As H5N1 avian flu ramps up again across Canada with the fall wild bird migration, we’re likely going to see more situations where more unique populations of captive birds are affected, beyond the usual large or small poultry flocks. The CFIA’s standard response to highly pathogenic avian flu (like the current H5N1 strain) is “All infected flocks are humanely destroyed, and carcasses are disposed of in an environmentally acceptable fashion.

As would be expected, this response is based primarily on commercial poultry, where if one bird in a flock is infected, you can be pretty sure the virus is widespread in the group. However, other kinds of birds that can still be infected may be housed very differently, and management and infection control measures may affect their risk if an infected bird is found on the property.  Other factors to consider are the ability to contain the risk from potentially exposed birds (which may include financial costs), and the importance of some birds in terms of conservation or genetics.

The standard approach is understandable with poultry that are highly susceptible to the virus and housed in a manner that transmission can be rampant.

However, given how widespread this virus now is in the wild bird population in North America, and the wide variety of captive birds that can be affected, could a one-size-fits-all approach perhaps cause more harm than good in some situations?

  • Maybe.

The main issue relates to non-commercial birds (those not raised primarily for food), particularly pet birds and birds in rehab facilities.

Sometimes, exposure of the entire group is likely and a whole-group response (euthanasia) might be indicated.

  • Risk assessment should play a role in this.
  • If a rehab facility has waterfowl and raptors (and maybe some pet birds in the house), and those groups are kept separate, does it make sense to depopulate all the birds on the property? Maybe not, at least all the time.
  • It comes down to the risk of exposure. Often, there can be pretty good  physical and procedural separation.

We shouldn’t realistically aim for “is there absolutely, positively, no chance that the birds were exposed?”. We can never hit that bar. Rather, we should aim for “are these birds at any greater risk than any other birds in the area, when we know that H5N1 flu is circulating in wild birds?”. That changes things a bit and recognizes that there might be some degree of risk, but it might not be any more than is inherently present with a virus that’s currently fairly widely distributed in nature.

Another major concern I have with any strict policy is driving things underground. If bird owners know a positive test means all their birds will be killed, they’re more likely to try to ride out a problem and not get testing done. That means we lose valuable information, don’t get a chance to respond to help contain the issue, and we can miss the ability to manage disease properly if it’s something other than flu. I can absolutely see non-commercial bird owners avoiding testing if stories of mandatory depopulation of birds like theirs increase. We need to know the extent of the spread of this virus so we can take other steps to control it, and driving things underground doesn’t help.

I fully admit it’s a tough situation. It requires people to change standard approaches, make decisions on the fly and do risk assessments without much information (at a time where CFIA is certainly not overflowing with resources).

We want to contain this virus for both human and animal health reasons. But, we have to realize this is an unprecedented avian flu situation in North America. We’ve never had this degree of sustained and widespread infection of wild birds.

I’m certainly not saying we should surrender and say ‘it’s endemic, we’re done, good luck.” but a more risk-based approach than has previously been considered is warranted, based on the risks to people and animals in the broader context. That’s a challenge, and we absolutely have to prioritize protecting human health.  As we enter human flu season (which is shaping up poorly), we don’t want mixing of avian and human flu viruses. We also need to minimize the risks to commercial poultry operations, which can affect thousands of birds at a time and can have ripple effects on the entire agri-food supply chain.

However, does a one-size-fits-all approach that requires euthanasia in every situation make sense? I don’t think it does.

What are the challenges to a risk based approach?

  • Lack of data to guide risk assessments in more unique situations.
  • Overloaded regulatory personnel (most veterinarians can relate to this right now too).  Case-by-case assessments and tailored responses typically take more time and resources than using the one-size-fits-all approach.
  • Often unclear or unreliable information on premises about what’s done and what the risks might be.
  • Perceived risk and risk aversion (e.g. the safest thing for regulators to do from a risk standpoint is euthanize any potentially exposed birds. Doing something different, even for good reasons, can increase risk to people or other animals, and then regulators may get blamed if things go wrong).

I think it’s time to try to implement some risk assessment-based approaches to control of this virus in some unique collections of captive birds. Often, euthanasia will still be the result if transmission between birds cannot be controlled, and that makes sense. However, there may be some situations where it can be argued that not all groups on a property are at the same risk for exposure.

What about hold and test?

  • That’s a consideration but it brings in some challenges.
  • Some tests are better than others, but no test is 100%.  It’s not as straightforward as it may first appear. However, with serial tests in a group, I think we can be pretty confident whether flu is or isn’t in a group of birds. That’s what we’re really looking to say. We want to say “are one or more birds in this specific group infected” vs “is this individual bird infected’. We can design logical testing strategies to do that.
  • Testing takes time, and also involves cost and that would be on the facility/owner. Cost concerns might certainly prevent this approach from being viable in some situations, and risks need to be carefully controlled while testing is undertaken.
  • Human exposure risks also need to be considered. If we’re collecting samples from birds, we’re potentially getting exposed to this flu virus. We can do things to reduce the risk, and any testing approach would require people that are properly trained and willing to accept the risk (like we, as veterinarians, do regularly). Fortunately, the current H5N1 seems very low risk for causing disease in people, but flu viruses are notorious for changing and low risk doesn’t mean no risk. We also want to make sure potentially infected birds don’t get exposed to human flu, another reason for careful handling during sampling.

Although the safest (and easiest) way to eliminate an “animal problem” is to eliminate the animal, euthanasia of any captive birds that might possibly be exposed to this virus when a positive is detected may not be the “best” approach in all cases.  It will be interesting to see if and how the CFIA adjusts their approach as they continue to deal with this unprecedented outbreak of avian flu.