1912 The Lancet  
wish to thank the President and the Council of this section for their courtesy in asking me to open this discussion. There can be no doubt that such a discussion is required at the present time, for on the one hand we have a school which to my mind may be regarded as placing the value of tuberculin too high and of focussing attention on this remedy to the exclusion of other important factors in treatment, and on the other we have a school-daily becoming smaller and less important-which denies
more » ... tant-which denies that tuberculin is of value. The latter view is largely based on the experience obtained when Koch first introduced his discovery, but that experience has subsequently been shown by patient work to have been vitiated chiefly by an improper system of dosage and an improper selection of cases. Both schools have right on their side, but both are too dogmatic in their generalisations. To-day, I take it, we want not dogmatic statements, but a practical and judicial examination of the results or impressions of our individual experience, so as to make others appreciate the fact that the truth concerning tuberculin lies midway between the conclusions formed by the two existing schools. I propose, therefore, to introduce the subject under a series of headings with the object of facilitating discussion. THE RESULTS OBTAINED BY TUBERCULIN TREATMENT. It is held by some that statistical evidence on this point should be conclusive on one side or the other, and so far as statistical results go they show that tuberculin in association with other methods gives better results than any other known method of treatment. A larger proportion of patients so treated certainly appear to maintain their working capacity for a longer period, whilst it would also appear to be proved that tubercle bacilli disappear from the sputum in a larger proportion of cases treated by this method than with any other method. There are, however, difficulties, some of which I regard as insuperable, with regard to statistical proof in medical problems. It is difficult, if not impossible, to compare the results of one observer accurately with those of another, as the original data are not strictly comparable. i For example, our views as to diagnosis differ. Some rely largaly, if not entirely, on the presence of tubercle bacilli in the expectoration, some will make a positive diagnosis on signs and symptoms which will not convince others, and some, again, rely in many instances upon a positive reaction to old tuberculin. It may be said, in passing, that if the problem of pulmonary tuberculosis can be reduced so far as diagnosis is concerned to a positive reaction to old tuberculin, or so far as treatment is concerned to treatment with one or other of the numerous preparations of tuberculin, then in this particular medicine has become an easy art. Certainly if we rely on an old tuberculin reaction as showing the presence of tuberculosis which requires active treatment, we shall have a large measure of success with tuberculin, but our results will be vitiated by the fact that a very considerable proportion of our cases would have shown equally good results by other methods, or, indeed, without any treatment. Then we are met by the different views of different observers as to what we mean by what is usually expressed by the unfortunate term co an early case." So-called early cases differ largely in their characteristics and as regards their progress. The fact that the disease appears to be of recent origin does not afford a reliable basis on which to calculate results, and I do not think that we shall be able usefully to correlate the results obtained by different observers until we are in possession of a better classification of the different forms or what are called stages of the disease. The best I know is that of Dr. A. C. Inman, which is based upon the facility with which auto-inoculation occurs or is produced. A further difficulty about statistical proof is the difficulty in keeping our patients, and especially our hospital patients, under observation for a sufficient length of time after discharge to form a useful opinion of the real results. There is also the difficulty when relapse occurs of knowing whether the relapse was not inevitable under the sanitary conditions, conditions of work and so forth, whatever the treatment. We cannot achieve the impossible, and if we make the attempt and fail it cannot fairly be laid at the door of tuberculin. For these and other reasons I hold that statistical information is too full of fallacies to be really useful, but I would refer those who are interested in this part of the subject to Dr. Lawrason Brown's article in Klebs's volume on " Tuberculosis. " Another method-also open to many fallacies-of estimating the value of tuberculin is the a priori method. For example, we have now accumulated a good deal of information concerning the value of vaccine-therapy in a number of other diseases, and although this method has its limitations, I think we shall be all agreed in holding that vaccine-therapy is a great addition to our therapeutical resources. We would by analogy expect tuberculin to be of value. Again, during recent years we have been able, as the result of the work :of Sir Almroth Wright and, especially in this connexion, of Dr. Inman, to formulate a working hypothesis which makes a strong appeal to many of us, as an explanation of the effects of sanatorium treatment.
doi:10.1016/s0140-6736(01)67413-0 fatcat:a6pszgsrtfhzzf5escjitmdeze