Warning: Long post. Lots to unpack here.

As H5N1 avian influenza continues to spillover into domestic animals and people, the question of when to test animals (especially cats) keeps coming up. Testing isn’t just a matter of swabbing a bunch of animals and sending the samples to the lab (which unfortunately is not uncommonly done, and it often leads to problems). There are lots of things to consider before we start testing, which I’ll bin into 4 major questions: what is the goal, who will be tested, what test will be used, and what will be done with the results?

What is the goal of testing?

This should always be question #1. We need to know why we’re testing to figure out how to do it and what to do with the results. Specifically are we testing:

• to understand more about how common the disease is, what it looks like clinically, risk factors, virus variants and transmission risks? Those are surveillance goals.

• OR to inform how we’ll care for and manage individual animals? That’s clinical testing.

These goals can overlap, but they are not the same. If we think first about what we want to achieve, we run the risk of wasting time and money, and not answering the right questions.

Who will be tested?

In the case of H5N1 testing in companion animals, do we test all cats and dogs? All the cats? All cats and dogs with certain signs of illness? All cats and dogs with certain risk factors? Any pet that looks abnormal?

What we learn will vary a lot based on who we test. It will also impact efficiency and costs. More testing can provide more information but also costs more (time and money). For rare diseases (and I still consider spillovers of H5N1 influenza into pets to be rare overall at this point), you can do a lot of testing and get no positive results. That’s still informative in some ways, but broad testing can be pretty low yield in this type of scenario.

What test will be used?

For H5N1 influenza, PCR is our standard go-to test for a number of reasons, but there are also some “rapid” point-of-care tests that get talked up periodically (more on that below).  Different tests perform differently, so we need to pick the right test for the right population based on the goal of the testing.

What will be done with the test results?

This is the biggest question.

• If we’re doing surveillance testing, is the design of the testing scheme robust enough to actually allow us to infer anything relevant beyond the animals sampled? Poor study design leads to poor (or no) conclusions.

• If we’re doing clinical testing, what will be the response be, and will it differ if the test is positive or negative?

Too often, people do a test, and only think about what to do after they get the result. That can be problematic. What will we do with pets that test positive for H5N1 influenza?

Will we euthanize positive pets?

I hope not, but it’s a knee jerk response we handle sometimes. I can’t see any value in euthanizing infected pets. We can typically isolate them effectively, and then they’ll either get better and eliminate the virus, or die. Either way, the infection is contained, and there’s no concern about long term shedding of the virus. Yes, there’s some short-term risk from the pet while it’s infected, so we need to do it right, but there’s risk in most of what we do, and it can be managed.

Will we treat positive pets?

If we have an effective treatment for a given disease, early clinical testing can be really useful. For H5N1 influenza, it’s less useful since we don’t have a treatment we’d typically use in a pet, and by the time we get test results back, it might be too late for it to have any effect anyway. We can potentially use antivirals (e.g. oseltamivir) in dogs and cats, but that would be reserved for rare situations where the animal has early and severely progressive disease, or where it’s at high risk for severe disease (e.g. an immunocompromised animal on chemotherapy). Most often, we’re just going to give them supportive care, whether they have influenza or not.

Will we isolate positive pets?

This would be useful to help prevent spread of the virus. However, if we think there’s a reasonable chance that a pet has influenza (of any kind), we should isolate them anyway. Since tests are not 100% sensitive and we don’t get results back right away, we have to err on the side of isolating sick animals from the start. If the case is high risk for flu (i.e. compatible clinical signs and realistic risk of exposure), I would want to test it, but even if it tests negative I would still consider it a flu suspect and manage it accordingly to be on the safe side.

Will we do nothing different for positive vs negative pets?

That’s probably the case most of the time. So then why would we want to add complexity and cost by testing pets?

• There are some limited surveillance benefits (discussed below).
• If the animal is in the clinic, it’s easier to do the testing and the cost/benefit is more reasonable.
• If the animal is at home, we need to consider whether or not the stress to the animal (i.e. taking it to the clinic, sample collection), the risk to the owner transporting it to the clinic, the risk to clinic personnel handling the pet, and the cost of sampling and testing are worthwhile, if instead the pet could just stay quiet at home (and hopefully recover)?

What I would like to see/do in terms of H5N1 influenza testing in pets

Surveillance testing

1. Testing of cats/dogs with clinical signs suggestive of H5N1 influenza infection.
   • Right now, the focus is on pets with severe neurological and/or respiratory disease. Testing these animals lets us know how common the infection is in these cases, lets us investigate risk factors and transmission risks, and lets us do viral sequencing.
   • Currently the most widespread testing for H5N1 flu in pets is probably secondary testing of rabies suspects, whereby cats that are being tested for rabies (which often have neurological disease) get tested for flu if the rabies test is negative. This is a convenient and useful surveillance method that can contribute useful infomation, especially if other organized surveillance is lacking, but the sample is limited.
2. Testing of dogs/cats with more common respiratory or flu-like illness signs.
   • We need this to understand the scope of disease caused by H5N1 influenza in pets. Are severe cases all that occur or are they all that we’re currently diagnosing? We need to know if there’s a fraction of cases that only have milder disease, particularly to help us assess and mitigate transmission risks.
   • We can’t test every sneezing cat, but we can do targeted surveillance by focusing on animals with a realistic risk of exposure to the virus. So, I’d like to test sick cats and dogs in areas where H5N1 flu is active (i.e. in wild birds) and where there’s a reasonable risk of exposure (e.g. outdoor access). This balances throwing a wide net to get a better broad understanding, and being efficient with our time and money.

Clinical testing

This is a tougher call. Most often, H5N1 flu testing would not impact patient care. It could help inform us about risk to the clinic or household, so it’s good information to have, but we can’t dismiss risk based on a negative test, so it doesn’t have a huge impact on what we do regardless.

The less surveillance testing there is, the more valuable opportunistic clinical test is for quasi-surveillance purposes, but that’s usually going to be at the owner’s cost. Spending a couple of hundred dollars to test a severely ill animal with little expectation that the test will change the response or the outcome is hard to justify. If money is tight, I’d rather have owners spend that on treatment.  Some people will pay for the testing because they want to know, or because they are worried about exposure of family members, but relying on owners to cover testing isn’t a great approach. It’s still worth discussing in any case since there can be some limited benefits.

What about rapid in-clinic tests for H5N1 influenza in pets?

It’s been suggested that veterinarians should start stocking rapid in-clinic H5N1 tests and use them liberally when presented with sick dogs and cats. I’m a bit wary of this because we don’t know how good those tests are. If we test a lot of low-risk animals and the test isn’t 100% specific, most of our positives are likely to be false positives. That’s alright if we’re just using it as a screening test and we follow-up with a proper test ASAP, but sometimes people rely too much on the screening test, and that can lead to over-reactions and other problems. On the other hand, if the test isn’t very sensitive, we run a good chance of missing the small number of true positives. If the test is no better than a coin toss, it won’t help.

Collecting good samples can be a challenge (especially trying to swab cats) so that drops the functional sensitivity of any test (i.e. we get more false negatives because of poor samples.

To be useful, we would need:

•  a rapid test with high sensitivity (reliably detects all the infected animals) AND high specificity (gives us very few false positives)
• AND good samples
• AND the capacity to follow up positive tests with PCR testing
• AND a communication plan for animals that are confirmed positive

But we’re not necessarily there yet. I’d consider rapid testing appealing, but probably not ready for prime time.

What will I do with my own pets?

I have two cats. Milo’s an indoor cat and Alice was a completely feral cat who’s gradually become very affectionate and just recently made the jump to being largely indoor. But we’re still not sure she uses a litterbox, and she’ll be at least partially outdoor for a while, I assume. She’s therefore at greatest risk of exposure to influenza from contact with wildlife. If she had a sudden onset of neurological or other severe disease, I’d test her, primarily because I’d really want to know, given my infectious disease mindset.

• If she was negative, I’d still assume she might have flu, but that we missed it with the test.

• If she was positive, then I’d have an early warning of potential exposure in the house. If anyone (human) started to get sick, flu would be on the radar and odds are high that we’d get tested ASAP and would probably be offered antivirals.

Barriers to surveillance

Money $$$

There’s VERY little funding available for companion animal infectious disease research. We try to scrape by with leftovers and token amounts, and we still manage to get good things done, but more support is needed. We can’t test if we don’t have the funding, and we rarely have funding.

Access to animals

You’d think it’d be easy, but field studies are actually very challenging. Veterinary clinics are really busy. When clinics are asked to remember that we have a study, to determine if their patient fits the inclusion criteria, and to contact us about sample collection, that takes time that just might not be available on a busy day. It took us a long time to get a good sample size for our SARS-CoV-2 research project and that was a lot higher profile.

• Lack of money compounds the issue, since we usually can’t pay clinics for the time required to communicate with us and any extra effort to collect samples (including getting consent forms completed, collecting samples, collecting other data and shipping samples. or helping us arrange to collect samples). We try to do as much as we can to take the burden off the veterinary clinic team. With local cases we can often do almost everything, but the farther away they are, the more we rely on clinics to help out.

There’s no single approach to surveillance. We need combined approaches both because we have different questions to answers and for practical purposes. We want as much well-designed formal surveillance as possible, but since that’s not always going to be available, we rely on opportunistic information from clinical testing. We’re learning more as more work gets done, but we have a long way to go before we have a reasonable understanding of the dynamics of H5N1 influenza in companion animals and the potential for transmission to and from cats and dogs.