A Study published in June 2022 reported results for dogs detecting COVID compared to other testing methods. It’s pretty neat. Let’s see what the dogs found.
Study on dogs recognizing COVID
Essentially, in this blinded, multi-center study, between March 16 and April 9, 2021, researchers asked people at two different COVID testing sites in Paris, France to submit a sweat sample to test how well dogs detect COVID infections in humans be able.
If the 516 people asked, 335 were eligible for the analysis of the results. Reasons for disqualifications included:
- People who refused the sweat test (n=113)
- Sweat test somehow not valid (n=2)
- Sweat test somehow not tested in dogs (n=40)
- Sweat sample tested from 1 dog only (n=26)
How do dogs recognize COVID?
Test participants placed 2 sterile surgical pads in their armpits and held them there for 2 minutes. Let’s not get bogged down in how the samples were handled and protected up until the dog test, but suffice it to say that the research team took a lot of precautions with the samples (so-called “odor hygiene”) and how the sample testing room was set up and cleaned and such.
The detection room set up looked like this, with 10 metal scent cones mounted on the walls for dogs to work through. It’s similar to the bin finding dogs do in scent contests, where they have to identify those with a target scent.
From the photo below, the dogs offer what’s called a “final response” — in this case, a sit. Essentially, dogs that detected COVID were trained to sit near samples that smelled like COVID.
The theory goes that they can detect the odor of VOCs (volatile organic compounds) that the human body produces in response to disease.
How long have the dogs been training to detect COVID?
It depends. For dogs that are already familiar with scent detection, for example to find drugs or bombs, the training time lasted about 3 weeks. For dogs that are inexperienced or new to the whole process of alerting to specific odors, the training took more like 5-6 weeks.
Dogs recognizing COVID results
Of the 335 people enrolled, 143 reported a variety of symptoms and 192 reported no symptoms but wanted to be tested due to recent/possible exposure to COVID. The dogs – provided by French fire brigades and the Ministry of Interior of the United Arab Emirates:
- Reported positive for COVID in 126 cases
- In 209 cases NOT alarmed
So how did the dogs fare compared to the other diagnostic tests currently in use? The dogs:
- Alerts on 20 cases where the other tests were missed
- Missed 3 cases that picked up the other tests
Without getting lost in the weeds, Test SENSITIVITY means how well a test method correctly identifies diseases. In other words, the test result has few false negatives. Test SPECIFICITY means how well a test method correctly identifies those without the disease. In other words, a true/trustworthy negative. These two measurements work in reverse. If one is high then the other is low and so on.
The serious scientific data
“335 outpatient adults (143 symptomatic and 192 asymptomatic) were included. A total of 109/335 participants tested positive on nasopharyngeal RT-PCR in either symptomatic (78/143) or asymptomatic (31/192) participants The overall sensitivity of detecting dogs was 97% (95% CI, 92 to 99) and even reached 100% (95% CI, 89 to 100) compared in asymptomatic individuals to NPS RT-PCR. That Specificity was 91% (95% CI, 72 to 91), achieve 94% (95% CI, 90 to 97) for asymptomatic people.”
How does this compare to other tests?
Good question. The other COVID test types included:
- PCR (nasal swab) tests as the “reference standard,” meaning they trust it the most
- Saliva PCR Tests
- Nasal swab antigen testing (how to do it yourself at home)
They also tested humans for several other viruses, including influenza strains and pneumonia strains, to rule out the possibility of dogs being warned about diseases other than COVID.
What about at-home nasal swab testing?
“That The sensitivity of detecting dogs was higher than the nasopharyngeal antigen test (97% CI: 91 to 99 versus 84% CI: 74 to 90, p=0.006), but the The specificity was lower (90% CI: 84 to 95 versus 97% CI: 93 to 99, p=0.016).”
When compared directly with nasal swab PCR testing, dogs found 17 additional cases (126 versus 109 with the PCR test).
Comparison between dogs that detect COVID and other types of tests
Here’s the full comparison chart for all the tests if you want to dive into the data for yourself.
Dogs Detect COVID, What Does It Mean?
Well, the study concludes that dogs could be “an alternative” to typical PCR tests, especially when humans need quick results.
However, they didn’t have enough people who were negative for COVID but positive for other types of coronavirus to “know whether canine detection is specific to SARS-CoV-2 or to all coronaviruses.”
Could we one day be sniffed directly by dogs?
Researchers say this is a challenge for several reasons:
- Many people are afraid of dogs
- Dogs could be contaminated by direct contact with sick people (The scent cones used protected dogs from this risk.)
- Lack of available dogs trained in COVID detection
Wondering if I could teach my dogs to find COVID?
Part of me would love to try and teach my dogs to be aware of the smell of COVID – because I’m pretty sure I had it at the end of May, but all home and official PCR tests have come back negative.
Nonetheless, I felt really bad for 2 full weeks, with an insane exhaustion that I’d only felt before with the honest flu (once 20+ years ago) and when I had mono as a teenager. Yes they tested me for flu and mono and even strep. These tests came back negative.
What makes me think it was probably COVID is that I suddenly had phantom liquorice smells a few days before the symptoms started. I was just sitting at my desk working and mysteriously smelling of liquorice back and forth for several hours. What the hell?
So who knows? I am very excited that the better/more targeted booster vaccine for the main strains will be available soon. I got my first booster in November 2021 and postponed the second booster until the better one was available. Whooooo! ;O)