It can be unwise to pay attention to research published on a pre-print server before it has been through a peer-review process. Although passing through peer-review is itself no guarantee of quality, the process represents at least one layer of scrutiny. Nonetheless, some pre-prints just catch the eye. Like
this one from France, titled
“Screening for SARS-CoV-2 persistence in Long COVID patients using sniffer dogs and scents from axillary sweats samples.”
So here’s the short version: In this study, dogs trained to suss out the odors of chemicals generated by a SARS-CoV-2 infection identified them in sweat samples from 23 out of 45 patients with long Covid but in none of 188 samples from healthy controls. If the canine olfactory skills are accurate and these results hold up under review, the lack of any false positives in particular is especially remarkable. The findings imply that a sub-group of these long Covid patients–around half–could be experiencing a chronic SARS-CoV-2 infection.
Besides the fact that this is a pre-print, here’s another cautionary note–only the abstract appears to have been posted. Given many unknowns about the study methodology, including how the long Covid patients were selected, evaluating the results is difficult if not impossible. Yet it still grabbed the attention of Eric Topol, a well-known cardiologist and author. Topol’s
tweet about the study noted: “It’s one thing for dogs to pick up Covid. It’s another to discriminate people with #LongCovid.”
With those caveats, below is the abstract.
“
Objectives: Dogs can be trained to identify several substances not detected by humans, corresponding to specific volatile organic compounds (VOCs). The presence of VOCs, triggered by SARS-CoV-2 infection, was tested in sweat from Long COVID patients. Patients and methods: An axillary sweat sample of Long COVID patients and of COVID-19 negative, asymptomatic individuals was taken at home to avoid any hospital contact. Swabs were randomly placed in olfaction detection cones, and the material sniffed by at least 2 trained dogs. Results: Forty-five Long COVID patients, mean age 45 (6-71), 73.3% female, with prolonged symptoms evolving for a mean of 15.2 months (5-22) were tested. Dogs discriminated in a positive way 23/45 (51.1%) Long COVID patients versus 0/188 (0%) control individuals (p<.0001). Conclusion: This study suggests the persistence of a viral infection in some Long COVID patients and the possibility of providing a simple, highly sensitive, non-invasive test to detect viral presence, during acute and extended phases of COVID-19.“
