Teaching concepts in epidemiology.

In the United Kingdom most medical students are first introduced to epidemiology as a formal discipline during their clinical years-usually during a course in social or community medicine. The amount of time allocated to the subject varies greatly between schools, but everywhere epidemiology is a recent addition to an already overcrowded curriculum and, like all parvenus, has a need to justify its position in terms that everyone can understand. Such justification must of course be to clinicians who have often had to yield time and who are rightly suspicious of any new arrival; but, perhaps more importantly, it must also be to students who reasonably demand that any new burden must be not only intellectually satisfying but of obvious relevance to their immediate clinical interests. Moreover, having tasted at last the delights of clinical doing, students have a natural distaste for anything that seems to demand even a partial return to some form of systematised classroom learning. The placing of epidemiology in the clinical years has, of course, great advantages for the teacher. He can build an understanding of the uses of epidemiology, as well as its methods, on the particular clinical interests and insights of individual students, so that they can feel it is they who have provided the starting points for teaching. This helps to bring a potentially dull subject alive and to allow epidemiology quite quickly to assume its essential role as a unifier of clinical, social, and preventive concepts of health and disease, instead of appearing as the mere servant of one. There are, however, ways in which the teacher of epidemiology stands greatly disadvantaged next to his clinical colleagues. The clinician's justification'of his activities, the immediacy of his time scales, the spiritual rewards of his action, need no conscious transmission to the student; they enhance (when they do not obscure) the scientific as well as the humanitarian basis of his thinking. But the epidemiologist has no such allies for his science-his activities seem remote, his time scales long, his rewards intellectual, and every step he takes, every conclusion he reaches, must be exposed to the cold logic and debate that he rightly insists are the very essence of his subject. Yet it is just this dependence on scientific reasoning that exposes epidemiology's weakest flank. For however deeply they have explored epidemiological methods and the statistical techniques used to interpret them, teacher and student together may still emerge wondering whether the whole subject of epidemiology is sufficiently strongly enough developed conceptually to bear that very process of critical reasoning that it demands for its practice. It seems to me that there are two components of this doubt. The first is that the analytical methods essential for the application of epidemiological techniques to the diseases of our age are generally more complex and of a later time than the simpler data-gathering processes of epidemiology's infancy. This makes it hard to join the two phases in a single coherent story, let alone to trace the development of the theoretical concepts that would bind them together; the apt student is right to be curious about the relatively silent years of epidemiology earlier this century. The second component of the doubt is, however, more fundamental and extends beyond the teaching of the subject into the organisation of its concepts. It is simply this: the able student is too often asked to accept as a science a discipline justified only by its techniques. The teaching of epidemiology, certainly as illustrated in many of the good textbooks available to students, tends to depend too much on detailed description of methods and the uses to which they may be put without concentrating equally on the theoretical concepts that relate those methods and which alone can justify any claim epidemiology may have to be called a basic science (whatever that may be). Nor is there a chance that a mere collection of methods, however precise, can ever predict the pattern of its own future development, which is surely a legitimate demand on any science. If epidemiology were content to remain the mere interpreter of recorded observations then perhaps the elaboration and perfection of techniques would be enough, but it has higher aspirations than that and if they are to be realised it will only be by developing more fully the theoretical concepts that unite its methods. Only then can the bright student be expected to accept the subject as intellectually satisfying and to see its relevance to the many aspects of clinical and laboratory medicine which have a just and increasing claim on its interest. To be useful, a unifying concept must start by being simple, and then be capable of increasingly complex and exciting development in any of the several