The 101 Most-Cited Papers from the British Medical Journal Highlight The
Total Page：16
File Type：pdf, Size：1020Kb
EUGENE GARFIELD INSTITUTEFORSCIENTIFICINFORMATION- 3501MARKETST,PHILADELPHIA,PA19104 The 101 Most-Cited Papers from the British Medical Jownd Highlight the Important Role of Epidemiology in Medicine Number 7 February 16, 1987 The British Medical Journal began pub- Medical Journal, but of these only 20 per- lishing in 1840 as the organ of the Provin- cent are eventually published.s (p. 243) The cial Medical Association (PMA), a group current editor, Stephen Leek, who authored of physiciartspracticing in the provinces out- a 1985 book on peer review in medicine,4 side England’s major cities. They had follows extensive refereeing practices (a re- banded together in 1832 to share their indi- cent review of which appears in a 1986 ar- vidual medical experiences with one another ticle in Schokwly Publishing.s) He alw sub- in much the same fashion as the metropoli- scribes to the rule promulgated by Franz tan physicians who had earlier formed their Ingelfinger, former edkor of the New En- own society and journal. The PMA”s first gland Journal of Medicine. The essence of periodical was called 7he Provincial Medi- the Ingelfinger rule is that articles submit- cal arrd Surgical Journal; in 1853 it merged ted to a journal cannot have been previous- with the Lmdon Journal of Medicine to be- ly published in, or be simultaneously sub- come the Association Medical Journal. Four mitted to, other joumals.b The British Med- years later the PMA acknowledged its grow- ical Journal is also a founding member of ing memkaship throughout Great Britain by the frstemational Committee of Medical renaming the society the British Medical Journal Editors-the Vancouver Group— Association. The journal was rdso re- which is responsible for issuing the Uniform named. 1Today the British Medical Journal Requirements for the Submission of Manu- is the third largest general medical journal scripts to Biome&cal Journals and guidelines in the worldz and, as the official organ of on multiple publication. These requirements the British Medical Association, is sent free are now followed by over 300 biomedical to its 70,000 members; the journal also journals throughout the world. has over 20,000 nonmember subscribers Since 1840 the British Medical Journal throughout the worlds (p. 239) has published over 50,000 papers, letters, This weekly journal primarily publishes and other items. In this study my ISI” col- peer-reviewed, clinically oriented papers in- leagues and I identified and then examined tended to “educate, inform, and entertain” the 101 British Medical Journal items most physicians from academia to general prac- cited in the 1955-1985 Science Citation tice to administration. These papers consti- Index@ (NY ). (See Bibliography.) Of tute about half of the journal’s pages each these, 99 are research papers and 2 (W.J. week. The remaining pages of the journal Irvine and A.W. Liley) are notes published comprise letters, book reviews, social-pol- in the preliminary communications section icy articles, topical items, and “leading ar- of the journal. The citations received by the ticles” commissioned to keep physicians up- 101 items range from 182 to 803. The me- to-date with the latest advances in medicine dian citation rate is 240. The notes received and to explain their uses in clinical settings. 186 (Irvine) and 264 (Liley) citations. Every year over 5,000 items, including The articles in the Bibliography represent 2,900 letters, are submitted to the British over 5 percent of those.British Medical Jour- 40 mzl items cited at least 50 times in the Worcester Royal Infh-mary, the paper dis- 1955-1985 SCI. Table 1 provides the cita- cusses campylobacter enteritis. tion breakdown, by groups of 100 citations B.N.C. Prichard and P.M.S. Gillarn, or less, for those British Medical Journal University College Hospital Medical articles cited 50 or more times. Only 9 items School, University of London, authored have received 400 or more citations; 28, “Treatment of hypertension with propran- more than 300; and 73, at least 200 citations. 0101,” the third most-cited paper, with 520 The majority of the articles-1 ,029—were citations. This 1969 paper continues to be cited between 50 and 74 times. cited— 12 times in 1986. As yet, the authors of these three papers Methods have not published commentaries in the As mentioned earlier, the 101 articles Cita~ion Ckzrsics series that appears in Cur- were chosen by examining citation data rent Contents” each week. These are auto- culled from the 1955-1985 SCI. Only cita- biographical accounts of informal events tions from journals and other serials are in- leading up to a paper’s publication. How- cluded. While the SCZdoes include citations ever, Alice M. Stewart, Department of So- to textbooks and other monographs, they are cial Medicine, University of Oxford, one of not treated as sources. The purpose of this the authors of the fourth most-cited article, analysis was simply to ident@ the 1(K)most- did comment on the survey of childhood ma- cited “classic” articles from this prestigious lignancies she wrote with Josefine Webb and medical journal. Using citation counts as the David Hewitt, also then at Oxford. Accord- sole selection criterion, we previously iden- ing to Stewart’s commentary, their work tified a similar series of Citation Classicsw was “a triumph for a small group of epi- for the Annals of Internal Medicine,7 demiologists... [who] were anxious to dis- JAMA-Joumal OJthe American Medical cover why the post-war increase in leuke- Association,8 Z%e Lmcet,g and the New mia had produced an early peak of leuke- England Journal of Medicine. 10No hypoth- mia mortality consisting only of lymphatic eses were proposed about the types of items cases, but.. even with the increase, leuke- that might appear in the list. mia remained a rare cause of childhood Selection of papers was limited by the 31 deaths. ” An association between fetal irra- years of citation data covered by the diation and cancer was later identified by the 1955-1985 SC1. Understandably, older pa- group. 12 pers that received the bulk of their citations The fifth most-cited work, by Martin G. prior to 1955 were not identifkd. This prob- Lewis, Chester Beatty Research Institute, lem will be remedied, however, once the Institute of Cancer Research: Royal Cancer SC1 is extended back to include citations Hospital and Royal Marsden Hospital, Lmn- from papers published in the first half of the don, and six colleagues, has been referenced twentieth century. 445 times since 1969. In his commentary Lewis says that this work “resulted in one Five Most-Cited Papers and Citation of the first detailed studies of humoral im- Ckssics munity in a human tumor system. ” 13At the The most-cited paper in the Bibliography, end of his commentary, which illustrates the by Andrew W. Kay, Western Intlrmary, type of information not usually included in University of Glasgow, Scotland (803 cita- scientific papers and that authors are encour- tions), describes the ‘‘Effwt of large doses aged to share in Citation Classics, he also of histamine on gastric secretion of HC1,” relates the sad news that three of his or’ ‘the parietal cell response in man to large coauthors-D.C. Bodenham, R.L. Ikono- doses of histamine. “~ 1 It is now over 30 pisov, and G. Hamilton Fairley–are de- years old but continues to be cited occasion- ceased, the latter the victim of a terrorist ally. The second most-cited work is of mom bomb attack in London. 13 recent vintage. Published in 1977 by Seven additional papers in the Bibliogra- M .B. Skirrow, Public Health Laboratory, phy, two coauthored by W .H. W, Inman, Table 1: Citation-frequency distribution of Bn”tish Research at Mill Hill. His autobiographyl’t Medical Joumd articles cited 50 or more ties in the has recently appeared. While the number of 1955-1985 SCF. A total of 1,969 articles were distinguished physicians in the Bibliography examined. Percent of Tutat is obvious, the Nobel Prize reflects basic Number of Number of Articles preclinical research. Citations Artidea Examined Exactly 303 authors appear on the 101pa- pers. Doll is listed on 11of these, while Ves- 5(KI 4 0.2 400-499 5 0.3 sey wrote 5, G.M. Besser and A.B. Hill 300-399 19 1.0 each are represented by 4, K.G.M.M. Al- 200-299 45 2.3 berti, R. Hall, Inman, and A.S. McNeilly 100-199 432 22.0 75-99 435 22.1 authored 3 each, and 23 authors wrote 2 pa- 50-74 1,029 52.3 pers. Many of the papers by these authors are companion pieces on the same subject Committee on Safety of Medicines, London, about the relationship of oral contraceptive published back-to-back in the British Medi- use to thromboembolic disease, have been cal Journal (N. Hurwitz, J.I. Mann, Speizer) or similar papers published within discussed in Citation Classics commentar- ies. The other five papers were authored by a year of one another (I. Aird, Doll, Ves- D.J.R. Laurence, M.O. Thorner, F.E. sey); three of Doll’s works contain the re- Speizer, M.P. Vessey, and B.M. Wright. sults of an originrd study (two 1964 papers) Vessey’s paper, coauthored with Sir Richard and a follow-up published in 1976. Doll, also discusses oral contraceptives and However, each of these “paired” papers venous thromboembolism. Vessey, inciden- contained enough unique information that tally, is also coauthor on both of the Inman researchers cited them more often by them- articles. selves than in conjunction with their com- panion papers. For example, 101 articles co- Author Information cited both of Aird’s papers that examine the Numerous studies at 1S1have demonstrat- relationship of blood groups to various types ed that Nobel Prize winners consistently of cancer.