Sunday, June 12, 2011

Fukushima Fallout Continues

The world is full of well intentioned people with not grasp of statistical probabilities. The Fukushima radiation fallout will continue because statistically misguided, but well intentioned people seem to have to repeat poorly contemplated probabilities. It is not just radiation, it is every part of our lives that statistics are involved that suffer.

On CCN, a well educated professor spoke on the risk of minute levels of radiation causing health problems. He could not say that there is zero probability of one or two cases in millions that MAY result because of Fukushima fallout, because there is always a CHANCE. How do you quantify the chance for the population to understand?

In the professor's case, Fukushima Iodine 131 fallout in the United States as of the first of April 2011 could cause a person on the west coast to absorb 3 to 5 decays or counts per day. Five counts per day is equivalent to 0.000058 Becquerel or counts per second. Compare that to an average background radiation of 12 counts per minute or 720 counts per second and you have a 0.000% chance of any health impact. What? Not enough decimal places? How about 0.00000806 percent chance of cancer risk over the normal background? If that percent risk frightens you, play the local lottery. Someone has to win right?

Even that estimate, 0.00000806 percent is high. It is only the percent increase in radiation. The radiation threshold is approximately 500 times normal background, with 100 times normal background showing no increase cancer risk. So the risk is verging on astronomically small. There is a chance though.

In Japan, the risks are much higher. Still, the risk is very low if the cause were anything but radiation. The 500 times normal background threshold is a conservative estimate. Studies for individual radioisotopes place limits in the range of 1000 times before there is any statistically significant (i.e. possible) chance of cancer. Those studies generally use linear no threshold (LNT) methods to determine risk. LNT in itself is overly conservative as it does not consider nonlinear factors and can confuse other risks with radiation levels. So the gray area can be 50 times greater. So being 20 pounds over weight has roughly 60 times more risk that having radiation levels 100 times normal and about 20 times the risk of radiation levels of 500 times normal. There is still a risk.

What is acceptable risk? That is the question of the millennium. There will never be zero.

No comments: