MEDLARS - Medical Literature Analysis And Retrieval System

M EDLARS, Medical Literature Analysis and Retrieval System.
Newest, most sophisticated, and costliest method yet devised for controlling the vast flood of medical literature published throughout the world.

1. Medical Literature Analysis and Retrieval System (MEDLARS) By Pallab Das

2. MED ical L iteratures A nalysis R etrieval S ystem

3. M EDLARS, Medical Literature Analysis and Retrieval System. Newest, most sophisticated, and costliest method yet devised for controlling the vast flood of medical literature published throughout the world.

4. Time coverage MEDLINE includes literature published from1966 to present, and selected coverage of literature prior to that period.

5. COVERAGE ITEMS  25 million references to journal articles in life science 18,000 separate journals were initially received at the National Library of Medicine. 2,300 are indexed into MEDLARS

6. COVERING AREAS Medicine, Nursing, Pharmacy, Dentistry, Veterinary Medicine, And Health Care.

7. Development of MEDLARS In 1957 the staff of the NLM started to plan the mechanization of the Index Medicus . By 1960 a detailed specification was prepared and by the spring of 1961 a request for proposals was sent out to 72 companies to develop the system. MEDLARS cost $3 million to develop and at the time of its completion in 1964, no other publicly available, fully operational electronic storage and retrieval system of its magnitude existed.

8. NATIONAL LIBRARY OF MEDICINE (USA)

9. NLM OFFICIAL WEBSITE

10. MEDLARS Online In late 1971 , an online version called MEDLINE ("MEDLARS Online") became available. This early system covered 239 journals and boasted that it could support as many as 25 simultaneous online users This situation continued through the beginning of the 1990s and the rise of the World Wide Web. a free public version of MEDLINE was instigated This system, called  PubMed , was offered to the general online user in June, 1997

11. Official Website of MEDLINE

12. Utility of MEDLARS : MEDLARS is known as the biggest bibliographical database of international level. It is important not only for the experts of medical science, but also for other scientists , sociologists, trades and businessmen etc. It has become more and more useful and its scope is very much wide. It also has been useful for the people which are not concerned with medical sciences.

13. National Library of Medicine classification System The National Library of Medicine (NLM) classification system is a library indexing system covering the fields of medicine and preclinical basic sciences . The NLM classification is patterned after the Library of Congress (LC) Classification system

14. INDEXING The MEDLARS data base is created by indexers who analyze and describe the contents of journal articles by means of carefully selected terms. These terms, called subject headings or main headings, are contained in an authorized list.

15. MeSH : Key to MEDLARS Known as MeSH , for Medical Subject Headings,this controlled vocabulary is essential for effective use of the system, since only MeSH terms are used to index articles for MEDLARS. MeSH is updated and published annually.

16. MeSH types

17. The 14 broad subject categories of MeSH Anatomical Terms Organisms Sociology, and Social Phenomena Diseases Chemicals and Drug Analytical, Diagnostic, and Therapeutic techniques and Equipment Psychiatry and Psychology Biological Sciences Anthropology, Education Technology, Industry and Agriculture Humanities Communication, Library Science and Documentation Named Groups of Persons Health Care Physical science.

18. In-House MeSH This is the more current and updated version of MeSh This in-house MeSH is revised auarterly and it contains about 1,000 more terms than the published MeSH .

19. Types of In House MeSH This are of two types : geographical headings provisional headings

20. The geographical headings are used to describe articles with epidemiological, political, sociological, or geographic interest. For instance, consider articles . Example: Public Health in China. Provisional headings Are the terms available for indexing and machine searching which have not yet been approved as main headings for inclusion in Index Medicus . To describe a drug under the heading ACTINOSPECTCIN - The term ANTIBIOTICS can also be assigned in the Index Medicus .

21. Index Medicus Library's most widely known MEDLARS publication. Approximately 18,000 articles are cited in each monthly issue. under about 8,000 MeSH headings. Each issue contains an author index. Cumulated Index Medicus is published yearly.

22. Various subsets of Index Medicus Abridged Index Medicus It is a monthly listing of references from 100 key English- language journals in clinical medicine . Monthly Bibliography of Medical Reviews This publication contains references to those journalarticles which are well documented surveys ofrecent literature .

23. Essential Requirement for indexers Most analysts are able to read at least one foreign tongue. Many were able to index literature in four or five languages. One can index efficiently without the perusal of every single word of a text.

24. Analysts use a read/scan method based on the following instructions Read and understand the title. Read the text down to the point at which the author states the purpose of his paper. Read every word of the summary. Closely scan the abstract. Scan the bibliographic references.  

25. Check- Tags A set of routine items that must be accounted for in the indexing of every article by the indexers. These items are known as check-tags. eg : Studies describing controlled clinical research on human beings are identified by the check-tag clinical research , while, studies of the comparative effects of two or more drugs, or two or more procedures or techniques, are identified by the tag comparative study.

26. Coordinate Indexing In MEDLARS complex concepts are expressed by combinations or coordination’s of two or more terms . Example : An article discussing " community health services under Regional Medical Programs " would be indexed with the main headings – COMMUNITY HEALTH SERVICES and REGIONAL MEDICAL PROGRAMS , both to be printed in Index Medicus .

27. Indexing Revision Conducted by a fully experienced indexer. known as a"reviser .“

28. The reviser rapidly scans the bibliographic description and then concentrates on the assigned subject headings , by following questions: Do the main headings reflect the true content of the article? Do the Index Medicus ( print ) headings cover the central points of the article? Are the headings spelled correctly, in exactly the form appearing in MeSH ? Are the correct subheadings used? Are the relationships expressed correct?

29. The Analyst's Reference Tools Best authority for his indexing of an article is the text of the article itself.. The Integrated Authority File

30. Example of the Indexing Procedure Sample article is entitled :-- " Positive Sputum Cytologic Tests for Five Specific Detection of Bronchial Carcinoma ” The indexer decides after reading this article this is primarily related to the diagnosis of pulmonary and bronchial neoplasm’s by means of cytological tests of the sputum. Heading/subheading combinations :- X SPUTUM*cytology X CARCINOMA, BRONCHOGENI C *diagnosis X BRONCHIAL NEOPLASMS*diagnosis X LUNG NEOPLASMS –diagnosis. (Indicating by the X that these are print terms, representing headings under which the citation should be printed in Index Medicus . )

31. The additional terms BRONCHIAL NEOPLASMS* pathology CARCINOMA, EPIDERMOID*pathology CARCINOMA, BRONCHOGENI C*pathology ADENOCARCINOMA*pathology CYTO DIAGNOSIS

32. PROCESSING THE CITATIONS The magnetic tapes produced by these input procedures ,in the initial stages. For each citation the following data are recorded: citation number Author title of article

33. CONDUCTING THE DEMAND SEARCH Retrieval in MEDLARS, as in any other information system, is essentially a matching process. A request for literature on a particular topic is matched against the file of citations to biomedical articles that have been input to the system. When a match occurs, an article is retrieved.

34. Request Analysis A search analyst goes through the same two-stage process: The analyst must decide what the request is about The analyst must translate his interpretation of the request into a search statement, in MeSH terms, that can be processed against the citation file.  

35. In the analysis of a request, it is logical to begin by breaking the request into its various aspects , or facets. Example , Consider a search for literature on the subject Renal Amyloidosis . This request has two facets: (1) the organ facet (kidney) and (2) the disease facet ( amyloidosis ). The requester is not interested in all articles on the kidney, and he is not interested in all articles on amyloidosis . He is interested only in articles that discuss both facets of his request, i.e., kidney and Amyloidosis , and which presumably deal with renal amyloidosis . This relationship between the two facets may conveniently be represented by a diagramof overlapping circles, as

36. This relationship between the two facets may conveniently be represented by a diagram of overlapping circles, as Renal Amyloidosis

37. A Simple Search Formulation This is a very simple search formulation for a request on renal amyloidosis . By giving each term in this strategy a unique identifying number, and by using symbols that are recognized by the MEDLARS computer, we can reduce this strategy to a simple algebraic search equation . Considering the following : M1 KIDNEY M2 KIDNEY GLOMERULUS Related to Kidney M3 KIDNEY PELVIS M4 KIDNEY TUBULES M8 AMYLOID  M9 AMYLOIDOSIS Kidney Related Dieseses

38. Contd ’…. Here it shows that M1 to M4 all the terms are related to Kidney and M8 and M9 is related to Amioloidosis

39. Developing the Lists of Search Terms The most specific available term is always used to describe a concept. Thus, an article on sunburn is indexed under SUNBURN and one on eye burns under EYE BURNS rather than under the more generic term BURNS .

40. Cross-references The cross-references in MeSH lead the searcher to many of those headings which might be considered for searching . for example : burns (nonspecific) in children We find four terms: BURNS BURNS, CHEMICAL BURNS, ELECTRIC BURNS, INHALATION The great advantage of the tree structure is that it minimizes the time needed by the search analyst to conduct comprehensive generic searches.  

41. The Machine Search and Its Output Several searches are batched together and processed simultaneously, The search formulation is matched sequentially against the complete data base of indexed citations. A citation is retrieved from this data base when its index terms match one of the combinations of index terms demanded by the search strategy. Qualifying citations are copied from the main citation file onto a retrieved citation tape.

42. Web Versions Around mid-1990s, the development of MEDLINE took another new dimension. Several web entrepreneurs such as HealthGate etc. purchased MEDLINE data and mounted it on their site with a search engine for free use for its visitors. The underlying aim of this move was to make the site popular. Among the different databases available, it was the obvious choice because of its low cost and wide popularity free MEDLINE sites that are available on the Web are, - HealthGate (http://www.healthgate.com); N4eds.cape (http://www.medscape.com); Biomednet (http://www.biomednet.com).

43. Health Gate Website

44. Medscape

45. Biomednet

46. It uses a simple search interface. The search features are It uses a simple search interface. The search features are Basic search - a simple keyword search. Advanced search - allows use of Boolean operators and other search aspects and limit options . Searches can be made using different fields like, author, text word in title, subject term, etc MeSH terms can be selected with the help of MeSH browser Mapping terms to MeSH is limited Searches can be limited to clinical queries Search strategy is automatically displayed Search results can be marked and printed or saved Default display of 20 citations per page

47. Paid MEDLINE SERVICE Knight Ridder (now Thompson-Dialog), Silver Platter, Ovid, and STN. Thus a MEDLINE plethora of online options are available; most web versions of MEDLINE have introduced links to full-text items wherever available.

48. MEDLINE Today Scope and Coverage 40 online databases in the field of bio-medical sciences with more than 20 miiiion records. It is estimated that about 20000 citations are added.

49. OVID Technologies (Now available from Wolters Kluwer ) Ovid is a search software designed to search more than 80 science, medical, technological databases including MEDLINE, Current contents, BIOSIS, ABI/Inform, and Psycholnfo . It is described as fully integrated full text linked database. Ovid is available on stand-alone, Windows, Novell NetWare, WinNT, Unix, via Internet . As per the assessment by DUKE UNIVERSITY MEDICAL CENTER LIBRARY , Ovid search engine js _ an excel lent..interface to MEDLjNE .

50. FACTORS AFFECTING PERFORMANCE IN MEDLARS The success or failure of MEDLARS, or any other information retrieval system, is dependent less on machine capabilities than on intellectual effort. This intellectual effort is exerted by the indexers, the search analysts, the vocabulary specialists and - very importantly - by the user himself. The great majority of articles in the data base will not be relevant. There are 999,980 non-relevant articles.

51. Evaluation of operating efficiency of MEDLARS conducted in 1966-67 Purpose: To study the demand search requirements of MEDLARS users. To determine how effectively and efficiently and the present MEDLARS services are meeting its requirements . To determine factors adversely effecting the performance. To discover ways in which more effectively or more economically user requirements can be satisfied.

52. Prime requirements of demand search users The coverage of MEDLARS Its recall power. The precision power. The response time of the system. The format in which the search results are presented. The amount of effort the user must personally expend in order to achieve a satisfactory response from the system. The evaluation program was conducted in order to meet user requirements and tolerances in relation to these various factors and to determine MEDLARS performance in regard to the user requirements .

53. The two most critical problems faced in the evaluation of MEDLARS were Ensuring that the body of test request was as far as possible representative of the complete spectrum of kinds of requests processed. Establishing methods for determining recall and precision performance.

54. Methodology Queries were submitted to the MEDLARS and on receipt of the queries; MEDLARS staff prepared a search formulation (i.e. query designation) for each query using an appropriate combination of MeSH terms. A computer search was then carried out in the normal way. At this stage each user was also asked to submit a list of recent articles that he judged to berelevant to his query .

55. Result of a search was a computer printout of references. Since the total items retrieved might be high (some searches retrieved more than 500 references), 25 to 30 items were selected randomly from the list and photocopies of these were provided to the searcher fro relevance assessment .

56. Each searcher was asked to go through the full text of the articles and then to report about each article on the following scale of relevance: H1 – of major value (relevance) H2 – of minor value W1 – of no value W2 – of value not assessable (for example, in a foreign language).

57. EACH SEARCHER WAS ASKED TO GO THROUGH THE FULL TEXT OF THE ARTICLES AND THEN TO REPORT ABOUT EACH ARTICLE ON THE FOLLOWING SCALE OF RELEVANCE If L items were in the sample, the overall precision ratio was 100(H1 + H2)/L and the ‘major value’ precision ratio were 100H1/L. It was obviously not feasible to examine the whole MEDLARS database in relation to each search in order to establish recall ratio. Therefore, an indirect method was adopted for calculation of recall ratio: Each user was asked to identify relevant items for his query before receiving the search output and then search was carried out to find out whether those items were indexed in the database and retrieved along with other items that are both relevant and irrelevant. If t such relevant items were identified by the user and available on the database for a given query, and H were retrieved in the search, the overall recall ratio and ‘major value’ recall ratio was estimated as 100H/t and 100H1/t1respectively .

58. The next stage of the evaluation was an elaborate analysis of retrieval failures, i.e., examining, for each search, collected data concerning failures include : Query statement Search formulation; Index entries for a sample of ‘missed’ items ( i.e.relevant items that are not retrieved) and ‘waste’ items (i.e. noise—retrieval of irrelevant items);and Full text (c).

59. RESULTS The average number of references retrieved for each search was 175, with an average or overall precision ratio of 50.4%; that is, of the average 175 references retrieved, about 87 were found to be not relevant. The overall recall ratio was 57.7% as calculated by an indirect method. Taking the average search, and assuming that about 88 of the references found were relevant, with an overall recall ratio of 57.7% implies that about 150 references should have been found, but 62 were missed. However, the recall and precision ratios for each of the 302 searches were analyzed and individual ratios were then averaged in the MEDLARS test. The results were :

60. The results were over all Major value Recall ratio Recall ratio 57.7% 65.2% Precision ratio Precision ratio 50.4% 25.7%

61. CONCLUSION MEDLINE functions as an important resource for biomedical researchers and journal clubs from all over the world. Along with the Cochrane Library and a number of other databases, MEDLINE facilitates evidence-based medicine. Most systematic review articles published presently build on extensive searches of MEDLINE to identify articles that might be useful in the review MEDLINE influences researchers in their choice of journals in which to publish.

62. Bibliography LANCASTER, F. W. ( n.d .). MEDLARS: Report on the Evaluation of Its Operating Efficiency. MEDLINE on CD-ROM: An Analysis of User Search Behaviour . ATWOOD, R. ( n.d .). A Grass-Roots Look at MEDLARS. CHARLES J. AUSTIN, H. ( n.d .). Data Processing Aspects of MEDLARS. National Library of Medicine . (1968). EVALUATION OF MEDLARS DEMAND SEARCH SERVICE. U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Lancaster, F. W. ( n.d .). EVALUATION OF ON-LINE SEARCHING IN MEDLARS. LANCASTER, F. W. ( n.d .). MEDLARS: Report on the Evaluation of Its Operating Efficiency. MEDLINE on CD-ROM: An Analysis of User Search Behaviour . NATIONAL LIBRARY MEDICINE. THE PRINCIPLES OF MEDLARS. NATIONAL LIBRARY MEDICINE. ROGERS, F. B. ( n.d .). The Development of MEDLARS. Swargiary , D. ( n.d .). Assignment on MEDLARS, Pondicherry University.

63. THANK YOU