Factors influencing the attack rate of pulmonary tuberculosis.

The most useful knowledge when attempting to prevent a chronic disease is that of the factors influencing its attack rate. A knowledge of the factors relating to mortality rates and prevalence rates is less valuable for this purpose, as both are secondary indices, mortality varying directly with attack rate and case fatality rate, and prevalence varying directly with the attack rate and indirectly with the case fatality rate. It is therefore regrettable that the attack rate of tuberculosis has been so little studied in comparison with the vast literature on mortality. Such studies as have been published have been summarized in the Prophit Report (Daniels, Ridehalgh, Springett, and Hall, 1948). This work consisted mainly of follow-up studies on medical students and nurses, and the results of studies on such selected material are unlikely to be generally applicable. The only study of the attack rate of tuberculosis in the general population is that of Springett (1951) which, in spite of its general excellence, can be criticized on the grounds that the population on which the attack rate was studied was somewhat ill defined. It therefore seemed of value to see what information could be obtained on this point from the material obtained from the two Rhondda Fach surveys (Cochrane, Cox, and Jarman, 1952, 1955).