The application of formal risk assessment to animal quarantine is still in its infancy but the obligations of membership in the World Trade Organisation mean that the Ministry of Agriculture is having to conduct quantitative risk assessments with increasing frequency. In recent years MAF has conducted quantitative assessments of the risks posed by the importation of a variety of animals or animal products. In one such assessment, conducted as part of a review of policies regulating the importation of dogs, MAF developed a Monte Carlo-type simulation model to assess the risk of releasing a rabid dog from quarantine under each of a number of import policies based on quarantine periods of different duration, with or without verified vaccination status.
The risk of selecting a rabies-infected animal, without any safeguard being in place, was estimated to be a function of the incidence of rabies amongst domestic dogs in the exporting country and the incubation period of the disease. The incidence and incubation period were described by triangular distributions.
Once the magnitude of the unrestricted risk of introducing a rabies-infected dog had been estimated for each country under consideration, the effects of vaccination and quarantine were assessed, again using a range of values in a simulation model. The effect of each safeguard was considered to be the product of the unrestricted risk and the estimate of failure of the safeguard.
On the basis of this risk assessment it was concluded that vaccinated dogs imported without prolonged quarantine pose no greater risk of introducing rabies than dogs entering through a 6 month quarantine. Prolonged quarantine could thus be replaced by vaccination without any reduction in security, reducing significantly the cost of importation and eliminating the prolonged separation which was usually distressing for pets and their owners.
Case study; the risk of introducing rabies through the importation of dogs
The application of formal risk assessment to animal quarantine is still in its infancy but the obligations of membership in the World Trade Organisation mean that quarantine authorities are having to conduct quantitative risk assessments with increasing frequency(9). In recent years the New Zealand Ministry of Agriculture has conducted quantitative assessments of the risks posed by the importation of a number of animals or animal products(8, 9). One such assessment is that carried out to determine the risk that the importation of dogs could introduce rabies.
Rabies is an invariably lethal viral disease which can affect all mammals, including humans. Infection is transmitted in the saliva of rabid animals. Fortunately, rabies has never occurred in New Zealand, having been kept from our shores by a combination of isolation and stringent quarantine measures.
Until recently, importation of dogs was restricted to those which could be demonstrated to have lived for a minimum of 6 months in one of a small group of rabies-free countries comprising Australia, Sweden, Norway, the United Kingdom and Hawaii. Immigrants and New Zealanders returning to this country after sojourns overseas had frequently protested at the long, and expensive, period of separation from family pets required under this import policy. However, recent elucidation of the epidemiology of rabies and demonstration of the efficacy of modern vaccines led the Ministry of Agriculture to review import policies.
Developed countries in which rabies is endemic have in place programmes to control urban rabies. The disease is notifiable, vaccination is encouraged and, in some cases, attempts are made to reduce the incidence in wild animals. Modern vaccines are more effective than those used previously and serological tests capable of measuring the degree of immunity developed in response to vaccination are available. In recognition of these developments a number of countries have been moving away from dependence on prolonged quarantine towards policies based on verification of vaccination status as an additional or alternative safeguard.
It is against this background that an assessment was made of the risk of releasing a rabid dog from quarantine under each of a number of import policies based on quarantine periods of different duration, with or without verified vaccination status(3). The PC programme @RISK (Palisade Corporation) was used to construct a Monte Carlo-type simulation model(10) which calculated the "unrestricted risk" of importing a rabies-infected dog, in the absence of any safeguards. The model then calculated the risk-reducing effects of vaccination and quarantine.
Wildlife rabies
Rabies is maintained in wildlife in various parts of the world. Where several species are involved in wildlife cycles, strain variants of the rabies virus tend to be restricted to a particular host species in any one geographic area. It is unlikely that a rabies strain adapted to a particular wildlife species would become established should it be introduced into a new species of wildlife in a new geographic area(5).
In countries in which rabies is not endemic in urban dog populations, infection of domestic animals or humans is a spillover phenomenon and does not usually result in the establishment of a new cycle. For example, there have been no documented cases of dog-to-dog, or cat-to-cat transmission of rabies within Europe.
Urban rabies
The cycling of the disease in stray and feral dogs and cats, so-called "urban rabies", is more dangerous to humans than rabies in wildlife, and accounts for an estimated 96 % of all recorded human cases(11).
While wildlife rabies is more common in the temperate regions of the world, urban rabies predominates in tropical countries. Dogs imported from such countries in Africa, Asia and South America present the greatest risk of introducing a rabies strain that could establish a cycle in indigenous dogs. Despite this, it is unlikely that such a strain of rabies would establish if introduced into New Zealand because the dog population in this country is subject to a relatively high level of control. There would, however, be considerable outrage and distress if an imported dog developed rabies and transmitted the disease to other animals and humans.
Incidence of rabies in pet dogs
The incidence statistics for rabies usually do not distinguish between pets and stray animals. It is obvious that the incidence in stray animals will be under-reported, because the clinical signs of rabies are variable and if a stray dog becomes sick or is found dead it is not likely to receive veterinary examination. The incidence of rabies in pet animals has, in general, been decreasing as a result of widespread immunisation with effective modern vaccines. In the United States, for example, approximately one third of pets are immunised against rabies.
Immunisation of dogs
Despite the differences in strains of rabies virus, immunisation with modern rabies vaccines will protect against infection with any virulent wild rabies virus. Dogs respond to rabies vaccination by producing neutralising antibody titres varying from 0 to 20 IU/ml or more(1). There is good correlation between antibody titres and protection.
Although challenge experiments indicate that any detectable neutralising antibody gives a degree of protection, a minimum titre of antibody of 0.5 IU/ml is considered desirable.
Effectiveness of vaccination in the field
Vaccine failure is not as common as it was in the past. In 1988, 33 million dogs and cats were vaccinated in the United States and only four cases of vaccination failures (that is, rabies developing in vaccinated animals) were recorded(4).
In France, only ten vaccination failures in dogs and four in cats have been recorded over 23 years(1).
Other studies have demonstrated the effectiveness of vaccination against rabies (2,7). In Lima, Peru vaccination of some 270,000 dogs (around 65% of the dog population) resulted in rabies incidence falling from an annual mean of 1,233 cases to three in the year following vaccination. Immunisation of dogs with modern rabies vaccines is clearly very effective in preventing the disease.
The vaccination status of an imported dog can be readily checked by measuring the antibody titre of its blood serum.
Incubation period
The incubation period of rabies can vary from 7 days to more than a year(1) but is usually between 3-8 weeks. From data presented in published reports, minimum, most likely and maximum estimates of rabies incubation period were made. The values used in the simulation model were 10, 56 and 365 days.
Risk of importing infected animals without safeguards
The unrestricted risk ® of selecting a rabies-infected animal, in the absence of any safeguards, was considered to be a function of the incidence of rabies amongst domestic dogs in the exporting country and the incubation period of the disease.
A Monte Carlo-type simulation model was developed to assess the unrestricted risk. This risk was calculated as;
R = ( i x d)/365
where i is the incidence of rabies and d is the incubation period in days. The incidence and incubation period were described by triangular distributions.
The incidence of rabies in several countries and in the city of Lima (Table I) was used to estimate the risk of randomly selecting a dog infected with rabies.
As it is probable that the number of cases reported underestimates the real incidence of the disease, an attempt was made to address this by creating a range of incidences. For example, the number of cases of rabies in dogs in the United States in 1988 was 91. In the risk assessment it was considered that the "most likely" number of cases was actually 50% higher than this (137 cases). Had the incidence been grossly under-reported, it was considered that the actual number of cases could have been five times greater (455 cases). These values are used to produce the triangular distribution used in the simulation model.
It can be seen that the unrestricted risk of importing an infected dog directly, without safeguards, from the United States, France, Canada and Germany is small, being fewer than 40 cases per million.
Effects of vaccination and quarantine
Once the magnitude of the unrestricted risk had been estimated for each country under consideration, the risk-reducing effects of vaccination and quarantine were assessed, again using a range of values in the simulation model. The final risk estimate for importations under each policy is the product of the unrestricted risk and the probability of failure of each safeguard.
The effectiveness of vaccination against rabies was estimated to be represented by a triangular distribution ranging from 0.8 to 0.99, with the "most likely" value being 0.94. The probability of vaccination failing to protect was thus represented by 1-0.99, 1-0.94 and 1-0.8.
Estimates for quarantine alone failing to prevent the introduction of rabies were based on a study by Gooch(6). Triangular distributions were estimated for quarantine periods of different duration (Table II).
The effect of each safeguard, or combination of safeguards, is the product of the unrestricted risk and the likelihood of failure of the safeguard.
The simulation model was run for 5,000 iterations. Table III shows the risks of introducing rabies following importation of dogs from selected countries after applying different safeguards. In 95% of iterations the risk estimate was less than that shown in Table III.
On the basis of the results of the simulation model it was concluded that vaccinated dogs imported without prolonged quarantine pose no greater risk of introducing rabies than dogs entering following a 6 month quarantine.
Selecting appropriate safeguards
The risk of introducing rabies with imported dogs is related primarily to the incidence in dogs in the country of origin. In North America, epizootics in dogs are related to the disease in skunks, raccoons and foxes, with no evidence of a canine cycle. There is good evidence that in continental Europe fox rabies has never been introduced into a rabies-free area through the movement of domestic animals, even when the administrative rules for control have been broken(1).
In countries with a canine cycle, the incidence of rabies can be expected to be higher and the consequences of introducing this strain are of greater significance. Importation of dogs from such countries should be subject to a longer quarantine period.
There is a thousandfold difference in risk between importing a dog from the United States and Lima, Peru. It is appropriate, therefore, for countries to be ranked according to the estimated risk, and for safeguards to be applied appropriately, either singly or in combination, to reduce the risk of rabies importation to a negligible level.
Conclusion
The risk assessment concluded that the risk of introducing rabies through the importations of dogs can be managed by the application of appropriate safeguards. The risk of introducing rabies under a policy of confirmed vaccination is no greater than under a policy of prolonged quarantine alone.
As a result of this risk assessment, the Ministry of Agriculture now permits the importation of dogs from countries in which rabies occurs in wildlife. Importation is still not permitted from countries in which rabies is established in the canine population.
Inactivated vaccines must be used according to the manufacturer's recommendations and dogs must be at least 3 months of age before vaccination. Microchip identification and certification is performed at the time of vaccination. In the case of first-time vaccination there should be a minimum 6 month post-vaccination period before export to New Zealand. To confirm that vaccination has been successful, animals are tested to insure presence of a minimum titre of rabies antibody of 0.5 IU/ml at least 6 months before export. The test for rabies antibodies is repeated within 30 days of export to New Zealand.
References
(1) Aubert MFA. Practical significance of rabies antibodies in cats and dogs. Revue Scientifique et Technique de l' Office International des Epizooties 11(3), 735-60, 1992.
(2) Chomel B, Chappuis G, Bullon F, Cardenas E, David de Beublain T, Lombard M, Giambruno E. Mass vaccination campaign against rabies: Are dogs correctly protected? The Peruvian experience. Reviews of Infectious Diseases 10, supp 4, Nov-Dec, 1988.
(3) Corrin KC, MacDiarmid SC. The risks of introducing rabies through the importation of dogs. Surveillance 24(1). In press. 1997.
(4) Eng TR, Fishbein DB. Epidemiological factors, clinical findings and vaccination status of rabies in cats and dogs in the United States in 1988. Journal of American Veterinary Medical Association 197(2), 201-9, 1990.
(5) Forman AJ. The threat of rabies introduction and establishment in Australia. Australian Veterinary Journal 70(3), 81-3, 1993.
(6) Gooch JM, Willers EH, Wong HL. 1971. Cited in Sasaki DM, Gooch JM. Cost effectiveness of Hawaii's antirabies quarantine program. Hawaii Medical Journal 42, 157-60, 1983.
(7) Hawaii Department of Agriculture: Rabies serosurvey of quarantined pets and mongooses. Presented to the 16th State Legislature, March 1992.
(8) MacDiarmid SC. Risk analysis and the importation of animals and animal products. Revue Scientifique et Technique de l'Office International des Epizooties, 12, 1093, 1993.
(9) MacDiarmid SC. Risk analysis, international trade, and animal health. In Fundamentals of Risk Analysis and Risk Management. V. Molak (ed), CRC Lewis Publishers, Boca Raton, 377, 1997.
(10) Vose, D. Quantitative Risk Analysis: A Guide to Monte Carlo Simulation Modelling. John Wiley & Sons, Chichester, 1997.
(11) WHO. World Survey of Rabies 26 (for year 1990). WHO/Rabies/93.208. World Health Organisation, Geneva, 1993.
TableI: Annual incidence of rabies and risk of selecting an infected dog
Country or Reported Most likely Maximum Risk of
city/total cases number of number of selecting
number of dogs cases cases an infected
animal per
Reported Reported million
number number
multiplied by multiplied by
1.5 5.0
USA 91* 137 455 1.6
(54.6 million)
(5 million)
France 38 57 190 3.8
(9.8 million)
(5 million)
(6 million)
Lima 3** 1,260*** 4,200**** 1,700
(0.42 million)
* These are the rabies cases in pet dogs, ie 84% of the reported canine rabies cases.
** The incidence fell to this following a vaccination programme.
*** The incidence a year before the vaccination programme.
**** 1% of the population was used as the maximum.
Table II: Risk that quarantine alone will fail to prevent the introduction into New Zealand of a dog incubating rabies
Minimum Most likely Maximum
1 month quarantine 0.47 0.5 0.55
4 months quarantine 0.22 0.25 0.28
6 months quarantine 0.09 0.11 0.13
TableIII: The risk of importing a dog incubating rabies (number of introductions per million)*
USA France Canada Germany Philippines Lima
1 month quarantine 1.92 4.83 24.1 45.5 101 2,400
4 months quarantine 0.95 2.39 12.1 21.9 50 1,200
6 months quarantine 0.42 1.03 5.14 9.76 21.9 506
Vaccinated 0.39 1.00 4.89 9.25 20.6 455
Vaccinated and 1 month quarantine 0.20 0.51 2.42 4.66 10.2 233
Vaccinated and 4 months quarantine 0.097 0.25 1.20 2.32 5.04 111
Vaccinated and 6 months quarantine 0.042 0.11 0.53 1.02 2.32 48.9
*In 95% of iterations the risk was equal to or less than the value shown.