tion, and that each carry out the plan now successfully operated by the Society of the County of New York, in collaboration with Boards of Health. 7. Incompetency of Health Officers.-Finally, your committee feels that the present imperfect condition of Sanitary Science in our country is largely due to the fact that health officers are too frequently appointed for political reasons rather than because of their peculiar fitness for the positions which they occupy. almost no medical school of our State or in the country is there more the most superficial teaching in State Medicine, and no incentive is held out to men entering the medical profession to devote their lives to this most important field of labor. We may well take lesson from some of our older sister nations, and now, when we are just becoming recognized as a world power in finance and diplomacy, we may well consider whether it is not right that we should elevate the plane of preventive medicine. The contributions of American medical men have not been inconsiderable. The achievements of the Medical Department of the Army in banishing yellow fever and cholera from our insular possessions should but be added incentive to the performance of great work in all departments of State medicine. Since 1866 England has had a law making it necessary for every man occupying the position of health officer to a community to have a special Diploma in Public Health, granted upon examination by one of several recognized schools, in subjects covering the whole field of State medicine. Your committee, therefore, asks you to call the attention of the Regents of the State University to this matter, and to ask them to carefully consider this subject and provide for the examination of all physicians candidates for the position of health officers in the State, and to grant to the successful candidates a special Diploma in Public Health. And to the end that this may operate for the permanent welfare of the whole State, it asks that you request our Committee on Legislation to frame such laws as in their judgment seem necessary to make it imperative upon all communities in the State that have health officers to require the proof of especial fitness for the duties of the office to which they aspire. Respectfully submitted, JNO. F. HEFFEN, JOSEPH D. CRAIG, D. V. STILL. CO-OPERATION ESSENTIAL TO PROGRESS IN VITAL STATISTICS.* By CRESSY L. WILBUR, M. D., Chief of Division of Vital Statistics, Department of State, Michigan. This paper is an attempt to show some of the ways in which cooperation betwen the national, State and municipal officers engaged in the collection, tabulation and study of vital statistics may be very useful, or even indispensable, to further progress. Vital statistics, and especially mortality statistics, are universally acknowledged to be the absolutely necessary foundations of modern public-health work. Notwithstanding this fact, and although scarcely a State or city in the country is without its public health administration, the United States is almost alone among the civilized nations of the world in having no national system of vital statistics. This condition has arisen from the constitution of our government, whereby the collection of vital statistics has been usually a part of the State administration, although in some cases data have been obtained by city ordinances without correlation with the State administration. The function of the national government has been confined to an effort to obtain vital statistics in connection with the regular decennial censuses, at first by an attempted enumcration of the facts coincident with the enumeration of population, and latterly by an increasing use of the material collected by the States and cities comprising the "registration area." The statistics collected by the individual States and cities have varied greatly in quality. The most diverse methods of collection and compilation have been employed, and only a small proportion of the States that have attempted to collect such data have made a practical success of it, even on the somewhat low standard of 90 per cent. of accuracy for mortality statistics alone. There has been little co-ordination, and a State undertaking to install a new system of vital statistics has been quite as likely to make all of the blunders that have involved other States in disaster as to follow the methods that have been tested by experience in other States and found essential to satisfactory results. *Paper read at the Meeting of the American Public Health Association in New Orleans, December 8-12, 1902. Abstract from the Proceedings by permission of the Secretary. After the statistics have been collected, the same hopeless lack of uniformity and approved system may be found in the registration reports published. Tables regarded as essential in the reports of one State may not be represented at all in those of another, and the details of age periods, etc., vary so greatly that comparisons are frequently difficult or impossible. No uniform methods of computing and stating death rates are in general use, and up to a very few years ago no general agreement existed in even such a vital feature of mortality statistics as the adoption of a uniform classification of causes of death. The latter difficulty-the adoption of a uniform classification of causes of death-stood at the very threshold of any course of action whereby vital statistics might be made more uniform and comparable than formerly. If registration officials could agree on the use of a uniform and fairly satisfactory classification, it was likely that they would afterward come to a basis of satisfactory union with respect to other statistical features of scarcely less importance. But for over fifty years the lack of a uniform classification of causes of death had stood in the way of progress in vital statistics, and it seemed hopeless to attempt much in this direction until this obstacle could be removed. As you know, the American Public Health Association was largely instrumental in bringing about this greatly needed reform. At the meeting held at Ottawa, in 1898, the Association urgently recommended the adoption of the International Classification, as it is now known, by all of the registration offices of the three countries, and especially by their census offices. The general plan of a decennial revision proposed by this Association was indorsed by the International Statistical Institute, the revision was held at Paris in 1900, a special invitation being extended by the French Government, through the usual diplomatic channels, to all of the countries participating, and the results were presented for general use beginning with the first year of the present century, 1901. All of the registration States of the Union have adopted this uniform system, as well as the United States Department of Labor and the United States Census. The Classification is likewise supreme in Canada and Mexico, so that the constituent countries of this Association are a unit in this respect. It is proper to say that the preliminary work of revision was not as completely carried out in this country as it should have been, owing to the fact that the recommendations of your committee on Demography and of your special Commission on the Revision of the Bertillon Classi fication, made to you at the meeting at Minneapolis in 1899, were not carried out, owing, perhaps, to the late period of the session. at which the report was made. As far as possible, under the circumstances, the United States Commission of Revision performed the necessary work up to that time, and all of the suggestions received from American registrars were given consideration at Paris. No invitation to the International Commission of Revision, however, reached this Association or its special Commission, by whom all of the preliminary work in this country had been done. The United States was represented by the United States Marine Hospital Service, which had not previously been associated in this work, except as represented by delegates to this Association. For this reason the Commission and the Association may disclaim responsibility for the actual results of the revision, although it seems desirable that we now heartily join in the use of the revised version until an opportunity shall come, with the next regular decennial revision, to further perfect the work. As an indispensable step in this direction, the United States Census Office has recently issued a Manual of the International Classification, which contains not only the assignment of all terms found in the French version, but also many hundreds of terms actually occurring in the practice of American registrars, but for which no provision had been made by the International Commission. Some of the more important of these terms would have been regularly provided for had not this Association failed to authorize the completion of the work undertaken by it, but as the exigencies of the Census compilation would not permit delay or reference to committees, the assignments of all such terms were necessarily made with the hope that they would prove, for the most part, in accordance with the general judgment of American registrars. Another matter of great practical difficulty in the compilation of mortality statistics is the treatment of returns in which the physicians have assigned two or more causes of death for the same case. The International Commission did not consider this subject in detail, although a method of procedure was prepared by Dr. Bertillon and appended to their report. Many of the decisions, however, would hardly be acceptable to American registrars, and the whole subject demands the most earnest attention in order that some rational, definite and practicable plan of treatment be formulated and generally adopted. It may be feasible to introduce a system in which certain weights. or "ratings" shall be assigned to each cause of death as reported by physicians. We ordinarily consider two factors in determining the precedence of one cause of death over another, in the absence of any other information as to duration, primary or secondary occurrence, etc. These factors are (1) the importance of the title to which the given term would be assigned in the classification, and (2) the degree of certainty that the term really belongs under that particular title. Thus, using a decimal scale, pulmonary tuberculosis is a title of first rank, say 10, as compared with pulmonary congestion, say 2 or 3 on the same scale. If chronic Bright's disease be returned in connection with pulmonary tuberculosis, we can compare the two by assigning the same number to Bright's disease if we consider it of equal statistical importance, or by assigning a lower number on the scale, as 8 or 9, if we consider it less important. Again, whatever rank we give to Bright's disease, it is evident that a death so returned is much more certainly due to the disease in question than if returned from uremia, which is likewise compiled under this title, perhaps improperly. Uremia should be marked, 1, 2 or 3, with respect to its vitality as a synonym of Bright's disease. Now the products of the numbers representing the importance of the titles into the numbers representing the validity of each term under its title, both taken on the easily denoted decimal scale, will give an index number for each term on a scale of 100, by which the preference of assignment between any two or more independent terms may be decided. The occurrence of diseases as frequent complications or sequelae would be taken into consideration in a general way in rating the relative importance of the titles, besides which complete lists should be prepared showing such relations. Such a plan as that suggested for the proper disposition of jointly returned causes of death would not only be of service in determining such assignments, but it is evident that a series of rates obtained for this purpose would be directly in the line of a detailed and specific criticism of the classification itself. Should a given title be generally rated as of "o" importance, or a certain term be generally marked as not properly pertaining to the title under which it has been included, then we would have exactly the information that we require for proposing and supporting a change in the International Classification. In other words, we would substitute a definite numerical rating for a more general expression of opinion. To make the proposed plan a success it is necessary that the |