Given the recent rises in Brucella canis in the UK, many vets are now requiring owners of imported dogs to have their dogs tested for this potentially dangerous and zoonotic bacterium. However, concerns have been raised about how reliable the tests are. If vets are genuinely going to refuse treatment for positive dogs, can we trust the tests enough to make that decision?
Table of contents
Brucella canis – an emerging disease in the UK
Brucella is a bacterium that primarily attacks the reproductive tract. Dogs have their own specific species, Brucella canis, which behaves broadly similarly to other species. But it has its own quirks (particularly when it comes to diagnosing it and its impact on humans). In dogs, the disease normally causes reproductive disease (infertility, pregnancy loss, testicular inflammation); but infections elsewhere in the body can cause spinal disease, meningitis and eye problems too.
The disease is primarily spread by contact with stillborn or miscarried puppies, semen and urine, to uninfected dogs. Infection via nasal discharge and through broken skin is possible. And the organisms have even been detected in saliva and tears of infected dogs. Unfortunately, dogs may carry the infection for many months or years before, or even without, developing symptoms. These carrier dogs are a major concern; as they can be a source of infection to other dogs and to humans.
In humans, fortunately it is milder than some other strains, with most healthy non-pregnant adults developing moderate flu-like symptoms. However, more serious signs including heart and joint infections, lung disease, and even nerve damage or spinal abscesses can also occur. The risks are much higher in children and those people with weak or damaged immune systems (e.g. due to disease or chemotherapy); and pregnant women are likely to be at risk from pregnancy loss.
You can read more about the disease in our companion article, here.
Until recently, the UK seems to have been free of this infection. However, in recent years – probably fuelled by increased importation of dogs from Eastern Europe, where it is much more common – cases have increased dramatically. In 2022, the UK had its first confirmed human infection from an imported dog.
What are the treatment options?
Unfortunately, they’re very limited. There is no known way to eliminate all the bacteria from a dog’s body. We can treat the active infection, but the dog will always be a carrier, with a risk of spreading the bacteria to other dogs or other people.
The best way to minimise the risk is to neuter, as the bacteria tend to concentrate in the ovaries and the testicles; removing those removes most (but never all) of the bacteria. It also prevents the reproductive routes for transmission, newborns/stillborns and semen. However, it can only reduce, not eliminate, the risk.
What tests are there?
There are three basic types of test that are widely available and practically useful.
Bacterial culture
This test relies on use being able to grow the bacteria in the lab from a swab – usually taken from a miscarriage, or a vaginal swab in female dogs. In males, it is sometimes possible to capture bacteria in semen samples or even blood samples.
PCR tests
As we all know post-Covid, PCR tests are a very powerful diagnostic tool. They work by detecting tiny fragments of the DNA of a pathogen, usually in a blood test, although again swabs can be used too.
Serology tests
These are blood tests that look for antibodies in the blood, and are the most widely used. They will tell us whether the dog has ever been exposed to Brucella. And given that dogs cannot clear the infection, if they ever were infected, we know that they are at least a carrier now.
There are several different types; the most common are lateral flow tests (also known as immunochromatography tests), iELISA (indirect enzyme-linked immunosorbent assay) and SAT (serum agglutination test). The lateral flow tests can be used in your vets. But the others need to be sent away to an external laboratory – usually the government’s APHA laboratories.
How reliable are they?
Here’s where the problems start. The tests are often described as over 90% accurate. Which is strictly true, but not in the way most people think.
Culture and PCR
In an ideal world, we’d rely on bacterial culture – if we grow the bacteria, we know that they were there! Unfortunately, culture misses many – probably most – cases. This is for two reasons. Firstly, the bugs are very hard to grow in the lab, and often seem to die off in a petri dish. Secondly, carrier animals often don’t spread bacteria all the time. You might take a swab at a time when the dog wasn’t shedding bacteria, assume that they’re negative, only for them to start shedding again the next day.
The same problems apply to the PCR tests: the numbers of bacteria in the blood of a carrier are very low, so we often don’t get enough DNA for the test to react to.
As a result, a dog who is negative on culture and/or PCR may well actually be positive; we just didn’t capture enough of the organisms to give us a positive finding. We call this a “false negative”, where the test says negative, but actually the dog is a carrier. That said, any dog who is positive is definitely positive – because we’ve seen the bacteria themselves.
Serology tests
This is why serology tests are so widely used: because a false negative is quite rare (using the APHA tests, it’s about 0.18% false negatives). In fact, the only time false negative results are likely is when we blood test the dog less than 3 months after infection. Most dogs will develop antibodies to Brucella canis within 2 weeks of infection, but sometimes – especially in puppies – it can take up to 13 weeks.
False positives are a bigger problem – where the test says an animal is positive but actually they aren’t. This can happen with any test, but it becomes more and more of a problem the rarer a disease is.
Even in imported dogs, it’s estimated that only about 1.6% are positive for Brucella canis. This means that the number of “true positives” is often smaller than the numbers of false positives.
The lateral flow tests (quick and easy though they are) are usually about 91% specific (meaning they will accurately rule out infection in 91% of patients). That’s actually pretty good, but when you crunch the numbers, only about 14% of dogs with a positive lateral flow test actually have Brucella canis – the other 86% don’t.
If you use an external lab, it’s better. Combining the iELISA and SAT test together (which is what is recommended) gives a reported 99% specificity. So much better! But even so, you’ll still get about 38% of dogs testing positive even if they don’t have the disease.
So what do we do?
Dogs who are confirmed positive for Brucella canis should always be isolated from negative dogs, children, and people with weaker immune systems. It is probably wise to ensure that they do not come into contact with pregnant dogs or humans either. So there is a serious issue of animal welfare here, as they have to be isolated from other dogs. This is why euthanasia is usually the recommendation for human health, and often for animal welfare too.
But given that we’ve just shown that false positives are fairly common, can we rely on the tests if the dog’s life is at stake?
The best way forward seems to be to use the three tests in sequence. Dogs who test negative on a lateral flow almost certainly don’t have Brucella, so we don’t need to worry about them any more. Dogs who test positive, however, have a 14% chance of carrying the bug. So, I’d recommend waiting 4-5 weeks and then sending the results off for the combined iELISA/SAT test. When combining both tests, we can be 96% sure that positive results mean a positive dog. And then we are on much firmer footing when deciding what to do next.
What about those vets refusing to treat dogs after a single positive test?
I don’t know what their situation necessarily is – perhaps they have an immunocompromised member of staff; perhaps there are other reasons why the risk is too high. But it would make me very uncomfortable if this was a general policy, rather than because of some specific local situation or risk analysis. Leptospirosis is much more dangerous to humans in many ways than Brucella canis, and yet we treat that, using barrier nursing and extra precautions.
At the very least, though, I think we shouldn’t be writing dogs off after a single test, especially a lateral flow: we need to use a suitable combination of tests, and then have that conversation with the owner.
In most cases I think the dogs are better off euthanased rather than locked away all the time, and yet still posing a risk to their families. But I’d rather we didn’t have this problem at all because we tested before we brought them into the UK rather than trying to catch up later!
For more details…
Check out the article recently published by Veterinary Voices UK. In it, we looked at the disease and the different tests in some detail, and did all the calculations so you can see how we worked out the numbers I’ve talked about above.
Further reading:
Discussion