The USDA has issued announced the first confirmed case of SARS-CoV-2 infection in dog in the US. The affected dog was a German shepherd from New York state, and  its owner had COVID-19. Interestingly, the dog had respiratory disease,  something we haven’t seen reported in dogs. It tested positive by PCR and antibody testing. Another dog in the household was healthy, but also had antibodies against SARS-CoV-2, indicating it had been infected too.

This one is noteworthy for a few reasons.

  • The dog was sick. It’s not guaranteed that the dog had COVID-19. It could have had some other disease and more information may come to clarify that. However, if this dog was sick because of SARS-CoV-2 infection, it would run contrary to evidence to date that has indicated dogs can be infected but are not likely to get sick.
  • Both dogs were infected. While the other dog didn’t get sick, it was infected. We’ve gone on the assumption that human-dog transmission is uncommon, and that still might be the case, but finding transmission to multiple dogs in a household is interesting.

Overall, it doesn’t change our main talking points much, and it highlights the need for more study.

A new experimental study in cats (Bosco-Lauth et al) re-inforces information from earlier studies and provides some important new insights. The standard disclaimer that the information is from a pre-print (non-peer-reviewed) paper applies, but the science seems pretty sound.

For this study, seven adult cats and three adult dogs were studied.

Cats…group 1

  • One group of 3 cats infected intra-nasally and then tested on days 1-5, 7, 10 and 14 after infection, looking for the virus. Blood was collected on days 7, 14, 21, 28, 35 and 42, looking for an antibody response.
  • This group was then re-exposed to the virus on day 28, with samples for virus detection collected 1, 3, 5, 7 and 10 days after 2nd exposure.
  • None of the cats got sick.
  • Cats shed the virus for up to 5 days after exposure, with peak shedding at day 3.
  • Nasal viral levels (from nasal flush samples) were higher than oral viral levels. That’s important information for surveillance studies.
  • Fecal samples don’t seem to have been collected. That’s unfortunate since fecal shedding of virus seems to be a potential issue worth exploring more.
  • Infected cats developed detectable antibodies as early as day 7, with all cats reaching a threshold titres by day 14. Antibody titres stayed stable or increased between days 28 and 42.
  • After re-exposure, a moderate increase in antibody titres was noted in the 14d after exposure. No viral shedding was noted after re-exposure.

Cats….group 2

  • Two cats were infected, then mixed with two unexposed cats 48h after infection. Exposed cats shed virus like Group 1 cats.
  • Interestingly, the other cats started shedding virus within 24h of being housed with infected cats, but had a more prolonged shedding period, with peak shedding occurring at 7 days post-exposure.
  • Experimentally infected cats and those infected via contact all developed an antibody response.


  • Three dogs were exposed and tested as per the cats in group 1. None developed any signs of disease and viral shedding was not detected. Antibody response wasn’t evaluated.

The short duration of shedding by infected cats is encouraging (although a bit different from some natural infections that have been followed). However, the longer duration in the in-contact (vs directly infected) cats may be more relevant to a natural exposure situation. Shorter duration reduces the risk of passing the virus onto another animal or person. The relatively short period of shedding also shows how active surveillance studies like we’re doing, where we sample pets of infected people, can underestimate transmission because of the narrow window of time that infected animals may shed the virus. This show again why antibody testing will be useful to characterize how much human-pet transmission has occurred.

Resistance to repeated viral shedding after second exposure is also encouraging, as it suggests that previous infection provides protection. There are many caveats, such as the very small sample size, experimental nature of infection and short interval between first infection and re-exposure, but it’s a start and provides more hope that people are resistant (to at least some degree, and for at least a while) after infection.

The consistent production of antibodies and their presence over at least 42 weeks is good for future antibody surveillance studies that can look back at previous infection. It’s also encouraging from an immunity standpoint since rapidly disappearing titres would suggest less of a protective effect from subsequent infection.

The lack of disease in this small group of cats is a bit different than other studies and we can’t say much about disease based on a small experimental study, since it seems like disease can occur in naturally exposed cats. This study shows that not all exposed cats get sick, as expected. Experimental studies can provide useful information but we have to be careful extrapolating too much to the real world situation, where animal health status, other risk factors and different types of exposure can occur.

As SARS-CoV-2 reminded everyone, there are lots of undiscovered viruses out there, and some of them can cause disease in humans, animals, or both. It’s not very hard to find a new virus, actually. Figuring out what it means is often the challenge, because simply finding a virus in a sick person or animal doesn’t mean that virus is the cause of the illness. Our bodies contain many completely harmless viruses, and viral discovery is one field in which information advances much more rapidly than knowledge.

A new feline virus was found during investigation of an  outbreak of vomiting and diarrhea in cats three animal shelters in British Columbia (Canada) between Nov 2018 and Jan 2019 (Li et al. Viruses 2020). The outbreak started with vomiting and diarrhea in 8 of 12 cats in the adoption room of one shelter (Shelter 2). The first affected cats had come from another shelter (Shelter 1) and were sick around the time of transfer. Upon further investigation, they found 11 sick cats at Shelter 1. Then, they found 13 affected cats at Shelter 3, after a cat from Shelter 2 was transferred there (see image below). No clear method of transmission was apparent, as most affected cats had not been housed with another (known) affected cat.

They did a very nice job of contact tracing (which is often very difficult in a shelter situation) and characterizing the disease. It was estimated the incubation period could have been as short as 24 hours, and as long as 5-7 days. Vomiting usually started 1-2 days before diarrhea, and the median duration of illness was 4 days.

Usually, outbreaks like this aren’t investigated too carefully because of cost and the low likelihood that something relevant will be found. Here, they used sequence-based methods to sequence all the viral bits found in fecal samples from the cats. This makes it possible to assemble most or all of the genomes of a range of viruses that are present. In doing so, they found a new virus, which they named “fechavirus”, in 8/17 affected cats. It is a type of chapparvovirus, and genetically it’s most closely related to canine cachavirus (a fairly recently discovered virus of questionable relevance). They also found three different bocaviruses in 9/17 cats. Other testing was unremarkable – a few different potential causes of gastrointestinal disease were found in some cats, but nothing convincing.

Based on the timing of infection and disease, the investigators suspected that this new virus was the cause of the outbreak, but the hard part is confirming it. Finding a new virus is much easier than figuring out its clinical relevance, and that can take time. Various “new causes of disease” have been found in many different species, but many are subsequently determined to be only minor players or irrelevant, so we should take this report as an indication to study this virus further. The fechavirus was found in less than 50% of the affected cats. We don’t expect to find the causative pathogen in all affected individuals during an outbreak, but this rate of detection raises more questions. Maybe it’s because the virus isn’t shed for very long by affected animals. Maybe it’s because the testing isn’t very sensitive. Maybe it’s because it wasn’t actually the cause of the illness and it was just an incidental finding.

Importantly, there was no study of healthy cats in the shelters or otherwise to test how commonly fechavirus is found in the general cat population. So care needs to be taken to not over-interpret the results. (Anyone remember canine circovirus? It got lots of attention (to the point of paranoia) when it was found in sick dogs, but we eventually figured out it’s a minor pathogen, at best).

The big questions…..

Is this virus an important, emerging cause of disease in cats?

Is it a rare virus that can cause sporadic severe disease?

Is it a virus that can contribute to disease only when another infection is present?

Is it a harmless virus that’s common in healthy cats and was found incidentally?

Time (and study) will tell, but it’s still important to know about these “new” viruses so we know what to investigate.

A few quick updates on some recent SARS-CoV-2-related stories.

North Carolina dog: Positive result not confirmed

This case, a pug in North Carolina that had an oral swab that supposedly tested positive for SARS-CoV-2 as part of a Duke University household surveillance study, was reported a few weeks ago.  It was strange that there was no confirmation of the test result since then. Now we know why.  It appears the original test result was actually “inconclusive” (i.e. not strong enough to be truly considered positive). Follow up PCR testing was negative. That in itself doesn’t mean the dog wasn’t infected, since it could have been a short term infection that wasn’t sampled during peak shedding. However, no antibodies were detected in the dog. That indicates the immune system didn’t recognize the presence of the virus. It’s not completely definitive, but supports this not being a true infection.  Perhaps transient contamination of the dog’s mouth with virus from its infected owners could have caused the original inconclusive test result.

Hydroxychloroquine/chloroquine study debate

This topic is a bit outside the animal health-related area, but is still interesting (and relates to some of my recent Twitter ranting about the state of scientific publication). The high profile Lancet paper that reported increased deaths associated with hydoxychloroquine use in COVID-19 patients, and led to WHO suspending that arm of its study, has been challenged because of numerous concerns about data and data availability. It doesn’t mean the results are necessarily wrong, but questions about the data mean that things need to be clarified, which the authors have apparently been reluctant to do so far. An open  letter to the journal outlines various problems with the report and has a large and reputable list of signatories.

Tiger SARS-CoV-2 whole genome sequencing

There’s not really anything notable here in the big picture, but anyone with an inclination towards whole genome sequencing data might be interested in the sequence results from the virus isolated from one of the Bronx Zoo tigers.

SARS-CoV-2 in a cat in Russia

Just one more report of a cat with SARS-CoV-2 infection, presumably from its infected owner. Not surprising.

I wrote about issues with N95 respirators with exhalation valves the other day, and decided to do a quick demonstration of the concerns.

Exhalation valves on some N95 masks are designed to make it easier to breathe out, because these one-way valves release exhaled air without forcing it through a filter.

  • The mask still protects the wearer from breathing things in, but it does very little to prevent an infected person from spreading infectious droplets when they breathe out.
  • That’s a big problem when the mask is meant to protect others FROM the wearer, which is why masks are recommended outside of specific healthcare situations in the first place.
  • In particular, cloth masks are becoming widely used in these situations, as they’re meant to protect others. They reduce the risk by containing the wearer’s respiratory droplets within the mask.

However, I’ve seen ads for cloth masks with valves (see picture right). That’s a bit like someone marketing an umbrella that is less likely to get caught by the wind because they’ve cut big holes in it.

To make the concept a bit more visual, take a look at the picture below. I put on a cloth mask, a surgical mask, an N95 mask and N95 mask with an exhalation valve. I then held a culture plate in front of my face and exhaled forcefully and coughed. As you can see, there was no bacterial growth, except from the mask with a valve (where there was a lot).

So, unless you are using an N95 mask during a high risk healthcare procedure (where the only concern is what you’re inhaling, and not what you’re exhaling), don’t use a mask with a valve.

One concept that we’ve recommended for COVID-19 control in veterinary clinics is staff cohorting. That involves keeping staff groups together to limit the risk of transmission should someone be infected. If groups (i.e. shifts, or teams that stick together and don’t interact with others) are formed, any single infected person would have contact with a smaller number of other people, thus reducing the risk of disease overall and protecting the clinic by making it less likely everyone would be sent home to self-isolate for 14 days at the same time.

In the ideal world, we’d strictly cohort permanently.  However, cohorting can be very difficult, depending on staffing, clinic layout and clinic operations. It can cause major scheduling hassles, limit the amount of patient care, and interfere with work in the clinic in general. Like a lot of infection control practices, the cost-benefit has to be considered, and in many clinics, optimal cohorting isn’t sustainable.  But that doesn’t mean the concept should be dropped completely. The goal should be to cohort as much as possible. There is increased risk with decreased cohorting. However, some of that risk can be offset by making sure other things are done really well.

Ultimately, it comes down to two key concepts (as for control of many infectious diseases): keeping the virus out of the clinic, and limiting its ability to spread if it sneaks in.

Keeping SARS-CoV-2 out of the clinic:

Restricting clinic access

Curbside drop off and pick up (including animals and products like food and medications) is very common now.  Anything that keeps clients out of the clinic will reduce risk, as the fewer human-human contacts occur, the lower the risk of exposure to the virus. If only staff come into the clinic, there’s less risk of introduction.


Hopefully everyone’s gotten the point that coming to work when you’re sick isn’t showing you’re a dedicated employee – it’s showing you’re irresponsible. All clinic staff should self-screen every day. Basically, that means paying attention to their own health and not coming to work if they have anything that could be suggestive of COVID-19 (including decreased smell and taste). It’s not a guarantee that no one who’s infected will come to work, because some may have asymptomatic infections, but keeping actively sick people away is still a huge factor.

The demise of the “waiting room”

This will likely apply to many professions, like dentistry and human medicine. Having people congregate in a small waiting area is not likely to be acceptable, possibly ever again. (It never made sense to me in human medicine anyway – lots of people, including sick people, crammed into a waiting area is just a recipe for disease transmission). Rethinking how reception areas are used is good for the long term. Rather than “waiting rooms” they may ultimately be “check-in” and “check-out” areas, where there’s quick, one-way flow on the way into or out of an appointment (for the subset of situations where an owner has to be in the clinic). Check-in by phone, direct admission to an exam room, waiting outside, dropping off the pet and coming back later, and similar strategies can reduce the need for people to hang around together in a small room, and in turn reduce the risk to themselves and others.

The end of walk-ins

This will probably be a common theme in many professions too. Unexpected arrivals disrupt order and measures to carefully schedule how and when people arrive. That doesn’t mean someone can’t spontaneously decide they should swing by the clinic to pick up a bag of pet food, it just means they have to do it differently. That could simply be calling and saying “I’d like to stop by to get some food, I can be there in 5 minutes.” Staff can then have the food waiting on the doorstep, or know you’re coming in at that time so they can make sure the area is clear. Or, they can say “How about in 10 minutes? We’ll have space then.” Yes, it’s still a bit of a disruption, but it’s minor.

Reducing the risk of in-clinic exposure to SARS-CoV-2:


If an infected person is in the clinic but they are wearing a mask, the risk of them transmitting this virus is lower. Routine cloth mask use whenever a 6-foot gap can’t be maintained between people is emerging as a key infection control tool. Cloth masks are far from perfect, but they can do a good job containing most infectious droplets, which are probably the main source of exposure.


Fewer people = less exposure. Working from home needs to be considered whenever possible. That can include tasks like management, completing medical records and calling owners. It can also include telemedicine. Even in situations where an animal has to be seen in person, telemedicine can reduce the duration of contact. For example, for a new puppy appointment, we’d like to have a detailed discussion of a variety of issues, then do an exam and usually vaccinate. We can do the discussion part by telemedicine (with the vet at home), so it just needs to be followed up by a quick appointment for the exam and vaccination. The owner doesn’t need to be there, so by covering the discussion topics first, an owner-less visit or short owner visit can be achieved without compromising animal care.

Clinic flow

Some things can be done to reduce contacts between staff in the clinic as well. We need to be within 6 ft of others at times (e.g. placing a catheter) but tweaks to clinic flow and operations can reduce the likelihood of crossing paths with someone, or having to work in close proximity, or the duration of time that needs to be spent in close proximity. Appropriate measures to do this will vary greatly between clinics, but there are a variety of things that can be done with both procedures and layout (e.g. moving tables, changing office space).

How will these changes be received by staff, and by clients? Some people will complain about anything new. However, there’s enough awareness now that these measures won’t likely look odd to the average person. If anything, I think people are more likely to raise concerns about failure to take steps like this, since they will be the norm in many situations outside of veterinary medicine too. We’re less likely to hear “my vet is doing some really strange things at their clinic”. Rather, if we don’t take reasonable measures, we’re likely to hear “why isn’t my vet doing the things my doctor, dentist, physiotherapist and everyone else is doing?” In addition to creating risk, failure to act may actually drive clients away.

As described by ProMedMail, the Dutch Agriculture Minister has provided another update on the outbreaks of SARS-CoV-2 that have affected at least 5 mink farms in the Netherlands to date (click here for the original Dutch version of the letter).

Another suspected mink-to-human transmission of SARS-CoV-2 has been identified (with potentially infections in  an additional two mink farm staff). These cases appear to be from a different farm than the first suspected mink-to-human infection. The route of transmission is presumed to be mink-to-human based on the gene sequences (and the illness in the mink preceding infection in the people). The sequence data I saw earlier seem consistent with that, but it’s hard to be 100% certain.

There’s also some more information about barn cats. On mink farms, cats would rarely have direct contact with mink (because mink would try to eat any part of a cat that was within reach), but the cats would have access to mink manure, which typically falls from wire cage flooring to the ground below. They have now identified antibodies against SARS-CoV-2 from 7/24 cats on one farm, indicating the cats were previously infection. The virus itself was also found in the samples from one cat, indicating it likely still had an active infection. Whether all 7 cats got infected from the mink (or mink manure), or whether there was subsequent cat-to-cat transmission will be pretty much impossible to figure out at this point.

While a lot still needs to be determined with these outbreaks, information to date highlights some important themes:

  • SARS-CoV-2 is predominantly a human virus but it can spill into other animals.
  • While most transmission is human-to-human, some infected animals can send the virus back to people, and infect other animals. (That shouldn’t come as a surprise, although sadly the One Health response to this virus has been pretty disappointing.)
  • Keeping infected people away from animals, as well as away from other people, is important. It’s better to prevent human-to-animal infection than to have to figure out how to deal with infected animals and worry about spread into wildlife.
  • Reducing the number and closeness of interactions, be they human-to-human, human-to-animal or animal-to-animal (within reason) and using practical precautions when distancing can’t be maintained (e.g. masks, gloves and other protective equipment when handling animals in high risk situations) are the key control measures for this virus.

The fact that there are multiple affected farms in the Netherlands but no reports elsewhere needs to be considered. It’s unlikely Dutch mink farmers are more likely to be infected or have closer contact with their mink. There’s reluctance in some countries to consider or test for infection with SARS-CoV-2 in animals, so whether this is a uniquely Dutch situation or a more common problem that’s not been diagnosed or reported elsewhere remains a question. Hopefully mink farmers everywhere are paying attention to this situation and implementing some control measures. It’s tough to use really good infection control practices in some of these facilities, considering how mink farms are managed and how many animals may need to be handled on a given day (e.g. when thousands of mink are being vaccinated), but measures to reduce human-mink contact whenever possible, use appropriate PPE, identify problems early and keep wildlife (and cats) away from mink barns are important.

COVID-19 derailed our plans for some backyard chicken work (e.g. research and education) this spring, but the emergence of COVID-19 doesn’t mean all other infectious disease issues have disappeared. Some problems will be reduced by the precautions put in place to control COVID-19, but other problems may actually get worse. Backyard chickens continue to be popular and I anecdotally may actually be more common now, at least in some areas, as people spend more time at home (and others worry (unnecessarily) about ongoing access to eggs and chicken at grocery stores).

I’m not anti-backyard chickens. I’m anti-“spending the weekend on the toilet” and anti-“seeing people hospitalized unnecessarily” and, I guess, just anti-Salmonella and anti-Campylobacter in general. I can’t see any redeeming qualities of those bacteria, at least in people.

That’s a rambling lead-in to a CDC investigation notice about Salmonella Hadar infections linked to backyard chickens. As always, these investigations markedly underestimate the scope of the outbreak, since most people who get sick don’t get tested, and chicken-associated infections with other strains that don’t cause enough widespread disease to get tracked don’t get any attention.

Regardless, it’s a reminder that this remains a significant problem.  As of the May 20, 2020 update:

  • 97 people had been diagnosed with the outbreak strain, with disease starting between February 26 and May 1 (see graph below).
  • People from 28 states have become sick (see map below).
  • 34% were hospitalized. None died.
  • 30% were kids younger than 5 years of age.
  • It looks like 4% of Salmonella Hadar isolates were extended spectrum cephalosporinase (ESC) producers – this is characteristic of certain bacteria that leads to resistance to some important and commonly used antibiotics (e.g. 3rd generation cephalosporins).
  • The likely source of the outbreak strain is backyard poultry, both chickens and ducks. These were often purchased at places like agricultural stores, directly from hatcheries or (probably worst case scenario for various reasons) over the internet.

The bias towards young kids is totally expected since that group is more susceptible to infection and probably more likely to be tested if they get sick. It’s also a group for which there is clear messaging: kids less than 5 years of age (and elderly people, pregnant women and people with compromised immune systems) should not have contact with young poultry. That’s a major education and/or compliance gap that’s seen in most animal-associated Salmonella outbreaks.

I won’t get into a full discussion of preventive measures, but the CDC notice includes a good list. They’re all common sense and very practical, but compliance is probably variable and often bad.

Wash your hands and don’t eat poop. Good general advice, but even more relevant if you have backyard birds.  And don’t make chicken diapers your sole infection control plan.

In the midst of outbreaks of COVID-19 on at least 5 mink farms in the Netherlands, a Reuters article reports that Dutch Agricultural Minister Carola Schouten issued a letter to parliament indicating that a farm worker was infected with SARS-CoV-2 from the mink. That’s a bit surprising to me, with the surprising aspect being the apparent ability to identify mink-to-human transmission. How this was determined isn’t clear and more details are needed.  The nuances of what was said also are unclear. A Google translation of a Dutch news report about the case says mink-to-human transmission was “plausible” (aannemelijk), while the English Reuters report is more definitive (“A person who worked on a farm where mink are bred to export their fur contracted the coronavirus from the animals.“).

More clarity is needed.

From a biological standpoint, mink-to-human transmission wouldn’t be surprising. If mink can infect other mink, it makes sense they could also spread the virus to people in close contact (although “close contact” with farmed mink is much less common, and much less close, than human contact with pets, for example). However, identifying animal-to-human transmission when there’s widespread human-to-human transmission is a challenge, especially when people can be infected by other people with asymptomatic infections.

Figuring out exactly how a person got infected can be a challenge in the community.  If someone on a farm gets sick, does that mean they got it from a co-worker, an animal or somewhere off the farm? Evaluation of the genetic sequences of the virus can help figure out who’s linked to who, as subtle changes in the virus occur over time.  Finding an identical virus in two individuals supports a link, but it doesn’t tell us in which direction the virus was transmitted, or rule out the potential that both individuals were infected by the same source. The Dutch report indicates there are similarities in the gene sequences of the viruses from mink and the worker, but that still doesn’t answer the question of “who infected who.” More information about contacts between the infected worker and other workers, contact between the worker and mink, timing of contacts and disease, and genetic sequences of strains found in people off the farm in that region is needed to better understand the situation. I assume much of that will be coming, so it will be interesting to see how this story unfolds.

I keep saying I’m going to stop talking about sporadic new SARS-CoV-2 infections in animals unless there’s something noteworthy. I’ll mention some recent cases in a dog and some cats in the Netherlands because I think there are some unique aspects that fit that bill.

Infected dog

This infected dog was euthanized on account of severe respiratory disease. So far, it has appeared that dogs don’t get sick if they are infected with SARS-CoV-2. Disease of any sort, let alone fatal disease, would therefore be noteworthy in a dog.  One report said “The American bulldog’s blood tested positive for SARS-CoV-2 antibodies, but the dog had tested negative for an active case of Covid-19.” I assume that means it was PCR negative, but that doesn’t mean it wasn’t infected. They added “It was thus unclear if the dog’s worsening condition was as a result of the infection, or due to other health issues.”  Hopefully more testing is being performed  to see if there were other problems that could have accounted for severe disease or whether SARS-CoV-2 might have been the cause.

Infected cats on a mink farm

During our national working group discussions of the outbreaks of SARS-CoV-2 on Dutch mink farms (now 5 affected farms), the question of whether there were other animals like barn cats on the properties was raised. The answer to that is apparently “yes.”

Three of 11 tested cats on the farms had antibodies against the virus, indicating they had been infected. That leads to questions about how they were exposed. Investigating that involves interviewing farm staff to see how much human contact they had, to get some idea whether contact with infected people or indirect contact with infected mink (e.g. droplets/aerosols from being in the affected barns, contact with potentially virus-contaminated manure) was the likely source.  This highlights the importance of preventing exposure of other animals and containing exposed/infected animals. We want to keep this virus confined to humans as much as possible, and not create opportunities for animals to pass it back to people or for animals to spread the virus to other domestic animals or wildlife.