Small Pox (Probability, but not poker)

The threat of the terroristic use of smallpox is in the news these days. I have often heard that the probability that a terrorist could create 10 small pox cases in the U.S. is very difficult to estimate and it is hard to justify vaccinating the U.S. population without a good estimate. So there are a few probability questions that come to mind:

1) What is your best guess of the probability that a terrorist will cause 10 people in the US to get a smallpox virus in the next 30 years?

2) How large does that probability need to be to justify vaccinating the US population.

Assume that 10 people die per million vaccines. Assume that if a terrorist succeeds in infecting 10 people and we do not vaccinate, then 100,000 people in the US are likely to die of small pox before smallpox is eradicated (again).