Andreas Rentz/Getty Images

As the COVID-19 pandemic continues to spread, early screening of SARS-CoV-2 infected individuals is key in interrupting infection chains.

Over the past months, several teams of researchers in Germany, the UK, and the U.S have been training dogs to identify the coronavirus in humans.

Coronavirus-sniffing dogs could be a solution for screening public places like airports and stadiums where dogs can detect infected individuals—including those who are asymptomatic—potentially screening as many as 250 people an hour. Individuals positively identified by dogs can then be further tested by serological tests to confirm infection. In Dubai airport, dogs are already being deployed since August to sniff sweat samples taken from randomly selected travelers on arrival.

When researchers train dogs to recognize COVID-19, they’re training them to recognize smells that people with the illness release, not the virus itself, which is odorless. As your body works to fight a disease, it releases specific odors that scientists call “volatile organic compounds”.

To train dogs, researchers must first safely collect samples from people—their saliva or sweat—and test them for COVID-19. To make sure dogs are not exposed, samples are treated to render the virus inactive, and a filter is applied to them that acts like an N95 protection mask.

Dogs have already proven to be extremely efficient at disease detection for cancer, malaria, or diabetes, thanks to their amazing olfactory abilities. Experienced detection dogs can be trained on a new odor compound within 6 to 8 weeks, but in the case of COVID-19, several  logistical challenges need to be figured out.

Disease detection dogs are an elite bunch, trained by specialized dog handlers. Currently, these dog and trainer teams are few and far between, and on a large scale, a screening program would require far more dogs and trainers than we can currently supply.

Another issue is that dogs in research labs are protected, but in a real-life situation they would potentially be exposed to infections. At present, very few dogs have contacted the coronavirus through human contact (mostly through their owners), but on a large scale, the risks of contamination could increase dramatically. Dogs, however, seem to have a high degree of immunological resistance to COVID-19. In the rare cases of dogs who did get infected, the animals were already suffering from an illness before they contracted the virus.

Also, there is a difference between sniffing out isolated samples in a lab—with a high success to sample ratio—and detecting the virus randomly on people in large crowds with unpredictable rates of success. The testing protocols will need to be adapted to real-life conditions to keep dogs motivated, and realistically, that could take months.

A pilot study in Germany has shown average detection rates of 94 percent. But dogs’ abilities are not perfect—neither are the standard nasal swab COVID-19 tests—and the sensitivity and accuracy of the dogs might vary depending on viral load. In Dubai Airport, the dogs are a “backup” precaution, not a direct determinant of whether people can get on planes.

We have yet to see dogs being deployed among the public on a large scale for coronavirus screenings. By the time we do, the benefits of training specialist dogs may not outweigh the logistical problems. Soon vaccines and better diagnostic tools will become available, but detection dogs may still be a viable option for early screening if scientists can reliably reproduce the detection methods in real-life scenarios without putting the dogs at risk.