According to a proof-of-concept study, dogs can sniff SARS-CoV-2 in urine samples with 96% accuracy.
Do you have a cotton swab stuck in your nose anymore? The “dog test” is far from the actual application, so it’s not perfect. This was because we could only distinguish between positive and negative results for dog-trained samples. Detection of SARS-CoV-2 failed when a completely new sample was presented.
Dogs are known to be able to sniff out scents specific to a variety of illnesses, and previous studies have shown that SARS-CoV-2 also has powerful features that dogs can detect in saliva and sweat samples. Is shown. In fact, research shows that dogs have already been deployed to detect COVID-19 at Dubai Airport. However, studies have shown that it was unclear whether dogs could detect the virus in urine samples. Viral load is usually low in urine samples.
Relation: Quick Guide: COVID-19 Vaccine in Use and How It Works
To understand this, a group of researchers first trained eight Labrador retrievers and one Belgian Malinois, known as Universal Detection Compound (UDC), an odor not found in nature in the environment. I recognized the scent of synthetic substances. They placed the compound in one of the 12 ports of the “Scent Ring” and rewarded the dog each time it responded to a port containing UDC.
After they learned to recognize UDC, researchers used a scent wheel to train dogs to respond to urine samples taken from SARS-CoV-2 positive patients. Samples were taken from 7 individuals (2 adults and 5 children) who were positive for SARS-CoV-2 and 6 children who were negative for SARS-CoV-2. In the training, the dog was given two scenarios. One port contains the target scent on the scent wheel, the other port contains the control or distractor scent, and the other the scent wheel contains all control or distractor scents. This is a scenario. The virus was inactivated by heat or detergent and became harmless to dogs.
After three weeks of training, the researchers found that all dogs were able to identify positive SARS-CoV-2 samples with an average accuracy of 96%. The overall specificity was 99%. This means that there were few false positives. However, the overall sensitivity was 68%. That is, there were some false negatives.
The reason for the low sensitivity may be due to the rigorous method of performing the test. If the dog did not respond and once passed through a port with a positive sample, it was labeled as a mistake, According to the statement..
“This is not the simple thing we are looking for in a dog,” Cynthia Otto, senior author and director of the Veterinary Working Dog Center at the University of Pennsylvania, said in a statement. “Dog needs to be specialized in detecting the odor of infection, but it also needs to be generalized to the overall odor of different people’s backgrounds.”
Indeed, the dog had a little trouble in the trial. According to the statement, they tended to distinguish the scent of the actual patient rather than the infected state, and were confused by samples from patients who had recently recovered from COVID-19 but had a negative test result. “The dogs continued to react to the sample and we kept telling them no,” Otto said in a statement. “But obviously, the patient’s sample still had something that the dog was keying in.”
However, the dog was repeatedly trained with the same sample from the same patient and could not be generalized to a completely new sample that is key to the actual application.
According to the statement, in future studies, researchers should train dogs with diverse samples and should not repeatedly test dogs with samples from the same individual. “Currently, researchers train dogs to detect if someone is infected with SARS-CoV-2 and if someone is vaccinated based on the odor remaining on the shirt, We carry out so-called “T-shirt research”. People wore it overnight.
“This study has collected far more samples (hundreds or more) than this first study and hopes to get closer to what dogs may encounter in their communities,” Otto said. Mr. says.
The findings were published in the journal on April 14th. PLOS One..
Originally published in Live Science.
