Materials and Methods

Annual rabies surveillance reports^6,11–14 prepared by the CDC were reviewed to identify states that reported a rabid dog or cat between 1997 and 2001. For each state that was identified, the person responsible for rabies surveillance in the state health department was contacted by e-mail or telephone and invited to participate in the study. Questionnaires were subsequently sent to individuals who indicated a willingness to participate. The questionnaire listed the number of rabid dogs and cats reported to the CDC by the state for each year of the study (1997 through 2001) and asked respondents to provide the numbers of dogs and cats tested for rabies during each of those years.

For any dog or cat confirmed to have rabies on the basis of results of virus isolation or direct fluorescent antibody testing of brain tissue¹⁵ that also had a history of rabies vaccination, respondents were asked to provide details of the vaccination history (including age when vaccine doses were administered), age of the animal at the time of death, any history of exposure to a rabid animal, time between exposure and the onset of clinical signs, clinical signs exhibited by the animal, duration of clinical signs, whether the animal had been examined by a veterinarian, and whether the animal had died or was euthanatized. In addition, respondents were asked to indicate whether a particular virus variant had been identified and how many people had received postexposure prophylaxis as a result of exposure to the animal. Data that were collected did not include identifying information on any human subjects.

Dogs and cats for which information on history of vaccination was provided were classified as currently vaccinated if the first dose of rabies vaccine had been administered ≥ 28 days prior to exposure to a rabid animal and vaccination had been performed in compliance with guidelines from the National Association of State Public Health Veterinarians (ie, initial vaccination no earlier than 12 weeks of age, booster vaccination 1 year after initial vaccination, and additional regular booster vaccinations as directed by the vaccine manufacturer).¹⁰ Animals that had a history of vaccination but were not considered current on vaccination were classified as not currently vaccinated (eg, animals for which too much time had elapsed since the last booster vaccination) or as being within 28 days of the first vaccine dose administration at the time of rabies onset. Animals that had a history of vaccination but for which specific details were unavailable were classified as having unknown or unclassified vaccination status.

Data analysis—Information from returned questionnaires was entered into a spreadsheet^a and imported into a statistical software program^b for analysis. Descriptive statistics were calculated. Incidence of rabies among vaccinated dogs and cats was calculated on the basis of total number of rabid dogs and cats with a history of vaccination during the 5-year study period. The χ² test was used to determine whether species (dog vs cat) was significantly associated with having a history of vaccination (yes vs no). One-way Kruskal-Wallis analysis on ranks was used to determine whether the number of people requiring postexposure prophylaxis as a result of exposure to rabid dogs with a history of vaccination was significantly greater than the number of people requiring postexposure prophylaxis as a result of exposure to rabid cats with a history of vaccination. For all analyses, values of P < 0.05 were considered significant.

Results

Forty-one states reported ≥ 1 rabid dog or cat between 1997 and 2001. No rabid dogs or cats were reported during the study period by Colorado, Hawaii, Idaho, Illinois, Indiana, Mississippi, Nevada, Utah, or Washington. Of the 41 states with ≥ 1 rabid dog or cat, 21 (51%) agreed to participate in the study. New Mexico was unable to provide total numbers of dogs and cats tested for rabies during 1997 and 1998, and Pennsylvania was unable to provide total numbers of dogs and cats tested during any year of the study period.

Participating states indicated that 170,987 dogs and cats were submitted for rabies testing during the study period (Table 1; only partial data were available for New Mexico, and data were not available for Pennsylvania). A total of 264 rabid dogs and 840 rabid cats were identified during this period. Alaska (12/175 [6.86%]) and West Virginia (13/980 [1.33%]) had the highest percentages of rabid dogs among all dogs tested, and Virginia (124/5,211 [2.38%]) and Maine (18/1,152 [1.56%]) had the highest percentages of rabid cats among all cats tested. States with the most dogs and cats tested during the study period were Texas (18,741 dogs and 16,690 cats) and California (13,339 dogs and 11,481 cats).

States reporting rabid dogs and cats with a history of vaccination included Texas, California, Pennsylvania, South Carolina, Virginia, New York, and Missouri. Thirty-five of the 1,104 (3.2%) rabid dogs and cats had a history of vaccination, including 13 of the 264 (4.9%) rabid dogs and 22 of the 840 (2.6%) rabid cats. Rabid dogs were not significantly (P = 0.06) more likely to have a history of vaccination than were rabid cats.

Information on vaccination status at the time of death was available for 28 of the 35 (80%) rabid animals with a history of vaccination. Of the 13 rabid dogs with a history of vaccination, 2 were classified as currently vaccinated, 5 were classified as not currently vaccinated, and 4 were classified as being within 28 days of the first vaccine dose administration at the time of rabies onset. In the remaining 2 dogs, vaccination status was unknown or could not be classified. Of the 22 rabid cats with a history of vaccination, 3 were classified as currently vaccinated, 8 were classified as not currently vaccinated, and 6 were classified as being within 28 days of the first vaccine dose administration at the time of rabies onset. In the remaining 5 cats, vaccination status was unknown or could not be classified. Two of the 13 rabid dogs and 3 of the 22 rabid cats had received booster doses of rabies vaccine after exposure to a rabid animal.

Overall, 6 rabid animals (1 dog and 5 cats) had a history of receiving 2 doses of rabies vaccine in their lifetime, including 2 cats that were classified as currently vaccinated. The dog received its first vaccination in 1998 at 3 years of age and was given a booster dose in 2001 after suspected exposure to a skunk. This dog lived 8 days before dying of rabies and was not classified as currently vaccinated because it had not received the initial booster dose 1 year after first vaccination. One of the 2 cats that were classified as currently vaccinated had received its first vaccination at 1 year of age and received its initial booster dose exactly 1 year later. Rabies was diagnosed 10 months after the booster dose was given. The other cat had been vaccinated at 3 months of age and received its initial booster dose at 15 months of age. This cat began to show clinical signs of rabies 3 days after the initial booster dose was given.

The 3 remaining cats that had a history of receiving 2 doses of rabies vaccine were classified as not currently vaccinated at the time of death. One cat had received its first vaccination at 1 year of age, received a second dose at an unknown age, and developed rabies at 8 years of age. The second cat received its first vaccination at an unknown age, received a booster dose after being exposed to a rabid raccoon, and died of rabies < 1 month after exposure. The third cat received its first vaccination at 3 months of age, received a booster dose after exposure to a rabid animal at 2.5 years of age, and developed clinical signs of rabies shortly after receiving the booster dose.

Of the 35 rabid animals with a history of vaccination, 12 (34%) reportedly were known to have been exposed to a rabid or potentially rabid animal. This included animals from Texas (2 dogs), Pennsylvania (8 cats), and New York (2 cats). Both of the dogs from Texas were reportedly known to have been exposed to rabid skunks. One of these dogs was infected with a Texas fox variant of the rabies virus; the other was infected with a skunk variant. One cat from Pennsylvania reportedly was known to have been exposed to a rabid raccoon, but virus identification was not performed. The other 7 cats from Pennsylvania and the 2 cats from New York reportedly were known to have been exposed to rabid animals, but the species of rabid animal was unknown. Thirteen of the 35 (37%) rabid animals (9 dogs and 4 cats) with a history of vaccination did not have any known exposure to rabid animals. For the remaining 10 (29%) animals (2 dogs and 8 cats), information on exposure to rabid animals was not available.

For the 2 dogs from Texas known to have been exposed to rabid skunks, the time from exposure to initial clinical signs was < 1 month. For the 10 cats known to have been exposed to rabid animals, the time from exposure to initial clinical signs was < 1 month in 4 cats, between 1 and 3 months in 2 cats, and unknown in the remaining 4 cats.

The virus variant was identified in 12 of the 35 (34%) rabid animals with a history of vaccination, including 10 dogs (6 from Texas, 1 from Missouri, and 3 from California) and 2 cats (both from Virginia). Three of the dogs from Texas were infected with the Texas fox rabies virus variant, and 3 were infected with skunk rabies virus variant. The dog from Missouri and all 3 dogs from California were infected with skunk rabies virus variant. Both cats were infected with raccoon rabies virus variant.

Information on clinical signs was available for 34 of the 35 (97%) rabid animals with a history of vaccination (information on clinical signs was not available for 1 cat), with 25 animals having > 1 clinical abnormality. Clinical signs in the 13 dogs included paralysis (n = 6), lethargy (5), and aggression (4). Six dogs had additional clinical signs, including restlessness, hypersalivation, anorexia, ataxia, limping, and visible wounds. One dog reportedly did not have any clinical signs. Clinical signs in the 21 cats included aggression (n = 9), paralysis (6), and lethargy (2). Nine cats had additional clinical signs, including excitability, change in voice, loss of appetite, choking, difficulty swallowing, ataxia, tremors, and anorexia.

Duration of illness was reported for 22 of the 35 (63%) rabid animals with a history of vaccination (10 dogs and 12 cats). Duration of illness was 0 to 3 days in 7 dogs and 9 cats, 4 to 7 days in 2 dogs and 3 cats, and > 7 days in 1 dog. Eight of the 13 (62%) dogs and 13 of the 22 (55%) cats were euthanatized, which likely shortened the duration of illness in these animals. Five dogs and 7 cats died from rabies. Outcome was reported as unknown for 2 cats. Five of the dogs were < 5 years old at the time of death, 5 were between 1 and 5 years old, and 2 were > 5 years old. Age at the time of death was unknown for the remaining dog. Two of the cats were < 5 years old at the time of death, 8 were between 1 and 5 years old, and 4 were > 5 years old. Age at the time of death was unknown for the remaining 8 cats.

Postexposure prophylaxis was required for humans exposed to 12 of the 13 (92%) rabid dogs and 19 of the 22 (86%) rabid cats with a history of vaccination, with a total of 169 people receiving postexposure prophylaxis. Median number of persons requiring postexposure prophylaxis following exposure to dogs was 7 (range, 1 to 12), and median number of persons requiring postexposure prophylaxis following exposure to cats was 4 (range, 1 to 15). Number of people requiring postexposure prophylaxis following exposure to rabid dogs with a history of vaccination was not significantly (P = 0.08) different from number of people requiring postexposure prophylaxis following exposure to rabid cats with a history of vaccination.

Discussion

Results of the present study suggested that rabies is uncommon in vaccinated dogs and cats but can occur. Among the 1,104 laboratory-confirmed cases of rabies reported by the 21 states participating in the present study, 35 (3.2%) involved dogs and cats with a history of rabies vaccination, including 2 dogs and 3 cats in which vaccinations were considered current. In a similar, unpublished retrospective study of data for 1992 through 1996, we found that 40 of 485 (8.2%) rabid dogs and 17 of 1,152 (1.5%) rabid cats had a history of vaccination. Differences between results of the present study and results of this unpublished study most likely relate to a decrease in the canine variant of rabies in the southern United States resulting from active oral vaccination programs and an increase in cat rabies secondary to the raccoon variant epizootic in the northeast.^16,17

In a previous prospective study¹⁸ of data for 1988, 14% of rabid dogs and 5% of rabid cats had a history of vaccination. It is likely that the prospective approach in this study was the reason for the higher percentages of rabid dogs and cats with a history of vaccination, compared with percentages in the present study, which was retrospective in nature. Collectively, these findings strengthen the observation that rabies can occur in vaccinated animals, albeit uncommonly.

Because domestic animals can serve as a bridge between wildlife rabies reservoirs and humans, vaccination of pets is one of the most effective public health tools available to safeguard human health. Currently, at least 18 rabies vaccines are licensed for use in domestic dogs or cats.¹⁰ Proof of potency, safety, purity, and efficacy is required for vaccine licensure. Currently, the USDA requires that at least 86% of vaccinated dogs or cats be protected against a lethal rabies virus challenge before the vaccine can be licensed for use.¹⁹

For rabies vaccines, vaccination failure can be defined as any instance when an animal that has been vaccinated in accordance with protocols recommended in the Compendium of Animal Rabies Prevention and Control¹⁰ develops rabies following exposure to the virus. A variety of factors can lead to vaccination failure. As an example, use of a vaccine that has not been properly stored or that is past its expiration date can lead to vaccination failure. Host factors that result in a failure to mount an appropriate immune response, such as immunocompromise secondary to disease or corticosteroid treatment, can also lead to vaccination failure. All of these factors should be examined in cases of suspected vaccination failure.

Vaccination failure can also occur as a result of exposure to a particularly virulent virus variant. Vaccine trials typically involve determining efficacy against a single virus variant, when in actuality, each variant exhibits a different degree of virulence. Thus, it is unknown how effective vaccines would be against other variants, unless multiple trials are performed. In the present study, there was no apparent overrepresentation of particular virus variants in those animals with vaccination failure, but additional studies and continued surveillance are needed to determine whether a problem exists.

An important limitation of the present study was the variability in data collected by each state, which affected our ability to make comparisons among states. States differed with regard to their surveillance systems and methods for investigating potentially rabid animals; criteria for submission of potentially rabid animals for testing, which could have reduced the number of rabid animals identified; population size, which affected the number of owned pets and chances for human and domestic animal interactions with wildlife; and the presence of various rabies virus strains. For example, Alaska tested as few as 175 animals during the 5-year study period, whereas Texas tested 18,741 animals. One reason for the difference in number of animals tested among states could be that the criteria for submission of animals for rabies testing may vary from state to state. Regions that are affected by > 1 predominant rabies virus variant, such as Texas and Arizona,¹² were more likely to have more extensive surveillance programs, which would increase the number of animals tested each year. Another reason for the difference in number of animals tested could be the difference in population size among participating states.

Rabies virus variants identified in dogs and cats in the present study were typical of the geographic region from which the animals came. This indicates that spillover from one terrestrial reservoir to another did not occur frequently during the study period and suggests that rabies control programs implemented by the states to control enzootic rabies in their specific region were successful.

Typically, rabid cats are reported to have aggressive behavior, whereas rabid dogs are more likely to have lethargy and paralysis,¹⁸ and findings in the present study were similar. This suggests that regardless of vaccination status, signs of rabies will be similar for a particular species.

In attempting to collect data for the present study, we were struck by the degree of difficulty some states had in acquiring information on rabid animals. In particular, only 51% (21/41) of the states that were contacted had the information available and were willing to participate. This was a major weakness of the study, as it affected the quality of the data and could have affected the conclusion that vaccination failure is a rare event. Many states experienced difficulties in completing the study questionnaire because these data are not kept or routinely monitored. This is problematic in that it is difficult to identify problems or trends in vaccination failure when data are not kept on a regular basis. In addition, it is recommended in the Compendium of Animal Rabies Prevention and Control that rabies in vaccinated animals be reported promptly to state public health officials, the vaccine manufacturer, and the USDA.¹⁰ The compendium also recommends that the virus variant be identified by a reference laboratory in every case of rabies in a vaccinated animal as part of a thorough epidemiologic investigation. In the present study, in contrast, the virus variant was known for only 10 of 13 dogs and 2 of 22 cats with history of rabies vaccination. Despite these limitations, the present study was effective in documenting the existence of rabies in vaccinated animals. It is our hope that similar studies will encourage states to develop an improved system to effectively record and report this information to public health authorities. For instance, as each state identifies a rabid animal, a questionnaire similar to the one used in the study could be administered to collect relevant information regarding exposure and vaccination status.