5

I read on https://www.businessinsider.com/sniffer-dogs-answer-to-the-covid-19-testing-crisis-mirror-2020-4:

Medical detection dogs able to sniff 750 people an hour could help identify coronavirus cases, researchers say

How accurate are dogs to detect whether a human has COVID-19?

This question is a repost of How accurate are dogs to detect whether a human has COVID-19? [closed], which got deleted because some people complained no study has been done yet on that matter. Since there is no at least one study on it, I'm reposting it.

Improve this question

2 Answers 2

Reset to default
2

I wouldn't be surprised if dogs could detect COVID-19, as they can help detect hypoglycaemia in diabetics, warn people who are about to have an epileptic seizure and have been used to sniff out some cancers.

This article on Nature's website (23 November 2020) has the following

Canines seem to detect coronavirus infections with remarkable accuracy, but researchers say large-scale studies are needed before the approach is scaled up.

Improve this answer
-3

The 2020 study {1} found that dogs can detect COVID-19 with a sensitivity of 82.63% and a specificity of 96.35%:

The dogs were able to discriminate between samples of infected (positive) and non-infected (negative) individuals with average diagnostic sensitivity of 82.63% (95% confidence interval [CI]: 82.02–83.24%) and specificity of 96.35% (95% CI: 96.31–96.39%). During the presentation of 1012 randomised samples, the dogs achieved an overall average detection rate of 94% (±3.4%) with 157 correct indications of positive, 792 correct rejections of negative, 33 incorrect indications of negative or incorrect rejections of 30 positive sample presentations.

Sample size: 8 dogs, 1012 samples. Note that, as expected, there is variability between dogs in terms of sensitivity and specificity, as shown in Table 2.

YouTube video from the authors: https://youtu.be/lzDYsZfd-fY

Reddit thread commenting on that study and its caveats: https://redd.it/hz3d9m (mirror).

References:

  • {1} Jendrny, Paula, Claudia Schulz, Friederike Twele, Sebastian Meller, Maren von Köckritz-Blickwede, Albertus Dominicus Marcellinus Erasmus Osterhaus, Janek Ebbers et al. "Scent dog identification of samples from COVID-19 patients–a pilot study." BMC Infectious Diseases 20, no. 1 (2020): 1-7. https://doi.org/10.1186/s12879-020-05281-3
Improve this answer
4
  • 3
    As the very top comment in the reddit thread you linked, all of this is quite misleading. The positives all had positive symptoms and the positive and negative cases had different types of samples, so the sensitivity is vastly overestimated if you consider asymptomatic/presymptomatic cases. Additionally, a specificity of 96% is not good at all, especially in a test population with lots of uninfected people. This study is complete junk, and I think it's really misleading to post an answer about it here without caveats.
    – Bryan Krause ♦
    Jul 28, 2020 at 19:42
  • @BryanKrause "Additionally, a specificity of 96% is not good at all, especially in a test population with lots of uninfected people". This doesn't change the quality of the study (unlike the other valid points that you made).
    – Franck Dernoncourt
    Jul 28, 2020 at 19:48
  • 2
    It does when the authors report it as "high diagnostic specificity". The confidence intervals are also quite suspect, they seem to not consider the repeated tests with the same dogs, pretending like they have a large N when they only have 8 dogs and only samples from 7 people in each group.
    – Bryan Krause ♦
    Jul 28, 2020 at 19:58
  • 2
    +1 on the question, -1 on the answer. That really is a junk study. It won't surprise me a bit if dogs are capable of detecting COVID-19, but that study doesn't do it.
    – Carey Gregory ♦
    Jul 29, 2020 at 4:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.