Round two of my COVID-19 in animals summaries: Dogs

Are dogs susceptible to the SARS-CoV-2 virus?

Yes, but not very… maybe.  It depends what you mean by “susceptible.”

Nice and clear, eh?

There’s a difference between being infected and getting sick. Dogs can be infected by the SARS-CoV-2 virus (which is the virus that causes COVID-19 in people), but they don’t seem to be as susceptible as cats, and it’s debatable whether dogs get sick (more on that below).

Infection in dogs has been shown in a few different experimental studies, and through identification of infected pet dogs that were exposed to people with COVID-19.  In one small study, SARS-CoV-2 was detected by PCR in experimentally infected dogs, but the researchers could not isolate any “live” virus from the animals, suggesting the virus was present at a low level and the dogs were probably not infectious. The dogs remained healthy, but some developed antibodies against SARS-CoV-2, supporting the idea that they were truly infected and their immune systems responded accordingly. They did not pass to virus to other dogs with which they were co-housed. In the end, some or all of the exposed dogs got infected, but none got sick and they didn’t infect any other dogs.

Another experimental study yielded similar results, in that dogs were infected and mounted an antibody response, but didn’t get sick and were probably not infectious.

How often do dogs get infected with SARS-CoV-2?

We don’t know. Surveillance has been limited, so the scope of human-to-dog transmission isn’t clear. In Hong Kong, early in the pandemic, they quarantined pets of COVID-19 patients who could not care for them (e.g. owner lived alone and had to be hospitalized), and the pets were all tested at the quarantine facility. Hong Kong authorities identified SARS-CoV-2 in nasal, oral and/or rectal swabs from  2/15 dogs that were quarantined following exposure to their infected owners. Neither of the positive dogs had signs of infection, both developed antibodies to the virus, and gene sequencing of showed that the virus from the dogs was the same as that of their respective owners. Of particular note was they were able to isolate live virus from one of the dogs, which suggests the dog could have been infectious to others, at least briefly.

Additional data has been limited, in large part because it’s a logistical challenge to sample dogs in households with infected people during their isolation period. One small study in Spain didn’t detect SARS-CoV-2 in any of the 12 exposed dogs tested.  An investigation of pets from a cluster of infected and exposed veterinary students in France also failed to identify the virus in 12 other dogs, although it wasn’t clear how many of the dogs were actually exposed to an infected person.  A study from Italy reported no detection of the virus in 64 dogs from households with previous human COVID-19 infections, including 3 dogs that had respiratory disease.

Our Canadian study didn’t initially find the SARS-CoV-2 virus in any of 18 dogs (more to come on the expanded version).

There are still numerous reports of individual infected dogs from different countries. In the US, approximately 23 dogs have tested positive for the virus so far. That’s not a lot in the context of the dog population, but remember that not many dogs have been tested. Furthermore, testing has focused on looking for the virus by PCR. That will underestimate infections, because based on what we’ve seen so far there’s only a short window of time when you can get a positive PCR result from an infected dog. Dogs seem to only shed the virus for a few days after infection, so sampling dogs in infected households (after the people are no longer infectious and it’s safe to do so) runs the risk of a lot of false negatives simply based on the timing of sampling.

Studies looking at antibodies in dogs (and other animals) will be more informative, if the tests are accurate.  Antibodies are an indicator of past infection, and they tend to hang around significantly longer than the virus itself.  So unlike PCR-based surveillance, we don’t have to get into the household right away during the time of human illness – we can test dogs later to see if they were infected.

Not a lot has been reported yet on antibody testing (also called serology) in dogs. A study in Italy found antibodies to SARS-CoV-2 in 3.4% of dogs; 6/47 (14%) dogs from known-positive households, 1/7 (14%) dogs from households of suspected cases, and 2/133 (1.5%) dogs from other households. Whether the 1.5% prevalence in other dogs is from dogs that were infected by owners that were never diagnosed, or it represents the false positive rate of the test isn’t clear. A French study found antibodies in 2/13 (15%) exposed dogs and 0/22 dogs from households with no known cases of COVID-19.  Those results are similar to our preliminary 20% (2/10) prevalence in dogs from positive households in Canada so far. Obviously, we need to test a lot more dogs to get better estimates, and the study is ongoing.

Do dogs get sick from SARS-CoV-2?

That’s still unclear. I’d say that evidence is still far from convincing. There are a few poorly documented reports of sick dogs, but the question largely unanswered in those cases is “were they sick from infection with SARS-CoV-2, or were they sick with something else and coindicdentally happened to have been infected by this virus at the same time?” My guess is that disease is rare in dogs, but not impossible, especially in animals that may have other comorbidities that make them more prone to severe disease from many other pathogens as well.

Can dogs infect other animals or people with SARS-CoV-2?

Probably not, but that’s unclear too. Dogs are likely much lower risk that cats in terms of transmission. The fact that live virus was isolated from a dog at one point raises concern, because if there was live virus in the dog’s nose, you have to assume there was some risk of exposure to in-contact individuals. Whether the dog was shedding enough virus to actually infect someone is completely unknown. Lack of transmission in experimental studies isn’t a guarantee (because of the artificial environment and very small animal numbers) but provides more support of limited risk.

Overall, I’d say the risk of transmission of SARS-CoV-2 from dogs is very low. I don’t think we can say it’s zero, but I think it’s unlikely that a dog would pose a realistic risk.  That said, why chance it? If a dog is infected or at risk of being infected (i.e. living in a household with an infected person), it should be kept away from other people and pets. Dogs interact nose-to-nose and nose-to-bum a lot, and we have a lot of contact with their faces. We’ve seen transmission of other respiratory viruses between neighbouring dogs through fence-line contact, so keeping exposed dogs under control and away from others is reasonable and practical.

Could dogs be an important reservoir of SARS-CoV-2 once it’s controlled in people?

No. Dogs are not susceptible enough to the virus. For dogs to be a reservoir, they’d have to be able to keep spreading it dog-to-dog. That’s not going to happen because of the low susceptibility and short shedding time. You’d need a very large number of dogs in regular close contact to even begin to get a risk.

Could dogs be a bridge to transmit SARS-CoV-2 to wildlife?

Probably not, or at least they’re much less likely to be a bridge than cats. Their low susceptibility, short period of infection, limited (if any) infectivity and limited direct contact with wildlife mean the odds of them being infected by their owners and then infecting wildlife are pretty negligible.

So, we shouldn’t worry about COVID in dogs?

Worry, no. But, we should pay attention.

What should be done with dogs?

Do the same things recommended for cats:

  • If you are infected, try to stay away from animals – all animals, human and otherwise.
  • If your dog has been exposed, keep it inside and away from others.

Ultimately, dogs are part of the family – so if your family is being isolated, the cat needs to be a part of that.


Relax. This is almost exclusively a human virus. With a modicum of common sense, the risk posed from pets approaches zero.