Dogs are taught to distinguish infected with SARS-CoV-2 in biological fluids from uninfected. As stated in the article, published in BMC Infectious Diseases, detection sensitivity was 83% and specificity of 96 percent. According to scientists, current study results should be considered preliminary, the suitability of this method for detection of infected people in the field need to be explored in the future.
It is now known that using volatile organic compounds the dog can detect people have malaria, bacterial and viral infections and even cancer. However, this method of diagnosis of diseases is not widely applied in clinical practice.
Sex Gendry (Paula Jendrny) and her colleagues from the University of veterinary medicine Hannover suggested that in the case of COVID-19 dogs can help in preventing the spread of virus and new outbreaks of the disease — dogs-detectors could be used in public areas such as airports or sporting events, as an alternative or Supplement to laboratory tests to detect infected people.
The study involved eight dogs. After two weeks of habituation to the automated training system on average, dogs require five days of training, so that the detection rate was higher than random guessing. After a seven-day course of study, the scientists conducted a double-blind randomized control study. This study used samples from seven infected and seven healthy individuals. Two positive sample was tracheobronchial secret, the rest saliva.
Of all 1012 images of dogs were able to correctly identify 949 (94% ± 3.4%): 157 positive and 792 negative. The number of false positives was 33, and the false positive rate — 30. Dogs distinguish between infected and control samples with a sensitivity of 82,63 per cent (confidence interval of 82.02–83,24) and specificity of 96.35 percent (96,31–96,39). There was no significant difference in the ability of virus detection in saliva and the secret of the trachea.