Plague is a nationally notifiable condition in the United States (4). State and local health jurisdictions report human cases to the Centers for Disease Control and Prevention. Case records typically include supplemental information on possible sources of exposure, clinical course, and outcome. We reviewed data from all reported human plague cases during 1970–2017 that were characterized by a clinically compatible illness and presumptive or confirmatory laboratory evidence as defined previously (5). For this analysis, we considered only the primary clinical manifestation of illness.

We created a data extraction tool to capture details on patient–animal interactions in the 2 weeks preceding illness onset. If animal exposure had occurred, we classified the type of animal(s) involved as domestic or wild and the interactions as directly or indirectly associated with exposure to Y. pestis. We grouped animal exposures into categories based on authors’ judgment regarding risk for transmission. From high-risk to low-risk, the categories were animal bite, scratch, lick, or cough; skinning of a deceased animal; providing care to or handling a sick or deceased animal; co-sleeping; casual handling or touching; and other (walking, feeding, or contact type unspecified). If a patient had >1 animal interaction, the interaction recorded is that of the higher-risk category.

During 1970–2017, a total of 482 human plague cases were reported in the United States. Median case-patient age was 31 (range <1–94) years; 58% were male patients (Table 1). Bubonic plague was the predominant primary clinical manifestation of illness (n = 364, 76%), followed by septicemic plague (n = 91, 19%) and pneumonic plague (n = 15, 3%) (Table 1). Outcomes were known for 465 patients; 65 (14%) reportedly died from their illness.

Figure

Figure. Frequency of animal exposure among human plague cases, by decade, United States, 1970–2017.

Animal exposure that was plausibly related to plague transmission was identified in 258 (54%) records. The median case-patient age was greater among those with animal exposure (33 years) than those without animal exposure (24 years) (p<0.05). The frequency of known flea bite and mortality rate did not differ between patients with animal exposures and those without animal exposures (Table 1). After peaking in the 1980s, frequency of human plague decreased (Figure). However, the proportion of plague cases with animal exposure seemingly increased over time, from 52% in the years before 2000 to 63% since 2000 (p = 0.07) (Figure).

Of the 258 plague patients with animal exposures, 154 (60%) had contact with domestic animals before illness, including 121 with dogs and 102 with cats. The types of interactions included casual handling or touching (n = 55, 36%); co-sleeping (n = 31, 20%); caring for or handling a sick or dead animal (n = 29, 19%); bite, scratch, lick, or cough (n = 20, 13%); or other (n = 19, 12%) (Table 2). Among those with domestic animal contact, 65 (42%) had exposure to a domestic animal that brought home dead wild animals and 21 (14%) to a domestic animal with evidence of fleas.

A total of 134 (52%) patients had exposure to wild animals before illness. Common wild animal exposures were to sciurid rodents (e.g., squirrels, prairie dogs, gophers) (n = 58), lagomorphs (n = 50), other rodents (n = 40), wild carnivores (n = 15), and cervids (e.g., antelope, deer) (n = 9). Types of interactions identified were skinning (n = 54, 40%); handling a sick or dead animal (n = 37, 24%); casual handling or touching (n = 29, 22%); other type of contact (n = 12, 9%); and bite, scratch, lick, or cough (n = 2, 1%) (Table 2). Wild animal interactions were generally higher-risk, more direct exposures.

Pneumonic plague occurred more frequently among patients with animal exposure (n = 13, 5%) than among those without animal exposure (n = 2, 1%) (p<0.05); most patients had a history of contact with domestic animals (n = 11, 73%). Of 6 pneumonic plague cases associated with occupational exposures, 5 were among veterinarians or veterinary technicians providing care to plague-infected animals. The proportions of bubonic (n = 205, 77% vs. n = 162, 75%) and septicemic (n = 42, 16% vs. n = 49, 23%) cases were similar between patients with and without these exposures (Table 1).

More than 50% of patients in the United States with plague since 1970 had animal interaction that might have directly or indirectly led to their exposure to Y. pestis. Animals are associated with human plague transmission in varied ways, ranging from direct exposure, such as caring for a plague-infected animal, to more subtle indirect encounters with infected fleas, such as by co-sleeping with a flea-infested pet in an area with epizootic plague. Nearly all patients with pneumonic plague had animal interaction before illness, and several occurred in an occupational setting. Although the frequency of human plague in the United States has decreased, the proportion of human cases potentially related to animal exposure has concomitantly increased.

Cat-associated and wild animal–associated human plague have been documented in previous reports (6–9). More recently, the severity of plague illness in dogs and the role these animals might play in human plague have been recognized (10). A cluster of pneumonic plague in Colorado was linked to a dog with pneumonic plague (11), and a recent case of canine plague resulted in the potential exposure of >116 persons at a veterinary clinic (12). Gould et al. found that co-sleeping with a dog occurred more frequently among human plague case-patients than among neighborhood controls (13).

Limitations of our analysis include the possibility that human plague cases might have gone undiagnosed and thus were not captured. Our findings might underrepresent animal-associated plague because case records contain variable levels of detail. Thus, some patients might have had animal exposures that were not captured. In many instances, we could not determine which exposure contributed to human illness, if any at all. Therefore, this analysis is meant to describe the potential rather than definitive scope of animal-related human plague.

This report offers perspective on frequency and diversity of animal interaction as possible means of human exposure to Y. pestis in the United States. Given that most human plague worldwide is caused by flea bites, animal-associated prevention messages have been geared toward hunters and trappers, including the use of gloves when handling or skinning wild animals. Our findings highlight One Health–oriented opportunities to maximize plague prevention through communication with veterinarians in plague-endemic areas. Veterinarians play an integral role in plague prevention for animals and humans by increasing use of flea prevention products, promoting basic precautions among pet owners caring for sick pets, and encouraging use of appropriate personal protective equipment in the veterinary community.