Tuesday, January 12, 2010

Absolute versus Relative Risk

I wrote this piece for a newspaper:
These days many people have their eyes lasered. They have a LASIK operation to get their short-sightedness corrected. The surface of the eye is cut, a laser burns off a small layer, and the surface is glued back again. The healing process is relatively straightforward and painless, and the results are typically superb. I had LASIK myself. However, the risks are non-zero: 1 out of 10'000 operations is associated with severe problems according to one study. Would you recommend LASIK to a friend? How do you decide?

What is your answer? Maybe you said that it depends on the risk aversion of your friend or how much short-sightedness is affecting life quality. If so, you have committed a logical fallacy like most people do when evaluating risks. Scientific knowledge (here the risk of the operation) is useless unless you actually know how to use this information in taking rational decisions. You are using absolute risk instead of relative risk.

If you decide not to have LASIK, you must follow an alternative. Not having LASIK means either wearing no glasses, glasses, or contact lenses. Most people who opt for LASIK are wearing contact lenses. I looked up the following risks: 1 in 100 contact lense wearer risks a severe eye infection within 10 years. I read a story where this macrobe was eating out the
cornea of the eye. Terrible. May be you get my point? LASIK might actually be safer than contact lenses, because contact lenses carry a lot of risk. How about wearing glasses? Wearing glasses has probably a lower risk than LASIK, because you don't operate on the eye or risk infections but what if you have an accident and the glasses get pressed against your eye? Still glasses are probably safer. Another option is not wearing glasses but this would be a significant reduction of quality of life. You need to look at the relative risks between the alternatives you have.

Here is another example. Jumping out of the window of a third floor risks severe bone breaks or even death. Would you do it? If your answer is no, you have still not understood! You are still deciding based on absolute risk. Your answer should be: How risky is the alternative i.e. not jumping! Still not convinced? OK. Let me lock you in a third floor apartment and set the whole building on fire. 20 minutes later you are on the window sims. Still not jumping? Now you will jump because the alternative is worse i.e. burning alive. It is all about relative risk. Of course, in most circumstances you do not jump out of the window, but only because the alternative is just watching out the window. Of course, the higher the absolute the risk, the more likely the alternative is the better option.

Ever wondered why soldiers in the 18th century move towards the fire of the enemy as opposed to run away? Simple, because the alternative was even worse, i.e. the 100% probability of being shot by your officer for desertion in combat! Walking towards the fire of the enemy is actually the much safer option! I realized this when playing paint balling with friends at university and we simulated the battle of Gettesburgh!

A more serious topic - vaccinations. Take the flu jab scare. A study shows that vaccinations have risks: a few kids will get severely handicapped. The snap reaction of some concerned parents, mostly mothers, is not to do the vaccination to avoid the risk. What they do not realize is that therefore they do the alternative which is having a child without vaccination. And those kids are also handicapped or worse die. So the question should be that of numbers: which alternative harms the least children. This is a good example of how a good intention of parents actually leads to a bad outcome. Acting in the best interest might be the worst outcome!

So to summarize, whenever you are scared of a risky choice, check whether it is more risky with respect to the alternatives you have to follow.

No comments: