Posts Tagged ‘NCI’

Thinking about risk

March 21, 2009  |  General  |  24 Comments

I’ve recently been looking over the statistics about this blog. The most popular posts are those which talk about common issues — croup and concussions, for example. But one of the most often read posts gets its popularity from people using search engines like Google to answer this question: how risky are x-rays, especially CT scans, to children? You can read my actual posts about that here and here, but what struck me most about the popularity of this topic is what it tells us regarding how we think about risk. In particular, how do we tend to think about the risk of events occurring which are very rare, but which carry grave consequences if they happen? Lawyers call these events “small probability — large loss events.” Economists have studied the subject quite a bit, too, especially as it relates to investment decisions people make.

We humans are not entirely rational when we think about risk. We tend to focus on low-probability but high seriousness events, particularly if we are thinking about them in the context of choosing to do or not do something. So, for example, if your child needs a CT scan, as a parent you many think about how the radiation in that test increases your child’s chances of getting cancer. What we don’t think about is that your child is far, far more likely to suffer harm in a car accident while you are driving to the CT scanner than he is to suffer harm from the scan. But since we drive our children around every day, we don’t think much about that risk.

According to the National Cancer Institute, a child’s overall risk of developing any form of cancer is 1-2/10,000 children, or 0.01-0.02%. Also according to the NCI, this number has changed very little, if at all, over the past 30 years. The use of diagnostic x-rays in children, especially CT scans, has increased enormously during that time, so we should be reassured by these statistics. Even so, radiologists are increasingly vigilant about how they can reduce radiation exposure when they use x-rays.

Bottom line — it is always worth asking if the risk of a test exceeds the value of the information the test will give. But for x-rays, the benefit virtually always outweighs the risk.