As concerns about canine infectious respiratory disease in the US have taken up most of my time lately, let’s merge that issue with what I had hoped to be the focus of the week: World Antimicrobial Resistance (AMR) Awareness Week.


Despite lots of media attention and associated fear, we’re still not sure what’s going on with all these coughing dogs, or even if there’s really a story at all. This could be something new, but more likely, it’s the usual suspects doing their usual thing (possibly at higher rates in some areas, as fairly commonly occurs periodically).

In the unlikely event this is something new, it’s likely viral. It’s much less likely to be something bacteria.

Either way, we have to think about how that might impact treatment. The short answer is: it probably doesn’t affect our treatment approach.

Viral respiratory illness can’t be treated with antimicrobials. Some affected dogs will develop secondary bacterial pneumonia, and antimicrobials are indicated in those cases. But it doesn’t matter what virus triggered it.

Primary bacterial respiratory infections in dogs are less common. The bacterium Bordetella bronchiseptica is typically the number 2 or number 3 overall cause of canine infectious respiratory disease complex (CIRDC), after canine parainfluenza virus and maybe canine respiratory coronavirus. Streptococcus zooepidemicus is a rare cause of CIRDC and usually causes sporadic but really nasty (often rapidly fatal) disease, most often in shelters or other high stress settings. Secondary infections (i.e. things that move in after a virus has already caused some damage) can be caused by a variety of different bacteria.

When considering antimicrobial therapy, we need to think about the disease we’re targeting. Cough isn’t a disease. It’s a sign of disease. Cough can be triggered by infection, be it bacterial or viral, and often persists even after the infection is over. Too often, we get into a mindset of “the dog is coughing really badly” or “the cough isn’t going away” and we unnecessarily reach for antimicrobials, hoping they will somehow help, when in reality we just need to give the dog more time to fully recover, or we can use other approaches to decrease inflammation and suppress the cough if that’s the part that’s still a problem.

Our 2017 Antimicrobial use Guidelines for Treatment of Respiratory Tract Disease in Dogs and Cats from the International Society for Companion Animal Infectious Disease are a good start for thinking about how to manage these sick dogs. We’re starting a revision, and I think we’ll see a few changes to the guideline, but most of the original content still applies. Some newer approaches to care of these cases are already incorporated into the antimicrobial use guidelines available to veterinarians through the Firstline app (see image below).

Here are some of the basic recommendations:

Basic upper respiratory tract infection: cough, runny eyes and nose, maybe a fever and a bit quiet, but dog is usually pretty bright overall.

  • No antibiotics.
  • This is likely viral, and if it’s bacterial, it’s mild and should resolve on its own. Tincture of time and supportive care are recommended.

More serious upper respiratory tract infection that is probably bacterial: more advanced signs of disease, mucopurulent (yellow, goopy) nasal and ocular discharge, but lungs are clear.

  • Consider antibiotics but most cases probably don’t need them.
  • As these cases are more severe it’s easier to justify antibiotics, but I can often go either way and have a fairly high threshold to say “start antibiotics” (at least right away). However, it’s not unreasonable in many cases.
  • Doxycycline is the drug of choice for treatment. It’s lower tier, effective, safe and works against the main bacterial pathogens of concern.
  • If there’s no response to initial treatment, we need to back up and think about whether that’s because it’s a bacterial infection is not responding, or whether what we’re seeing is more likely viral or non-infectious. That’s often hard to sort out, but we need to consider it carefully rather than just jumping to another drug every time we don’t get the response we expect on the first try.

Mild/moderate pneumonia: Varying upper respiratory signs, but with signs of lung involvement, such as audible crackles and wheezes, and radiographic evidence of pneumonia. These dogs are sicker but are stable. They are breathing reasonably normally, are quiet but alert and do are not crashing.

  • Antibiotics are definitely indicated.
  • Doxycycline is still the drug of choice. Along with the points listed above, it achieves good drug levels in the lung and is a great first line choice for pneumonia. If it’s a rare, milder or earlier Strep zooepidemicus pneumonia (mainly we’d suspect this because it’s part of an outbreak), amoxicillin would be fine instead, but usually we want a broader spectrum drug than that and one that gets better levels in the lung.
  • Some people are suggesting that enrofloxacin seems to work better in some of the more recently reported cases. That could be a true reflection of better activity against certain bugs, or issues with resistance to other drugs, but could also just be a function of using enrofloxacin as a second line option later in disease (where its use corresponds to natural resolution of disease, versus a true effect of the drug). It’s a good observation and I don’t dismiss anecdotes like that, I want to explore it more to try to tease out the reasons versus making a full switch to regularly using a higher tier drug like that. It could be that enrofloxacin is a better drug (overall or in specific areas), but given the potential issues with use of this higher tier drug, we’re best to be cautious and try to make sure we really have a firm indication that it’s necessary. If we put every dog with pneumonia on a fluoroquinolone like enrofloxacin, we won’t be able to use the drug for long because resistance will quickly become an even bigger problem.
  • Azithromycin is another option for treatment of pneumonia, as it also achieves great levels in the lung.
  • The more convinced I am of a true treatment failure, and the more severe the disease, the more I’d escalate, but sometimes we can be mislead by our observations. If it’s clear that doxycycline isn’t working in one area but another drug is, it’s logical to use that other drug, but we want to make sure we’re limiting changes in approaches and use of higher tier drugs as much as we can, because the more we use them, the quicker we lose them.

Severe/septic pneumonia

  • One of my big considerations when deciding whether to use more broad spectrum treatment in any patient is “What’s likely to happen if my drug choice is wrong?” If the answer is “the animal will probably die,” I can justify using a higher tier drug or combination to help ensure it’s effective on the first try. For the cases above, I wouldn’t typically jump to a broader spectrum combination, but with severe septic pneumonia we are dealing with a subset of dogs that are really sick with significant lung disease. They are not oxygenating well. They have low blood pressure or other signs of severe systemic inflammation. They’re at risk of crashing hard and fast, and I need to get the infection under control pronto. So, I can justify a broad spectrum antimicrobial – nothing crazy (e.g. not meropenem), but a broad spectrum drug/combination that’s higher tier and something I generally avoid, but am comfortable using in a situation like this.
  • Intravenous clindamycin & enrofloxacin, ampicillin & enrofloxacin, or an intravenous 3rd generation cephalosporin (e.g. ceftiofur, cefotaxime) would be reasonable choices in these cases.

How long do we treat a dog with pneumonia?

We have very little duration of treatment data for most infections we deal with in veterinary medicine, especially companion animals. We tend to be quite risk averse and therefore default to really long antimicrobial treatment courses.  Based on the short durations of antimicrobial use in people with similar conditions (where there’s lots of evidence that shorter is better), and even in cattle (where there’s lots of evidence and desire for shorter courses because it’s a hassle treating them), we need to be aim for shorter courses of antimicrobials in pets too.

  • Five days is what I’m recommending now. We don’t have data for that, but we also don’t have data for using any longer duration of treatment, and since we have good comparative data from other species, increasing anecdotal evidence and a duty to consider a “least harm” approach, I’m happy with five days. We can always go longer if needed, based on patient response and complicating factors, but short durations are often effective, come with fewer adverse event risks, and are cheaper and easier for owners (who, realistically, often don’t complete long treatment courses when their pet is doing well anyway).

We have to remember that antibiotics are there to help resolve bacterial infection, but eliminating the infection doesn’t immediately fix everything. Signs like cough and radiographic changes can linger, and more antibiotics don’t help those things resolve any faster.

So, while we’re not sure what’s going on with all the coughing dogs in the US right now, we can be reasonably confident we know how to treat them. Our usual approaches will still work. We need to be conservative with antimicrobials, but also ready to use appropriate drugs (including broad spectrum, higher tier drugs) when indicated. The right drug at the right dose for the right patient still applies.