Frequently Asked Questions
What are the aims of this project?
To train dogs to learn whether they can reliably detect the odour of COVID-19, and if so to deploy COVID-19 detection dogs to passive screen, i.e. without physical contact, any individual, including those who are asymptomatic, and indicate to dog handlers whether they have detected the odour of COVID-19. This would be fast, effective and non-invasive, and make sure the limited NHS testing resources are only used where they are really needed.
Which organisations are involved?
Medical Detection Dogs (MDD), the London School of Hygiene & Tropical Medicine (LSHTM) and Durham University.
Why use dogs?
Medical Detection Dogs has spent over ten years successfully researching the science behind a dog’s sense of smell, always adopting a rigorous scientific approach to its work. It has published research on canine detection of the odour of cancers, changes in blood-sugar levels in people with diabetes and bacterial infections in a number of peer reviewed journals. Together the collaborators on this project (MDD, the London School of Hygiene & Tropical Medicine and Durham University) have shown that dogs can detect malaria, the results of which were published in Lancet Infectious Diseases. These results have further supported the belief that diseases have their own unique odour that dogs have the ability to detect and we believe that dogs have the potential to detect the odour of COVID-19.
Why do you think COVID-19 has an odour?
All our research to date has indicated that every disease we have worked on has a unique odour. There is some evidence from chemical analysis that viruses have a unique odour and there is an existing research paper which demonstrates that dogs can be trained to detect viruses in cattle. Therefore, there is very good reason to believe that COVID-19 has an odour that will be detectable.
How will we gather the samples?
Training samples will be collected from NHS hospitals in an exercise coordinated by our collaborators at the London School of Hygiene & Tropical Medicine.
What sort of samples will they be?
We will capture the odour of the virus in a number of ways, including breath and body odour samples we expect to contain the virus.
When they say samples what do they mean, skin cells, live patients etc?
We will be taking samples of patient body odour, not human tissue. For example, collected on face masks. Material that would otherwise be incinerated as hospital waste.
If using patients, how will you protect trainers against COVID-19 infection?
We will not be testing directly on patients. Instead, frontline NHS research staff will be fully trained and equipped to safely collect odour samples, which will be safely prepared so they are non-infectious. Only when they are ensured to be non-infectious will the dogs and trainers be exposed to them.
How sure are you that the dogs will be able to sniff out people infected with COVID-19?
Dogs have been shown to be able to accurately identify other conditions and infections in humans such as fluctuating blood-sugar levels in diabetic patients, some cancers, Parkinson’s disease and malaria. We do not know if COVID has a specific odour, but we think it will, and if it does, we are confident that they will be able to detect COVID19.
What other diseases have dogs been trained to detect, and how successful have they been?
Dogs can sense small changes in odours produced by the human body associated with disease. Dogs have been trained to diagnose cancer in patients, alert diabetic patients that their blood sugar is low and many other diseases. In a study published in the Lancet Infectious Diseases last year we demonstrated that dogs can even detect the presence of the malaria parasite Plasmodium falciparum, in human odour with a sensitivity and specificity greater than 90%, which is above the World Health Organization’s standards for a diagnostic test.
Would this be instead of testing kits?
No, this will be in addition to testing. What this method offers is a very rapid screening of large numbers of people. The benefits of having a rapid screening method such as detection dogs could ensure the proper allocation of our limited test kit supplies, by identifying those who are most likely to be positive before they become symptomatic and infectious.
How do dogs detect a virus?
Dogs can detect small changes in body odour linked with infection.
How many dogs will you train?
The initial phase will see 6 Bio Detection dogs, the Super Six, undergo intensive training so that within 8 to 10 weeks we will be able to see if they can reliably detect the odour of COVID-19. If this is successful we will then move into a second phase to test them in live situations, following which we hope to work with other agencies to train more dogs for deployment.
How will the dogs be trained?
Dogs searching for COVID-19 would be trained in the same way as those dogs MDD has already trained to detect diseases like cancer, Parkinson’s and bacterial infections – by sniffing samples in the charity’s training room and indicating when they have found it. Following this stage of training we would then transfer to detecting the odour on individuals and the detector dogs would work in a similar way to our Medical Alert Assistance Dogs, which independent research has verified are highly effective.
When will we know if this is successful?
We expect to have early data on whether COVID-19 has an odour that the dogs can detect within 2 to 3 weeks of samples arriving at the Medical Detection Dogs training centre and in the following 6 to 8 weeks we will establish the dogs’ sensitivity and specificity over a large sample size.
How soon could dogs be screening people in airports and other public places?
The sooner we are able to start the research the sooner dogs can be deployed. We expect to be able to move from testing on samples to practical testing of the passive screening ability of dogs in no more than 3 months from the start of the project. We would hope to be working with potential deployment partners shortly after that so that ‘in theatre’ testing can be done alongside those who would help scale up the capacity of detector dogs needed. This could mean detector dogs deployed in under 6 months.
How would it practically work to have dogs at airports for example?
Dogs already work in a number of roles at airports including explosives and drugs detection. COVID-19 Detection Dogs would be deployed in a similar way working with trained handlers to passively screen, i.e. without physical contact, passengers at agreed points in the airport.
Would the dogs in the airports be the ‘super six’ or will you be training more if trial is successful?
We will train more if the trial is successful and work with other agencies who have experience to expand the deployment.
Will dogs spread the virus?
No, our dogs will be trained on a dead virus and then have no contact with the individuals they are screening but will sniff the air around the person. The dogs will only be permitted to be touched by the handler, which therefore means there will be very low risk of spread of the virus from the dog to their handler or to the people they live with.
The dogs live in family homes, how will you ensure these families don’t get infected with COVID-19 when the dog comes home from training?
The dogs will be working with samples that don’t contain live virus, so we believe the risk to be minimal. Nevertheless, we have a strict protocol to minimise any risk there might be which involves socially distanced handover of dogs by their fosterer, changing collars and leads between fosterer and the training centre, and cleaning dogs on arrival and departure. Strict procedures will be followed when dogs are deployed which will follow best practice and veterinary advice.
How much will it cost?
We estimate that the cost of research design, sample collection, proof of principle research and training to be approximately £500,000 and the cost of deployment training so that we are able to work with partners to roll out a deployment model to be approximately another £400,0000.
Who funds this research?
We are grateful to have received financial support from the UK government as well as philanthropic funding but we are currently looking for more funding and considering fast tracked funding calls on COVID19 from research councils.
What exactly will the money be used for?
Funding will be used for the work involved in research design, obtaining ethical and safety approvals, collection of samples and making them safe, training the dogs, analysing the data and developing a deployment model if the proof of concept research is successful.
Once you have funding, how long will it take to deploy the dogs?
We will first need to collect odour samples from infected patients in NHS hospitals with proven cases of symptomatic and asymptomatic COVID19 alongside odour samples from people how have been tested who definitely do not have COVID19 (following a recent negative test). Once all the samples are collected we will need to make sure the samples are safe (not able to infect dog training staff). We will use a randomised subset of the samples to train the dogs so that they can be taught the difference between infected and not infected people. From this point we anticipate it will take around 6-8 weeks until our dogs will be ready to be tested in double blind tests. We will then know how effective the dogs are at detecting the disease. They will then be able to be deployed.