The National Library of Medicine and Information on Drugs, 1964
Page  1

THE NATIONAL LIBRARY OF MEDICINE AND INFORMATION ON DRUGS Martin M. Cummings, M.D. Director, National Library of Medicine The marshalling of resources, public and private, to insure better health for the American people has been a phenomenon of the post-World War II decades. Medical research, professional education and training and"the practice of the health sciences -- all have reached unprecedented levels. Since 1945 the $97 billion spent by the Federal Government alone for health and medical services exceeds the total funds thus spent in all previous years of the nation's history. Many individuals concerned with the handling of documentation and published literature have proposed a centralized system in which all scientific and technical knowledge would be systematically stored, and to which all scientists could apply for the most recent information on their subjects. This concept of ready access to a comprehensive store of recorded knowledge has constituted an ideal toward which many librarians have worked. Libraries have as their primary functions the acquisition of reports of new scientific knowledge wherever recorded, their indexing, cataloging and classification, and their storage for use in both future and present time, and their dissemination. The medical library thus serves medicine as a guarantor of the organization of its scientific and professional knowledge, and of accessibility to what is already known.

Page  2 Traditionally the medical library plays a central role in the interchange of published biomedical information. It contains the record of the past and present contributions of scientists and practitioners. It is essential to, and ultimately relates to, all other biomedical communication systems. The library and its functions are interwoven into the basic fabric of research, teaching, and practice, and are indispensable to progress in these pursuits. The National Library of Medicine has participated in the scientific revolution of the past century in a very real way. It produced the first bibliographic control of the medical literature through the development of Index Catalogue (1876) and Index Medicus (1879). It provided the first large-scale photoduplication service to the scientific community (1940). Through MEDLARS it now provides the first computer-based demand searches and recurring bibliographies in the field of medicine. NLM backstops these reference and bibliographic services by a large interlibrary loan service to all domestic and foreign libraries. The National Library of Medicine is concerned with the collection, organization, and distribution of biomedical published literature generated in all countries and all languages. As the world's largest medical library, it now contains more than 1,200,000 volumes of biomedical literature. Last year its acquisitions of books, monographs, and serial periodicals reached a level.of 100,000 items. More than 250,000 scientific reports and articles were contained

Page  3 in the "journal literature" alone. Of this number, 150,000 articles in 2,500 journals considered to contain original or high quality reports are indexed for publication in Index Medicus annually. It has been estimated that almost half of this volume of literature contains material related to chemical and drug information. To achieve bibliographic control of this vast volume of literature NLM developed MEDLARS, which is an operational computer system with a capacity of performing 100 demand searches daily and 50 recurring bibliographies annually, plus the production of Index Medicus. The major concepts of each published article are analyzed by professional indexers and are described as subject headings. These descriptors are stored with an English-language citation or reference on magnetic tape as part of the computer-based MEDLARS system. The depth of indexing has been increased from a level of three subject tags per article in 1962 to a current level of eight descriptors in 1964. Our goal is to achieve a depth of indexing of 15 to 20 subject headings per article, for 200,000 articles to be analyzed annually, by 1966. Consider the magnitude of the task of indexing when seventy percent of the literature which is being analyzed has been published in forty foreign languages. Careful translation of titles and indexing of concepts is required to make the information useful to the English-speaking and English-reading scientific community. The variabilities of the forms and quality of publication are enormous.

Page  4 Thus, when one tries to segment or fragment this totality of literature to give special emphasis or consideration to something as ill defined as "drug information," serious problems are presented. It has been suggested that specialized information centers be created to manage the handling of special-purpose information. We agree that there is such a need. However, it should be remembered that these instrumentalities also face library-type problems of collection, organization, and distribution of information. The new dimension which must be added is that of "evaluation." Here objectivity and reliability of information are more important than speed of reporting. It is our view that the biomedical community, especially the pharmaceutical industry and the Food and Drug Administration, must share the responsibility for literature evaluation when specifics such as drug toxicity and adverse reactions are being considered. More basic information requirements dealing with chemical structure and biologic activity seem to be controlled by current and proposed services provided by Chemical Abstracts,. We view the role of the National Library of Medicine to be one of acquiring, organizing, and managing the distribution of the published medical literature. However, we do not promise to provide the kind of refinement and evaluation of the drug literature which we believe requires the application of biomedical science rather than library science. Nor, at this stage of systems development, do we propose to undertake the management of structural chemical information

Page  5 in this literature. Thus, we view the Library to be a clearinghouse rather than information center. Recognizing the importance of the drug information problem, the National Library of Medicine has taken steps to give special attention to this portion of biomedical literature. It has developed a collaborative agreement with the Food and Drug Administration, through which the technical competence of the FDA may be applied to the deep indexing of the published literature concerned with drugs. In return, the National Library of Medicine will provide demand searches and recurring bibliographies to FDA. The Library is prepared to enter into similar agreements with other organizations who may wish to provide scientific manpower for depth analysis of the drug literature, to be stored and retrieved through MEDLARS. Our aim is to provide a "first cut" of the literature which will make it possible for groups with special interests to do the refined analysis required to meet their particular needs. These groups should not have to perform the primary library tasks of acquisition, indexing, and cataloging of the literature. The National Library of Medicine has also assumed the responsibility of supporting translations and abstracts of the Russian and Japanese biomedical literature through contracts with the Federation of American Societies for Experimental Biology, and with Biological Abstracts and Excerpta Medica. Important information on drugs and chemicals is brought to the attention of the American scientific community through these services.

Page  6 In addition we have just negotiated an agreement with the Government of Israel and Hebrew University (Hadassah Medical School) to provide drug digests (evaluative abstracts) of fifty non-English-language journals containing articles on drugs and chemicals thought to be of interest to the FDA and the U.S. pharmaceutical industry. One thousand copies of these drug digests will be distributed to interested groups for evaluative purposes, beginning January 1965. Consumer feedback will be evaluated by Dr. Ellis Kelsey in the Office of the Surgeon General to give us guidance regarding future directions which this abstract service should take. No matter how efficiently a library or specialized information service may operate, the information derived from published reports will always be more or less stale. Consider the long delays in observing and confirming toxic effects of chemicals or drugs in animals or man. Add to this the delays in having these reports prepared, reviewed, edited, and accepted for publication. Then add the time required for printing and distribution of the findings. It is not uncommon to see the passing of several years between the original observation and the widespread dissemination of the pertinent information. For this reason I am impressed by the need for a network or system which collects and disseminates rapidly the informal, unpublished reports of chemical information, biologic screening data and, most importantly, data derived from controlled clinical trials.

Page  7 In my view the FDA, with the cooperation of industry, university groups as well as clinicians who use drugs and chemicals for the management and treatment of disease, represent the appropriate organizational elements to be included in the development of such an information system. The magnitude of expense of such a system is so great that it will require Federal subsidy if not Federal operation. Modern library techniques and facilities should be incorporated in such an undertaking. The National Library of Medicine is prepared to participate in such a scheme, recognizing clearly that its contribution will be one of providing computer methodology of storing and retrieving information rather than one of analysis or evaluation of data. Problems of codes, terminology, and classification already exist. Such nomenclature and structural problems should be dealt with by scientific groups such as the American Chemical Society and Chemical Abstracts Services. Meetings such as this should assist not only in identification of problems but also in the standardization of courses of action and integration of existing resources. The recently published "Survey of Chemical Notation Systems," produced by the National Academy of Sciences, represents a most comprehensive evaluation of methods of handling chemical information. The report separates theory from practical experience a most valuable contribution. Its recommendations deserve careful study.

Page  8 Finally, I have serious philosophic reservations concerning the provision.of drug information for profit when the availability of such information may affect the nation's health. We must be prepared to face the moral issue involved in making such vital information available only to those who can afford to purchase it. In this context I commend the American Medical Association, which provides drug information gratis to its members and to subscribers of the J.A.M.A. I question those special information services which claim to separate clearly "significant" drug information from observations which they deem to be of minor value. In the long run such "selected" information may be misleading. Also, those who believe that only English-language publications contain sining the provision.of dning the provision.of drug information for profit when the availability of such information may affect the nation's health. We must be prepared to face the moral issue involved in making such vital information available only to those who can afford to purchase it. In this context I commend the American Medical Association, which provides drug information gratis to its members and to subscribers of the J.A.M.A. I question those special information services which claim to separate clearly "significant" drug information from observations which they deem to be of minor value. In the long run such "selected" information may be misleading. Also, those who believe that only English-language publications contain sining the provision.of drug information for profit when the availability of such information may affect the nation's health. We must be prepared to face the moral issue involved in making such vital information available only to those who can afford to purchase it. In this context I commend the American Medical Association, which provides drug information gratis to its members and to subscribers of the J.A.M.A. I question those special information services which claim to separate clearly "significant" drug information from observations which they deem to be of minor value. In the long run such "selected" information may be misleading. Also, those who believe that only English-language publications contain significant information ignore the lessons of history, as well as the current international nature of research. The responsibility for professional evaluation of content must rest with physicians and scientists, not with documentalists. We must not sacrifice quality of information for the desiderata of speed, scope, and coverage. Control on quality of drug information should be as stringent as that placed on those concerned with drug development and use. Only through a partnership between those responsible for the management of literature and those who have the responsibility for its evaluation will it be possible for a national drug information center and clearinghouse to be established. Under such an arrangement the National Library of Medicine is prepared to utilize its resources to participate in such a national purpose. Read at the 148th National Meeting of the American Chemical Society, Chicago, Illinois, September 1, 1964.