The psychology of human error

Sir: I am responding to your invitation for comments on the article by Green on the above topic (Eur J Anaesthesiol 1999; 16: 148-55). Whatever the merits of the analogy between giving an anaesthetic and flying a modern aeroplane (and there must be strong doubts as to its reality [1]), there is one undeniable difference, the methods by which one can undertake either activity. Any doctor, provided (s)he can pass the necessary intellectual hurdles and sweet talk his/her way through two or three amateur selection interviews, can become an anaesthetist. That will not be enough for those wishing to be commercial pilots. What the aviation authorities recognise is an aspect which Green totally ignores. Although all human beings are prone to error, some are more error-prone than others. This quality of being 'accident prone' is one which must be screened out in potential pilots: it is not one which can be safely left to be discovered 'on the job'. To this end, rigorous use is made of personality testing and a highly refined personality profile exists which trainee pilots must resemble in key ways. We, on the other hand, airly dismiss such tests as 'Women's magazine psychology', all the more arrogant since we make no attempt to counter this by becoming expert in selection by interview. We, as a specialty, are apparently quite content to allow a propensity to commit errors to emerge either during training or after (depending on when the appropriate mix of events occurs), to the clear danger to patients. The range of personality traits which have no detectable effect on performance in Anaesthesia is quite wide, but a few unfavourable characteristics can be identified [2]. Having conducted an 8-year study of the selection of registrars accompanied by personality profiling [3], and observing the relative lack of accuracy of long-term predictions made by selection panels against those made by a professional psychologist, I remain convinced that selecting future anaesthetists with an appropriate attitude and instinctive responses is a necessary preliminary to the success of the sensible suggestions made by Green for reducing errors to a minimum. M. D.