People with COVID have a subtle scent that dogs can detect in under a second

Originally published at: People with COVID have a subtle scent that dogs can detect in under a second | Boing Boing


There have been so many cases of dogs’ reactions being misinterpreted or manipulated that I’m taking this news with a grain of salt. Hopefully, more studies and testing will prove that my skepticism was unwarranted.

ETA link:


Can we please come up with a better name for those breeds? Maybe good-nose breeds?

Who’s a good boy? With a really good nose? You, that’s who!


If a dog can smell it, it’s a chemical compound that can be identified, no? And a test delivered? Unless the dog is picking up other cues as well.


I’d bet the vast majority of those cases involve police officers.
Let’s hope they aren’t involved in Covid testing of any kind.


tenor (1)


The chemical tests exists, the dogs did better than one and just slightly worse than the other.

It’s very possible that the dogs are detecting different compounds than these standard tests in wide use and that could lead to better tests. It might also be that the dogs are not sniffing something unique to COVID-19. The Guardian article switches between “Coronavirus” and “COVID-19” if the dogs are just detecting something related to Coronavirus then they could have an unacceptable false positive rate.


Considered that a PCR test requires to get a long stick put in your nose, having a dog sniffer, even if less sensitive,I think it has advantages.


The researchers claim that the dogs are detecting an odour given off by people who are infected with COVID-19.

This conjured up a vision of dogs getting a good whiff from your nose and “stealing your breath” like cats used to be accused of doing. It’s a witch!

Now, as to whether a retriever floats or not… :stuck_out_tongue_winking_eye:


In the US they’re referred to as “Sporting” dogs. Sporting Group – American Kennel Club


Yeah, maybe they can detect covid-19 or variants with a statistically significant rate greater than 50% in a controlled setting. And if it’s really a medically useful role to help make diagnoses then I’m all for that.

But I could easily see this being used to discriminate against marginalized groups by “finding” covd in shoppers, or not “finding” covid in patients. I hate to be cynical, but as you say, the track record of the misuse of drug-sniffing dogs suggests some healthy skepticism.


I say we develop a biological machine to sense and process small quantities of airborne chemicals then quickly refer against reference panels. We should make it mobile, self fuel and eject waste products. It should be able to quickly detect novel target conditions. It should also demand snuggles and have a lousy user interface.


… after waiting your turn in what can be a long line (or a short but damn slow line) when you’re thinking “a ton of these other waiting people are here because they have covid 19 symptoms”. So aside from the discomfort of the swab, there’s also the fact that they could have had some goob doggos come sniffing down the line a number of times and still also have time to stand around looking cute before the PCR could be performed.

Also when the machine shuts down to self-recharge energy all heck breaks loose and it emits the creepiest sounds.


I’ve seen similar articles from over a year ago. How long does it take to test and implement this stuff, anyway?

That sort of thing was claimed a year ago or so. It sounded fishy at the time already, and doesn’t seem to have caught on to any significant degree in the meantime.

I wonder about differences in performance between individual dogs, and whether any given dog is eager or bored at the time in question. There are also studies showing that in many cases involving drug-sniffing dogs and such, the dogs don’t react to actual smells as much as they react to subtle cues given off (however unconsciously) by their handlers. Colour me sceptical.

How many dogs would it take? That would be the real issue; breeding and training five hundred thousand dogs, by which time the need has passed. I suppose once a ‘force’ of dogs had been deployed new scents could be introduced quickly. That dogs can detect illness and perhaps even a specific disease is not surprising. The cancer detecting abilities of dogs have been reliably established. Motivating a dog to do it for an extended period of time might be a challenge :slight_smile:

That, too – the approach doesn’t scale well. Remember that you’d not just need to find and train dogs, you’d also need to find and train dog handlers. You can train people to perform antigen or PCR tests for COVID-19 fairly quickly; taking proper care of dogs, not so quickly.


Of course, if it is found to work reliably and effectively, legislatures in QOP-run states will pass laws banning dogs.