A basis for health strategies. A classification of disease.

A few years ago attention was drawn to a remarkable paradox: at a time when medical knowledge is greater and technology more powerful than ever before, medicine is assailed by doubts about its role and purpose.' Some reasons for the doubts are clear enough: disagreement about the respective responsibilities of public and private agencies for the finance and delivery of health services; rapidly rising costs of medical care and the lack of an acceptable basis for limiting them; gross inequalities in health between continents, between countries, and between different sections of the population of the same country; ethical issues which arise particularly from the ability to prolong or terminate life; and formidable problems of litigation attributable to the difficulty of distinguishing clearly between errors of clinical judgment and'negligence. But perhaps the most telling source of uncertainty is the possibility that we have overestimated what has been achieved, indeed what can be achieved, by treatment of the sick. We begin to suspect that some disease problems may prove to be, as has been said of the universe, "not only queerer than we suppose but queerer than we can suppose."2 The doubts are reflected in ambiguities and differences of opinion about the best approach to the control of disease, particularly the balance between preventive and therapeutic measures. For example, it has been implied that treatment must take priority over prevention ("Prevention is better than cure only if you have not got something that needs curing"3), that the contribution of prevention and treatment should be considered separately ("Prevention is needed for its own sake and not as an alternative to someone else's cure"3), and that all medical activities should be guided strictly by tests of their effectiveness and efficiency.' Equally remarkable are conclusions concerning the relative effectiveness of different methods of prevention. In coronary heart disease, for example, it was stated "When one goes beyond stopping smoking and the detection and treatment of hypertension, tangible evidence is not only difficult but is also liable to be influenced by commercial pressures."" A contrary view was expressed in a report of a recent meeting, where "Professor Geoffrey Rose and other speakers from the United States seemed agreed that in population studies dietary changes had the most important effect on the incidence of coronary heart disease-much more than did changes in smoking."6 The reasons for such differences in viewpoint are undoubtedly