Training dogs to sniff out COVID-19
They can sniff out counselors late blood sugar levels in diabetics, drugs, explosive chemicals used in bombs, and as many dog owners know any food in one hundred meter radius. dokes have notorious powerful noses with hundreds of millions of central sceptres that can pick up traces of substances at just one pop trillion. And so now teams around the world from Lebanon to the UK attesting out dog's olfactory abilities when it comes to sniffing out cubic nineteen. One of those putting hounds on the viral hunt is Dominic Cork a professor at the National Veterinary School of. In front first phase is to train the dog to put his nose in coon and sniff. So we knew that if he story then we put some positive sample in this goal and dogs are going to one whole week but they'll in the cones and everything is made as a game I'm Nichole Davis, and this is science weekly. We Got Dominique on the line to ask him a bit more about how you actually train dogs to sniff out a disease. Unfortunately, the audio isn't great so about that but the first question I wanted to ask Monique was exactly when he first decided to ton his dogs noses towards K. Nineteen well, it's I'm a I'm head of a Canine Sports Medicine unit that the vet, school in our fault. And and we are working a lot on working dog I'm also involved in search and rescue dogs instead thirty five years as firefighters. And I've always been working on Doug affection actually. So we also have a big program in the. Vet. School, which is Naza. He's in the goal of the program is to develop the medical detection dogs in in France and so when when the COVID did show up, we had a meeting It was on the ninth of March I remember and. The. First question was, what are we going to use samples? So we checked everything in the graffiti and we saw that the the the sweat under the armpit that would be very few chance of bessie of contamination and actually has no passive condemnation. The dog is not sensible. So we make so that the dogs do not tach at any moment, the samples than we started with such rescue dog from different fire departments. Minute Ducasse what two weeks to consider that it was working in the. And that's what we've been doing for six months. So, let's get to the nuts and bolts here. What is it that the dogs are sniffing Anita? You say you take samples from People's armpits. Similarly, people use an awful lot of deodorants and other toiletries at does that get in the way of things dogs sniffing the virus sniffing the? Effects of the virus when the virus enters add a sale, the viruses replicating also using the Senate. To produce his own proteins he's on molecules and these chemical molecules they have to go out of the buddy. They can go out food the European through the feces for the tears and through the sweat. So that's what the dogs are looking for, and that's been a quite a few studies in the past showing that insalled cultures different virus were producing different others. Let's go nemo. Valetta. Organic compounds. And that's what we are looking for now to answer questions regarding the utterance in perfumes and so on. The key point for these dogs is to have some top quality and fresh positive samples in all the to make the in printing. So we're GONNA need roughly eighteen positive samples that are fresh. We don't rely only on a on A. Positive results also asked samples to hospitals coming from people who have chemical symptoms. The scan that he's typical etcetera etcetera, and if you do it this way while the dogs reading in memory, the specific other and you can put any type of the. or perfume this is not a problem. It would be a problem if the people who are using only one brand of the audience in the same product. But the Zillions of different types of the in perfume. So the only Common Otter, the dogs are finding in the eighty something samples that they are sniffing at the beginning is the the covid and that's what we want to focus on that. Domini, how do you train the dogs I'm what do they do? Once they find a positive sample today sit down or bark or how do they signal that they found something To train, the dogs are using some what we call a faction guns that. Developed that's a good way to not have the dog in contact with the samples and so the first phases to train the dog to put his nose in the cone in sniff, and so we do that too. If he story then we put some positive sample in this in the dogs are going to for one whole week. Now going to put down those in the cones so that the imprinted with the specific honor of the positives and then with more. With some neutral, which means the swabs without anything, and then we put some negative samples and wing prisoner amount of corns. We increase the amount of negative samples. We put some some lines. We've only negatives on the positives and everything is made as a game. In other words, the dog is getting his story when he finds the positive most of the time, we asked the doctor sit in front of the. That's pretty easy to obtain. But if the dog is marking differently barking or scratching or whatever I don't care because it's the important thing is to have the dog marking correct simple. What is important is to keep the motivation of the dog and the motivation of the dog is coming through the interaction with the duck hander and through the fact that he's rewarded when he works good and plays with his a duck handler. Happiness is to keep going for working dogs. A key issue that some skeptics of this approach have raised is that you might be able to tell someone with covert from somebody who doesn't have cable but can you really tell apart someone who has covered from someone who has another virus like flu? For example what do you say to that? We are starting right now to check if the dogs are mocking? Some people with other types of virus infections or other type chronic disease like lung cancers, auditees, and so on. But there has been some studies a lot of studies trying to identify the volatile organic compounds coming from different types of virus that have been put in south counter and each time it shows that the other print of the virus coming from these vetting are any compounds is specific to a virus. Would we see in terms of practical results is that there's a lot of times where the dog has been more accurate than the PCR. We've got some people with negative that were marked by the dogs. Samples were remarkably the dogs we send back the anonymous number of these samples to the hospitals, the remade the PCR, the our positive. We also have some negative people that were marked by the dogs. We have a refugee at ten cases like this where we told the hospital. Okay. These people are positive for us and they couldn't get in touch with these people but these people went back to the hospital a few days later and they were clinically covid nineteen and most of the time they were with digestive simple. You have to keep in mind also that that when you when you look for virus in the nose. You don't look for the virus at the other end of the buddy. Just. How accurate is this at? What kind of results are you getting with dogs? The accuracy of the dogs is measured for two terms of sensitivity which means that the doug doesn't miss some positives and specificity which means deduct doesn't miss some negatives. Sensitivity is the most important and. The values that we obtained are between ninety up to nine hundred, nine point five percent in Dora sensitivity and the specificity is always close to a hundred percent. So this entity is that say ninety five percent while it means that you might have some false positive, but it's no big dipped. One or two percent of positives. The big deal would be to miss some positives and actually this is not something that happens with the dog. So you can take that in any sense results are good as long as the dogs are well trained.