Vincent Li
Dogs and their sense of smell have been lauded throughout history. First, they were used as tracking animals for hunters seeking prey. Then, various breeds of dogs served as rescue animals to pinpoint survivors in natural disaster scenarios or as detection dogs to sniff out illicit goods. More recently, our canine companions are sniffing out the faintest of biochemical signals to detect illnesses with extreme accuracy. Considering their successful track record in detecting various forms of cancer, with a 97 percent accuracy rate in smelling out positive cases through blood samples (ScienceDaily 2019), some researchers have begun testing whether their olfactory capabilities extend to the realm of COVID-19 (Else 2020). If proven to be efficacious, dogs could represent the next frontier of screening for COVID-19 due to their cost-effectiveness and overall efficiency.
How exactly are dogs capable of smelling such faint volatile compounds in the first place? Unlike humans, who have approximately 5 million olfactory receptors, dogs have over 200 million (Rozenbaum 2020). Furthermore, canine noses are far more sensitive: being able to detect concentrations as low as one nanogram of a compound in one liter of solution, or one part per trillion (Rozenbaum 2020). This combination of increased olfactory receptors and sensitivity confers their keen sense of smell. Normally, the reason why dogs are so astute regarding the presence of diseases or onset of sicknesses is because they can smell volatile organic compounds (VOCs) that are released by individuals with particular disorders. Although not known what biochemical compounds are being released by COVID-19 positive people, dogs have shown an outstanding predisposition towards accurately detecting SARS-CoV-2.
A pilot study performed by Jendrny et al. in July of 2020 found that the trained dogs correctly detected 157 positive and rejected 792 negative cases, but incorrectly detected 33 positive cases that were actually virus-free and falsely rejected 30 positive ones. Overall, the dogs possessed an “average diagnostic sensitivity of 82.63%... and specificity of 96.35%” (Jendrny et al. 2020). Sensitivity is defined as being able to correctly identify a positive sample, and specificity is the ability to correctly reject a false sample (Swift et al. 2019). The prospect of using canines to sniff out infected individuals seems promising, but per the authors, will require broader tests to confirm the sensitivity and specificity rates. Furthermore, the methodology by which they train the dogs may have to be refined in order to increase the accuracy of the dog’s abilities even further. In the case of this pilot study, the detection dogs spent one week training with salivary and tracheobronchial secretions from infected human patients (Jendrny et al. 2020). Whether a superior regimen for improving detection rates exists will require further experiments beyond those of Jendrny et al. (2020).
Even with these drawbacks, how exactly do these results compare to the current gold standard of covid testing: RT-PCR? RT-PCR, which refers to real-time reverse transcription polymerase chain reaction, detects the level of viral mRNA that is present within the patient sample. Reverse transcriptase uses these mRNAs as templates to produce complementary DNA (cDNA), which is then amplified in number using polymerase chain reaction, and the cDNA produces a fluorescent signal that confirms that the virus is actively present within the patient once a threshold has been reached (Jawerth 2020). RT-PCR has a sensitivity of approximately 70% and a specificity of 95% based on the “lower end of current estimates” reviewed by Watson et al. (2020). Although the sensitivity reaches upwards of 98%, this demonstrates that trained dogs possess a comparable level of accuracy as does the current standard used for COVID-19 detection (Watson et al. 2020).
Despite their potential as a screening tool for COVID-19, the trained dogs exhibit a number of drawbacks that need to be addressed before their implementation may be finalized. The first issue is that the study design was highly regimented and not reminiscent of a real-life scenario in which the dogs must screen dozens of people within a short time frame. All samples were inactivated using beta propiolactone to prevent the dogs and the handlers from becoming infected with SARS-CoV-2 during training and the subsequent experiments, and this was followed by an incubation period of about three days (Jendrny et al. 2020). This safety precaution is not feasible if the point of using trained dogs is for the explicit purpose of timeliness and cost-effectiveness. If not feasible, then the concern is that the dogs may eventually become infected with SARS-CoV-2 due to their potential to be exposed to high viral loads. Such infection then raises questions as to whether the accuracy of their detection diminishes as a result because the dogs’ olfactory systems may become altered (Jendrny et al. 2020). Most importantly, the positive samples procured for Jendrny et al. pilot study were obtained from symptomatic patients, so whether the dogs display such high levels of sensitivity and specificity for asymptomatic or presymptomatic carriers has yet to be confirmed (Jendrny et al. 2020). The work of Jendrny et al. establishes dogs as potential screeners for COVID-19, but serves as a reminder that further experimentation is necessary before their results can be practically applied.
Still, dogs can provide a faster, less invasive, and more cost-effective way to screen people for COVID-19. Even with the extensive vaccination efforts across the world, the need to continue screening for sick people with various respiratory illnesses presents a very real need for diagnostic methods. If dogs can be trained to do so within a matter of weeks or even months, they can provide a strong supplement to more traditional molecular or antibody-based tests. For regions with limited testing kits and resources available for COVID-19 diagnostics, these disease detection dogs may represent an alternative that can handle a large volume of people in public areas such as airports, reducing the impact of potential superspreaders. If not for COVID-19 now, dogs and their shrewd noses may be applied for a myriad of diseases.
References
Else H. 2020 Nov 23. Can dogs smell COVID? Here's what the science says. Nature News. [accessed 2021 Mar 14]. https://www.nature.com/articles/d41586-020-03149-9#:~:text=Around%20the%20world %2C%20canines%20are,virus%20with%20almost%20perfect%20accuracy
Jawerth N. 2020 Jul 16. How is the COVID-19 virus detected using real time RT–PCR? IAEA. [accessed 2021 Mar 14]. https://www.iaea.org/bulletin/infectious-diseases/how-is-the-covid-19-virus-detected-using-real-time-rt-pcr
Jendrny P, Schulz C, Twele F, Meller S, von Köckritz-Blickwede M, Osterhaus AD, Ebbers J, Pilchová V, Pink I, Welte T, et al. 2020. Scent dog identification of samples from COVID-19 patients – a pilot study. BMC Infectious Diseases 20.
Rozenbaum M. 2020 Jun 19. The Science of Sniffs: Disease Smelling Dogs. [accessed 2021 Mar 14]. https://www.understandinganimalresearch.org.uk/news/research-medical-benefits/the-science-of-sniffs-disease-smelling-dogs/#:~:text=Dogs%20are%20most%20famously%20known,cancer%20based%20on%20breath%20samples
Study shows dogs can accurately sniff out cancer in blood. 2019 Apr 8. ScienceDaily. [accessed 2021 Mar 14]. https://www.sciencedaily.com/releases/2019/04/190408114304.htm
Swift A, Heale R, Twycross A. 2019. What are sensitivity and specificity? Evidence Based Nursing 23:2–4.
Watson J, Whiting PF, Brush JE. 2020. Interpreting a covid-19 test result. BMJ:m1808.
Image: https://snappygoat.com/b/4d2435ba96aa1a80366d54e68bc15be4803c487d
How exactly are dogs capable of smelling such faint volatile compounds in the first place? Unlike humans, who have approximately 5 million olfactory receptors, dogs have over 200 million (Rozenbaum 2020). Furthermore, canine noses are far more sensitive: being able to detect concentrations as low as one nanogram of a compound in one liter of solution, or one part per trillion (Rozenbaum 2020). This combination of increased olfactory receptors and sensitivity confers their keen sense of smell. Normally, the reason why dogs are so astute regarding the presence of diseases or onset of sicknesses is because they can smell volatile organic compounds (VOCs) that are released by individuals with particular disorders. Although not known what biochemical compounds are being released by COVID-19 positive people, dogs have shown an outstanding predisposition towards accurately detecting SARS-CoV-2.
A pilot study performed by Jendrny et al. in July of 2020 found that the trained dogs correctly detected 157 positive and rejected 792 negative cases, but incorrectly detected 33 positive cases that were actually virus-free and falsely rejected 30 positive ones. Overall, the dogs possessed an “average diagnostic sensitivity of 82.63%... and specificity of 96.35%” (Jendrny et al. 2020). Sensitivity is defined as being able to correctly identify a positive sample, and specificity is the ability to correctly reject a false sample (Swift et al. 2019). The prospect of using canines to sniff out infected individuals seems promising, but per the authors, will require broader tests to confirm the sensitivity and specificity rates. Furthermore, the methodology by which they train the dogs may have to be refined in order to increase the accuracy of the dog’s abilities even further. In the case of this pilot study, the detection dogs spent one week training with salivary and tracheobronchial secretions from infected human patients (Jendrny et al. 2020). Whether a superior regimen for improving detection rates exists will require further experiments beyond those of Jendrny et al. (2020).
Even with these drawbacks, how exactly do these results compare to the current gold standard of covid testing: RT-PCR? RT-PCR, which refers to real-time reverse transcription polymerase chain reaction, detects the level of viral mRNA that is present within the patient sample. Reverse transcriptase uses these mRNAs as templates to produce complementary DNA (cDNA), which is then amplified in number using polymerase chain reaction, and the cDNA produces a fluorescent signal that confirms that the virus is actively present within the patient once a threshold has been reached (Jawerth 2020). RT-PCR has a sensitivity of approximately 70% and a specificity of 95% based on the “lower end of current estimates” reviewed by Watson et al. (2020). Although the sensitivity reaches upwards of 98%, this demonstrates that trained dogs possess a comparable level of accuracy as does the current standard used for COVID-19 detection (Watson et al. 2020).
Despite their potential as a screening tool for COVID-19, the trained dogs exhibit a number of drawbacks that need to be addressed before their implementation may be finalized. The first issue is that the study design was highly regimented and not reminiscent of a real-life scenario in which the dogs must screen dozens of people within a short time frame. All samples were inactivated using beta propiolactone to prevent the dogs and the handlers from becoming infected with SARS-CoV-2 during training and the subsequent experiments, and this was followed by an incubation period of about three days (Jendrny et al. 2020). This safety precaution is not feasible if the point of using trained dogs is for the explicit purpose of timeliness and cost-effectiveness. If not feasible, then the concern is that the dogs may eventually become infected with SARS-CoV-2 due to their potential to be exposed to high viral loads. Such infection then raises questions as to whether the accuracy of their detection diminishes as a result because the dogs’ olfactory systems may become altered (Jendrny et al. 2020). Most importantly, the positive samples procured for Jendrny et al. pilot study were obtained from symptomatic patients, so whether the dogs display such high levels of sensitivity and specificity for asymptomatic or presymptomatic carriers has yet to be confirmed (Jendrny et al. 2020). The work of Jendrny et al. establishes dogs as potential screeners for COVID-19, but serves as a reminder that further experimentation is necessary before their results can be practically applied.
Still, dogs can provide a faster, less invasive, and more cost-effective way to screen people for COVID-19. Even with the extensive vaccination efforts across the world, the need to continue screening for sick people with various respiratory illnesses presents a very real need for diagnostic methods. If dogs can be trained to do so within a matter of weeks or even months, they can provide a strong supplement to more traditional molecular or antibody-based tests. For regions with limited testing kits and resources available for COVID-19 diagnostics, these disease detection dogs may represent an alternative that can handle a large volume of people in public areas such as airports, reducing the impact of potential superspreaders. If not for COVID-19 now, dogs and their shrewd noses may be applied for a myriad of diseases.
References
Else H. 2020 Nov 23. Can dogs smell COVID? Here's what the science says. Nature News. [accessed 2021 Mar 14]. https://www.nature.com/articles/d41586-020-03149-9#:~:text=Around%20the%20world %2C%20canines%20are,virus%20with%20almost%20perfect%20accuracy
Jawerth N. 2020 Jul 16. How is the COVID-19 virus detected using real time RT–PCR? IAEA. [accessed 2021 Mar 14]. https://www.iaea.org/bulletin/infectious-diseases/how-is-the-covid-19-virus-detected-using-real-time-rt-pcr
Jendrny P, Schulz C, Twele F, Meller S, von Köckritz-Blickwede M, Osterhaus AD, Ebbers J, Pilchová V, Pink I, Welte T, et al. 2020. Scent dog identification of samples from COVID-19 patients – a pilot study. BMC Infectious Diseases 20.
Rozenbaum M. 2020 Jun 19. The Science of Sniffs: Disease Smelling Dogs. [accessed 2021 Mar 14]. https://www.understandinganimalresearch.org.uk/news/research-medical-benefits/the-science-of-sniffs-disease-smelling-dogs/#:~:text=Dogs%20are%20most%20famously%20known,cancer%20based%20on%20breath%20samples
Study shows dogs can accurately sniff out cancer in blood. 2019 Apr 8. ScienceDaily. [accessed 2021 Mar 14]. https://www.sciencedaily.com/releases/2019/04/190408114304.htm
Swift A, Heale R, Twycross A. 2019. What are sensitivity and specificity? Evidence Based Nursing 23:2–4.
Watson J, Whiting PF, Brush JE. 2020. Interpreting a covid-19 test result. BMJ:m1808.
Image: https://snappygoat.com/b/4d2435ba96aa1a80366d54e68bc15be4803c487d
Proudly powered by Weebly