Ten-year audit of randomized trials in digestive surgery from Europe K. Slim, M. Haugh*, P.-L. Fagniez², D. Pezet and J. Chipponi DepartmentofGeneralandDigestiveSurgery,Hotel-Dieu,University HospitalofClermont-Ferrand,Clermont-Ferrand,*FrenchCochrane Centre,CentreLeonBerard,Lyonsand ²DepartmentofDigestive Surgery,HospitalHenriMondor,Cre Âteil,France Correspondence to: DrK.Slim,DepartmentofGeneralandDigestive Surgery,Hotel-Dieu,BoulevardLeonMalfreyt,F-63058, Clermont-FerrandCedex1,France Paperaccepted2May2000 Randomized trials remain scarce in the ®eld of digestive surgicalinvestigationbecausetheypresentuniquefeatures andproblems 1 .Thisstudywasareviewofrandomizedtrials inthe®eldofdigestivesurgerytoassesstheevolutionin Europeoverthepastdecade. Materials and methods A comprehensive computerized search was conducted in three databases Medline, Embase and the Cochrane Library)between1January1990and31December1999. Amanualsearchwasalsoperformedinthepastsixissuesof 14 major journals American Journal of Surgery, Annals of Surgery, Archives of Surgery, British Medical Journal, British JournalofSurgery, DiseasesoftheColonandRectum, European Journal of Surgery, Gastroenterology, Gut, Journal of the AmericanCollegeofSurgeons, Lancet, NewEnglandJournalof Medicine, SurgicalEndoscopy, Surgery).Onlyoriginalarticles comparingtwosurgicalproceduresorasurgicalandanon- surgicaltreatmentwereretained.Whentheauthorswere from different countries, the trial was considered as international. Thetrialswereclassi®edintothreecategories:category 1, trials evaluating the core procedure itself similar to drugs);category2,trialsevaluatinganimportantstepofthe surgical procedure; and category 3, trials evaluating a technicaldetail. Results Twohundredand®fty-onetrialswereretrievedapproxi- mately two trials published per month). Fig.1 shows the evolution of the number of trials per year in Europe comparedwiththatinNorthAmerica,where52trialswere recorded over the same interval references for all these trialscanbeobtainedfromtheauthorforcorrespondence). TheUKwasthecountryofaf®liationin51trials20per centoftrials),Francein38,Germanyin34,Swedenin22, Finlandin20andothercountriesin82.Therewerefour internationalmulticentretrials.Whenthecontributionof each country was calculated per million inhabitants, the numberofrandomizedtrialswassigni®cantlyhigherinthe Nordiccountries,namelyFinland,SwedenandDenmark Table1). Although the number of trials of conventional surgical procedures remained stable mean about 15 per year), the observed increase was due to trials evaluating laparoscopicproceduresalmosthalfofthepublishedtrials overthepast4years).ForthewholeofEurope,210trials 84percent)wereincategory1or2Table1).Inthewhole sample,22percentofthetrialsweremulticentre;inFrance 59percentofthetrialsweremulticentrewitharangeinthe othercountriesfrom7percentinSpainto33percentinthe Netherlands. Discussion This study showed that European digestive surgeons design, conduct and publish increasing numbers of randomizedtrials,mainlysincetheintroductionoflaparo- 0 1999 40 1990 35 30 25 20 15 10 1991 1992 1993 1994 1995 1996 1997 1998 Europe USA 5 No. of trials Fig.1 Evolutionofthenumberoftrialsperyearoverthepast decadeinthe®eldofdigestivesurgeryinEuropeandtheUSA Table1 Numberoftrialswiththeircategory)foreach Europeancountryandtherespectivenumberpermillion inhabitants Category No. of trials Total no. per million Country 1 2 3 of trials inhabitants UK 20 22 9 51 0´8 France 12 18 8 38 0´7 Germany 14 15 5 34 0´5 Sweden 9 10 3 22 2´8 Finland 2 16 2 20 4´0 Italy 3 13 1 17 0´3 Netherlands 6 6 4 16 1´0 Spain 2 8 4 14 0´4 Denmark 8 3 1 12 2´4 Other 10 13 4 27 0´3 Whole of Europe 86 124 41 251 0´6 Short note ã 2000BlackwellScienceLtd BritishJournalofSurgery2000, 87, 1585±1586 1585