The Spread of Rheumatic Fever through Families.

A number of epidemiological studies on rheumatic fever have been made, but in this disease, this field of endeavor is in its infancy as compared with other lines of approach. It has seemed to us, however, that here is another valuable weapon which may be brought to bear, not only because we would like to know more about the epidemiology of rheumatic fever, but it may tell us more about the nature of this obscure disease. Knowledge concerning the actual distribution and spread of human disease, that is, its epidemiology is obtained in two ways: by direct observation, and by statistical methods, the latter being wholly dependent upon the former. There has been no dearth of direct observations of one sort or another in rheumatic fever, but in spite of them, actual facts which lend themselves to statistical analyses along epidemiological lines are few. The difficulties seem to lie primarily in the nonspecific character of many of the clinical aspects of this disease. Some of its forms seem to be recognized with reasonable accuracy, but unfortunately no specific test has yet been devised to tell us when an individual actually has rheumatic fever, and we are thus unable to determine its true incidence or distribution. With these limitations in mind we have made an attempt to study the spread of rheumatic fever through families, as an epidemiological problem. We have chosen the family as a unit for study, (a) because it represents a small group in which an intimate investigation may be carried out; and (b), because one of the striking things about rheumatic fever is its high familial incidence, of which there is ample evidence (1) (2) (3). Cohn (4) has assembled figures obtained by different observers on this point and calls attention to the fact that in so-called rheumatic families, 8-10 per cent of exposed persons are