Health & Management / Effects of Raccoon Biology and Behaviour on Raccoon Rabies / List of hyperlinked Techniques & Protocols:

< > Surveillance for Rabies in Raccoons and for Raccoon Rabies:

Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-screen view with caption Click here for full-page view with caption 

Introduction and General Information

Surveillance is important to provide information on the level of raccoon rabies, the occurrence of epizootics, and any increases in geographical range of this disease. Surveillance has been used to track the spread of raccoon rabies virus variant from its original focus in Florida (1950s onwards) and from the new focus starting on the Virginia/West Virginia border in the late 1970s.
  • Surveillance for a disease may be active or passive. The main form of surveillance which has been used for rabies in animals in the USA is passive surveillance, based on suspected rabid individuals being reported to health authorities and subsequent testing of the animal for rabies. In many localities, testing is limited to animals that have exposed a human or a domestic animal. Additional surveillance has been carried out in some locations using road-killed raccoons, examination of hunted raccoons etc., and serological surveys have been carried out in trapped raccoons.
  • Note: There may be considerable differences in reporting of rabid animals at a county level, making comparisons between counties difficult. (B358.4.w4)

Results of surveillance studies may be analysed statistically or using mathematical models or computer models; modelling capabilities have been increased by the use of computer programs.

Published Guidelines linked in Wildpro

Return to top of page

Raccoon Populations: Distribution and Density

Raccoons are found in southern Canada; throughout the United States, except for parts of the Rocky Mountains; Mexico and in Central America through to central Panama. Their range has expanded in recent decades both farther westward into the Rocky Mountains in the US and northward farther into Canada than their historical range. For further information see: Common Raccoon Procyon lotor - Distribution & Movement (Literature Reports) .

Raccoon population densities vary considerably depending on habitat, for example as low as 0.5 - 3.2 per km² in northern prairies, but 250 per km² in a woodland marsh and as high as 333.3 raccoons per km in parts of an urban national park. Population densities may vary greatly in a given location over time. For further information see: Common Raccoon Procyon lotor - Social Behaviour - Territoriality - Predation - Learning (Literature Reports) - Dispersion, Territoriality and Population Densities

Measuring population size and density
  • It is difficult to determine accurately the absolute raccoon population size and density for an area because: (B406.38.w38)
    • It is difficult to determine the percentage of the population that has been counted; (B406.38.w38)
    • It is difficult to know how large an area the individual counted raccoons are using. (B406.38.w38)

A variety of methods may be used to estimate population sizes and population trends, including direct counts, transects or spotlighting surveys, scent marking indices, fur traders' or sealing reports, field trial surveys, radioisotope tagging, road-kill data and mark-recapture techniques. Different methods have different advantages and disadvantages. 

  • Scent marking indices: attracting raccoons to scent stations (sifted soil mixed with a scent attractive to the animals to be counted). By assessing the number of visits to scent stations (footprints are visible in the sifted soil) it is possible to estimate raccoon abundance in an area. Scent stations can be set up along a roadway or on transects on trails. (B402.10.w9, B406.38.w38, J59.11.w1) Scent marking stations may be sufficient to compare gross population levels between areas or between years, but not to census the population. (B406.38.w38)
  • Transects/Spotlight surveys: researchers travel along transects or drive along roads at night with spotlights, night-vision equipment or infrared nightsights, counting the raccoons which can be seen at the side of the roads. (B402.10.w9, J1.40.w3) 
    • Surveys along roads are less likely to be affected by the observer disturbing the animals than are off-road transects. (J1.40.w3)
  • Fur traders' or sealing reports: the number of raccoon pelts harvested in a particular area is a traditional method of assessing raccoon abundance in an area, but is dependent on the accuracy of trapper's records (e.g. on how many were taken and where the animals were trapped) or the accuracy of fur buyers' records (e.g. on numbers of pelts bought) and additionally, double-counting may occur if a buyer gets furs from another buyer included in the survey, even if requested only to include furs bought directly from hunters, and is also dependent on the price of fur and other socioeconomic conditions. (B402.10.w9, B406.38.w38)
  • Field trial surveys: hunters in field trials observe and tree raccoons, but do not kill them. This data can be collected and converted into an index of raccoon abundance. (B402.10.w9)
  • Mark-recapture techniques: this method which involves live trapping, marking, releasing and then trapping again provides a good estimate of population size and population structure, and allows weighing, vaccination and health checks of the animals. (J40.48.w3, B402.10.w9) These techniques are used commonly but may be biased, for example if not all members of a population have an equal chance of being caught and marked. (J1.6.w5, J40.34.w1, J40.49.w1)
    • Note: failure to identify differences in trapability may produce biased estimates of disease prevalence. (J59.29.w1)
  • Radioisotope tagging: this variant on mark-recapture involves injecting live-trapped raccoons with a radioisotope which will be released in the faeces, then collecting and analysing scats found in the area, estimating the population based on the assumption that the proportion of isotope-containing scats is equal to the proportion of tagged raccoons. (J40.49.w1)
  • Road-kill data: This can be used to indicate relative population size and trends or fluctuations in population size, if it is assumed that the number of animals removed as road-kill per year (or other fixed period) is correlated with the size of the population. (J1.26.w5, J59.6.w1)
    • In Baltimore, Maryland, USA, where a raccoon epizootic arrived in March 1985 and ended in May 1987, analysis of road-kill data for 1984 to 197 showed that the numbers of raccoons found as road-kill declined in 1986 and 1987 compared to 1984-1985, whereas road-kill figures for other species did not decrease; this suggested that a decrease in the population of raccoons occurred over this period, and the rabies epizootic was thought to have contributed to this decline. (J1.26.w5)
  • For further information see: Common Raccoon Procyon lotor - Social Behaviour - Territoriality - Predation - Learning (Literature Reports) - Methods of Marking and Following Raccoons 
Associated techniques linked from Wildpro

Return to top of page

Diagnostic Tests for Rabies in Raccoons

Two types of diagnostic tests may be used: tests for rabies virus or tests for antibodies to rabies virus. Raccoons can produce antibodies to rabies as a result of exposure to rabies, or following vaccination.

Rabies diagnosis/detection of rabies virus or rabies virus antigen

A variety of tests have been developed for diagnosis of rabies and detection of rabies virus. For many years, detection of Negri bodies, and/or mouse inoculation were used. However, immunological methods are now used routinely for rabies diagnosis.

  • Histopathological findings in animals with rabies are not diagnostic. Infection with rabies virus generally results in nonsuppurative encephalitis characterised by perivascular mononuclear cell cuffing and multifocal gliosis, but these lesions are not pathognomonic for rabies. Additionally, these lesions may be absent, as was noted in a rabid raccoon with concurrent distemper virus infection. (J351.80.w1) 
  • The presence of Negri bodies on examination of histological sections or brain impression smears was the standard test for rabies for many years. It has the advantages of simplicity and requiring little equipment: microscope slides, standard histological dyes and a microscope. However, while the presence of Negri bodies is pathognomonic for rabies, other inclusion bodies may be mistaken for Negri bodies, leading to false positive diagnoses, Additionally, Negri bodies are not always visible with standard stains, leading to false negative diagnoses. In the USA, other methods are now used for routine rabies diagnosis. See: Rabies (Viral Disease): Diagnostic Criteria
  • Immunological detection of rabies virus is more sensitive and specific than is detection of Negri bodies.
    • The direct fluorescent antibody test is presently the standard test for rabies virus detection. See: Rabies virus (with special reference to raccoon rabies variant) - Detection and Identification Techniques (Viral Reports)
    • An immunoperoxidase method using the streptavidin-biotin complex (ABC) and the monoclonal antibody (mAb) 802-2 was shown in a study to detect rabies antigen in paraffin embedded tissue sections from rabid raccoons at 1:40,000 dilution (and also, but fainter, at 1:80,000 dilution). (J3.136.w4)
    • The direct rapid immunohistochemical test (dRIT) is a newly developed test which can be used on brain touch impressions, producing a reaction product which is visible under an ordinary light microscope (mangeta inclusions, while the neuronal background is blue). The test allows diagnosis within one hour. (J84.12.w1)
  • The mouse inoculation test may still be used when an animal which is suspected to have exposed a human to rabies is negative with the fluorescent antibody test. (J270.10S4.w1)
  • For further information on rabies diagnosis see: 
Tissues to be tested for diagnosis of rabies in raccoons
  • Testing of raccoons for rabies diagnosis should involve examination of the cerebrum, brain stem and cervical spinal cord: areas of the CNS in which rabies lesions have been found consistently in raccoons with naturally acquired rabies. (J212.4.w1)
    • Note: CNS infection is usually bilateral, but not necessarily symmetrical. (J212.4.w1)
  • Testing of saliva or salivary gland may or may not reveal the presence of rabies virus. Rabies virus or viral antigen may be detected in saliva/salivary gland of raccoons more commonly in raccoons infected with raccoon rabies virus variant than in raccoons infected with other rabies virus variants. See: Rabies - Detailed Pathological Findings - (Necropsy-Post Mortem) (Disease Reports)
Confirmation of raccoon rabies virus variant

Raccoon rabies virus variant is a recognised rabies virus variant. This variant was first recognised as a clinical entity in the 1950s in Florida. (B358.4.w4, B360.16.w16) The use of antigenic typing or genetic typing makes it possible to distinguish between rabies virus variants and to determine whether raccoon rabies virus variant is involved in a given case of rabies. 

  • Several studies using panels of monoclonal antibodies (antigenic typing) have shown that raccoon rabies virus isolates from all over the eastern US states where raccoon rabies has spread are the same as one another and different from other terrestrial rabies virus isolates and from rabies virus isolates from bats. (J1.26.w6, J93.24.w1, J100.149.w1, J237.36.w1)
  • Genetic typing has also confirmed the close relationship between raccoon rabies virus isolates from different places in the eastern US; isolates of rabies virus from raccoons with raccoon rabies (rabid raccoons in the eastern US from Florida to Maine) share about 99% sequence homology. The next most closely related variant is that found in skunks in the south central US states. (J308.6.w1)
  • Recent work on more than 100 raccoon rabies virus isolates from all along eastern North America, analyzed phylogenetically using the neighbour-joining method, indicates an ancestral lineage in the southeastern states, another lineage found only from a few samples in Florida, and a third lineage broadly distributed through the remainder of the raccoon rabies area, tracing back to Virginia and appearing to be evolving separately from the ancestral viral clade. (P102.16.w1)
  • Further information on rabies virus strains is provided in Rabies virus - Viral Type Diversity (Viral Reports)
Serological tests for rabies antibody

Many early serological surveys for rabies in raccoons looked for serum neutralisation of rabies virus. Sera which neutralised the virus were generally referred to as containing serum neutralising (SN) antibodies, but it has been acknowledged that a simple neutralisation test did not prove the identity of the neutralising agent (i.e. did not prove that neutralisation was due to the presence of rabies virus specific antibodies). Nevertheless, several surveys have shown differences in the percentage of raccoons which were SN positive, and in SN titres of individuals, between areas which were raccoon rabies free and areas in which an epizootic was taking place or had recently taken place. These data do suggest that at least part of what is being measured is a specific response to rabies virus. However, the interpretation of low SN titre sera is difficult since a non-specific inhibition may be measured. (B358.4.w4, B360.16.w16)

  • The presence of serum neutralising (SN) antibody in wild animals indicates that exposure to rabies virus has occurred. (J13.23.w1, B358.4.w4, J63.5.w1) 
    • Raccoons (and skunks, foxes and opossums) in counties in Alabama in which fox rabies was absent were all negative for serum neutralising (SN) antibodies, whereas four of 102 raccoons (3.9%) (and four of 118 foxes) from counties where fox rabies was epizootic were SN positive (in counties in which fox rabies was enzootic, two of 56 foxes but none of 19 raccoons, 35 skunks, nine bobcats or 12 opossums were positive). (J13.23.w1)
    • In areas with recent outbreaks of fox rabies, 11/196 (5.6%) raccoons, as well as 12/262 (4.6%) foxes, 2/285 (1.8%) opossum, 5/27 (18.5%) bobcat and 7/48 (14.5%) skunk sera were serum neutralizing positive (undiluted serum neutralising at least 32 LD50), while 300 animals from rabies-free areas were all serum neutralizing negative. (J63.5.w1)
  • Unfortunately, non-specific cytotoxicity interfering with viral growth may affect results in tests such as the rapid fluorescent focus inhibition test (RFFIT) or the fluorescence inhibition microtest (FIMT). (J1.24.w6)
  • Neutralization of rabies virus by serum is not always due to the presence of rabies virus-specific antibodies; nonspecific virus neutralising substances may also be present in serum of several species. (B357.17.w17)
    • Nonspecific inactivation of virus may occur in virus neutralisation tests with low dilutions of raccoon sera. (J1.24.w6, J126.57.w1, J270.10S4.w1)
    • Complete neutralization at titres greater than or equal to 1:25 should be specific. (J126.57.w1, J270.10S4.w1)
  • Use of an indirect ELISA avoids problems with non-specific cytotoxixity. (J1.24.w6)
  • Further information on serological tests is provided in: Rabies virus (with special reference to raccoon rabies variant) - Detection and Identification Techniques (Viral Reports)
Comparison of serological tests for rabies antibody in raccoons
  • Tests which have been used for the detection of anti-rabies antibodies in raccoons include the mouse protection test (mouse serum neutralization test), indirect fluorescent antibody test (IFA), rapid fluorescent focus inhibition test (RFFIT), fluorescence inhibition microtest (FIMT) and ELISA
    • A study comparing antibody titres in the sera of 38 raccoons from Florida by mouse inoculation and RFFIT found that 17 sera were positive by mouse inoculation at titres of >1:2, while these plus a further six sera were positive by RFFIT at a titre of >1:5 and the remaining 14 sera were negative in both tests. "The average point prevalence of neutralizing antibody was 20.9% (range, 15.5% [1970] to 25.0% [1974]." It was suggested that the RFFIT may detect more individuals with rabies antibodies than detected by the mouse inoculation test. (J100.148.w1)
    • A study compared rabies antibodies in sera from Iowa raccoons, as detected using three serological tests: rapid fluorescent focus inhibition test (RFFIT), mouse serum neutralisation (MSN) and indirect fluorescent antibody (IFA) test. With the RFFIT test, 51 of 985 raccoons (5%) showed serum neutralising titres of >3.0 (titres 3.0 - 24.2). Testing of 24 of these sera in the MSN test gave 23 as positive (titres 3.2 - 17.9); one raccoon with a RFFIT test titre of 4.8 was negative with the MSN. With the IFA, six raccoons were positive; these animals had RFFIT titres of 4.3 - 17.6 and the three which had been tested by MSN had MSN titres of 5.2-6.0. It was noted that the IFA detects antibodies to the internal nucleocapsid proteins whereas the SN antibody tests detect antibody to surface glycoprotein. The fact that both types of antibodies were detected was considered to confirm that the wild raccoons had been exposed to rabies antigen. (J1.28.w9)
    • Further information on these tests is provided in: Rabies virus (with special reference to raccoon rabies variant) - Detection and Identification Techniques (Viral Reports)

Further information on the diagnosis of rabies and on serological tests is provided in:


Trapping raccoons
  • Raccoons can easily be live-trapped in cage- or box-type traps, set in arrays, or near den sites, feeding areas or scats, or on travel routes such as obvious runways. Traps need to be strong and well constructed to hold raccoons and the trap should be at least 10 by 12 by 32 inches (25.4 by 30.5 by 81.3 cm) in size. Foot-hold traps can be used also, but there is a risk of raccoons chewing the trapped foot (sometimes resulting in loss of the foot). A wide variety of foods have been used to bait traps for raccoons, including fish, dog food, table scraps, sorghum, jellies, pastries, oatmeal, peanut butter, marshmallows or combinations of these. (B405.w1, B486.10.w10, J1.25.w5, J1.26.w9, J1.28.w11, J1.29.w16, J40.34.w1, J331.114.w1, J331.148.w1 J332.58.w1, J335.11.w1, P69.14.w1)
  • For further information see: Common Raccoon Procyon lotor - Social Behaviour - Territoriality - Predation - Learning (Literature Reports) - Methods of Marking and Following Raccoons 
Anaesthetising raccoons
Raccoons can be anaesthetised using a variety of anaesthetic combinations, including:
  • Ketamine and xylazine by intramuscular injection:
    • Ketamine 30 mg/kg with 3 mg/kg xylazine. (P69.14.w1)
    • Ketamine 10 mg/kg bodyweight plus 0.4 mg/kg xylazine intramuscularly based on estimated bodyweight. (J4.213.w4)
    • Ketamine 10 mg/kg bodyweight plus 1 mg/kg xylazine intramuscularly. (J1.26.w8)
    • Ketamine 20 - 30 mg/kg bodyweight and 2 - 3 mg/kg xylazine (10:1 ketamine: xylazine). (J1.26.w9)
    • Ketamine 10 mg/kg plus xylazine 1 mg/kg, intramuscularly. (J1.28.w11)
    • Ketamine 5 - 10 mg/kg plus 1 - 2 mg/kg xylazine. [Data for captive animals](B407.w18)
    • Ketamine and xylazine in a 5:1 ratio (doses 22.0 to 38.2 mg/kg ketamine and 4.4 to 7.6 mg/kg xylazine). (J180.40.w1)
  • Ketamine and acepromazine by intramuscular injection. (J59.29.w2, J332.79.w1, J334.55.w1)
    • Ketamine 8- 10 mg/kg preceded by acepromazine 2.2 mg/kg (five to 25 minutes before the ketamine). (J40.38.w1)
    • 8 - 10 mg/kg of a 3:7 ratio mixture of ketamine and acepromazine. (J59.29.w1)
  • Ketamine 3 - 5 mg/kg plus medetomidine 50 - 100 g. [Data for captive animals](B407.w18)
  • Ketamine 2.0 mg/kg, xylazine 2.0 mg/kg and butorphanol tartrate 0.2 mg/kg by intramuscular injection. (J1.28.w8)
  • Tiletamine hydrochloride and zolazepam hydrochloride (Telazol) at 5 mg/kg. (J59.29.w1)
  • Tiletamine hydrochlorde and zolazepam hydrochloride (Telazol) and xylazine, in a 3:2 mixture of Telazol and xylazine, to give dose rates of 3.2 +/- 0.6 mg/kg (mean +/- SD) Telazol and 2.1 +/- 0.4 mg/kg xylazine. Suggested doses to be used were 3.0 mg/kg Telazol and 2.0 mg/kg xylazine, to give up to 60 minutes for handling, blood sampling, radiotagging and tooth extraction, with full recovery in about 120 minutes. (J1.40.w4)
  • Ketamine alone: 
    • Ketamine 8 - 10 mg/kg intramuscularly. (J40.38.w1)
    • Intramuscular injection of 0.5 - 1.0 mL of 100 mg/mL ketamine was used for sedation sufficient for handling and inoculating raccoons. (J1.28.w12)
    • Intramuscular injection of 20 mg/kg ketamine. (J135.83.w1)
    • Intramuscular injection of ketamine at 10 mg/kg estimated body weight. (J40.63.w2, J40.67.w1, J331.148.w1)
  • Gaseous anaesthesia; (J331.114.w1, J332.58.w1) this may be administered with the raccoon in an anaesthetic box. (J332.58.w1)
  • Note: different drugs used may affect the rate of re-trapping. (J59.29.w1) See: Common Raccoon Procyon lotor - Social Behaviour - Territoriality - Predation - Learning (Literature Reports) - Methods of Marking and Following Raccoons
Taking blood samples from raccoons
  • Blood may be taken from the jugular vein or by cardiac puncture in anaesthetised raccoons. (J1.25.w5, J1.26.w9, J1.26.w8, J1.34.w8, P69.14.w1)
    • Use a 20 gauge, 3.8 cm needle, into a 10 mL Vacutainer evacuated blood collection tube (Becton Dickinson, Mississauga, Ontario, Canada L5J 2M8) (J1.26.w9) or a 5 mL syringe on a 20 gauge needle. (J1.25.w5)
    • Separate the sample by centrifugation, (J1.26.w9) or remove serum once the blood has clotted. (J1.25.w5)
    • Store the serum at minus 20 C in 2 mL serum vials. (J1.25.w5, J1.26.w9).
Associated techniques linked from Wildpro

Return to top of page

Surveillance for Rabid Raccoons

"Detecting the true distribution of rabies in an area depends on the sensitivity of the surveillance system." (N7.38.w1) 

Most surveillance data is derived from the number of animals which are found positive for rabies by state health departments. The number and types of animals submitted for testing may affect these figures. Reasons why animals may be submitted for testing include routine surveillance in a particular area, special interest of a health department or a member of the general public, or due to an animal having bitten a human or another animal. (N7.37.w1) Surveillance efforts and criteria used to determine which animals should be tested may vary between areas, and over time in a given area. (B358.4.w4, J1.26.w5)

  • Note: a recognised limitation of studies of naturally occurring raccoon rabies is a lack of knowledge of the proportion of rabid raccoons detected, even when intense surveillance is used. (J67.56.w1)
Passive surveillance

The main form of surveillance used for detecting rabies in the USA is passive surveillance. (J1.28.w10, N7.37.w1) Individual animals are tested for rabies if they are noted by a human, submitted for testing, and meet the criteria for rabies testing in that locality.

  • In general, surveillance programmes based on testing suspect animals to which humans have been exposed are probably good indicators of the existence of rabies epizootics, but poor indicators of the incidence of enzootic rabies. (B352.5.w5)
  • Passive surveillance measures when the animal is removed from the population, however it does not provide information about when it was infected or how many infected animals are not reported. Surveillance reports can provide information on geographic, environmental and temporal trends in rabies occurrence. They are better at providing information on epidemics (> 15% prevalence) than on endemic incidence of disease (< 1% prevalence). (J1.28.w10)
  • Cases of rabies, even epidemics, may occur but may not be detected by surveillance. (J67.71.w1)
  • Many individual animals will not expose or potentially expose a human or domestic animal to infection, and therefore will remain untested and undetected. (B358.4.w4, N7.38.w1)
  • Seasonal variations in reported cases of rabies may not only be due to variation in the occurrence of rabies. There may be seasonal variations, greater in some geographical areas than in others, in the amount of contact between humans and animals. (N7.37.w1)
Criteria for submission

Factors affecting submission of animals for rabies diagnosis include regional differences in laboratory resources and testing policy, public awareness of rabies, the stage of a rabies epizootic in an area, and the human population density (more raccoons being submitted from areas where they live close to humans than from areas with sparse human populations). (J1.28.w10)

  • "Positivity rates are very dependent on the submission policies of the locality involved." (P66.1.w1)
  • Criteria used by different state health departments to accept animals for testing may vary considerably. (N7.37.w1, N7.38.w1)
    • Often only animals which expose a human or domestic animal will be tested. "Many rabid animals, especially wildlife, never expose a human being or domestic animal and, therefore, remain undetected." (N7.38.w1)
    • Variations in availability of local animal-control services and transportation to state laboratories may cause variations between counties within a state in the numbers of animals examined. (N7.38.w1)
    • Some animals are more likely than others to be captured for testing (e.g. dogs versus bats), therefore "the proportion of reported cases by species may not reflect the true distribution of rabies in animals." (N7.38.w1)
  • One study in Loudoun County, Virginia, in which large numbers of raccoon rabies cases were reported in 1981-1982, tested road kill raccoons and those found dead, as well as those which had had contact with humans or pets (normally, road kills and animals found dead would not have been tested. (J1.21.w6) See below: Enhanced surveillance
  • A study in 1982 - 1983 based on submissions to various counties of Pennsylvania, Maryland and Virginia, noted that limitations of the study included variable submission policies between localities, as well as possible inaccuracies in reporting of the age, health and behaviour of submitted animals. (J101.126.w1)
Effects of human population density and activity on surveillance
Many rabid animals are never observed by humans and therefore are not tested, not confirmed as rabid, and are unreported. For this reason the cases reported to CDC by state, territorial and the District of Columbia health departments can only give a crude indication of the occurrence of rabies and do not indicate the extent of rabies virus infection in either wild or domestic animals in any geographical area. (J4.207.w1)
  • The number of rabid raccoons submitted for testing is known to be affected by human population size and human population density. The magnitude of reported initial raccoon rabies epizootics is strongly influenced by human population size. (J67.68.w1)
    • In the Mid-Atlantic states, in urban areas, the large human population living in close proximity to raccoons provides an increased opportunity for observation, capture and testing of rabid raccoons. (P66.1.w1)
    • In the early 1980s, when raccoon rabies was extending through the Mid-Atlantic states, it was noted that urban and suburban areas were most intensely affected and that this was related not only to suitable habitat for raccoons but also "the high concentrations of people and pets in these areas leads to more wildlife encounters that result in submission of wild animals for testing." (D239)
    • During a study in Connecticut, 1991 - 1994, while raccoon rabies was advancing through the state, it was noted that transmission of raccoon rabies may have been underdetected in regions of the state with sparse human population, since the number of animals tested was highly correlated with human population density. (J91.57.w1)
  • The number of rabid raccoons submitted for testing will be affected by both human outdoor activity levels and raccoon activity levels:
  • As the number of individual animals submitted for testing increases, the likelihood that a rabid individual will be detected increases, but the number submitted has less effect on the percentage positive for rabies. (J1.28.w10)

It has been recognised for many years for both raccoons and other species that the number of individual animals reported infected with rabies does not necessarily reflect the prevalence of rabies in the population:

  • In Palm Beach County on the Atlantic coast of southern Florida in 1955-1956 eight of 22 raccoons either trapped or found dead on highways were rabies-positive (36.4%) but it was considered that this would not have been detected by the normal county reporting system since there were no known human exposures associated with the epizootic. (J1.6.w3, B358.4.w4)
  • In Mephitis mephitis - Striped skunks in Carroll County, Illinois, 30 of 329 skunks tested for rabies were positive (9.1%) over the period 1958 to 1964; only two rabid skunks were reported in the same period in this county. (J40.30.w1)
  • It was noted in 1970 that, compared with rabid foxes or skunks, the behaviour of rabid raccoons is inconspicuous, therefore it is probable that a lower proportion of rabid raccoons would be reported. (J1.6.w3)
Time to detection of raccoon rabies by passive surveillance
Detection of rabies by passive surveillance is likely to involve a time lag. (J1.28.w10) This may be a considerable time (months, possibly even years).
  • There may be a lag time of one to several months between the start of a rabies epizootic and when it is recognised. This is due to the requirement for the interaction of a rabid animal and a human, collection, submission for testing and testing of the suspect animal, and reporting back of results, with several such results required before the epizootic is recognised. (J91.27.w1)
  • There are no tools available at present to calibrate detection times which are obtained through passive surveillance. (J363.3.w1)
    • In order to calibrate this, a second, highly sensitive surveillance system would be required, such as combined active sampling of road-killed raccoons, purposeful hunter collections plus data from kill-trapping. (J363.3.w1)
Examples of differences between rabies presence and rabies detection by passive surveillance
  • The genetic stability of the raccoon rabies virus variant, and its divergence from variants in other reservoir species, suggests long-standing isolated transmission of this variant in raccoons. It is therefore probable that enzootic raccoon rabies existed in the Florida raccoons long before it was first reported but that the disease was only noticed in the 1950s with changes in human demographics leading to greater contact between humans and raccoons. (J308.5.w1)
  • In Florida, an early study commented that following one report of a rabid raccoon, it was possible for researchers to detect other rabid raccoons in communities in the same general area, but without any cases having been reported in those communities. It was noted that since raccoons infected with rabies were not necessarily aggressive, they may be less likely to be noticed and reported. [1960](J330.75.w1)
  • In Florida in 1968, a study found equal levels of enzootic raccoon rabies in two counties (Manatee and Sarasota) but the reported number of cases in raccoons in the two counties were dissimilar. (J1.6.w3)
    • In raccoons trapped in Florida in September 1968, 5/18 (27.8%) from Sarasota County had serum neutralising (SN) antibody for rabies while in Manatee County 5/25 (20%) had SN antibodies; one raccoon from Manatee county and none from Sarasota County were rabies virus positive. Populations of raccoons (and opossums), as shown by trapping data, were similar in the two counties, but 31 cases of rabies had been reported in wildlife in Manatee county in 1968 while only four cases had been reported in Sarasota county. (J1.5.w4)
      • The totals for 1968 for raccoon rabies were 26 cases in raccoons in Manatee County and 8 cases in raccoons in Sarasota County. (D222.AppII.w11)
Use of positivity rates for analysis of passive surveillance data

Different methods of analysing reports to indicate whether rabies may be increasing or decreasing in the population of an area include (a) the number of individuals diagnosed with rabies in a given area in a given time period; (b) the percentage of the total submitted which are positive for rabies infection.

  • Since the number of animals tested for rabies may show large variations [due to factors other than disease occurrence], the ratio of positive cases to number of animals tested (positivity rate) may be preferable, reducing the effect of variations in sample size. (J30.76.w1)
    • To detect trends, a three-month running average may be used to reduce the impact of any single month (e.g. with a small sample size in one month). (J30.76.w1)
  • Data from about 700 raccoons from six counties and cities in four states affected early in the mid-Atlantic raccoon rabies outbreak found that overall, 81% were positive but in different localities 15% to 80% were positive. It was noted that positivity rates vary depending on the locality's policy on submission of animals for rabies testing. (P66.1.w1)

It must be remembered that the number of cases of rabies reported is affected not only by the level of the disease in the animal population but also by the level of reporting by the human population. 

  • A study in Georgia found that the highest number of submissions of raccoons for rabies testing occurred March and April, but the highest proportion of rabies-positive animals was found in the summer months. (J101.98.w1)
Enhanced surveillance

In addition to passive surveillance of raccoons which have potentially exposed a human or other animal, additional surveillance for rabies may include testing of:

  • Individuals showing signs of illness;
  • Individual animals found dead of unknown causes;
  • Road killed animals;
  • Animals killed during the hunting/trapping season;
  • Individuals live-trapped as nuisance animals;
  • Individuals trap-killed or live-trapped and euthanased during research.
  • Individuals trapped during trap-vaccinate-release programmes.
  • N.B. with live-trapped animals:

(J1.21.w6, J1.34.w8, J1.42.w, J4.189.w8, J4.213.w4, J59.6.w1)

Associated techniques linked from Wildpro

Return to top of page

Serological Surveys

"Serum surveys of wild animals in nature could possibly be used, in conjunction with population density surveys, in predicting epizootics of rabies with greater accuracy than with either method alone." (J63.5.w1) 

  • Serological data such as an increase in the proportion of seropositive raccoons following a rabies epizootic indicate that not all raccoons exposed to natural infection die. Understanding of the relationship between serological findings and natural infection with rabies would allow better use of serological data for the epidemiological study of rabies in wildlife. (J100.123.w1)
  • From field data it is known that the point prevalence of serum neutralizing antibodies in a population of wild animals is generally much greater than the point prevalence of infection with the disease. In some circumstances only serology may give positive data on rabies. (B358.4.w4)
  • Note: Field investigations, in contrast to surveillance programmes based on examining individual animals which may have exposed humans to rabies, are likely to be biased towards individuals which are behaviourally normal and respond normally to stimuli (olfactory or food), allowing them to be trapped. (B395.2.w2)

Results of serological surveys have been very variable. Comparisons of result of different surveys are made more difficult because of the different serological tests used in different surveys.

Negative data
  • A study in an area of the southeast USA (Florida (six counties), Georgia (15 counties) and South Carolina (one county)) with fox rabies (but not raccoon rabies) in 1954-1957 failed to detect rabies virus in any of 299 raccoons examined for the virus. (B395.2.w2, J4.135.w1)
Data from areas with rabies variants other than raccoon rabies
  • A study in areas with a recent history of outbreaks of fox rabies detected rabies antibodies (neutralisation of at least 32 LD50 of rabies virus by undiluted serum) in 11 of 196 (5.6%) raccoons tested. (B395.2.w2, J63.5.w1) It was noted that for 670 animals of all species together, before the epizootic peak only 1.6% tested positive while after an epizootic peak 5.2% were positive and in enzootic areas, 5.1% were positive). (J63.5.w1)
  • A study in Iowa, 1971 - 1972, detected rabies antibody in 7.6% of 301 raccoons (as well as 68.3% of 82 Mephitis mephitis - Striped skunks, 26.3% of 19 spotted skunks and 1.2% of 84 Didelphis virginiana - Virginian opossum) but it was commented that the prevalence results may have been raised by non-specific reactions. (B395.2.w2, J1.28.w9)
Data from raccoon rabies areas
  • Data from Florida following early epizootics provided the following information:
    • During an epizootic, mean serum neutralising antibody titres were higher than after the epizootic, although prevalence of SN antibodies was slightly lower during than after the epizootic. (B358.4.w4)
    • Antibody titres and antibody prevalence were significantly higher (P<0.05) in virus-positive than in virus-negative raccoons, but only 50% of virus-positive raccoons were seropositive (B358.4.w4)
    • Prevalences and serum neutralising titres were significantly higher (P<0.05) in raccoons from enzootic areas than in raccoons from areas with only sporadic cases of rabies in raccoons and significantly lower (P<0.05) than in raccoons from epizootic areas. (B358.4.w4)
  • A study in Florida failed to detect any rabid raccoons among 103 "normal appearing" individuals tested (serum neutralising antibody was found in 4.4% of tested raccoons). In Clay County, Florida, April 1969 to February 1970, rabies virus was not detected in any animals but SN antibody was detected in three of 134 raccoons, including one individual about a year old, indicating the presence of the virus during the preceding year. (J101.98.w1)
  • In areas outside the known raccoon rabies foci, where only sporadic cases of rabies are recorded in raccoons, the prevalence of serum neutralising antibodies in trapped raccoons averaged 2.6%. For raccoons collected from within the raccoon rabies focus, but not in areas with known epizootics, (i.e. in enzootic areas) antibody prevalence was 3% to 12%, average 7.2%, significantly higher (P<0.05) than that in the sporadic areas. Prevalences in raccoon populations that have experienced epizootics of rabies were significantly higher (P<0.05) than those in enzootic areas: peak levels of 17-22% for four separate epizootics along the Gulf Coast of Florida in 1969 to 1971, and 12% of 75 raccoon sera collected from raccoons in South Carolina in 1970-71 were positive - after the epizootic front had passed, undetected, into the state from Georgia. (B358.4.w4)
    • During the epizootic on Marco Island, Collier County, Florida, 3 of 11 raccoons trapped in October 1970 and 8 of 45 trapped in November had serum neutralising antibodies (27% and 18% respectively). (B358.4.w4, J101.98.w1)
      • A study in Florida, on Marco Island, November 1970 to August 1974, following the epizootic in the fall of 1970, found that the average point prevalence of neutralizing antibodies in the raccoon population was 20.9%, varying from 15.5% in 1970 to 25.0% in 1974. Serial testing on three young raccoons (less than 15 weeks old, 30 weeks old and 52 weeks old) suggested circulating endemic rabies in or near the study area between June 1972 and February 1974. (J100.148.w1)
    • Thirty six raccoons collected during a rabies epizootic were held in captivity; out of seven animals which were seropositive at the time of capture, three remained seropositive for at least the next two years and three converted to seronegative by 21 months after capture (the seventh died from an unrelated cause). Of the 29 animals which had been seronegative when captured, one died of rabies 39 days after capture and another at 79 days, both having seroconverted prior to death; a further two animals had detectable antibodies periodically while in captivity. It was noted that if tested a few months after the epizootic, nine of the 34 live raccoons (i.e. not including the two which died of rabies) would have been seropositive (26.5%), which was similar to the prevalences of seropositive raccoons detected following the 1968 epizootic in nearby raccoon populations. (B358.4.w4)
      • In raccoons trapped in Florida in September 1968, 5/18 (27.8%) from Sarasota County had serum neutralising (SN) antibody for rabies while in Manatee County 5/25 (20%) had SN antibodies. (J1.5.w4)
  • For raccoons on Long Boat Key, Florida, in 1969, prevalence of serum-neutralising antibody in raccoons was 9% before, 15% during and 22% after (35% at eight months after and 18% at 14 months after) an epizootic of rabies in the population. (J100.123.w1)
    • On Long Boat Key, one mile off southwest Manatee and Sarasota counties, an epizootic spread through the island in 1969. Of the 103 raccoons trapped and sampled, antibody prevalence in March 1969 (during the epizootic) was 11%, in April 20%, in August 22%, in April 1970 35% and zero two years later. No virus was detected in any serum samples from raccoons collected after the epizootic and only three cases of rabid raccoons were reported from the island after the epizootic. The disappearance of antibodies was considered to reflect limited duration of antibody in individual raccoons plus mortality among seropositive individuals. (B358.4.w4)
  • A study of sera of raccoons captured in Washington, D.C., during and immediately after the epizootic, in the Roanoke-Salem area, Virginia, near the border of the epizootic and in an area of southeastern Virginia, at least 100 miles from the epizootic front, found none samples from the later two areas (36 and 15 samples respectively) to be positive in the rapid fluorescent focus inhibition test (RFFIT). Of the 291 samples from Washington, D.C., only eight (3% of the 291 samples) were positive at titres of > 1:25 (a further 20 had titres of 1:5 but this was not considered a significant titre due to concerns regarding possible cross-reactions or non-specific viral inhibition by sera at low dilutions). (J270.10S4.w1)
  • At the National Zoological Park's Conservation and Research Center, Front Royal, Virginia, 401 serum samples from 168 raccoons were collected August 1981 to August 1983, following a rabies epizootic which had peaked April 1980 to June 1981 (with no cases of raccoon rabies confirmed September 1982 to October 1983). Testing with the RFFIT found 168 raccoons to be seronegative, one to have a titre of 1:45 (but the animal was not found subsequently) and eight to have relatively low titres (1:9 to 1:13) and with periods with titres of zero; these were not considered to be meaningful. The single bled animal which was found to be rabid did not have detectable antibodies. (P103.1983.w1)
  • At the National Zoological Park, Washington, D.C. from June 1981 to October 1983, 40 animals were tested with the RFFIT. Only four individuals had positive titres and only two of these were considered to have meaningful titres (1:42 and 1:260); both died of rabies. (P103.1983.w1)
  • Development of virus neutralizing antibodies (VNA) was described in raccoons captured from Parramore Island, off Virginia, USA, following distribution of oral baits containing vaccinia-rabies glycoprotein recombinant vaccine. Seven of 18 raccoons trapped had rabies VNA > 0.5 IU/mlL (range 0.6 to 54.9 IU/mL. Out of the 17 animals with biomarkers indicating contact with vaccine bait, 14 (i.e. 82%) survived a severe intramuscular challenge with street rabies (from a naturally infected raccoon from Pennsylvania); only 9% of control raccoons survived the same challenge. VNA titre was not a good predictor of whether a raccoon would survive challenge. However it was suggested that an anamnestic response may act as a better predictor of immune status: all the vaccinated raccoons which survived virus challenge developed a greater than four-fold rise in VNA titre within seven days of the challenge, while no such antibody rise was seen in the animals which died. (J62.60.w1)
Comparison between serological data and detection of rabid individuals
  • Serological surveillance may detect rabies in an area before rabid individuals are reported:
    • In 1970, sera collected from raccoons along the South Carolina-Georgia border and around coastal Charleston in South Carolina, 130 miles from the reported front of raccoon rabies, were positive for serum neutralizing antibodies. Sera collected in these areas in 1969 had been negative for rabies antibodies. Cases of raccoon rabies were not reported from the border counties until 1972. (B358.4.w4)
  • Statewide surveillance in Florida, 1953 - 1972, found that none of 103 raccoons acting normally were rabies positive, but SN antibodies to rabies were detected in 4.4% of these raccoons. (J101.98.w1)
Associated techniques linked from Wildpro

Return to top of page

Possible effects of Rabies on Different Surveillance Methods

Results of surveillance for disease may be biased if the disease affects the likelihood of an individual being sampled, for example if behaviour affects the likelihood of an individual being sampled and the disease affects, or is otherwise linked to, behaviour. (B420.XII.w12, J1.36.w3, J1.36.w6) This may be advantageous for disease detection if the disease increases the likelihood of a diseased individual being sampled and the aim of the surveillance is to detect the presence of the disease. It is a disadvantage for disease detection if the disease decreases the likelihood of a diseased animal being sampled. 
  • Failure to identify differences in trapability of individuals within a population may produce biased estimates of disease prevalence. (J59.29.w1)
  • If surveillance aims to monitor disease prevalence or changes in prevalence, then it is necessary to calculate the bias associated with the disease and correct for it. (J1.36.w6)
Effect of rabies on trapping
  • Trapping, in contrast to surveillance programmes based on examining individual animals which may have exposed humans to rabies, is likely to be biased towards individuals which are behaviourally normal and respond normally to stimuli (olfactory or food), which allows them to be trapped. (B395.2.w2, J4.189.w8)
  • Olfaction may be reduced in rabid raccoons, as indicated for rabid skunks, and this could reduce the likelihood of rabid individuals being caught in baited traps. (B420.XII.w12)
  • The likelihood of traps catching rabid individuals may be less affected for traps set in natural pathways, which may be followed by rabid as well as non-rabid animals, due to their providing easier travel. (B420.XII.w12)
  • Animals which are lethargic due to rabies may be less likely to be caught in traps, due to reduced movement. (B420.XII.w12)
  • A study in New Jersey found that the prevalence of rabies in 172 raccoons snared in the Cape May Peninsula vaccination area during the trapping season was less than 1%, compared with 15% in 20 raccoons collected by other methods. A suggested explanation was that rabies-associated behaviours did not enhance the success of snaring, but rabies associated behaviours such as aggression might enhance the probability of animals being seen and killed by a home owner or pet, or struck by a vehicle. (J1.34.w8)
Effect of rabies on road-kill
  • Road-kill data might overestimate rabies prevalence if behaviours associated with rabies increased the likelihood of a rabid individual being killed on the road. (J4.189.w8)
  • Road-kill data might underestimate prevalence if behaviours associated with rabies decreased the likelihood of a rabid individual being killed on the road, e.g. by decreasing general activity level. (J4.189.w8)
  • Prevalence in 51 raccoons collected by the public and 41 collected along roads during one year in the New Jersey Cape May Peninsula vaccination area was 10% for each method. It was therefore considered that the results from the two methods were equivalent. (J1.34.w8)
  • In Ontario, Canada, during 1999 - 2003, it was noted that 9% of rabid raccoons were road kill, while 21% were found in or associated with barns. (J1.42.w3)
Associated techniques linked from Wildpro

Return to top of page

Temporal and Spatial Analysis of Epizootics

The patterns of occurrence of raccoon rabies have been the subject of a number of studies and analyses. Earlier studies described the outbreaks in terms of the timing (season and duration), age and sex of the rabid raccoons, reasons for submission (e.g. contact with humans) and environmental and population factors which may have been relevant. Starting in the late 1980s, raccoon rabies epidemics have been subjected to mathematical modelling and later to computer modelling. Computer models allow complex analyses and integration with GIS data. However, it is important to remember that the reliability of the output of any model is dependent on the quality of the data put into it, including the right parameters being used to develop the model: i.e. the model must correctly model the real world. 
Descriptive and statistical analyses
  • On the Atlantic coast of southern Florida, in a small area of Palm Beach County in 1955-1966, eight of 22 raccoons (i.e. 36.4%) either found dead on highways or trapped were rabid. Neither the size nor the geographical limits of the epizootic were determined. (B358.4.w4)
  • On the Gulf coast of westcentral Florida, in Manatee county, an area in which only a few cases of raccoon rabies had been recorded for the previous 15 or 20 years, 40 cases were reported in late1967 to early 1968. At the termination of the epizootic in September 1968, one of 35 trapped raccoons was rabid and six of 35 (17.1%) were antibody-positive. In the adjacent coastal county, Sarasota, five of 18 raccoons (27.8%) were seropositive but none were virus positive and only three rabid raccoons had been reported in the previous 16 months; differences in reporting procedures were considered to be responsible for the discrepancy in the number of raccoon rabies cases reported. (B358.4.w4)
  • Outbreaks on Long Boat Key, Marco Island and Tyndall Air Force Base, all in the coastal habitat of the Gulf Coast of Florida, were seasonal (winter and early spring); involved a greater proportion of adult females being seropositive; showed high levels of immunity during and immediately after the epizootics; and were associated with disturbance and habitat destruction, close association with humans, feeding from garbage cans and concentrations around dumps and landfills. (B358.4.w4)
    • On Long Boat Key, one mile off southwest Manatee and Sarasota counties, an epizootic started at the northern end of the island where an unlimited food supply was provided by a local restaurant and residents, with the population being maintained at an "overcrowded" level. During the first seven months of 1969 the epizootic spread throughout the island with 95 reports of "symptomatic" raccoons; 51 of 64 raccoons (80%) submitted for examination were rabies positive. It was considered that rapid collection of reported symptomatic raccoons greatly reduced the risk of exposure of local residents and reduced both the intensity and duration of the epizootic. Raccoons were trapped and examined for both rabies virus and serum neutralising antibodies. Of the 103 animals, during March and April about 10% of the collected raccoons were rabies positive but none of those collected in August, after the epizootic, were positive. Antibody prevalence in March was 11%, in April 20%, in August 22%, in April 1970 35% and zero two years later. No virus was detected in any serum samples from raccoons collected after the epizootic and only three cases of rabid raccoons were reported from the island after the epizootic. On an adjacent island, Anna Maria, raccoons trapped and sampled were all virus negative, but 7.5% were seropositive, despite no cases being reported during this period (one rabid raccoon had been found in June 1968). Raccoon density was similar on the two islands before the epidemic started. On Long Boat Key, natural habitats were rapidly destroyed by construction projects and large artificial feeding stations were available, providing an "overcrowded" population with the potential for increased competition and contact, which may have made the raccoons more susceptible to pathogens and increased transmission potential. (B358.4.w4)
    • On Marco Island, adjacent to the Everglades swamps, during the last quarter of 1970 a short epizootic occurred with an antibody prevalence of 27% in October 1970 reducing to 12% by January 1971. (B358.4.w4)
  • In Florida, up to 80% of the reported cases of rabies per year are in raccoons. Raccoon rabies spread in Florida in the 1950s and 1960s along natural waterways and canals. Periodic increases have been observed for raccoon rabies cases in all regions of Florida, including dramatic increases in the central peninsular region in 1968 to 1969 and in 1984 to 1988, however the pattern of cases indicate enzootic rather than epizootic frequency: during 1986 to 1988, 326 rabies-positive raccoons were reported from 52 of the 67 counties (78% of counties), 30% of cases came from counties reporting less than five cases per year, and annual reports of 10 or more cases per county occurred only five times in three years. The geographical distribution of cases suggests low level endemicity throughout Florida. The results of the study suggested that raccoons are the sole maintenance source for enzootic terrestrial rabies in Florida. (J1.26.w6)
  • In Virginia, about 42% of reported rabies cases for 1981 to 1982 came from Loudoun county in a rural area. Only seven raccoons were submitted January to June 1981, with no rabies-positive animals by the fluorescent antibody test, but 70 were submitted in July to December with 46 (66%) rabid and in 1982, 352 raccoons were tested with 268 (76%) positive. Submissions of raccoons for testing in Loudoun county (including road kills and individuals found dead as well as animals which had had contact with humans or pets), included 13% in winter (December to February), 23% in spring (March to May), 30% in summer (June to August) and 34% in fall (autumn - September to November); the percentage of positive raccoons remained at 74% over the seasons. Geographical data showed northward and slight eastward spread of the disease, with 98% of infected raccoons in 1981 coming from a single watershed, while numbers from the northern section of the county increased during 1982 and was highest in the last quarter of 1982. 60% of submitted animals were female and 40% were male; prevalence in females was 81% and in males 69%; 64% of positive animals were females. Prevalence in adults was slightly higher than prevalence in juveniles: 78% versus 73%. Prevalence of infection in kits increased in September 1982, peaked in October 1982 and then decreased in the next two months. It was noted that the outbreak may have been linked to environmental and geographical factors: rainfall had been low for September 1980 to July 1981, this would have decreased favourable habitat and led to an influx of raccoons into the Goose Creek watershed area, which may have led to physiological stress - increased intra-specific competition and hormonal and neurological changes - which may lower resistance to introduced disease. The disease was seen first in the western area of the county, in the Goose Creek watershed area and spread northwards and a little eastwards. The western and northern areas have rolling to steep topography with many streams as well as cultivated land (corn), pastures and hardwood areas. The low number of raccoons submitted from the southeastern section of the county was considered likely to be related to the lowland flat, less favourable habitat with few streams (this area contains about 40% of the human population of the country, therefore it is unlikely that low submission rates were due to lack of people to notice rabid raccoons and submit them for testing). Prevalence in juvenile females was 72%, suggesting that mating was not a major factor in transmission. (J1.21.w6)
  • Data for 1,610 raccoons submitted for testing between 1st January 1982 and 20th September 1983, from the mid-Atlantic raccoon rabies focus states of Virginia, West Virginia and Maryland, found that:
    • The highest numbers of raccoon submissions occurred in March to June (spring) and October to November (fall); peaks of positive animals occurred at similar times but this was less pronounced. (J101.126.w1)
    • The positivity rate was highest in August/September and December/January (80 to 92%) and lowest in March to June (22 to 44%). (J101.126.w1)
    • About 25% of age-identified raccoons were juveniles. About 39% of adults and 39% of juveniles were positive. (J101.126.w1)
    • For animals with sex identified, males and females were presented in about equal numbers; 52% of males and 62% of females were positive (chi squared = 9.97, p<0.01). (J101.126.w1)
    • The commonest site from which raccoons were collected was yards (66% of submitted raccoons, with 378/843, 44.8% being rabies positive); 14% were found in buildings (68/174, 39.1% positive); 7.5% on cultivated land (65/95, 68.4% positive) and 4.2% on unimproved land (15/53, 28.3% positive). (J101.126.w1)
    • The commonest time submitted raccoons had been seen was during the day (68%) and 61% of these were rabid; 56% of those seen at dusk, 39% of those seen at night and 31% of those seen at dawn were rabid. (J101.126.w1)
    • Submitted raccoons were described as abnormal (43%), normal (31%) or dead (26%); 71% of "abnormal", but only 12% of "normal" raccoons were rabies positive; 70% of aggressive, 64% of overly friendly, 75% of wobbly, 76% of paralysed and 57% of injured raccoons were rabies positive. Of raccoons which had interacted with domestic animals (mainly dogs), 82% were rabies positive. (J101.126.w1)
  • A study of 798 raccoons from Virginia found that: (J270.10S4.w1)
    • 33% (254 individuals) were rabies positive; (J270.10S4.w1)
    • the number of rabid raccoons detected peaked during the spring; (J270.10S4.w1)
    • the positivity rate for adults was 43% but for juveniles was 13%; (J270.10S4.w1)
    • similar numbers of males and females were submitted; 41% of males and 30% of females were positive; (J270.10S4.w1)
    • the highest positivity rates were found in raccoons captured during the day (38% positive) or at dusk (34% positive) [rather than at dawn or at night]; (J270.10S4.w1)
    • for different land characteristics, the percentage positives were as follows: residential, 35%, agricultural 38%, industrial 33%, recreational, 18%, commercial, 15%. Additionally, the positivity rates for those found in a yard was 38%, in a building 14%, on unimproved land, 29%, in forest 30%, and within 100 ft (30m) of water, 55%. (J270.10S4.w1)
    • only 16% of raccoons reported to be behaving normally were rabies positive, but 38% of abnormal raccoons, 49% of aggressive raccoons, 35% of "sick" raccoons, 35% of those with a wobbly gait, 33% of "indifferent", 30% of "overly friendly", 23% of unresponsive and 22% of paralysed raccoons were positive; (J270.10S4.w1)
    • raccoons killed by dogs, municipal officials or members of the public had a positivity rate of 36% and those found dead had a positivity rate of 30%, but only 0-9% of those caught in traps or killed on the road were rabid. (J270.10S4.w1)
    • only 8% of raccoons reported to have potentially exposed a human to rabies were positive but 49% of those reported to have exposed other [domestic] animals were positive. (J270.10S4.w1)
  • Data from Florida and Georgia where rabies is endemic did not indicate regular oscillations of rabies prevalence (unlike the three-, four- or five-year oscillations seen for rabies in foxes). (J13.50.w1)
  • In Florida, for the period 1951 to 1988, periodic increases in cases of rabies in raccoons were observed in all areas (four "districts", with District 1 being counties of the Florida panhandle, District 4 being counties of southern Florida - Lake Okeechobee and the Everglades, District 3, central Florida, and District 2, northern Florida). In District 1, there was a gradual increase in cases of raccoon rabies through the 1980s. In District 2, peaks were seen in 1980-81 and in 1986, in District 3 dramatic increases were seen in 1968-69 and in 1984-88, while in District 4 there were increases particularly in 1969, 1974, 1977 and 1988. Raccoons made up only 16% of animals tested for rabies, but accounted for 68% of positive rabies cases (a total of 326 cases of rabid raccoons). However, the pattern of cases indicate enzootic rather than epizootic frequency: during 1986 to 1988, 326 rabies-positive raccoons were reported from 52 of the 67 counties (78% of counties), 30% of cases came from counties reporting less than five cases per year, and annual reports of 10 or more cases per county occurred only five times in three years. The geographical distribution of cases suggests low level endemicity throughout Florida. The results of the study suggested that raccoons are the sole maintenance source for enzootic terrestrial rabies in Florida. Cases in other terrestrial wild animals were sparsely distributed and generally occurred in counties with multiple reports of rabid raccoons. Similarly, cases in cats and dogs were sporadic and were usually reported from in or near foci with increased rabid raccoons. Of 84 rabies isolates (from 89 cases in terrestrial mammals during 1987) tested with a panel of monoclonal antibodies to the N protein to determine rabies virus variant, 83 were confirmed as being raccoon rabies virus variant (the sole exception reacted with all the MAbs in the panel, suggesting a laboratory or vaccine strain; it was not possible to discount laboratory contamination and there were no further samples from the animal for repeat testing). (J1.26.w6)
  • Rabies was reported in Maryland in 1981 after 20 years without rabies in terrestrial mammals. In 1985, raccoons represented 88.4% (672/760) of animals diagnosed as infected with rabies. The number of infected raccoons increased in late winter and early spring and the rate of infected raccoons was consistently highest in spring; this was considered likely to be related to breeding activity in late January and February increasing aggressive behaviour and interaction. In both agricultural and suburban communities rabid raccoons were found mainly in residential areas; 70% were found in private yards. Only seven of the rabies-positive raccoons were considered to be behaving normally; 183 were considered to be behaving abnormally (e.g. "sick", aggressive or paralysed) and a further 130 were found dead. It was noted that a mean of 1.6 other animals were exposed to each rabid raccoon; when raccoons exposed humans, 1.7 humans were exposed per rabid raccoon. It was noted that the disease had spread through Maryland raccoons at about 15 to 25 miles per year in an east-northeasterly direction, with spread facilitated by waterways but slowed by mountains. Spread to the Eastern Shore was delayed by the presence of two large thruways (major roads) within two miles of one another plus the 100-yard-wide Chesapeake-Delaware Canal, however once past these barriers in 1989 it had spread, reaching halfway down the Eastern Shore peninsula by 1992. Following the initial epizootic the disease had remained enzootic in the state with minor upsurges of rabies at intervals of three to four years as the population recovered after each outbreak. Spillover into other species, both wild and domestic, had occurred, but without any evidence of animal-to-animal spread occurring in any other species. (J4.201.w2)
  • In Virginia, from 1984 to 1989, 3,256 raccoons were submitted for rabies testing, of which 1,053 (32.3%) were rabies positive. Generally, more rabid raccoons were found in the northern counties of Virginia than elsewhere in the state. Starting in 1987, the number found in central and coastal counties increased. A seasonal pattern was noted, with both the percentage and absolute numbers of positive raccoons detected increasing in a bimodal pattern, in late winter and in early fall, with the percentage positive increasing one month before the absolute number of rabies positive raccoons; on a yearly basis, the percentage positive increased one year ahead of the absolute number positive: the total number of detected rabid raccoons peaked in 1987 but the greatest percentage positive occurred in 1986. It was suggested that the late winter rise may be associated with the breeding season (January to March) and associated increases in contact rates and aggression among raccoons, while the early fall rise may be related to juveniles dispersing into the adult population. Possible reasons suggested for the reduced rate in the summer included: raccoons becoming less communal in the spring, a smaller remaining susceptible population following rabies activity earlier in the year, and a lag before detection using passive surveillance. Peaks followed by reduced reported incidence may be due to raccoons having died of rabies, leaving a temporarily smaller population. In five counties in northern Virginia, in which rabies incidence had been high throughout the epidemic, the percentage positive showed a two- or three-year cyclic peak, while this was not evident in the absolute number positive. Suggested reasons for these cycles were increases in susceptible populations due to new raccoons being born or due to movement of raccoons. (J1.28.w10)
  • During the period 27 March 1991 (rabies first recognised in Connecticut) to 31st December 1994, 2,219 of 4,703 raccoons tested were positive (47%). From June 1991, only animals involved in a potentially infectious contact with a human or a domestic animal had been tested. The percentage positives increased rapidly and dramatically and then stabilised, with 20 to 40 positive raccoons (and 5-10 skunks) reported per month. There were no obvious seasonal trends between three-month reporting periods. Mapping of cases by three-month intervals showed a northeast spread from the initial case in southwest Connecticut near the New York border. [From the maps provided it appears that the Connecticut River delayed spread and that it may have been crossed in one or two limited areas]. The epizootic wave appeared to progress at about 30 km per year, with cases continuing after the wave but at a low level. It was noted that the number of animals, and the number of raccoons, submitted per town was positively correlated with the size of the town population. Variation between towns in number of animals or raccoons tested was NOT correlated with the land area of the different towns, suggesting that sampling was not evenly distributed. It was noted that no reliable data on raccoon density was available, so that it was not possible to make any estimates of the incidence of infection in the raccoon population, nor on any demographic changes which may have resulted. Most rabies-positive animals (89.5%) came from private properties, particularly the yards of properties. Most human exposures were indirect, from handling a dog or cat which had just fought with an animal. Only 93 raccoons caused 16.7% of human exposure incidents directly, to 109 humans (11.6% of those exposed). A few of these incidents were related to unprovoked direct attacks by rabid raccoons on people; most were due to people transporting or killing individuals which later turned out to be rabies-positive. (J91.57.w1)
  • During a study in New Jersey during testing of oral vaccination, 21/25 rabid raccoons for which age was determined were adults (while adults were only 65.1% of total captures during the study) and 13/19 of rabid raccoons with sex determined were female (68.4%, compared with 48.8% of total captures during the study). [Note: the main study involved live-captured animals; rabid animals were found dead (two), road-killed (nine) or were live-captured and euthanased because of suspicious clinical signs (15)]. (P102.4.w2)
Mathematical and computer models
  • An epizootic model for raccoon rabies was constructed using a set of coupled differential equations and involving parameters for birth rate, death rate, density-dependence in mortality, rabies-induced mortality, rabies transmission rate and the proportion of raccoons developing immunity when infected. The mathematical model indicated that the presence in the population of some individuals with naturally acquired immunity tends to result (compared to a population without immune animals) in: (J13.50.w1)
    • A higher host population density in areas with rabies; (J13.50.w1)
    • Oscillations in disease prevalence have a relatively low magnitude and short dampening time; (J13.50.w1)
    • Reduced likelihood of the disease becoming extinct, even at a local level. (J13.50.w1)
  • In Pennsylvania, a trend-surface analysis indicated that following initial entry of rabies on the south-central border of the state, rapid diffusion of the disease occurred to the north-east, following the corridors of the Appalachian mountains (ridge and valley) and the Great Valley, followed by slower spread westwards once the disease reached the northern counties and the high plateau areas. The fast initial spread occurred in the Great Valley and the southern edge of the Appalachian Mountains. It was noted that the Great Valley is limestone-floored fertile farmland and has been an important routeway for humans in eastern North America, while the Appalachian Mountains are thin-soiled and rugged. (J67.40.w1)
    • The study did not comment on raccoon habitat preferences nor provide detailed information on habitats in the different geological areas, however in general raccoon population densities are lower in upland areas and can be high in farmland areas, particularly in edge habitats. Densities of up to 20 raccoons per km2 have been noted in bottomland and marshland areas of the midwestern USA while densities in east Tennessee's Appalachian Mountains were 3.6 to 5.9 raccoons per km2.  (B402.6.w5) Additionally, a study of the "landscape epidemiology" of fox rabies in Virginia, 1954 to 1973 noted rapid spread up the Great Valley in 1971-72 and noted that this reflected high population densities of foxes, continuous suitable gray fox habitat and an absence of physiographic barriers, while ridges did act as physiographic barriers. It was also noted that "the positive correlation between densities of wild animal populations and soil fertility is well known." (J91.27.w1)
  • Two different analytical methods applied to data from Connecticut, looking at the time of first appearance of rabies in a given township, both showed that township-to-township spread of raccoon rabies was slowed significantly (by about 22% (Trend Surface Analysis) or about 29.7% (probabilistic simulation)) when a river lay between townships. Large rivers were considered to "act as semipermeable barriers to local transmission leading to a sevenfold reduction in the local rates of propagation across neighboring townships." The estimated velocity of the frontal wave of raccoon rabies was estimated to be about 46 km/year and it was suggested that the presence of the rivers had slowed the spread of raccoon rabies across Connecticut by about 12 to 14 months; spread across the state took about 48 months, while a probabilistic simulator indicated that in the absence of rivers spread would have taken about 37 months. [2002](J279.2.w3)
    • It is probable that the largest rivers [when encountered approximately perpendicular to the direction of spread of the disease] will have the greatest effects on the dynamics of rabies transmission in raccoons. (J279.2.w3)
    • It was noted that on a local scale, the raccoon rabies wave front "shows considerable variation in rate of propagation." (J279.2.w3)
    • This study looked at transmission dynamics on a relatively fine spatial scale, that of townships, affected by small scale variations in local habitat. (J279.2.w3)
  • A stochastic spatial model, also applied to data from Connecticut, looked at possible effects of (a) human density; (b) effects of rivers and (c) long-distance translocation. The model showed the best fit with observed data indicated that "a slower local spread of rabies was strongly associated with river crossings, the global spread by translocation was relatively frequent, and human population density had very little effect on the local spread of rabies." It was noted that early reports of rabies in townships noncontiguous to infected townships could be due to translocation, delayed diagnosis in adjacent townships, or misdiagnosis. It was noted that 21 townships reported their first rabies case when no contiguous township was infected, suggesting frequent long-distance translocations. River crossing produced a seven-fold slowing of rabies spread, and the first appearance east of the Connecticut River may have been due to a translocation event. It was noted that other variations in movement (e.g. travel of the disease north-east, and the later swing southeast, may have been associated with other habitat and physiographic effects not included in this model. [2002](J135.99.w3)
  • A spatially heterogenous rabies-spread model which had been constructed for Connecticut [J135.99.w3] was applied to data from the spread of raccoon rabies in New York. The researchers noted that rivers may act in either of two opposing ways: if parallel to the advancing wave front of rabies (as seen in New York) then they can enhance rabies transmission up the river valley, since riparian areas provide very good habitat for raccoons; conversely, if a [large] river is perpendicular to the path of a rabies wave front (as seen in Connecticut) then it may act as a barrier to further spread in the direction of the advancing wave. [2004](J179.271.w1)
  • A further analysis of data for Connecticut suggested that rabies moved more slowly through townships with heavier forest cover than through lightly forested townships and that while rivers slowed spread, this effect was greatest for rivers bordered by heavy forests: rabies did not cross rivers in heavily forested locations. Additionally, the data indicated that several probable long-distance translocation (LDT) events had occurred; some either were false-positives or did not become established and propagate, while others did become new foci, affecting the course of the epidemic in surrounding townships. Possible reasons for failure of spread from nascent foci include stochastic fade-out, if the reproductive rate of the disease is near or below the threshold required for establishment, or occurrence in an isolated host population, therefore the impact of the LDT may be limited by spatial heterogeneity. (J67.71.w1)
  • A study of the temporal structure of rabies epizootics in the mid-Atlantic and northeastern US states from the mid-1970s onward was based on data at the county level, involving more than 35,387 cases of rabies in raccoons in 390 counties. It was noted that the criteria for submission of animals for rabies testing varied somewhat between states, however, the method of analysis, looking at data from each county individually, was considered to minimise the effects of this. For each county, the median number of detected rabid raccoons over the whole period from the time rabies was first detected, through to December 1997 (for Maryland, to December 1998) was determined. An epizootic was defined as starting when the monthly rabid raccoon number was greater than the county median for two consecutive months, lasted (numbers higher than median) for at least five months and ending when it dropped to equal to or below the county median for two consecutive months. An epizootic period was defined as the time from the start of an epizootic to the start of the next epizootic. This study found that the median first epizootic period was 48 months (range seven to 158 months, mean 45.4 months, dominant mode 41 - 60 months). By linear regression, it was found that successive epizootic periods decreased by about five months for each epizootic period. Additionally, it was found that the size of the first epizootic was substantially higher than that of successive epizootics: mean 80.5 rabid raccoons per county during the first epizootic compared with a mean of 17.3 for all subsequent epizootics, therefore there was a significant difference in the size of the first epizootic compared to that of the second epizootic in each county. The data was compared to an epizootic model for raccoon rabies, previously constructed, on the basis of population dynamics (see J13.50.w1). The data was found to best fit the model if the percentage of raccoons surviving rabies infection and becoming immune was low: l% - 5%. This was true for three different values entered for the transmission rate. [2000](J135.97.w3)
  • An analysis of data from the mid-Atlantic/northeastern rabies area (35,387 rabid raccoons from 390 of the 475 counties in 11 states) showed that there were differences between northern and southern states within this area in (a) the size of epizootics (number of rabid raccoons per month); (b) the annual incidence of epizootics and (c) the duration of epizootics. In particular, values in North Carolina, Virginia and West Virginia tended to be different from values for Connecticut, Delaware, Massachusetts, Maryland, New Jersey, New York, Pennsylvania and Rhode Island: the southern states tended to have epizootics which were shorter, involved fewer raccoons and occurred less frequently. Some of the differences were associated with human population density (which affects likelihood of a rabid raccoon being seen) and with per capita health spending (which might reflect testing priorities), but allowing for these, differences still remained, indicating other factors. Increased raccoon rabies in areas of high human densities, in urban and suburban areas, could also be associated with high raccoon population densities in these areas, which could contribute to the development of larger epizootics. It was noted that further research was required to assess the effects of variables such as suitability of habitat types for raccoons. (J279.1.w2)
  • A study of raccoon rabies epizootics in Maryland, Pennsylvania and Virginia, considered environmental and human demographic features associated with the recorded epizootics. An epizootic was defined as starting when the monthly rabid raccoon number was greater than the county median for two consecutive months, lasted (numbers higher than median) for at least five months and ending when it dropped to equal to or below the county median for two consecutive months. The analysis looked at land use (according to USGS definitions (W184.Oct05.w1)) and human density. Various factors were found to be associated with whether or not a county had a large epizootic (defined as more than 24 rabid raccoons in the first epizootic): (J1.39.w4)
    • High human population density was significantly associated with a county having a large epizootic; (J1.39.w4)
    • High percentages of agricultural land, water, low-intensity residential, high-intensity residential, grasslands and emergent herbaceous wetlands were significantly associated with a county having a large epizootic; (J1.39.w4)
    • High percentages of deciduous or mixed forests were significantly associated with a county experiencing no epizootic or a small epizootic. (J1.39.w4)
    • The strongest associations were with agricultural land coverage, water coverage and mixed forest, as well as an interaction between human population density and water coverage. (J1.39.w4)
    • It was noted that "any report of animal rabies depends on a human participant who collects the animal and submits it for rabies testing. Up to some level, increasing the density of human observers should increase the likelihood of a rabid animal being detected and tested. Where human population density is low, a greater number of rabid raccoons may have to be present to result in a reported epizootic of equal magnitude." (J1.39.w4)
    • [Note: despite the comment regarding the requirement for human observers, this study did not appear to consider directly whether human population density actually produced artefacts of high and low rabies occurrence, i.e. whether large reported epizootics in areas of high human population density, and not in areas of low human population density is due simply to the presence of observers.]
  • An analysis of raccoon-variant rabies in raccoons and skunks in the Eastern USA showed that rabies in skunks had approximately a one-month lag behind rabies in raccoons, and indicated a four to five year cycle between epizootics. The model showed that reports of raccoons with rabies remained "fairly constant" through the year, while those for skunks consistently peaked during the fall months. A major consideration was whether there was any evidence for raccoon rabies cycling in skunks independently. The temporal and spatial analysis of data concluded that the number of rabid skunks was best predicted by the number of rabid raccoons in the previous month. There was thus no evidence of independent cycling of the raccoon rabies virus among skunks. (J84.9.w24)
  • Apparent seasonality may be related to the stage of the outbreak, or may be associated with the breeding season. (J270.10S4.w1)
Scales of habitats and data analysis
  • It is important to recognise and remember that although there are some areas where habitat is relatively uniform over large areas, there can be considerable, even dramatic, changes in habitat over very short distances - a few hundred metres or less, while locations of detected rabid individuals may be imprecise - e.g. to the nearest town. (B395.4.w4)
  • The effects of local ecogeography, human factors etc. affecting rabies cycles may be obscured if data is analysed at too large a scale; this was noted in Ontario when rabies outbreaks (in various species) were analysed on the scale of counties. Using smaller units of area (townships) was considered to be more informative. [2004](J1.40.w1)
Associated techniques linked from Wildpro
  • --

Return to top of page

Authors & Referees

Authors Debra Bourne MA VetMB PhD MRCVS (V.w5)
Referee Dr Robert G. McLean (V.w42), Rick Rossatte (V.w95)

Return to top of page