Wednesday, 27 May 2009

Measuring and comparing risks

Yesterday's newspaper had a feature about the risks associated with competing in a triathlon. A study presented to the American College of Cardiology Conference had reported that the death rate among competitors in triathlons was about twice that of competitors in marathons. (1.5 per 100,000 competitors compared with 0.8). That news report closed with the comment from one of the staff at my university that the idea that exercise is dangerous should be compared with sedentary life.

O.R. professionals ought to be able to see through the nonsense of the report and the comment. What are you trying to measure? How do you compare one activity with another? The death rate in the U.K. is about 1 per 100 per year, or 1000 per 100,000. Dividing that by 365 and then by 8, we get 0.35 per 100,000 in a three hour period. So the death rate in marathons (which last 3 to 4 hours for the majority of competitors) is about twice the national death rate. But the rate varies with age and gender and lifestyle. However, the national death rate includes deaths from accidents, which generally affect the more mobile sectors of the population. The people who die outside marathons include the terminally ill, the aged, etc. -- not the sort of people who compete in endurance sport. They probably have an extremely small chance of dying of natural causes in the next three hours. But they have that risk of accident. So we can conclude that the person who decides to enter an endurance sport increases their chance of dying during that event. But the actual risk is still very small; the half-marathon that I mentioned earlier has about 2500 competitors. If the figures for marathons and half-marathons are comparable, there will be an avaerage of one death every fifty years.

But, even more seriously, the reports about when the deaths occur in triathlons, as all but one of those recorded were in the swimming sport, should alert organisers to warn the competitors about the risks of not being prepared for a long, frantic swim in water that is colder than in heated swimming pools.

As for me, I shall cycle home today. My risk of an accident is about 1 in 4000 based on an average of 1 accidents per ten thousand person miles (here) and a journey of 2.5 miles home. (This is about my experience -- I have been hospitalised three times in 40 years of cycling, with an average annual mileage of a little over 1000 miles, giving 3 accidents in 40,000 miles.)

No comments: