Dogs and Cancer

[Megan Jones]

There is some evidence that dogs can be trained to sniff out cancerous cells in humans. What I am wondering is whether they instinctually sense the odour to indicate good or bad news? Surely if positive reenforcement is used to hone these skills, the animal would come to associate the smell with rewards?
And do they smell only human cancers or cancer in themselves and other dogs? If so, what would it mean to them (it's my understanding that they don't understand the concept of death)?

[Wendell Wagner]

Here's a scientific study which appears to confirm this:

http://www.sciencedaily.com/releases...0106002944.htm

Note that the dog doesn't see the person at all. The dog smells samples taken from a number of people and picks out the one who has cancer. The best guess is that there are chemicals in cancerous tissue which a dog can smell. Dogs have a very strong sense of smell. There's no reason to suspect that dogs normally can smell cancer in ordinary situations. This might only work in laboratory settings where the only smell around is the tissue itself, already removed from the person.

[Scarlett67]

I remember reading that the search-and-rescue dogs at the WTC were getting too discouraged by finding only dead bodies, so the human rescuers started taking turns hiding so the dogs could "find" a living person now and then. Or maybe it was the site of the OK City bombing.

A bit of Googling seems to indicate that this is SOP at disaster scenes with a high body count. Also, here's a site that backs up my hazy memory about the WTC and OKC (scroll to the bottom).

This would seem to indicate that dogs don't necessarily associate a negative "success" (finding a dead body) with something good like a reward. Whether they can extend this to smelling cancer is a whole 'nother question. I've heard of pets pawing and whining at someone's body part, only to have the person find out they had some disease that the dog apparently smelled. So perhaps they can tell that what they smell is "bad," even if they are trained and rewarded if they find it.

[Bijou Drains]

Here's a cat who seems to predict death in a nursing home

http://www.cbsnews.com/stories/2007/...n3097899.shtml

[t-bonham@scc.net]

Originally Posted by Bijou Drains

Here's a cat who seems to predict death in a nursing home

http://www.cbsnews.com/stories/2007/...n3097899.shtml

That one has been pretty well debunked.

A classic case of confirmation bias.

The cat sleeps somewhere every night. Do they have somebody die every morning? Of course not. But they only remember the times when someone died the morning after the cat slept in their room.