I watched an interesting TV programme yesterday, which showed various examples and tests, some under laboratory conditions, of dogs detecting various types of cancer by use of their incredible sense of smell. Cancer tissue, equivalent to what was described as '2 parts per trillion' was hidden in test apparatus, along with normal tissue hidden in similar apparatus. The test dogs, which of course had been trained to seek out such diseased tissue, checked the containers, then went immediately to the correct one. The trainers who accompanied the dogs didn't know which container held the diseased tissue until after their dog had made its choice, so could not give any signals to the dog. A general accuracy level with various dog types of 8 correct from 10 was quoted, which is far above anything which could be out down to chance or luck.
In some cases, pet dogs had become excited and stressed about lesions on their owners legs licking them constantly, until immediately after the owner underwent initial medical treatment, when they lost interest, even though the scar tissue was still there.
Some experts felt there was sufficient evidence to continue investigations which could prove beneficial to cancer detection, but others felt it was too unreliable to be of any practical use.
What was not in doubt was the incredible power of a dog's sense of smell. The figure quoted above, of '2 in a trillion' was not explained fully, and I know that a trillion means different numbers in Europe and USA. As an example though, one expert said it was the equivalent of 1 second in 320 years. I haven't worked this out, but it's a miniscule amount, and a pretty weak smell too!

Sounds like the end of the programme I caught on TV last night. The outstanding performance of one dog who correctly identified a breath sample taken from a woman who had cancer five years previously, and which had not been stored in lab conditions (ie, it was in a non-airtight container in the woman's garage) was astounding, and certainly adds weight to the theory that dogs could be a very useful tool in detecting this illness. I have to admit that I was sceptical when I read the blurb beforehand, but having watched some of the programme I have to admit to now being half-way convinced that this could actually work and be worth looking into much more seriously.