From Mr J. D. Douglas-Boyd Sir: I have seen it reported that "nearly one in five deaths on the road are caused by drivers who have been drinking". This statistic is used to prove that drinking and driving is dangerous and I don't doubt that it is. However, the statistic is meaningless unless we know what proportion of the population drinks and drives routinely. If one in five drink and drive, then the death rate caused by these drivers will be same as that for non-drinkers. If one in four drivers routinely drink and drive, the statistic could prove that drink-drivers are safer than sober drivers.

Since it is unlikely that we will ever have an accurate survey of the entire nation's drinking and driving habits, we will never know what the one-in-five figure acutally means.

Yours faithfully, J. D. DOUGLAS-BOYD High Ercall, Shropshire 20 December