Feed on

Understanding Risk

Given the choice between having a well-trained medical professional attend to my BASIC medical needs (i.e. not things like heart surgery) or having access to the internet, online medicines, and a professional that was well-trained in statistics and actuarial studies, I would easily choose the latter,

If I were to cite evidence for you that suggests a particular lifestyle increases the risk of dying among a particularly effected group by, say, 25%, how would you respond? Like me, I am sure you’d be startled. Of course, when data is reported in this particular way, it is a reporting of relative risks and not the actual risk of death that any particular person faces.

When you encounter such information, ask yourself the question, “25% over what?” For example, if a total of 4 people in a population die of some malady, and exposure to some other risk factor, such as mosquitos or cancer or …, causes one additional person to die, then the increase in relative risk is 25%. Indeed, with numbers this small, it is going to be VERY hard to know if this is caused by the outside factor, or is just random variation.

What matters then? To me, it is the absolute risk. If exposure to some outside factor causes one additional death then I care over what total population this occurs. If, for example, we are talking about the population of a small city of 10,000 people, then the data reported above is not really a 25% increase in mortality risk, but rather a 1 in 10,000, or a 0.01% increase in your risk of dying due to this factor. Those are vastly different things. You might thing this is trite, or obvious, but be careful when you are in the presence of such data and you will often see the two notions conflated, badly.

Leave a Reply