eradication of E. granulosus in Shiqu County would be very difficult due to the continental situation and thus immigration of infected animals into the area. Therefore, sporadic cases in humans would likely still occur even after near eradication and a long-term surveillance program would need to be maintained.
If one uses the above-mentioned E. multilocularis and E. granulosus control programs as a measuring stick for success of past programs, in addition to work done with mathematical models, it can be anticipated that a two times per year dog deworming scheme in conjunction with a sheep and goat vaccination program, assuming a compliance of approximately 75%, should decrease prevalence in the intermediate hosts, dog definitive hosts, and human aberrant hosts by 65–95% for E. granulosus.4 There are a greater number of unknown factors associated with how a dog deworming program will affect E. multilocularis incidence. However, based on the infection pressure calculated from mathematical models, a decrease in human incidence between 9% and 38% or between 13% and 50% is suggested, depending on the life span of E. multilocularis used.5 The large number of DALYs lost annually and per population, along with the economic impact associated with the disease in humans and livestock, makes echinococcosis a parasitic disease worth careful consideration. Assuming either 80% of DALYs due to CE and 21% of DALYs due to AE are averted or 80% of DALYs due to CE and 31% of DALYs due to AE are averted, the cost per DALY saved remains firmly within the World Health Organization second most cost-effective band of less than U.S.$150 per DALY saved if one organization or sector was wholly responsible for control costs.27 However, cost per DALY averted is quite deceptive since large monetary savings of up to U.S.$800,000, due primarily to savings in livestock production factors, pay for the program. If cost-sharing was implemented between the public health and agricultural sectors or between the public and private sectors, cost per DALY attributable to each sector would be far lower. For example, if responsibility for control costs were divided proportionally between the public health sector (human health-related benefits) and the agricultural sector (livestock-associated benefits) the cost to the health sector would fall within the World Health Organization most cost-effective band of less than U.S.$25 per DALY averted. This report has shown that by putting a limited amount of funding into a dog deworming and sheep and goat vaccination program, a large savings in human health and monetary losses due to both human morbidity and mortality, as well as losses in livestock production, can potentially be obtained.
Received October 26, 2004. Accepted for publication December 16, 2004.
Acknowledgments: We thank the local government officials and health services providers of Shiqu County for their assistance in facilitating the fieldwork associated with this project. The authors would also like to thank Dr. Jakob Zinsstag (Swiss Tropical Institute, Basel, Switzerland) for his valuable input.
Financial support: This work was supported by an Ecology of Infectious Diseases program grant from the U.S. National Institutes of Health (TWO 1565-02) and by the National Science Foundation.
* Address correspondence to Christine M. Budke, Institute of Parasitology, University of Zürich, Winterthurerstrasse 266a, CH-8057 Zürich, Switzerland. E-mail: budke{at}vetparas.unizh.ch 
Authors’ addresses: Christine M. Budke and Paul R. Torgerson, Institute of Parasitology, University of Zürich, Winterthurerstrasse 266a, CH-8057 Zürich, Switzerland, Telephone: 41-1-635-8535, Fax: 41-1-635-8907. E-mails: budke{at}vetparas.unizh.ch and paul.torgerson{at}access.unizh.ch. Qiu Jiamin and Wang Qian, Sichuan Institute of Parasitic Diseases, 10 University Road, Chengdu 610041, Sichuan, People’s Republic of China, Telephone: 86-28-555-3149, Fax: 86-28-555-8409, E-mails: qjm{at}mail.sc.cninfo.net and wangqian67{at}yahoo.com.cn.
Reprint requests: Paul R. Torgerson, Institute of Parasitology, University of Zürich, Winterthurerstrasse 266a, CH-8057 Zürich, Switzerland.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. S. Boufana, M. Campos-Ponce, A. Naidich, I. Buishi, S. Lahmar, E. Zeyhle, D. J. Jenkins, B. Combes, H. Wen, N. Xiao, et al. Evaluation of Three PCR Assays for the Identification of the Sheep Strain (Genotype 1) of Echinococcus granulosus in Canid Feces and Parasite Tissues Am J Trop Med Hyg, May 1, 2008; 78(5): 777 - 783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. SCHANTZ THE BURDEN OF ECHINOCOCCOSIS Am J Trop Med Hyg, July 1, 2005; 73(1): 1 - 1. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
