Screening individuals for COVID-19 with scent dogs at an airport has been shown to have similar accuracy to PCR testing
Pre-screening airport arriving passengers for COVID-19 with scent dogs has been found to be as accurate as PCR tests for the identification of infected individuals according to the results of a study by a team of Finnish researchers.
Currently, the identification of COVID-19 relies upon either a polymerase chain reaction (PCR) or the results of lateral flow tests. The PCR tests are very accurate, such that when someone has a positive result, there is a 98% probability that they have COVID-19. Although less accurate, antigen tests have a pooled sensitivity of 79% but a pooled specificity of 100%. Another potential yet rapid means of testing for COVID-19, involves using scent dogs. In fact, a 2016 study found that such dogs were able to differentiate between infected and non-infected cultured cells with bovine viral diarrhoea virus. The value of using dogs for the rapid detection of a virus is that the animals have an extremely sensitive olfactory system with a lower limit of detection of one part per trillion.
During the pandemic, a number of small scale studies demonstrated how scent dogs can detect COVID-19 in different body fluids and while these results have been encouraging, real-world data are needed to confirm the value of scent dogs, particularly in situations where rapid detection is needed; for example, the screening of people arriving at an airport.
For the present study, the team compared the accuracy of scent dogs to detect COVID-19 in airport arrivals with standard PCR tests. The dogs were initially trained to identify the COVID-19 scent and this process was subsequently validated by randomly presenting each of the dogs with both positive and negative samples over a period of a few days. In the final and real-world part of the study, the scent dogs screened skin swab samples from incoming passengers at an airport and for whom PCR tests were also performed. The main outcomes of interest were diagnostic accuracy (based on sensitivity, specificity and positive predictive value) and which were compared against the PCR test result.
Scent dogs and detection of COVID-19
A total of 4 dogs were used in the study and, during the validation part, the dogs had an overall accuracy of 92% with a sensitivity of 92% and a specificity of 91% compared to PCR test results.
For the real-world aspect of the study, the results for scent dogs and PCR tests were correctly matched for 97.7% of samples. However, there were few individuals actually infected with COVID-19; hence there was a very low (0.47%) positivity rate. As a result, the researchers included an additional 155 positive samples, of which the dogs accurately identified 98.7% of these as positive.
The authors concluded that while the validation part of the study was successful, the lack of positive samples in the real-world part made it difficult to confirm the sensitivity for detection although use of additional samples did offer evidence of convincing accuracy.
Kantele A et al. Scent dogs in detection of COVID-19: triple-blinded randomised trial and operational real-life screening in airport setting BMJ Glob Health 2022