Background
Pet ownership is common. Although the proportion varies by continent and country, studies indicate that in most countries the majority of households own pets [
1‐
8]. Cats and dogs are the most frequently owned pets, but other species are often reported. A recent study estimated that 56% of Canadian homes have at least one dog or cat, with a minority owning fish (12%), birds (5%), rabbits or hamsters (each 2%), lizards, guinea pigs, snakes, frogs, turtles, ferrets, or gerbils (each 1%) [
6]. In addition, people may come into contact with animals outside of their homes such as in public settings (e.g., petting zoos, schools, fairs), through work-related activities, or recreation (e.g., parks), yet little is known about the frequency and nature of this type of animal contact. For the most part, animal ownership and interaction is not discouraged by healthcare professionals, as numerous studies have confirmed the mental and physical benefits of pet ownership and companionship [
9], particularly among children, the elderly and immunocompromised individuals [
10‐
16]. Thus, it is not surprising that several studies have indicated that homes with children are more likely to have one or more pets than those without children [
3,
17‐
19], and animal ownership practices of immunocompromised individuals appear to be similar to those of the general public [
15,
20,
21]. Despite these benefits, there are also potential health hazards associated with pet ownership and contact.
Animal bites are a serious public health problem, with an estimated 316,000 U.S. emergency room visits for a dog bite-related injury in 2008 (rate: 104 visits per 100,000 population) [
22]. Dogs are responsible for approximately 80% of all bites, cats accounting for less than 20% and other pet species and wildlife responsible for the remainder [
23]. Bites may lead to painful disfiguring wounds, infection, altered function of the affected area, and rarely death. Animal scratches can also have important physical consequences depending on the area of the body affected. In addition to the physical damage of bites and scratches, pets can transmit pathogens to people (zoonoses). People can acquire pet-associated zoonotic organisms through the skin and mucous membranes (via animal bites, scratches, or other direct contact), contact with animal saliva, urine and other body fluids or secretions, ingestion of animal fecal material, inhalation of infectious aerosols or droplets, and through arthropods and other invertebrate vectors [
24]. Although any exposed person can become infected with a zoonotic pathogen, risks are particularly high for those with a compromised or incompletely developed immune system, such as the young (< 5 yrs), elderly (≥ 65 yrs), pregnant and those with immune function-reducing conditions or treatments, including diabetes, cancer, infection with human immunodeficiency virus (HIV), splenectomy and radiation therapy [
24,
25]. The increased disease risk for children is additionally imparted through closer physical contact with household animals, reduced hand hygiene and behaviours that include pica and exploration of the environment through mouthing. Not only are these groups at increased risk for infection with a zoonotic pathogen, infection with many zoonotic pathogens is more likely to result in severe disease in high risk groups [
24,
25]. Pets often have frequent, close interactions with household members, such as licking of hands and sleeping in beds [
26], which can further increase pet-associated disease risks.
Many of the disease risks that occur with pet contact can be eliminated or reduced through simple measures, such as hand hygiene, proper animal husbandry and altered animal-contact behaviours. In order for infection prevention to be successful, however, individuals in contact with animals must be aware of these disease risks. Thus, awareness of zoonotic disease risk is a prerequisite for effective disease prevention. To-date, few studies have evaluated the general public’s knowledge of pet-associated zoonoses [
27,
28]. A study conducted in 1986 revealed that many individuals in the general population lacked even basic knowledge about zoonotic diseases (e.g., only 63% of household heads in De Kalb County, Georgia, USA believed pathogens from pets could be transmitted to humans) [
28]. A more recent study (2008) supported the earlier study’s findings of poor zoonotic disease knowledge by the general public (e.g., only 56% of dog owners in Brazos County, Texas, USA knew intestinal helminths could be transmitted from dogs to humans) [
27]. Both studies were limited in their geographic scope and did not integrate human disease risk status (i.e., extremes of age, immunocompromised), animal contact-related attitudes, and zoonotic disease knowledge and risks. In addition, most previous studies have only measured animal contact through pet ownership and did not include animal contact outside of the household. The objective of this study was to further characterize the general public’s household knowledge, attitudes and risks related to pet ownership and animal contact. Furthermore, this study aimed to determine if those households with individuals at increased risk of disease differ in knowledge, attitudes, or practices from the remainder of the public.
Discussion
This study aimed to characterize household knowledge, attitudes and risks related to pet ownership and animal contact in Ontario, Canada. Despite the importance of this topic due to the potential injury and disease risks posed by pets, the high proportion of households that own pets, and the close interaction pets often have with household members, few studies [
27,
28] have evaluated this area. Furthermore, this study is unique in that it broadly addresses animal contact within and outside the home, and integrates household demographics, including human disease risk status (i.e., extremes of age, immunocompromised), animal contact-related attitudes and zoonotic disease knowledge and risks.
The animal ownership patterns we observed were consistent with those previously reported for the surrounding area [
19] and country [
6], with over 63% of households having one or more pets. Similar to a previous study [
33], over 75% of respondent households had one or more members that had frequent animal contact through household pets, animals outside the home, or both. These results highlight the common occurrence of direct animal exposure for the public. The perceived benefits from pet ownership and contact were evident as both pet and non-pet owners believed that the benefits of pet ownership outweigh disease risks.
In order to assess zoonotic disease knowledge, respondents were provided a list of infectious pathogens, diseases and syndromes and asked to indicate which were transmitted from pets to people. The decision as to which of the listed pathogens, diseases and syndromes were pet-associated was based on the best available knowledge at the time the survey was administered. Pathogens that could be transmitted from pets to people, even though the majority of such human infections are acquired from non-pet sources (i.e.,
Salmonella Giardia, MRSA) were considered pet-associated zoonoses. Hepatitis was not considered pet associated, as Hepatitis A, B and C are non-zoonotic and Hepatitis E has only recently been thought to be transmitted from animals and the scientific community has yet to resolve if pets play any role in human infection [
34,
35].
We found considerable variation in the respondents’ ability to correctly classify different pathogens. As has been previously reported [
27,
28], the public appears to be well-informed of the potential for rabies transmission from pets (95% answered correctly), however, is less informed about less severe but more common pathogens, diseases and syndromes, such as ringworm,
Salmonella and infectious diarrhea (53%, 33%, 24% answered correctly, respectively). Equally alarming, were the human or pet-specific diseases believed to be transmissible between species (i.e., HIV/AIDS, distemper, each 10%). As awareness of zoonotic disease risk is a prerequisite for effective prevention, the limited zoonotic disease knowledge of the public is a concern. Despite an increased awareness by researchers over the past several decades of the wide scope and magnitude of zoonotic diseases, the public’s knowledge appears to have changed little; our findings are consistent with those of a 1986 study [
28] that reported similar proportions for the public’s correct classification of rabies, ringworm and infectious diarrhea as pet-associated zoonoses.
The low zoonotic disease awareness observed by respondents was perhaps not surprising as less than one-third of respondents reported having ever received information about pet-associated diseases or precautions to reduce the risk of these diseases. As previously noted [
27,
36], veterinarians and the Internet were most frequently reported as providing this information to pet and non-pet owners, respectively. Only 25% of pet owners recalled ever being asked by a physician or their staff if they owned pets. The limited involvement of physicians and public health was not surprising. Several studies have indicated that physicians often rely on veterinarians for advising the public about the potential for zoonotic disease and thus discuss this topic with their patients less frequently than veterinarians [
37,
38]. Although physicians believe educating patients about pet-associated health hazards is important, time constraints and competing health messages were often cited for not doing so [
39]. Similarly, although veterinarians were an important source for zoonotic disease information in this study, there appear to have been many missed opportunities by the profession. Analogous to a previous study [
27], only 27% of the individuals who had been to a veterinarian in the past year indicated they had ever received zoonotic disease information from a veterinarian.
Although statistically significant, the mean zoonotic disease knowledge score of pet owners was only marginally greater than non-pet owners (6.5, 6.2 respectively). The similar scores between the two groups implies that the additional resources available to pet owners (e.g., veterinarians, targeted education provided by physicians and public health) may not effectively provide a large amount of additional knowledge. Further efforts by these groups, such as readily available educational materials on this topic provided in waiting rooms and during office visits are important. A recent study reported that 43% of surveyed veterinarians did not have client educational materials on zoonotic diseases available in their practices [
40]. Due to the ever-increasing use of the Internet for personal and animal health information [
41], it is critical to ensure reliable resources are also available on-line.
From our results, it was evident that both the human and veterinary fields have room for improvement in patient/client education. Furthermore, increased communication between professions is needed to improve overall zoonotic disease knowledge and develop optimal approaches for reducing pet-associated pathogen transmission and injury in households. These conclusions are supported by previous studies that indicate that the majority of veterinarians and physicians do not regularly discuss zoonotic disease risks with clients, patients or each other [
37,
38,
40,
42]. The utility of such discussions was evident in our study, as respondents who recalled having received zoonotic disease information in the past had a higher mean zoonotic disease knowledge score compared to those who had not. While there was no assessment of the type, timing and quality of information that was provided, this finding suggests that general zoonotic disease counselling can have a positive impact.
Pet and non-pet owners were, for the most part, not concerned about pet-associated zoonoses and were comfortable with their level of knowledge and methods to reduce zoonotic disease risk. As individuals who are not concerned, and are comfortable with their knowledge-base may be unlikely to seek additional knowledge from available resources, active methods may be required to improve awareness of pet-associated zoonoses. Such methods may include brief waiting room surveys to assess disease knowledge and high-risk behaviours, followed by improved physician and veterinarian-directed review of pet-associated zoonoses with patients/clients. For pet owners, a significant positive association was observed between an individual’s zoonotic disease score and their level of comfort with their knowledge on this subject. Thus, for pet owners, inquiring about their level of comfort with this topic may be a reliable screening tool for identifying those in greatest need of additional education.
Pet ownership was common in households with individuals at higher risk of infections (e.g., < 5 yrs, ≥ 65 yrs, immunocompromised). Based on the limited studies that have previously surveyed this topic for the immunocompromised [
15,
20] and extremes of age [
19,
33,
43,
44], this finding was expected, and given the positive aspects of pet contact, this in itself was not particularly concerning. However, it was concerning that households with higher risk individuals did not differ from the remaining households regarding their perceived risk of pet-associated disease, zoonotic disease knowledge score, recalling being asked by a medical provider if they owned any pets, or recalling having received information regarding pet-associated disease risks and preventive measures. This suggests a troubling (but perhaps unsurprising) lack of recognition or knowledge of the potential for pet-associated zoonoses. A history of contact with pets in the home and animal contact outside the home should be part of every physician’s wellness evaluation, especially for individuals at higher risk of zoonotic disease. Primarily relying on veterinarians for providing pet-associated disease information is problematic for several reasons. As noted by our study, animal contact frequently occurred in both pet and non-pet owning households; many individuals with animal contact in the latter group will likely not consult a veterinarian. In addition, veterinarians are often unaware of their clients’ immune status [
37,
40] and arguably other attributes of higher risk households (such as extremes of age). Without this information, veterinarians are unable to adequately inform and council clients on their household disease risks. Veterinarians may also be reluctant to venture into aspects of human health.
A relatively low proportion (4%) of households reported having had a family member acquire a disease from a pet. The accuracy of this estimate is unclear, since it is prone to recall bias and various other factors that make it difficult to determine if an infection was truly acquired from a pet (e.g., multiple routes of transmission for many zoonotic pathogens, subclinical shedding of pathogens by pets). The potential risk of zoonotic infection and injury, however, were clearly noted in our study with a high proportion (27%) of households reporting one or more dog or cat-derived bites or scratches during the previous 12 months. This result is alarming, as such injuries can have serious health consequences. Previous studies vary widely in the reported incidence of animal bites, with most variation likely due to the population sampled and sampling methods used [
23]. A U.S. study that used a telephone survey estimated a national annual incidence of self-reported dog bites of 1.8% (18 per 1000 population) [
45]; by far the highest incidence reported to-date [
23]. Our study found 1.6% of all household members sustained a self-reported dog bite during the previous 12 months. The high incidence of dog bites in our sample is unlikely to be due to differences in case definitions or sampling bias as we requested data only on bites that broke the skin; the proportions of pet ownership and presence of children in the households, two commonly reported risk factors for dog bites [
23], were similar to previous reports and census data for the region. This finding warrants further evaluation by additional studies to confirm the elevated rate and determine potential risk factors. In the interim, educational resources addressing safe pet interaction and bite/scratch prevention strategies are strongly needed for this population.
We acquired our data as a convenience sample from the waiting rooms of general practice physician offices in Ontario, Canada. All Canadian residents receive medically necessary healthcare services at no charge [
30], reducing the potential that variable access to physicians would result in a biased study population. Despite this, it is possible our source population or derived sample were not representative of the surrounding general population. Barriers in access to health care services that may disproportionately affect different groups, such as new immigrants [
46], over-representation of groups typically in need of increased health care visits, such as the young, elderly, or immunocompromised, or use of a non-randomized sampling approach were potential concerns. However, based on census data, our sample appeared to be representative of the region, with the exception of a lower percentage of visible minorities and over representation of higher income households. The finding that 59% of households had one or more individuals at higher risk of infectious disease was reasonable as the proportions of our sample < 5 yrs (8%) and ≥ 65 yrs (8%) were consistent with census data for the region (6% and 12%, respectively); the proportion of households with immunocompromised members (32%) may be expected based on estimates for the United States [
47] and the expected prevalence of immunocompromising conditions, such as diabetes [
48], for the population in the Waterloo region.
Bias may have been introduced into our study. Possible sources of bias include differences between respondents and non-respondents (however the high response proportion of 75% makes this less likely) and the self-administered nature of the questionnaire where questions may have been misinterpreted (misclassification). In addition, respondents’ ability to recall information or activities typically deteriorates as time elapses, with better recall for more recent experiences. For this reason, it is possible the reported medical interactions (i.e., ever asked by a physician if they owned pets, ever received zoonotic disease information from a physician) are more reflective of the staff at the two surveyed clinics than other medical staff used by respondents previously. Due to the descriptive nature of the study, analysis relied primarily on univariable statistics, not accounting for possible correlation between the variables. Furthermore, pet ownership was associated with several demographics, as previously reported [
27], that may have confounded the observed associations between pet ownership and attitudes, knowledge, or behaviours. Since this was a cross-sectional study, there are no data on the sequence of events relating to variables such as knowledge score.
We queried if respondents had pets in their household at the time of completing the survey and analyzed pet ownership as a dichotomous variable, not accounting for different pet species. Based on responses, it is likely that some of the non-pet owners previously had pets (i.e., 29% of non-pet owners who had previously received zoonotic disease information, had received this information from a veterinarian). It is possible this classification scheme may have biased our results, likely falsely increasing the zoonotic disease knowledge score attributed to non-pet owners and reducing apparent differences between pet and non-pet owners. Thus, differences between pet-owning households and those that have never owned pets may be more pronounced than we observed. In our analysis, we did not account for the ownership of different pet species. Different species have varying levels of disease risk for particular pathogens (e.g., high prevalence of Salmonella shedding by reptiles and amphibians), and owner demographics and other characteristics may vary by the species owned. Thus, it is possible that zoonotic disease knowledge, attitudes and practices may vary by species owned, and by classifying households simply as pet and non-pet owners systematic (bias) or nonsystematic error may have been introduced. However, since cats or dogs were owned by 93% of the pet owners and other species were infrequently reported, any such error was likely minimal.
Authors' contributions
All authors participated in the study concept, design and questionnaire development. JWS enrolled participating clinics and administered questionnaires to participating patients. JWS was responsible for data analysis, data interpretation and manuscript preparation. All authors provided input on data analysis, interpretation and final manuscript development. All authors have approved the final version of the manuscript.