Mad cow vs. car crashes, risk assessment:
In 1986, there was an outbreak in Great Britain of BSE where a total of 178,000 cows died.
Control methods were identified by 1988 and implemented in Britain from July that year, but it took several years for the measures to be fully applied in Britain and other countries which were at risk. 66.102.7.104
As of March 2001, 88 people--86 in Britain and two in France--have died from v-CJD (the human disease caused by eating meat contaminated by central nervous system tissue from BSE-infected cows).
It's not possible to calculate the rate of people who ate contaminated meat, who caught v-CJD. That's because we don't know how many people ate meat from infected cows, or what fraction of the meat from infected cows was contaminated. The prion is only present in central nervous system tissue, not muscle tissue. So, if you eat meat from an infected cow (only meat, not mixed with any brain, dorsal ganglia, or spinal cord tissue), the risk is zero.
Only the v-CJD deaths are from infected meat. There is a background rate of regular CJD, which is transmitted and/or inherited by other methods.
The numbers we do know, is that 178,000 cows were known to be infected, and only 86 people (so far) were killed by it, in Britain. The disease had spread widely in cow herds in Britain, for years, before it was detected and control measures adopted. Therefore, many people must have eaten meat from sick cows.
By contrast, the risk of dying of an auto accident:
Figures published today show there were 320,283 road casualties in Great Britain in 2000.
Risk from vehicle accidents:
3,409 people were killed (fourteen fewer than in 1999), 38,155 were seriously injured (down 2 per cent on 1999) and 278,719 were slightly injured (a similar level to 1999).
216.239.53.104
I assume the death rate from car accidents hasn't changed much, year to year. Notice, the 3409 number is deaths per year, while the 86 dead is the total for all years.
I have to admit, my initial claim is unsupportable, because it isn't possible to calculate the risk ratio (as I stated it). For auto accidents, you can calculate a risk ratio, of deaths/average trip length. My initial claim was that the risk of eating meat from an infected cow was less than the risk of an auto accident on the way home from the grocery store. For v-CJD, the risk ratio can't be calculated. I know how many people caught the disease, but I don't know (except very roughly, order of magnitude), how many people ate meat from infected cows. The numbers must have been at least in the hundreds of thousands.
So I will have to revise my statement, using the available facts. I'll have to use "the entire population per year" in my ratios, to calculate the risk of death, for both car accidents and V-CJD. Assume that, more or less, all people in Britain eat cow meat and take car trips regularly. This simplifys things a lot, because I can use the entire population of Britain in both ratios, as the "at risk" population. It also eliminates trying to calculate a "per event" ("per meal of meat from infected cow") risk ratio, which can't be done.
Assume that the 86 deaths all are attributable to a 2-year period when the disease was widespread in cows in Britain, before control measures were enforced. It had to be at least 2 years; if we use more years, the risk/year declines sharply. So, 86/2 = 43 deaths per year for the entire population of Britain. Compare that to 3409 deaths for auto accidents (same population, same time period). One risk is two orders of magnitude higher than the other. That's a difference in risk so large, I think it's meaningfull, even given all the assumptions and incomplete data I had to use. |
