1928 BMJ (Clinical Research Edition)  
MAY 112, '128 I BTAY MEDICAL JOURNAL 825 therefore, that a knowledge of when to use it and how to use 'it is more important than the selection of a particular antigen. (4) I-t is more difficult to make a comparison between cases treated by tuberculin and by ordinary methods tllani would be supposed. Attempts have been made to compare thie mortality amongst those who have had tuberculin anid those who have not. In order to convince the non-believer in tuberculin, statistics have been chiefly
more » ... ve been chiefly confined to the cases with tubercle bacilli in the sputum. Now tuberculin, to be of real ser-vice, should be used in the early stages before the bacilli have appeared in thle sputum. Even in the later stages, however, it would appear from the comiparatively few cases that have been recordled that life can be prolonged by tuberculin in a certain percentage of cases. In the earlier stages the diagnosis cannot be made with certainty without the use of tuberculin and skiagraphy. If the former shows hypersensitiveness and the latter indicates activity the condition is certainly tutiberculous. Eveni if there are no radiographic changes and no abnormal physical signs, and yet there are symptoms of early tuberculosis, the physician who uses tuberculin will be on his guard if he gets a strong positive reaction with it. He will not necessarily treat such a patient with tuberculin, *but if that patient is not progressing favourably under ordiniary methods he certainly will do so. To produce statistics wllich will -convince the unbeliever is, however, difficult, for he cuts the ground from under one's feet bysaying, " Everyone reacts to tuberculin," which is, of course, a lhalf truth, anddepends on what is meant by a tuberculin reaction. The comparison with " controls " is not so easy as it sounds. In my hands I find:
doi:10.1136/bmj.1.3514.825-c fatcat:adyuyly6sfed7low7xelfwh3ce