Suppose we want to estimate the risk of getting (hence dying from) a disease-inducing-serving of beef during the current "Mad Cow Scare" in the U.S. To do this, we could estimate the risk of getting a disease-inducing-serving of beef during the British Mad Cow Epidemic. Clearly, by doing this we are only interested in estimating a WORST CASE SCENARIO, and we should keep this in mind. This is because the number of cows infected in the U.S. is clearly much lower than the number of cows that were infected during the British epidemic, and this is extremely likely to remain so (due largely to our current awareness of how to stop the disease's spread and our will to stop it).
Also, I feel it is difficult to estimate the U.S. situation. It seems possible that the public is not hearing about every BSE case. One reason for this is that BSE is fairly rare even in cows and only one in, at most, 1700 cows is being tested in the U.S. Also, U.S. beef producers saw what happened in Canada. It seems unlikely that they would want themselves and our country to suffer the economic consequences associated to reporting every possible case. All this makes estimating the US number challenging, and perhaps the British number will suffice for our purposes.
Hence we must ask, "What is our goal?". My goal is to find out whether the chance of getting CJD from tainted meat is low. Hence if the British number is low, then it will certainly serve my purposes. We found that this chance is about 1/10^8, which is plenty low for me.