Skip Navigation

Bat Rabies in the United States and Canada from 1950 through 2007: Human Cases With and Without Bat Contact

  1. Gaston De Serres1,
  2. Frédéric Dallaire2,
  3. Mathieu Côte2, and
  4. Danuta M. Skowronski3
  1. 1Institut national de santé publique du Québec, Quebec
  2. 2Université Laval, Quebec
  3. 3British Columbia Centre for Disease Control, Vancouver, Canada
  1. Reprints or correspondence: Dr. Gaston De Serres, Institut national de santé publique du Québec, 2400 d'Estimauville, Quebec, PQ, Canada, G1E 7G9 (gaston.deserres{at}ssss.gouv.qc.ca).

Abstract

Background. Since the 1980s, rare cases of rabies in humans in Canada and the United States have been almost exclusively caused by the bat-variant virus.

Methods. We reviewed indigenously acquired cases of bat-variant rabies in humans in Canada and the United States from 1950 through 2007.

Results. Of 61 cases identified, 5 occurred after organ transplantation and were excluded from further analysis. A bite was reported by 22 (39%) of the case patients, 9 (16%) had a direct contact (i.e., were touched by a bat) but no history of a bite, 6 (11%) found bats in their home (2 [4%] in the room where they slept) but reported no direct contact, and 19 (34%) reported no history of bat exposure whatsoever. With the exception of California (8 cases) and Texas (7 cases), no state or province had >3 cases. Of the case patients, 76% were men, and 40% were 10–29 years of age. The median incubation period was 7 weeks (<10 weeks in 72% of cases). The incidence of bat-variant rabies cases increased from 2.2 per billion person-years in 1950–1989 to 6.7 per billion person-years in 1990–2007. Of 36 case patients with bat rabies described since 1990, 16 had no history of direct bat contact; 2 (13%) of the 16 would have qualified for rabies postexposure prophylaxis on the basis of exposure criteria expanded in 1995 to include bats that were in the same room as a sleeping person. The incidence of rabies for this type of exposure was 0.6 cases per billion person-years.

Conclusion. The true preventable proportion of cases and the number needed to treat with rabies postexposure prophylaxis to prevent 1 case would be useful information to inform the current guidelines.

Rabies is a dreaded disease. The rabies virus causes encephalitis and death in humans and in most other mammals. A distinct species-associated variant characterizes each of the major terrestrial animal hosts. Humans are not natural hosts but can become infected through contact with other rabid animals. With a single documented exception, human cases of rabies have been uniformly fatal [1]. According to the best estimates worldwide, >55,000 deaths occur annually, most of which are associated with dogs [2]. Prevention is through a protocol of rabies postexposure prophylaxis (RPEP) that consists of a single dose of immunoglobulin and 5 spaced doses of vaccine. RPEP is safe and effective; failure rates, typically associated with a delay or breach in protocol, have been estimated to range from 1 in 80,000 in developed countries to 1 in 12,000 in developing countries [3]. The reported incubation period varies from a few days to >19 years, but 75% of patients become ill in the first 90 days after exposure [3, 4].

Three principal global areas of rabies have been defined [4]. These areas are (1) countries with enzootic canine rabies (all of Asia, Latin America, and Africa); (2) countries in which canine rabies has been brought under control and wildlife rabies predominates (Western Europe, Canada, and the United States); and (3) rabies-free countries (mostly islands, including England, Australia, and Japan). In Canada and the United States, rabies is enzootic in foxes, skunks, raccoons, and bats; these animals are the only natural reservoirs in Canada and the United States. The first identification of rabid insectivorous bats in North America was in the early 1950s. Since then, the incidence of rabies in bat populations, although known to vary according to the species, its social preferences (colony forming vs. solitary), and migratory patterns, is not known to be on the rise [3, 4].

A program of rabies immunization for domestic dogs led to a dramatic decline in human cases of canine rabies in Canada and the United States in the 1950s. In its place, the bat-variant virus has become the dominant cause of rabies in humans. The advent of the detection of anti-nucleocapsid monoclonal antibody in the late 1970s and the more-recent nucleotide sequencing applied to virus strains recovered from humans has facilitated distinction between bat-variant rabies and virus strains originating in other terrestrial reservoirs; this has led to increased links with bats that may have been previously unknown or unrecognized [3, 5]. Transmission of bat-variant rabies through other terrestrial animals to humans is theoretically possible, but no such human cases have been documented [4]. A squirrel infected with bat-variant rabies was reported to have bitten a person, but RPEP was administered, and human disease did not follow [6].

Bats apply sophisticated echolocation to navigate, and healthy bats are easily able and prefer to avoid collision or other contact with humans. Bats are night creatures, and it is unusual for them to fly in daylight. Those active during the day are suspect, as are any found on the ground. The behavior demonstrated by rabid bats is not always dramatically altered [4]. Although some may become aggressive, others may simply become disoriented, lose their flying ability, and appear clumsy. Healthy juvenile bats acquiring new skills may also demonstrate such erratic behavior. Relentless attachment, particularly to the hands or heads of humans, is a clear sign of abnormality.

Bat teeth are very fine, and bat bites may be undetectable as pinpoint puncture marks ≤1 mm in diameter. Most bat-inflicted scratch marks are <1 cm long. Such minor evidence of contact may be difficult to recall or elicit from a person dying from rabies. In the past 2 decades, most human case patients with bat-variant rabies in the United States and Canada did not report a history of a bat bite [5, 7, 8]. The recognition that indigenous cases of rabies in humans in North America are largely associated with bats and that even direct contact with bats may not be perceived has led to the lowering of the threshold for RPEP administration. In 1995, the Centers for Disease Control and Prevention (CDC) urged “consideration of post-exposure prophylaxis for persons potentially exposed to bats even where a history of physical contact cannot be elicited” [9, p. 272]. This is reflected in guidelines published by advisory committees in the United States in 1999 and in Canada in 2002, which include scenarios in which a bat is found in the same room as a sleeping person, an unattended child, a mentally disabled person, or an intoxicated person [10, 11].

Individual case reports of bat-variant rabies have been published previously. This report summarizes all cases of bat-variant rabies identified in humans in the United States and Canada from 1950 through 2007, including those with and those without recognized bat contact.

Methods

We reviewed all human cases of rabies from January 1950 through September 2007 in the United States and Canada as described in the Morbidity and Mortality Weekly Report and Canada Communicable Disease Report. The number of human cases we found with bat-variant rabies corresponds with totals published by the CDC and the Public Health Agency of Canada. Only indigenously acquired cases were considered; cases acquired elsewhere but diagnosed and treated in the United States or Canada were excluded.

The incidence rate per person-year was calculated by dividing the number of cases by the sum of the populations of each year from 1950 through 2007. Population data were obtained from the US Census Bureau [12, 13] and Statistics Canada [14].

Results

From January 1950 through September 2007, the total number of human cases of rabies declined considerably, from a peak of 20 cases in 1952 to never more than 6 cases per year since the 1960s, which mostly reflects the control of canine-variant rabies (figure 1A). Even recently, however, year-to-year variation in the number of human cases of rabies is evident. There were 23 cases for which the source of rabies was unknown, including 21 (21%) of 102 cases from 1950 through 1959 and 2 (7%) of 29 cases from 1960 through 1979. Since 1980 and the advent of anti-nucleocapsid monoclonal antibody detection and nucleotide sequencing, distinct species-associated variants were identified in all reported human cases of rabies.

Figure 1

A, Number of human cases of indigenously acquired rabies in Canada and the United States since 1950, including cases involving organ transplantation, by type of source animal. N/A, not available. B, Number of human cases of indigenously acquired bat rabies in Canada and the United States from 1950 through 2007, excluding organ transplant cases.

Since 1950, 61 human cases of bat-variant rabies were reported in the United States (55 cases) and Canada (6 cases). A list of the characteristics and the bat-contact history of cases is presented in table 1. Five cases occurred following the transmission of the virus to organ transplant recipients from 2 infected donors; 3 of the cases involved recipients in Texas [34]. These 5 cases have been excluded from further analysis.

The 56 non—transplant-associated cases of bat-variant rabies were distributed across both countries in 27 states and 5 provinces, with no clear geographic clustering (figure 2). Each state or province had ≤3 cases, with the exception of California (8 cases) and Texas (7 cases). More than 75% of case patients were male (42 patients), and 41% (23 patients) were adolescents or young adults (age, 10–29 years). By decreasing order of frequency, bat-variant rabies was diagnosed in persons aged 20–29 years (24%), 10–19 years (18%), 40–49 years (18%), ⩾60 years (18%) , <10 years (11%), 50–59 years (7%), and 30–39 years (4%).

Figure 2

Geographic distribution and number of human cases of bat rabies in Canada and the United States from 1950 through 2007, excluding cases involving organ transplantation. Provinces and states with no cases are in white.

Of the 56 case patients, 31 (55%) had direct contact with a bat: 22 reported being bitten, and 9 had direct contact (i.e., were touched by a bat) but reported no history of a bite (table 1 and figure 1B). The remaining 24 case patients (43%) did not have recognized direct contact with a bat: 6 found bats in their home (2 while sleeping) and 19 reported no history of bat exposure whatsoever. Of the 12 case patients with a history of a bat in their bedroom (cases 4, 5, 6, 8, 16, 21, 26, 28, 31, 33, 34, and 35), 3 had no direct contact, 5 were bitten, and 4 were awakened by a bat having landed on them. Of the 56 cases not associated with organ transplantation, 36 (64%) occurred from 1990 through 2007. Of these 36 cases since 1990, 16 involved no history of direct contact with a bat; 2 (13%) of the 16 case patients would have qualified for RPEP on the basis of exposure criteria expanded in 1995 to include bats that were in the same room as a sleeping person.

Figure 3

Incubation period of indigenously acquired human cases of bat rabies with a history of a bat bite (black), direct contact with no bite (gray), or the presence of a bat in the house without direct contact (white), excluding cases involving organ transplantation.

Table 1

Characteristics of the 61 human cases of indigenously acquired bat rabies in the United States and Canada, 1950–2007.

Among the 6 cases involving pediatric patients <10 years of age (cases 2, 4, 5, 11, 33, and 35), only 2 occurred since 1990 (cases 33 and 35). A bat was found in the room of the first of these 2 case patients (case 33), but family members had examined the child and found no evidence of a bite. After the child's death, the bat that had been buried was recovered and tested positive for rabies. The second case patient had no known bat bite but showed his mother a 2-cm erythematous lesion on his upper arm 3 days after he slept in a cabin where 2 bats had been found. The lesion was not identified as a bat bite at the time. The remaining 4 pediatric cases occurred before 1990, and all 4 case patients had a history of bat bite.

Among the 25 persons with a discrete history of bat exposure, under the assumption that the infection was caused by that recognized exposure, the incubation period varied from 16 days (cases 1 and 33) to 180–270 days (case 22). The onset of symptoms occurred before 10 weeks in 18 (72%) of the 25 case patients (table 1 and figure 3). The diagnosis of rabies was considered for the first time at autopsy in only 6 (11%) of the cases (cases 11, 15, 18, 29, 34, and 38).

Excluding the 5 transplant-associated cases, the overall incidence rate of bat rabies in humans was 3.9 cases per billion person-years, with similar rates in the United States and Canada (3.9 and 4.4 cases per billion person-years, respectively). The incidence of bat-variant rabies cases increased from 2.2 cases per billion person-years in 1950–1989 to 6.7 cases per billion person-years in 1990–2007 ([table 2 [table 2 and figure 1B). It is likely that a proportion of the 23 cases occurring from 1950 through 1979 with no known source of exposure and an uncharacterized rabies virus were caused by the bat variant. Under the assumption that all cases were caused by bat rabies, the overall incidence would be 5.5 cases per billion person years, with 15.5 cases per billion person-years in 1950–1959, 0.5 cases per billion person-years in 1960–1969, and 4.6 cases per billion person-years in 1970–1979 ([table 2 [table 2). From 1990 through 2007, there were 2 case patients for whom the only known exposure was the presence of a bat in the room where they slept. The incidence for this type of exposure was 0.6 cases per billion person-years. During the same period, there were 11 cases (2.0 cases per billion person-years) with no recognized or reported history of contact.

Table 2

Incidence rate and number of human cases of bat rabies in the United States and Canada, excluding cases involving organ transplantation, by decade and type of contact.

Discussion

Rabies in humans in the United States and Canada remains dreaded but rare. With the near elimination of human cases caused by canine rabies, bats have become the animal source of the greatest concern. This is in part because of frequent opportunities for interaction through shared urban and rural distribution and because controlling enzootic rabies in nondomesticated animals is more difficult than in domesticated animals, as seen in the successful control of canine rabies achieved through immunization of dogs.

Overall since 1950, we found an incidence rate of bat rabies in humans of 3.9 cases per billion person-years, with similar rates in the United States and Canada. The introduction of anti-nucleocapsid monoclonal antibody detection at the end of the 1970s led to improved species attribution and the identification of the bat variant among persons with unrecognized exposure. This and the increased concern about bat rabies generally make it difficult to comment on temporal trends. Rabies in humans is dramatic in its clinical presentation and is accompanied by cardinal features. Nevertheless, the diagnosis may be missed; in this series, 11% (6) of the cases of bat rabies were diagnosed postmortem, and it is possible that additional cases without an autopsy performed were missed. The extent of this underestimation is impossible to determine but is likely to be higher among those cases without recognized and suspected exposure history. Although rabies is also enzootic among bats in Europe, only 5 human cases of bat rabies have been described there, 2 of which occurred in nonvaccinated professional bat handlers [57]. In Mexico, 2–3 human cases of bat rabies were identified per year since 2000, although the trend as a proportion of all human cases of rabies has been increasing in recent years [58]. For the rest of Latin America, data are fragmentary, but some cases have been reported [5962].

Among reported case patients, young-adult and adolescent males seem to predominate. They may practice activities that increase their exposure risk or may be less inclined to seek medical attention or prophylaxis should a contact, even a bite, occur. For this group in particular, it would be beneficial to emphasize through public messages the importance of maintaining a safe distance from bats, avoiding the handling of bats, and minimizing opportunities for contact with bats. Primary preventive measures include education and the barring of bats from human dwellings by sealing openings as small as 2 cm and by placing screens on windows. Professional bat-control advice may be necessary where human dwellings are already inhabited with bats. Although effective, secondary prevention through immediate wound washing and RPEP requires that exposure has been both recognized and brought to the attention of public health personnel. Given the frequently cryptic nature of bat contact, this is clearly not always possible.

Among the 56 human case patients with bat rabies described in this series, nearly half had no history of direct contact, and one-third had no known or reported bat exposure at all. The recollection of bite incidents can affect the analysis, and these cases were likely caused by an unrecognized bite and salivary contamination, as opposed to aerosol exposure. Since 1999, RPEP recommendations have been broadened in the hopes of preventing additional cases among persons with unrecognized bat contact, including “persons who were in the same room as the bat and who might be unaware that a bite or direct contact had occurred (e.g., a sleeping person awakens to find a bat in the room or an adult witnesses a bat in the room with a previously unattended child, mentally disabled person, or intoxicated person)” [10, p. 8]. Only 2 (13%) of the 16 case patients since 1990 who reported no direct contact had possible exposure while sleeping and, thus, would have qualified for RPEP had this been recognized as sufficient exposure. Although the remainder may also have had qualifying contact that was not elicited before illness or death, other types of exposure outside the eligibility criteria for RPEP may have been responsible. The true preventable proportion of cases and the number needed to treat with RPEP to prevent 1 case would be useful information to inform the current guidelines. In that context, evaluation of the frequency of different types of bat contact and the potentially at-risk but undeclared exposures in a representative sample of the population may be worthwhile.

Acknowledgments

Financial support. Ministère de la santé et des services sociaux du Québec.

Potential conflicts of interest. All authors: no conflicts.

  • Received October 12, 2007.
  • Accepted December 10, 2007.

References

| Table of Contents