Computer literature searches on dengue

Transcription

1 Computer literture serches on dengue G. Kuno1 Mny reserch workers, to sve time, rely entirely on either on-line or off-line dtbses offered by n incresing number of informtion services. The chrcteristics of eight dtbses, including five on-line services, were nlysed in the present study concerning the retrievl of informtion on dengue, the most importnt mosquito-borne virl disese of humns. Differences in the rte of retrievl mong dtbses were pprent, depending on the min subject of publiction s well s on the geogrphicl loction of the publisher. While rtes of retrievl of references in moleculr biology were generlly stisfctory (mostly >70%/o), those in clinicl medicine nd epidemiology were not (<50%/o). The ltter, s published in mny dengue-endemic tropicl countries, were found to be indequtely covered. For the globl surveillnce of dengue, which hs incresed in intensity nd spred to mny countries becuse of incresed interntionl trvel, the development of new dtbse emphsizing tropicl geogrphic medicine is highly desirble. Dengue, the most importnt virl disese trnsmitted by mosquitos to humns, ffects mny millions of people worldwide every yer. Although it is primrily confined to tropicl countries becuse of mosquito distribution, incresed intemtionl trvel hs cused significnt number of imported cses in temperte countries. Full-time investigtors of dengue, like other scientists, need to spend considerble mount of time serching for pertinent, scientific informtion. Fortuntely, mny computerized informtion dtbses nd retrievl services re now vilble. This rticle describes the chrcteristics nd problems of some of the dtbses frequently used for retrievl of informtion bout dengue. Mterils nd methods Definition of dengue source A dengue source item is defined in this study s ny scientific publiction in which dengue fever, denguelike illness, dengue viruses or their derivtives (such s ntidengue ntibodies, cdna, nd virl ntigens), virus-vector interction nd/or vector control in the context of dengue re discussed. Source items include originl reserch rticles, reviews, commentries, editorils, public helth bulletins, nd letters to the editor. In multi-uthored books ech rticle or chpter ws counted s n individul source item. The following were excluded becuse they re not considered s officil, scientific publictions: 1 Dengue Brnch, Division of Vector-borne Infectious Diseses, Ntionl Center for Infectious Diseses, Centers for Disese Control nd Prevention, 2 Clle Csi, Sn Jun, Puerto Rico Reprint No (i) rticles in newsppers or news-oriented periodicls for the generl public or scientific communities; (ii) bstrcts from scientific meetings without full text; (iii) periodic reports from reserch institutions or orgniztions which specified tht the contents did not constitute officil informtion; nd (iv) rticles in newsletters or informtion exchnges from scientific institutions or orgniztions. Articles with very limited ccess were lso not included, e.g., disserttions in most tropicl countries, lthough in the USA nd western Europe this informtion is vilble through bstrcting services. Most entomologicl rticles were excluded unless the mosquitos were studied with dengue virus or in the context of dengue fever. On the other hnd, epidemiologicl news in periodic public helth bulletins or reports, such s WHO's Weekly epidemiologicl record, the Epidemiologicl bulletin of the Pn Americn Helth Orgniztion, nd Morbidity nd mortlity weekly report (MMWR) of the Centers for Disese Control, were ccepted s legitimte publictions, lthough informtion is often preliminry in nture. However, smll number of the MMWR rticles reprinted in the Journl of the Americn Medicl Assocition nd the Weekly epidemiologicl record were excluded if the dtbse included the originl source item from MMWR. Similrly, when source item ppered in journl tht ws printed in severl lnguges, it ws counted only once. Primry dtbses Five on-line dtbses were obtined through Dilog Informtion Services, Inc. (Plo Alto, CA, USA): (i) Medline (U.S. Ntionl Librry of Medicine, Bethesd, MD, USA); (ii) Biosis Previews (Biosis) (Biosis, Phildelphi, PA, USA); (iii) CAB Abstrcts Bulletin of the World Helth Orgniztion, 71 (2): (1993) World Helth Orgniztion

2 G. Kuno (CAB) (CAB Interntionl, Wllingford, Englnd); (iv) Excerpt Medic Dtbse (EM-Dtbse) (Amsterdm, Netherlnds); nd (v) Sciserch (Institute for Scientific Informtion, Phildelphi, PA, USA). For on-line services the following keywords were used: dengue, dengue virus, hemorrhgic fever, flvivirus, Aedes egypti, Aedes lbopictus, nd mosquito control. Informtion ws retrieved three times to reduce vrition, if ny, mong two opertors. Two other dtbses compred were the printed version of Tropicl diseses bulletin (TDB) from the Bureu of Hygiene nd Tropicl Diseses, London, Englnd; nd Annotted bibliogrphy on mosquitoborne diseses in Southest Asi (SEAMEO) from the Museum nd Reference Centre, SEAMEO- TROPMED Ntionl Centre of Thilnd, Bngkok, Thilnd. One noncomputerized source included in this study ws Literture References for Mosquito nd Mosquito-Borne Diseses, which is published periodiclly s service to the recipients of the Journl of the Americn Mosquito Control Assocition (AMCA, Lke Chrles, LA, USA), in collbortion with the WHO librry, Genev, Switzerlnd. Other sources reviewed included CATLINE, n on-line service from the Ntionl Librry of Medicine, Virology & AIDS Abstrcts (Cmbridge Scientific Abstrcts, Bethesd, MD, USA); Abstrcts on Hygiene nd Communicble Diseses (Bureu of Hygiene nd Tropicl Diseses, London, Englnd); Index to Scientific Book Contents (Institute for Scientific Informtion, Phildelphi, PA, USA); Index Medicus Ltinomericno (Bibliotec Regionl Medic (BIREME), So Pulo, Brzil); miscellneous bibliogrphic publictions for Indonesi, Mlysi, Philippines, Singpore, nd Thilnd from the Southest Asin Medicl Informtion Center (Interntionl Medicl Foundtion of Jpn, Tokyo, Jpn); the lists of publictions in newsletters; periodic institutionl reports; ctlogues from book publishers; books listed in periodicls; cittions in publictions; nd personl communictions. For this study only the items published between 1983 through 1988 were covered for ll dtbses. Out of 962 source items tht met the definition of dengue rticle, originl rticles were obtined from 931 items. These constituted the body of literture nlysed in this study. Among the remining 31 items tht could not be personlly exmined, 2, 4, 1 nd 1 items were retrieved from AMCA, Biosis, Sciserch, nd EM-Dtbse, respectively. The serch ws terminted in October Mention of trde nmes, trde mrks, or orgniztions is for identifiction only nd does not constitute endorsement by the United Sttes Deprtment of Helth nd Humn Services. Clssifiction of source items All items were clssified ccording to the type of publiction, geogrphicl loction of the publisher (but not the uthor's ntionlity), subject, nd theme. The types of publictions were: (i) periodicls nd numbered publictions, (ii) books (including proceedings of scientific meetings), nd (iii) ll others which included lithogrphic documents such s booklets published for the generl public by orgniztions or institutions. The geogrphicl origins of periodicls nd books were identified, nd ll rticles were clssified to one of the following res: (i) Europe, (ii) Afric, (iii) Indi, (iv) south-est Asi (Indonesi, Mlysi, Philippines, Singpore, Thilnd, nd Vietnm), (v) est Asi (Chin, Hong Kong, Jpn, nd Tiwn), (vi) Austrli nd New Zelnd, (vii) USA nd Cnd, (viii) Mexico nd Centrl Americ, (ix) Cribben Bsin, nd (x) South Americ. The most importnt subject ws rbitrrily selected from ech source item nd clssified in one of the following subject res: (i) clinicl medicine nd epidemiology, (ii) lbortory dignosis, (iii) bsic virology nd immunology, (iv) moleculr biology, (v) vector biology nd control, nd (vi) generl subjects (review, editoril, public helth bulletin, commentry, nd historicl ccount). Results Comprtive retrievl The dengue literture published between 1983 nd 1988 consisted of 777 source items in periodicls nd other numbered publictions, 136 items in books (including proceedings of scientific meetings published s unnumbered publictions), nd 18 items in other publictions. The number of items retrieved nd the percentge for ech type of publiction re shown in Tble 1. Retrievl of items in periodicls mong five on-line dtbses rnged from 24.7% to 61.1%; nd mong the remining three dtbses, it ws less thn 30%. Percentges of relevnt source items retrieved rnged from 10.9% to 26.1% (Tble 1). In contrst, retrievl for the items in books ws generlly much less thn 10%, with the exception of the SEAMEO dtbse, which showed 47.0% retrievl becuse it ws the only dtbse tht covered 46 items in proceedings published in Asi. Twelve source items, not retrievble by Medline, were retrieved from CATLINE. When the five on-line dtbses were combined, 67.9% of source items in periodicls were retrieved (Tble 2). Among the items not retrievble by ny on-line dtbse but retrievble by t lest one of the 166 WHO Bulletin OMS. Vol

3 Computer literture serches on dengue Tble 1: Retrievl of dengue literture by primry dtbses. No. retrieved by dtbses Type of No. of source _ publiction items A B C D E F G H Periodicls nd num bered publictions (61.1)b (48.9) (38.1) (24.7) (40.0) (23.7) (27.2) (21.6) Books (1.5) (3.6) (1.5) (6.6) (0.7) (1.5) (5.1) (47.0) Others (5.6) (22.2) (11.1) (11.1) (11.1) Totl No. of items retrieved by on-line dtbse Percentge relevncy Abbrevitions of dtbses: A, Medline; B, Biosis Previews; C, Excerpt Medic Dtbse; D, CAB Abstrct; E. Sciserch; F. Tropicl Disese Bulletin; G, J. Am. Mosquito Control Assoc.; H. SEAMEO Bibliogrphy. b Figures in prentheses re percentges. three off-line dtbses were comprehensive report of dengue in Colombi (1), dengue-3 outbrek in Clcutt (2), nd blindness ssocited with dengue infections in Thilnd (3). Although combintion of ll eight dtbses further incresed the percentges of retrievble source items, bout 17%, 44%, nd 61% of the items in periodicls, books, nd other publictions, respectively, could not be retrieved. Although the missing 17% of source items in periodicls nd other numbered publictions my pper negligible to some people, this percentge included mny importnt rticles, such s clinicl tril of interferon efficcy during dengue hemorrhgic fever (DHF) outbrek in Cub (4), DHF reports in St. Luci nd Dominic (5) nd in Vietnm (6), report of intrvsculr hemolysis in Chin (7), neurologicl syndromes ssocited with dengue infections in children (8), the importtion of dengue from the Mldives to Germny (9), imported dengue in Polnd (10), dengue outbrek in Sonor Stte in Mexico (11), dengue serosurvey in Liberi (12), serosurvey for ntidengue ntibody in Kuwit (13), design of houses in reltion to infesttion by mosquitos (14), Tble 2: bses Retrievl by the combintion of primry dt- No. retrievble by: Type of publiction Five on-line dtbses All 8 primry dtbses Periodicls nd numbered publictions 607 (78.1) 647 (83.3) Books 21 (15.4) 76 (55.9) Others 4 (22.2) 7 (38.9) Totl 632 (67.9) 730 (78.4) Figures in prentheses re percentges. controversil report of dengue trnsmission by Culex ftigns in Chin (15), nd erly development of spinner culture system using mosquito cells grown t higher temperture for mss production of virus ntigen (16). Although the role of lnguge in the determintion of journl coverge ws not entirely cler (e.g., ll five periodicls in English published in the Phillippines were not covered), the retrievl rte of source items from non-english periodicls (prticulrly from Asi) ws much lower thn tht from English periodicls despite the fct tht mny items in the former group hd bstrcts in English. A similr low rte ws lso found in the South Americn publictions. Even in western Europe, where retrievl rtes re generlly much higher, the rte in Frnce, where 86% of source items were printed in French, ws only 55%, while tht in Englnd ws 82%. Retrievl by theme of publiction The influence of the theme of publiction on retrievl ws nlysed. Dengue literture ws divided into primry nd secondry themes. The source items of primry theme were those publictions in which the word "dengue" ppered either in the title or bstrct, or dengue could be redily identified s the most importnt subject of the text. In the items of secondry theme, dengue, dengue virus, or derivtives of dengue virus were typiclly studied in reltion to other viruses or diseses which were the min subject of the item. As shown in Tble 3, retrievl rtes of primry theme items mong 7 dtbses, lthough slightly lower, were similr to the overll retrievl for items in periodicls (Tble 1). Retrievl of secondry theme items ws generlly lower thn tht of first theme items, with the exception of CAB which hd higher retrievl rte. WHO Bulletin OMS. Vol

4 G. Kuno Tble 3: Retrievl ccording to the theme of source items No. of source No. retrievble by dtbses Theme items A B C D E F G Primry (53.1)b (44.3) (32.0) (19.7) (37.8) (23.9) (28.8) Secondry (44.9) (31.9) (31.9) (29.6) (19.4) (7.9) (6.5) See footnote in Tble 1. b Figures in prentheses re percentges. Retrievl by subject (Tble 5); from these figures, four dtbses Since ech dtbse hs clerly defined set of crite- (Medline, Biosis, EM-Dtbse, nd Sciserch) i for coverge, emphsis nd objectives, it is posshred between 23% nd 35% of source items. sible to nlyse the dtbses by subject ctegory As regrds sole coverge, 61 (12.8%) of 477 nd determine which re more useful. Retrievl of items retrieved by Medline were not shred by the six subjects mong seven dtbses is compred in seven other dtbses studied. Similrly, 32 (15.6%) Tble 4. In the re of clinicl medicine nd epi- of 205 items retrieved by CAB were not shred by demiology, Medline ws the best nd retrieved ny other dtbse. lmost hlf of the literture in this re. For lbortory dignosis, retrievl rtes between Medline nd Retrievl by publisher's geogrphicl loction Biosis were lmost equl. Percentges of retrievl for bsic virology/immunology nd moleculr biology Since dengue is mjor helth problem primrily in were much higher mong five on-line dtbses thn tropicl countries, the dtbse coverges of publicthose in clinicl medicine/epidemiology or lbortory tions generted in those countries nd in temperte dignosis. It is evident tht the strength of CAB lies countries were compred. In this nlysis, retrievl in vector biology nd control, since it registered ccording to the geogrphicl loction of the publithe highest retrievl in tht subject ctegory. shers (but not the uthors' ntionlities) ws studied. Tble 6 shows tht for publishers in western Europe, South Afric, USA, Cnd, Indi, Austrli Overlpping nd sole coverge Since mny publictions re covered by more thn nd New Zelnd, the retrievl rtes by two or more one dtbse, considerble mount of overlp of dtbses exceeded 50%. In contrst, the rtes for retrievble source items ws found. The degree of most Asin nd South Americn countries were overlp between ny two dtbses ws nlysed generlly much lower. Tble 4: Retrievl ccording to the subject of source items No. of source No. retrieved by dtbses Subject items A B C D E F G Clinicl medicine nd epidemiology (47.9)b (35.8) (22.4) (20.4) (19.5) (22.0) (23.0) Lbortory dignosis (50.0) (48.6) (33.8) (14.1) (30.3) (25.5) (29.6) Bsic virology nd immunology (65.4) (58.8) (54.3) (6.9) (62.8) (29.3) (22.9) Moleculr biology (81.3) (75.0) (68.8) (12.5) (72.9) (27.1) (33.3) Vector biology nd control (43.8) (36.2) (26.2) (57.7) (33.1) (3.8) (24.6) Generl subjects (33.3) (10.8) (9.9) (24.3) (10.8) (9.0) (13.5) See footnote in Tble 1. b Figures in prentheses re percentges. 168 WHO Bulletin OMS. Vol

5 Computer literture serches on dengue Tble 5: Overlp of coverge between two dtbses nd sole coverge by only one dtbse No. of source items retrieved by two dtbses combined A B C D E F A B 315 (34.1 )b C (27.8) (24.2) D (14.5) (11.8) (7.6) E (30.1) (24.7) (23.3) (8.7) F (17.0) (12.9) (12.8) (5.4) (13.7) No. of source items retrieved by only one dtbse See footnote in Tble 1. b Figures in prentheses re percentges of 931 dengue source items. Tble 6: Retrievl ccording to the geogrphicl loction of the publisher No. retrievble by dtbses- Region N o. of source items A B C D E F G Europe (64.5)b (39.0) (49.8) (26.4) (56.0) (25.1) (31.2) South Afric (75) (75) (50) (100) (50) (25) (25) Indi (62.5) (65.6) (15.6) (21.9) (59.4) (6.3) (12.5) South-est Asi (31.4) (27.5) (5.3) (15.4) (2.4) (6.8) (23.7) Est Asi (27.1) (48.8) (34.9) (10.1) (17.8) (17.8) (8.5) Austrli nd New Zelnd (75.0) (62.5) (37.5) (37.5) (62.5) (37.5) (25.0) USA nd Cnd (68.8) (53.6) (47.3) (30.3) (51.6) (30.8) (30.0) Mexico nd Centrl Americ (57.1) (14.3) (28.6) (14.3) (28.6) (21.4) Cribben Bsin (59.5) (32.4) (5.4) (5.4) (8.1) (16.2) (5.4) South Americ (20.0) (19.4) (12.9) (22.6) (6.5) (12.9) (12.9) See footnote in Tble 1. b Figures in prentheses re percentges of the totl number per region. WHO Bulletin OMS. Vol

6 G. Kuno Discussion Abstrcting services re not designed to be comprehensive becuse of differences in objectives or res of speciliztion, so tht comprtive study of the retrievl of dtbse subset, such s dengue references, should not be interpreted s n evlution of the qulity of the vrious dtbses. However, informtion on the chrcteristics of ech dtbse will gretly ssist investigtors serching for informtion on dengue or other disese. Indexing of informtion from books hs lwys been difficult re. The extremely low retrievl rte demonstrted in this study is nevertheless regrettble. The development of indexes specificlly designed for book contents, such s the Index to Scientific Book Contents (by the Institute for Scientific Informtion in the USA), is welcome ddition to informtion dtbse. However, check of this index for the period produced only two sources items, while CATLINE, book ctlogue of the U.S. Ntionl Librry of Medicine, hd 12 source items. As it is not importnt to collect ll the source items, the dequcy of retrievl depends on the proportion of wht ech investigtor considers s importnt source items tht re retrievble in given field of speciliztion. As shown in the results, moleculr biologists my be stisfied to rely on only one or two on-line dtbses becuse of high rtes of retrievl. However, it is highly likely tht mny investigtors in the fields of clinicl medicine nd epidemiology (nd to some extent entomologists) my be disstisfied, not only becuse of lower retrievl rtes (less thn 50%) but becuse of the omission of mny importnt documents. Adequte coverge of source items in these fields is importnt for the following resons. (i) Severe forms of dengue, DHF nd dengue shock syndrome (DSS), cses of which re incresing in number, hve mnifested themselves in vrious forms in different loclities. Becuse unusul symptoms nd syndromes hve been continuously reported (17-23), the numerous clinicl studies of dengue do not necessrily represent redundnt scientific observtions or knowledge. (ii) Uniform ppliction of criteri for grding the severity of DHF/DSS, estblished by WHO, hs been difficult becuse of clinicl vrition in some prts of the tropics (24, 25). Thus, documenttion of ll typicl cses is importnt. (iii) Dissemintion of epidemiologicl informtion is criticl to the estblishment of n effective globl dengue surveillnce system. If viruses re isolted, they could be used for moleculr epidemiologicl studies to monitor the movement of geogrphicl strins. 170 (iv) Antibody prevlence studies re vitlly importnt for mpping vccintion strtegy with n ttenuted vccine, since, ccording to one theory (26), pre-existing ntidengue ntibody is sid to be risk fctor in secondry infections for the development of DHF/DSS. (v) Proper indexing of epidemiologicl informtion on dengue helps physicins in temperte countries to better dignose imported tropicl diseses by mtching the trvel history of ptients with the epidemiologicl informtion in dtbses. Since dengue is tropicl disese, it is ironic tht dengue informtion published in tropicl countries is generlly retrievble t much lower rtes thn tht published in industrilized countries in temperte region. For exmple, no periodicl from the Philippines, where DHF/DSS ws first documented in the 1950s nd where it continues to be mjor helth problem, ppers in the List ofjournls indexed in Index Medicus (1990). Similrly, no periodicl from t lest 16 dengue-endemic countries, including Colombi, is covered in the Index Medicus. Apprently, this is not problem unique to dengue literture, since similr problem ws found in leishmnisis literture (27). In mny primry dtbses, informtion from "core" joumls (28), which re published in English primrily in highly industrilized countries in westem Europe nd north Americ, is covered more intensely thn tht from non- "core" joumls in other countries. This prctice is prtly bsed on the ssumption tht importnt contributions from less industrilized countries re eventully published in intemtionlly-recognized periodicls in the west. While this pplies lso to dengue reserch, the fct tht mny importnt rticles in clinicl medicine nd epidemiology re not submitted to those westem periodicls should not be overlooked. As described erlier, it is evident tht lnguge nd geogrphicl coverge re importnt for jouml selection. Low coverge of medicl joumls in tropicl countries not only ccelertes dependence on periodicls in highly industrilized countries in the west for publiction nd recognition by the investigtors in the tropics but lso impedes the growth of scientific publictions in developing countries. Bsed on lower rtes of retrievl nd considerble number of importnt source items not covered, it ppers tht, so fr s clinicl medicine nd epidemiology of dengue re concemed, mny improvements re needed. If it is difficult to modify coverge policies of existing dtbses (29), it is highly desirble to develop different dtbses designed for clinicl medicine nd epidemiology of tropicl diseses with the doption of new set of criteri. Geogrphicl medicine must be included s n essentil criterion for more complete coverge to WHO Bulletin OMS. Vol

7 Computer literture serches on dengue improve globl surveillnce of dengue which is spreding to mny countries becuse of incresed intemtionl trvel. It is emphsized tht nerly 60% (181/313) of the source items on clinicl medicine nd epidemiology were published only in the countries outside western Europe nd North Americ. Thus, removing lnguge brrier is lso importnt. Another pproch to improve coverge of source items from tropicl countries is to use informtion compiled in regionl informtion centres or interntionl orgniztions. BIREME is supported by the Pn Americn Helth Orgniztion nd is strong in Ltin Americn nd Cribben medicl literture. Miscellneous bibliogrphic publictions (such s Bibliomed-Singpore/Mlysi, Indonesin Index Medicus, Philippine Medicl Literture, Thi Index Medicus) published in English for south-est Asin countries by the Interntionl Medicl Foundtion of Jpn (Tokyo, Jpn) re lso indispensble becuse of unique dtbses not shred by most primry dtbses vilble to on-line users. Since some of those dtbses, such s those of BIREME nd ntionl informtion centres of some countries, re lredy vilble on CD-ROM, development of networks mong dtbses in the ner future is highly desirble. The recommendtions for improving coverge described bove were not ment to encourge mediocre publictions. Without continuous improvement in qulity, punctul publiction nd delivery to indexing institutions, nd n ggressive effort by the editoril stff to gin coverge in primry dtbses, recognition of mny contributions of dengue in some tropicl countries will, unfortuntely, remin low. At time when more nd more pperless electronic journls re being initited, the fct tht ccess to computerized informtion services is still difficult for mny investigtors in dengue endemic countries is mtter to be considered elsewhere. Acknowledgements I m deeply indebted to the stff of the Informtion Center of the Centers for Disese Control nd Prevention, Atlnt, GA, USA, for obtining originl rticles nd operting the on-line services. For invluble ssistnce in this work I should like to thnk Dr Yuki Eshit, Kurume University School of Medicine, Kurume, Jpn; Dr E. Leontsini, Johns Hopkins University School of Hygiene nd Public Helth, Bltimore, MD, USA; Dr Lu Bo-Lin, Institute of Microbiology nd Epidemiology, Acdemy of Militry Medicine, Beijing, Chin; Dr F. Rodhin, Institut Psteur, Pris, Frnce; Ms Sunthree Rojnsuphot, Virus Reserch Institute, Nonthburi, Thilnd; Dr R.S. Songco, Hospitl of Infnt Jesus, Mnil, Philippines; nd Ms E. Tunis, Ntionl Librry of Medicine, Bethesd, MD, USA. Ms Eileen Curtin (Institute for Scientific Informtion) nd Ms Lois Ann Colinni (Ntionl Librry of Medicine) provided useful informtion on journl selection. Resume Recherches bibliogrphiques informtis6es sur l dengue L'efficcite de l recherche d'informtions sur l dengue, mldie virle trnsmise pr les moustiques l plus importnte chez I'homme, 6te nlysee u moyen de huit bses de donnees, dont cinq interctives. Les r6sultts obtenus ont vri6 selon l'origine de l bse de donnees, le type de publiction, le sujet et l loclistion geogrphique de I'editeur. Si l recherche de references en biologie mol6culire de l dengue ete g6nerlement stisfisnte (plus de 70% de l litterture mondile sur l dengue), cel n' ps ete le cs en medecine clinique et en 6pid6miologie (<50%). Contrirement ux references bibliogrphiques prues dns des periodiques, les ouvrges publi6s etient peu repr6sent6s (retrouv6s dns <10% des cs dns l pluprt des bses de donn6es). L dengue etnt un probleme de snte importnt, surtout sous les tropiques, il prlt curieux que l couverture des spects cliniques et 6pid6miologiques de cette mldie dns les publictions des pys d'end6mie de l dengue soit insuffisnte pr rpport celle retrouvee dns les pys fortement industrilises d'europe occidentle et d'am6rique du Nord. En rison des vritions observ6es dns les mnifesttions cliniques de l dengue et de l n6cessite d'meliorer l surveillnce mondile de cette mldie dns le cdre de l pr6vention, l mise u point d'une nouvelle bse de donn6es qui mettrit I'ccent sur une m6decine g6ogrphique sns brriere linguistique est tout fit souhitble. References 1. Boshell, J. et l.[dengue in Colombi.] Biomedic, 6: (1986) (in Spnish). 2. Chkrborty, M.S. et l. Isoltion of dengue virus (type 3) from Clcutt in Bulletin of the Clcutt School of Tropicl Medicine, 31: (1983). 3. Supeernunth, A. Dengue hemorrhgic fever cusing blindness in Thi ptient: cse report. Bulletin of the Deprtment of Medicl Sciences (Bngkok), 9: (1984). 4. Limont, M. et l. [Use of leukocytic interferon during n epidemic of dengue hemorrhgic fever (virus type 11) in Cub.] Interferon y biotechnologi, 1: (1984) (in Spnish). 5. Dengue hemorrhgic fever in Sint-Luci nd the Dominicn Republic. Bulletin of the Pn Americn Helth Orgniztionn 21: (1987). 6 Igrshi, A. [Infectious diseses in Vietnm: with specil references to Jpnese encephlitis nd dengue hemorrhgic fever.] Infection, inflmmtion, WHO Bulletin OMS. Vol

8 G. Kuno immunity, 16: (1986) (in Jpnese). 7. Yun, C-F. et l. [Clinicl nlysis of nineteen cses of dengue fever with intrvsculr hemolysis.] Chinese journl of infectious diseses, 4: (1986) (in Chinese). 8. Mendoz, H. et l. [Virl infections ssocited with encephlitis in bbies.] Archivos Dominicnos peditri, 20: (1984) (in Spnish). 9. Pilski, J. et l. [Dengue in the Mldives.] Mitteilungen der Osterreichischen Gesellschft fur Tropenmedizin und Prsitologie, 7: (1985) (in Germn). 10. Dziubek, Z. et l. [A cse of dengue fever.] Przegld epidemiologiczny, 40 (4): (1986) (in Polish). 11. Ten Figuero, S. [Outbrek of clssicl dengue in the city of Hermosillo Sonor.] Border helth, 4: (1988) (in Spnish). 12. Vn der Wls, F.W. et l. Post-encephlitic epilepsy nd rbovirus infections in n isolted rinforest re of centrl Liberi. Tropicl nd geogrphic medicine, 38: (1986). 13. Al-Nkib, W. et l. Preliminry report on rbovirusntibody prevlence mong ptients in Kuwit: evidence of Congo-Crimen virus infection. Trnsctions of the Royl Society of Tropicl Medicine nd Hygiene, 78: (1984). 14. Schofield, C.J. & White, G.B. House design nd domestic vectors of disese. Trnsctions of the Royl Society of Tropicl Medicine nd Hygiene, 78: (1984). 15. Li, X.-D. & Zho, G.-F. Studies on the experimentl trnsmission of dengue virus by Culex ftigns. Chinese journl of microbiology nd immunology, 5: (1985). 16. Zhng, M.-J., et l. Suspension culture of mosquito cells for multipliction of dengue virus. Journl of the Acdemy of Militry Medicine (Chin), 10: (1986). 17. Sumrmo et l. Clinicl observtions on hospitlized ptients with virologiclly confirmed dengue hemorrhgic fever in Jkrt, Peditric Indonesin, 26: (1986). 18. Nimmnnity, S. et l. Dengue hemorrhgic fever with unusul mnifesttions. Southest Asin journl of tropicl medicine nd public helth, 18: (1987). 19. George, R. et l. Unusul clinicl mnifesttions of dengue virus infection. Southest Asin journl of tropicl medicine nd public helth, 19: (1988). 20. Chuhn, G.S. et l. "Encephlitis" s mnifesttion or coincidence in cse of clssicl dengue fever. Journl of the Assocition of Physicins of Indi, 35: (1987). 21. Chinuvti, T. Dengue hemorrhgic fever- fscinting disese of Asi. Journl of gstroenterology nd heptology, 5: (1990). 22. Innis, B.L. et l. Acute liver filure is one importnt cuse of ftl dengue infection. Southest Asin journl of tropicl medicine nd public helth, 21: (1990). 23. Hdinegoro, S.R. & Nthin, M.A. The chnging ptterns of clinicl mnifesttions in dengue hemorrhgic fever: ten yers observtions. Southest Asin journl of tropicl medicine nd public helth, 21: (1990). 24. Munir, M. & Rmpengn, T.H. Dengue shock syndrome: n evlution of clinicl experiences. Peditric Indonesin, 24: (1984). 25. Sirinvin, S. & Hthirt, P. Problems in severity grding of dengue hemorrhgic fever. Southest Asin journl of tropicl medicine nd public helth, 21: 696 (1990). 26. Hlsted, S.B. Immune enhncement of virl infection. Progress in llergy, 31: (1982). 27. Wrren, M. Foreword to "Leishmnisis in Arbi: n nnotted bibliogrphy", by Killick-Kendrick, R. & Peters, W. Americn journl of tropicl medicine nd hygiene, 44 (Supplement): 1 (1991). 28. Grfield, E. Journl selection for Current contents: editoril merit vs. politicl pressure. Current contents, 28 (11): 3-11 (1985). 29. Grfield, E. How ISI selects journls for coverge: quntittive nd qulittive considertions. Current contents, 33 (22): 5-13 (1990). 172 WHO Bulletin OMS. Vol