Case reports of infections acquired from pets are sporadically found in the human medical literature. A common problem with these reports is the failure to look for true evidence of zoonotic transmission. The typical thought process is "We found this bug in a person, it’s most often found in animals, so this person must have gotten it from their pet." Often, this likely is actually the case (although some evidence would be nice). In other instances, like the Pasteurella bone infection described in the Orthopedics paper I wrote about recently, this assumption is probably completely off-base.

Something that is usually missing from these reports is any investigation of the suspected pet. Finding the bacterium in question in/on the pet, and showing that it’s the same strain as was found in the person goes a long way to supporting the conclusion that the pet was the source. It’s not a 100% guarantee, since you can’t say whether it went from pet-to-person or person-to-pet, but with a typically pet-associated bacterium its much more likely to have come from the pet, so finding the same strain in both pet and person is pretty solid evidence. Most case reports don’t bother even trying to get this much information (but they still get published…).

A case report in an upcoming edition of Zoonoses and Public Health (Register et al 2012) is another example of a study that provides some information about a potential pet-associated infection, but stops short of the type of proof that is needed. To their credit, the pet wasn’t available for testing and they are clear that it’s a "possible" case of zoonotic transmission, which puts them a step up on other studies, but it’s still too bad the additional information couldn’t be reported.

The case report describes an 11-year-old girl with cystic fibrosis (CF). People with CF are at greatly increased risk of respiratory infections, including infections caused by microorganisms that typically don’t cause disease in other people. This child had a sputum sample collected during a routine doctor’s visit, when she didn’t have any signs of respiratory infection. Bordetella bronchiseptica, a bacterium most often found in dogs and cats (and one of the causes of kennel cough (aka canine infectious respiratory disease syndrome) in dogs), was isolated from the sample. When this was explored further, it was revealed that child’s family had obtained a new kitten three weeks earlier, and the kitten had signs of respiratory disease.

Unfortunately, the kitten had been removed from the household by the time B. bronchiseptica was diagnosed in the girl, so it couldn’t be tested. (They don’t say why the family got rid of the cat.) However, the authors at least took it a step further and did some detailed molecular characterization of the bacterial strain they recovered from the sputum sample. Different genetic traits suggested that the strain was feline in origin, and, combined with the fact that the child had contact with the kitten, this provided a little more support to the hypothesis that the cat was the source. It’s not proof, but still interesting. They also determined that the B. bronchiseptica strain was missing a gene that’s associated with helping cause disease, suggesting it’s not as able to make people sick. That might explain why the child was carrying the strain in the absence of disease, although people can carry lots of different bacteria that can potentially cause disease without being clinically ill.

In the end, it was an interesting little report and the authors conclude with a couple of good statements:

"Acquisition of detailed clinical and epidemiological data paired with discriminatory genetic comparison of case isolates and contact isolates is needed to more firmly establish transmission patterns and identify likely contact risks." (A fancier way to say what I said above.)

"…it seems advisable to counsel CF patients regarding adherence to practices that minimize opportunities for zoonotic transmission of B. bronchiseptica from family pets or other potentially infected animals."