Spot probability estimates

It is not too uncommon for information to be given by the press (and hence not very reliable) from which spot probability estimates may be made about things that go wrong.  One area in which this some times occurs is medically-induced illness and mortality (medical treatment errors) in the hospital system.

Another one that I heard recently concerned the death of a homeless man who apparently had sought a dry night’s sleep in a large rubbish bin.  This was checked by the driver of the waste removal vehicle (by banging the sides and moving it roughly, as was standard practice) before being lifted and the contents dropped into the body of the vehicle.  It was only later that the body of the man was found amongst the rubbish.  At a hearing into the death, the rubbish removal company cited the fact that in a recent period they had handled 13 million bins, in which 100 people had been discovered.  This was the first death that had arisen in this way.  Hence the probability of finding a person in a bin is 7.7E-6 and of a person dying in this way is 7.7E-8.  It is interesting and an indication that the person presiding over the hearing is enlightened on the subject of risk, that the waste removal company was not found at fault in anyway.  In the great majority of cases of which I am aware, judges, magistrates and coroners are not able to understand the implications of such figures.

Leave a Comment