My "there must be something wrong with this preprint" sense is tingling but it's hard to pinpoint.
medrxiv.org/content/10.110…
Summary: sniffer dogs can detect Long COVID patient sweat samples 51,1% of the time with a 0% false positive rate on controls.
medrxiv.org/content/10.110…
Summary: sniffer dogs can detect Long COVID patient sweat samples 51,1% of the time with a 0% false positive rate on controls.
Handlers are blinded to the sample locations, so not a Clever Hans effect. There's detail about how patients were sampled but not controls, though one presumes it's the same. I rather suspect unintentional scent marking distinctions, either in handling of samples...
... or in things that Long COVID patients tend to do (or not do) that the general public does not, rather than a dog-detectable odour eminating from their bodies. It's easy to accidentally mark samples with odours that humans cannot detect.
It's worth pointing out that even humans can do tracking to a surprising degree if only they put their nose to the ground:
researchgate.net/publication/66…
And even our inferior senses can readily detect the signs of recent traffic, such as newly crushed vegetation or disturbed soil.
researchgate.net/publication/66…
And even our inferior senses can readily detect the signs of recent traffic, such as newly crushed vegetation or disturbed soil.
But perhaps I'm just too pessimistic. Perhaps dogs actually can smell Long COVID on people. *shrug*
• • •
Missing some Tweet in this thread? You can try to
force a refresh
