What statistics should we teach medical undergraduates and graduates?

It is suggested that the emphasis on teaching statistics to medical undergraduates has usually been quite wrong: that the courses have become much too long, too detailed, and irrelevant to the needs of the majority. Examples are given which may help to introduce the basic concepts which all medical (and dental) undergraduates require, and which form a basis for the more traditional teaching of analytical methods to the appropriate subset who proceed to undertake medical research.