Dog Savior: Immediate Scent-Detection of SARS-COV-2 by Trained Dogs
Vesga, O.; Valencia, A. F.; Mira, A.; Ossa, F.; Ocampo, E.; Agudelo Perez, M.; Ciouderis, k.; Perez, L.; Cardona, A.; Aguilar Perez, Y.; Gonzalez, J. M.; Catano Correa, J. C.; Agudelo Berruecos, Y.; Hernandez-Ortiz, J. P.; Osorio, J. E.
Show abstract
Molecular tests for viral diagnostics are essential to confront the COVID-19 pandemic, but their production and distribution cannot satisfy the current high demand. Early identification of infected people and their contacts is the key to being able to isolate them and prevent the dissemination of the pathogen; unfortunately, most countries are unable to do this due to the lack of diagnostic tools. Dogs can identify, with a high rate of precision, unique odors of volatile organic compounds generated during an infection; as a result, dogs can diagnose infectious agents by smelling specimens and, sometimes, the body of an infected individual. We trained six dogs of three different breeds to detect SARS-CoV-2 in respiratory secretions of infected patients and evaluated their performance experimentally, comparing it against the gold standard (rRT-PCR). Here we show that viral detection takes one second per specimen. After scent-interrogating 9,200 samples, our six dogs achieved independently and as a group very high sensitivity, specificity, predictive values, accuracy, and likelihood ratio, with very narrow confidence intervals. The highest metric was the negative predictive value, indicating that with a disease prevalence of 7.6%, 99.9% of the specimens indicated as negative by the dogs did not carry the virus. These findings demonstrate that dogs could be useful to track viral infection in humans, allowing COVID-19 free people to return to work safely.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Validation of a rapid, saliva-based, and ultra-sensitive SARS-CoV-2 screening system for a pandemic-scale infection surveillance 94%
- Predictive values, uncertainty, and interpretation of serology tests for the novel coronavirus 94%
- Surveillance of Vermont wildlife in 2021-2022 reveals no detected SARS-CoV-2 viral RNA 94%
Similar papers in this journal
Similar papers in this journal
- High prevalence of SARS-CoV-2 antibodies in pets from COVID-19+ households 91%
- Evaluation of a multi-species SARS-CoV-2 surrogate virus neutralization test 91%
- Active surveillance of cats and dogs from households with human COVID-19 cases reveals over one quarter of pets infected with SARS-CoV-2 in 2020-2021 in Texas, United States 91%
Similar papers in this journal
- Investigating sensitivity of nasal or throat (ISNOT): A combination of both swabs increases sensitivity of SARS-CoV-2 rapid antigen tests 93%
- Comparison of Saliva and Mid-Turbinate Swabs for Detection of COVID-19 92%
- Evaluation of the INDICAID™ COVID-19 Rapid Antigen Test in symptomatic populations and asymptomatic community testing 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.