Comparison of survival and complication rates of tooth - PDF document

Comparisonofsurvivalandcomplicationratesoftooth-supportedÞxeddentalprostheses(FDPs)andimplant-supportedFDPsandsinglecrowns(SCs)BjarniE.PjeturssonUrsBra¨ggerNiklausP.LangMarcelZwahlenAuthors’afliations:BjarniE.Pjetursson,UrsBra¨gger,NiklausP.LangUniversityofBerneSchoolofDentalMedicine, possible,onformalquantitativeevidencesynthesisandmeta-analysis(Egger&Smith1997;Eggeretal.2001a,2001c).Reviewingtheliteratureinvolvesgradingoftheavailableandpublishedstudies.Of-ten,suchgradingsarebasedonstudyde-sign.Usuallyrandomizedstudiesareratedhigherthanobservationalstudies(GradesofRecommendation2004).Furthermore,thequalityofstudiesandtrialsisofcrucialimportance:iftheÔrawmaterialÕisßawed,thentheÞndingsofreviewsofthismaterialmayalsobecompromised(Egger&Smith1997;Eggeretal.2001a).Thetrialsandstudiesincludedinsystematicreviewsandmeta-analysesshouldideallybeofhighmethodologicalqualityandfreeofbias.Asaconseqence,thedifferencesinstudyoutcomesobservedbetweenpatientscanconÞdentlybeattributedtotheinterven-tionunderinvestigation.Iftherearenostudiesonthehighestlevelofevidence(randomizedcontrolledclinicaltrials),thesystematicreviewhastobebasedonthehighestlevelofevidenceavailable(Eggeretal.2001b;Glasziouetal.2004)andpointoutwhichadditionalresearchshouldbeconductedtostrengthentheevidencebase.Thestudiesinthedentalliteraturere-portingontooth-supportedandimplant-supportedÞxeddentalprosthesis(FDPs)aremostlyobservationalstudiesandsin-gle-centercasecohorts.Thesystematicreviewsconductedsofar(Pjeturssonetal.2004a,2004b;Tanetal.2004)indicatethattheobservedsurvivalofFDPsafter10yearsrangedbetween85%and95%.Tostrengthentheevidencebase,morestudiesshouldreportonthelong-termoutcomesofreconstructionsalreadyin-serted.Afewkeyoptionsshouldbetestedincomparativerandomizedcontrolledclin-icaltrials.Suchtrialswouldbestbeconductedinamulti-centersettingwithlongenoughfollow-upandsamplesizesthatallowestimatingrelevantdifferences.Forexample,todetectwitha80%powerandatthesigniÞcancelevelof5%,aclinicallyrelevantdifferenceofanannualrateoflossofreconstructionsof1%vs.2%,atwo-armstudywouldneedtorando-mizein1yearapproximately1060patientsandtofollowthemforatleast4yearsresultinginatotalstudytimeof5years.Withalongerfollow-upofabout10yearsitwouldbesufÞcienttorandomize500pa-tients.Forascenariocomparingtheannualrateoflossofreconstructionsof0.5%vs.2.5%,the5-or10-yearfollow-upstudywouldneedtorandomize260or130pa-tients,respectively.Therearenostudiesonthehighestlevelofevidence(RCTs)availableinthedentalliteraturecomparingtooth-supportedandimplant-supportedFDPs.Inaddition,onlyfewstudieshavereportedonthelongevityofreconstructionsonimplantswithsamedetails.Therefore,aseriesofsystematicreviews,basedonconsistentinclusionandexclusioncriteria,hassummarizedtheavailableinformationonsurvivalandsuc-cessratesandtheincidenceofbiologicalandtechnicalcomplicationsofconven-tionalFDPs,cantileverFDPs,combinedtooth-implant-supportedFDPsandsolelyimplant-supportedFDPsandimplant-sup-portedSCs(Langetal.2004;Pjeturssonetal.2004a,2004b;Tanetal.2004;Jungetal.2007).Theaimofthissystematicreviewistoextendthesereviewstoanalyzeandcom-parethesurvivalandsuccessratesbydifferentdesignsoftoothandimplant-sup-portedÞxedreconstructionsandtoassesstheincidenceofbiologicalandtechnicalcomplications.MaterialandmethodsSearchstrategyandstudyselectionThreeMEDLINE(PubMED)searcheswereperformedforarticlespublishedintheDentalLiterature.TheÞrstonecoveredthetimeperiod1966ÐApril2004andsearchedforarticlesreportingonconven-tionaltooth-supportedend-abutmentFDPsandcantilevertooth-supportedFDPspub-lishedintheEnglishlanguagebysearchingforÔÞxedpartialdenturesORbridges,ÕandÔpartialedentulismÕ(Pjeturssonetal.2004b;Tanetal.2004).ThesecondsearchcoveredthesametimeintervalandwasconductedforEnglish-languagearticlesreportingonsolelyimplant-supportedFDPsandcombinedtooth-implant-supportedFDPsusingthesearchtermsÔÞxedpartialdenturesORbridges,ÕÔpartialedentulism,ÕÔimplantsandÞxedpartialdenturesORbridges,ÕÔimplantsÕandÔcom-plications,ÕÔimplantsÕandÔfailures,ÕÔim-plantsÕandÔlongitudinalÕ(Langetal.2004;Pjeturssonetal.2004a).Thethirdlitera-turesearch,forMEDLINEfrom1966uptoandincludingJuly2006,wasconductedforEnglish-andGerman-languagearticlesinDentalJournalsusingthefollowingsearchterms(modiÞedfromBerglundhetal.2002)andlimitedtohumantrials:ÔimplantsÕandÔsurvival,ÕÔimplantsÕandÔsurvivalrate,ÕÔimplantsÕandÔsurvivalanalysis,ÕÔimplantsÕandÔcohortstudies,ÕÔimplantsÕandÔcaseÐcontrolstudies,ÕÔimplantsÕandÔcontrolledclinicaltrials,ÕÔimplantsÕandÔrandomizedcontrolledclinicaltrials,ÕÔimplantsÕandÔcomplications,ÕÔimplantsÕandÔclinical,ÕÔimplantsÕandÔlongitudinal,ÕÔimplantsÕandÔprospective,ÕÔimplantsÕandÔretrospec-tive.ÕAdditionalsearchstrategiesincludedthetermsÔsingle-tooth,ÕÔfailure,ÕÔperi-im-plantitis,ÕÔfracture,ÕÔcomplication,ÕÔtechni-calcomplication,ÕÔbiologicalcomplication,ÕÔscrewlooseningÕandÔmaintenance(Jungetal.2007).ÕAllthreesearcheswerecomplementedbymanualsearchesofthebibliographiesofallfull-textarticlesandrelatedreviews,selectedfromtheelectronicsearch.Furthermore,manualsearchingwasap-pliedtorelevantjournalsintheÞeldofinterest.Preparingthissystematicreview,allthreeoriginalsearcheswereupdatedandextendeduptoandincludingSeptemberInclusioncriteriaIntheabsenceofRCTs,thissystematicreviewwasbasedonprospectiveorretro-spectivecohortstudies.Theadditionalinclusioncriteriaforstudyselectionwerethat:thestudieshadameanfollow-uptimeof5yearsormore,theincludedpatientshadbeenexam-inedclinicallyatthefollow-upvisit,i.e.,publicationsbasedonpatientre-cordsonly,onquestionnairesorinter-viewswereexcluded,thestudiesreporteddetailsonthechar-acteristicsofthesuprastructures,publicationsthatcombinedÞndingsofbothFDPandsinglecrowns(SCs)describedatleast2/3ofthereconstruc-tionasFDPs.SelectionofstudiesTitlesandabstractsofthesearcheswerealwaysscreenedbyatleasttwoindepen-98|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaardPjeturssonetal.SurvivalofFDPsandsinglecrowns tooth-supportedFDPs,theearlieststudiesdatedback35yearsandthemedianyearofpublicationwas1995(Table1).Thestudiesreportingonimplantandtooth-implant-supportedreconstructionsweremorere-centandalmostexclusivelypublishedwithinthepast10years.Themostrecentarticlesreportedonimplant-supportedSCswith2002asamedianyearofpublication(Table1).Themajority,or23outof28,ofthestudiesonconventionalandcantilevertooth-supportedFDPswereretrospective.Ontheotherhand,themajorityofstudiesonimplant-supportedreconstructions(42outof57)wereprospectiveinnature.Thehighestproportion(81%)ofprospectivestudieswasfoundfortheimplant-sup-portedSCs(Table1).Toevaluatethein-ßuenceofstudydesign,10prospectivestudiesandÞveretrospectivestudiesre-portingonsurvivalofimplants,supportingFDPs,wereanalyzedseparately.Fortheprospectivestudies,basedon1576im-plants,thesummaryestimateofthesurvi-valwas95.6%(95%CI:93.3Ð97.2%)andfortheretrospectivestudies,basedon1973implants,thesummaryestimateofthesurvivalwas95%(95%CI:93Ð96.4%).FormallyinvestigatingthedifferenceineventratesinaPoissonregressionanalysisconÞrmedtheabsenceofastudydesigneffect(0.64)forthismaterial(Pjeturs-sonetal.2004a).Fromthe21studiesreportingoncon-ventionalFDPs,11reportedonbridgede-sign.TherelativedistributionofFDPsinthesereportswas6%metal-ceramic,54%gold-acrylic,whilefortheremainder40%,thebridgedesignwasnotreported.ForcantileverFDPs,17%oftheFDPsweremetal-ceramic,45%weregold-acrylicandfor38%ofthematerial,bridgedesignwasnotreported(Table2).Forimplant-sup-portedFDPs,52%oftheFDPswereme-tal-ceramic,33%weregold-acrylicandfor15%,thebridgedesignwasnotreported.ThecorrespondingÞguresfortooth-im-plant-supportedFDPswere32%,19%and49%.Forimplant-supportedSCs,15outofthe26studiesreportedonthematerialused.FourtypercentoftheSCsweremetal-ceramic,4%weregold-acrylicand9%wereall-ceramic(Table2).Abouthalfofthestudies,reportingonimplant-supportedreconstructions,describedthetypeofretentionutilized.Forthesolelyimplantandcombinedtooth-implant-sup-portedFDPs,themajorityoftherecon-structionswerescrewretained.Ontheotherhand,themajorityoftheSCswerecemented(Table2).SurvivalwasdeÞnedasthereconstructioninsituwithorwithoutmodiÞ-cationovertheobservationperiod.ConventionalFDPsFifteenstudiesprovideddataonthesurvi-valofconventionalFDPs(Table3).Thereportsweredividedintotwogroups:TheÞrstgroupwithatotalof2088FDPsandameanfollow-uptimeof5.7yearsandthesecondgroupwithatotalof1218FDPsandameanfollow-uptimeof11.9years.Intheformergroup,273outof2088FDPsandinthesecondgroup190outof1218FDPswerelost.Inmeta-analysis,theannualfailurerate(Table3)wasestimatedat1.28per100FDPyearsfortheformerand1.14forthelattergroup,translatingintoa5-yearsurvivalofconventionalFDPsof93.8%anda10-yearsurvivalof89.2%(Table3).Thestudieswerealsodividedaccordingtothematerialutilized:AgroupofÞvestudieswithatotalof1163metal-ceramicFDPsandagroupofsixstudieswithatotalof1756gold-acrylicFDPs.Thegroupwithmetal-ceramicFDPsdemostratedahigher10-yearsurvivalof89.1%(95%CI:82.9Ð93.2%)comparedwithsurvivalof(95%CI:72.6Ð93.5%)forthegold-acrylicFDPs.Thisdifference,however,didnotreachstatisticalsigniÞcance(CantileverFDPsTwelvestudiesprovideddataonthesurvi-valofcantileverFDPs(Table4).There-portswere,likefortheconventionalFDPs, Table1.CharacteristicsofincludedstudiesTypeofreconstructionYearofpublicationsNumberofStudydesignFollow-uptimeRangeMedianProspectiveRetrospectiveRangeMeanConventionalFDPs1968Ð20061995212191Ð257.1CantileverFDPs1970Ð20001991133102Ð237.1Implant-supportedFDPs1989Ð20022001241950Ð166.3Tooth-implant-supportedFDPs1989Ð2002200014951Ð196.1Implant-supportedSCs1996Ð20062002262151Ð135.4FDP,Þxeddentalprosthesis;SCs,singlecrowns. Table2.MaterialandtypeofretentionTypeofreconstructionNumberofreconstructionsMaterialRetentionceramic(%)resin(%)ceramic(%)reported(%)retained(%)reported(%)ConventionalFDPs3548654040ÐÐÐCantileverFDPs8161745038ÐÐÐImplant-supportedFDPs1336523301554352Tooth-implant-supportedFDPs538321904935245Implant-supportedSCs153040494739556FDP,Þxeddentalprosthesis;SCs,singlecrowns.Pjeturssonetal.SurvivalofFDPsandsinglecrowns100|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard copewithocclusalload.Theseevaluated10-or12-unitFDPsinthemandiblewithtwotothreecantileverunitsbilaterally,supportedbyonlytwocanines(Oetal.1991;Carlson&Yontchev1996).Iftheseextremereconstructionswereex-cludedfromtheanalysis,the10-yearsur-vivalofcantileverFDPswentupto81.7%(95%CI:77.2Ð85.3%).Thestudieswerealsodividedaccordingtotheveneermaterialutilized:Agroupoftwostudieswithatotalof136FDPswithceramicsasaveneermaterialandagroupoffourstudieswithatotalof211FDPswithacrylicveneerswereavailableforanalysis.TherewasnosigniÞcantdiffer-encebetweenthetwogroups.ThemetalceramicFDPshadasurvivalafter10yearsof85%(95%CI:81.4Ð87.9%),comparedwithasurvivalof84.9%(95%CI:75.1Ð91.1%)forthegold-acrylicFDPs.Implant-supportedFDPsTwentystudiesprovideddataonthesurvi-valofsolelyimplant-supportedFDPs(Table5).Thereportswereagaindividedintotwogroups:TheÞrstgroupwithatotalof1384FDPsandameanfollow-uptimeof5years,andthesecondgroupwithatotalof219FDPsandameanfollow-uptimeof10years.Intheformergroup67outof1384FDPsandinthesecondgroup27outof219FDPswerelost.Inmeta-analysis,theannualfailurerate(Table5)wasestimatedat0.99per100FDPyearsfortheformerand1.43forthelattergroup,translatingintoa5-yearsurvivalofimplant-supportedFDPsof95.2%anda10-yearsurvivalof86.7%(Table5).Thestudiesinthe5-yearsobservationgroupwerealsodividedaccordingtotheveneermaterialutilized:Agroupof13studieswithatotalof927FDPswithceramicasthenveneermaterialandagroupoffourstudieswithatotalof450FDPswithacrylicveneers.Thegroupwiththemetal-ceramicFDPsshowedasigniÞ-cantlyhigher(0.005)survivalafter5yearsof96.7%(95%CI:95.4Ð97.7%),comparedwithasurvivalof90.4%(95%CI:79.9Ð95.6%)forthegold-acrylicFDPs.Combinedtooth-implant-supportedFDPsTenstudiesprovideddataonthesurvivalofcombinedtooth-implant-supportedFDPs(Table6).Thereportsweredividedintotwogroups:TheÞrstgroupwithatotalof199FDPsandameanfollow-uptimeof5yearsandthesecondgroupwithonly72FDPsandameanfollow-uptimeof10years.Intheformergroupnineoutof199FDPsandinthesecondgroup14outof72FDPswerelost.Inmeta-analysis,theannualfailurerate(Table6)wasestimatedat0.92per100FDPyearsfortheformerand2.51forthelattergroup,translatingintoa5-yearsurvivalofimplant-supportedFDPsof95.5%anda10-yearsurvivalof77.8%(Table5).Implant-supportedSCTwenty-sixout465SCswerelost,andthestudyspeciÞc5-yearsurvivalvariedbetween89.6%and100%.Ten(45%)outofthe26SCswerelostwhilethesupportingimplantswerelost,butintheremaining16cases(55%),onlytherecon- Table5.Annualfailurerateandsurvivalofimplant-supportedFDPsStudyYearofpublicationTotalnumberofFDPsoffailuresTotalFDPsEstimatedfailurerate(per100FDPyears)Estimatedsurvivalafter5years(%)10years(%)5-yearfollow-upDegidi&Piattelli2005971561.7991.5Becker2004515.102610100Wennstrometal.200456532691.1294.6Preiskel&Tsolka2004786.625190.3998.1Anderssonetal.200336511640.6197Jemtetal.200263532951.0295Naertetal.20024095.51520490.7396.4Gotfredsen&Karlsson200152522360.8595.9ggeretal.200140511980.5197.5Mengeletal.2001750330100Behnekeetal.2000685.413720.2798.7Hosnyetal.2000186.501170100rtrop&Jemt199968533230.9395.5Wennerberg&Jemt1999133526080.3398.4Wyatt&Zarb1998975.4164983.2185.2Olssonetal.199523541023.9282.2Lekholmetal.19941975138891.4692.9Totalsummaryestimate(95%CI)138456769890.99(0.64Ð1.52)95.2(92.7Ð96.8%)10-yearfollow-upggeretal.2005331023200.6393.9Lekholmetal.1999163102113781.5285.9Gunneetal.1999231041912.0981.1Totalsummaryestimate(95%CI)219102718891.43(1.08Ð1.89)86.7(82.8Ð89.8%)BasedonstandardPoissonÕsregression,testforheterogeneityBasedonrandom-effectsPoissonÕsregression,testforheterogeneityFDP,Þxeddentalprosthesis.Pjeturssonetal.SurvivalofFDPsandsinglecrowns102|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard conventionalandcantilevertooth-sup-portedFDPsshowedhigherfailurerates.Moreover,forthecantileverFDPsthisdifferencereachedstatisticalsigniÞcance0.011)(Table8).Whenthestudiesreportingsolelyonmetal-ceramicreconstructions,excludinggold-resinandall-ceramicreconstruction,wereanalyzedseparately,thelowestan-nualfailureratewasseenforimplant-supportedFDP(0.66)followedbyim-plant-supportedSCs(0.92).Investigatingformally,therelativefailureratesofme-tal-ceramicreconstructions,usingagainimplant-supportedSCsasreference,bothcantilever(2)andconventionalFDPs(1.15)hadsigniÞcantlyhigherannualfailurerates0.026and0.041)(Table9).Analyzingthestudieswith10-yearsfol-low-uptime,theannualfailureratesran-gedfrom1.12to2.51,andthe10-yearsurvivalrangedfrom77.8%to89.4%.Aftera10-yearsobservationperiod,thelowestannualfailurerateswereseenforimplant-supportedSCs(1.12)andintheconventionalFDPs(1.14).CantileverFDPsandcombinedtooth-implantFDPshadsigniÞcantlyhigherannualfailureratesof2.20and2.51(0.043and0.045),re-spectively.Nevertheless,itmustbekeptinmindthattheresultsforcombinedtooth-implant-supportedFDPsandimplant-sup-portedSCsafter10-yearsfollow-uparebasedonasmallnumberofobservations,with60and69reconstructions,respec-tively(Table10).Onesingleclinic,theDepartmentofPeriodontologyandFixedProsthodontics,UniversityofBerne,Switzerland,pub-lisheddataonallÞvedifferenttypesofreconstructions(Table11).Withtheexcep-tionofconventionalFDPs(6.4years),alltheothergroupsofreconstructionshadameanfollow-uptimeof10years.Theresultsfromthiscenterweresimilartotheresultsobtainedinthemeta-analysisoftheprostheticdentalliterature.ThehighestsurvivalwasforconventionalFDPs(95.8%),followedbyimplant-sup-portedFDPs(93.9%)andimplant-sup-portedSCs(89.4%).LowersurvivalswerereportedforcantileverFDPs(84%)andcombinedtooth-implant-supportedFDPs(70.2%)(Table11).SuccesswasdeÞnedasanFDPthatre-mainedunchangedandfreeofallcompli-cationsovertheentireobservationperiod.Hence,suchareconstructiondidnotre-quireanyinterventionduringtheobserva-tionperiod.ConventionalFDPsFourstudies(Libbyetal.1979;Reichen-Graden&Lang1989;Fayyad&al-Rafee1996a,1996b;Walton2003)providedinformationonFDPsthatremainedintactovertheobservationperiod(Table12).Theestimatedstudy-speciÞcannualcomplica-tionratesrangedfrom1.34to7.07per100FDPyears.Inmeta-analysis,theannualcomplicationratewasestimated3.41at100FDPyearstranslatingintoa5-yearcomplicationrateofconventionalFDPsof15.7%(95%CI:8.5Ð27.7%)(Table12).Onestudy(Fayyad&al-Rafee1996a,1996b)reportingonpatientstreatedataUniversityclinicandbyaprivatedentistinSaudiArabiarepresentsanoutlierwitha Table8.Summaryofannualfailurerates,relativefailureratesand5-yearsurvivalestimatesTypeofreconstructionsTotalnumberofreconstructionsTotalfailurerate5-yearsurvivalsummaryestimate(95%CI)(%)failurerateConventionalFDPs208811,9985.71.28(0.64Ð2.59)93.8(87.9Ð96.9%)1.57(0.96Ð2.58)CantileverFDPs43221125.21.8(1.15Ð2.82)91.4(86.9Ð94.4%)2.15(1.19Ð3.89)Implant-supportedFDPs1384688050.99(0.64Ð1.52)95.2(92.7Ð96.8%)0.77(0.45Ð1.3)Tooth-implant-supportedFDPs19997650.92(0.5Ð1.7)95.5(91.9Ð97.5%)0.99(0.44Ð2.25)Implant-supportedSCs465228051.14(0.76Ð1.7)94.5(91.8Ð96.3%)1(Ref.)BasedonstandardPoissonÕsregression.Basedonrandom-effectsPoissonÕsregression.Basedonmultivariablerandom-effectsPoissonÕsregressionincludingalltypesofFDPs.FDP,Þxeddentalprosthesis;SCs,singlecrowns. Table9.Summaryofannualfailurerates,relativefailureratesand5-yearsurvivalestimatesfordifferenttypesofmetal-ceramicreconstructionsTypeofreconstructionsTotalnumberofreconstructionsTotalfailurerate(95%CI)(%)failurerateConventionalFDPs1163930181.15(0.71Ð1.87)94.4(91.1Ð96.5%)1.86(1.02Ð3.39)CantileverFDPs30419476.42(1.44Ð2.79)90.5(87Ð93.1%)2.01(1.09Ð3.7)Implant-supportedFDPs94850145.30.66(0.52Ð0.83)96.8(95.9Ð97.4%)0.89(0.45Ð1.75)Tooth-implant-supportedFDPs1247125.71.37(0.35Ð5.32)93.4(76.6Ð98.2%)1.35(0.59Ð3.15)Implant-supportedSCs25916366.30.92(0.66Ð1.27)95.5(93.9Ð96.7%)1(Ref.)BasedonstandardPoissonÕsregression.Basedonrandom-effectsPoissonÕsregression.Basedonmultivariablerandom-effectsPoissonÕsregressionincludingalltypesofFDPs.FDP,Þxeddentalprosthesis;SCs,singlecrowns.Pjeturssonetal.SurvivalofFDPsandsinglecrowns104|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard atahigherriskofhavingcomplications,thanpatientswithtooth-supportedcon-ventionalFDPs.BiologicalcomplicationsTooth-supportedreconstructionsThemostfrequentbiologicalcomplicationbytooth-supportedreconstructionswaslossofabutmentvitality.Onehundredninety-sixoutof1227abutmentteethconsideredvitalatthetimeofcementationpresentedwithalossofpulpvitalityovertheobservationperiod.Theannualcom-plicationrateswere1.26forconventionaland3.95forcantileverFDPstranslatinginto5-yearratesoflossofabutmentvital-ityof6.1%forconventionalFDPsandasigniÞcantly(0.017)highercomplica-tionrateof17.9%forcantileverFDPs(Table15).Whenthetwostudies(Oetal.1991;Carlson&Yontchev1996)reportingon12-unitFDPsinthemandiblesupportedbytwocanineswereexcludedfromtheanalysis,the5-yearrateoflossofabutmentvitalityforcantileverFDPsde-creasedto5.4%(95%CI:2.8Ð9.2%).Onestudy(Bergenholtz&Nyman1984),speciÞcallyaddressinglossofvital-ityinpatientsreconstructedaftersuccess-fultherapyforadvancedperiodontitis,reportedthehighest5-yearrateforlossofabutmentvitalityof8.2%(95%CI:5.9Ð11.1%).SigniÞcantlyhigherlossofvitalitywasobservedinabutments,whencomparedwithnon-preparedcontrolteeth.Thesecondmostcommonbiologicalcomplicationwasdentalcaries.Cariesrateswerereportedatthesurface,abut-mentandattheFDPlevels.Onlyonestudy(Karlsson1986)addresseddentalcar-iesonasurfacelevelandfound8.1%ofallsurfacesbeingdecayedwithin10years.From3176abutmentsanalyzed,290abut-mentsdevelopeddecayovertheobserva-tionperiod.Theannualcomplicationrateswere0.99forconventionaland0.95forcantileverFDPsgiving5-yearratesofden-taldecayatabutmentteethof4.8%forconventionalFDPsand4.7%forcantileverFDPs(Table15).SeveralstudiesreportedthenumberofFDPslostduetocaries.From1439FDPsexamined,27werelostduetodentalcaries.Theannualfailurerateswere0.32forconventionaland0.31forcantileverFDPs,respectively,translatinginto5-yearratesofFDPslostbecauseofdentalcariesof1.6%forconventionalFDPsand1.5%forcantileverFDPs(Table15).Thethirdbiologicalcomplicationwasthelossofthereconstructionduetorecur-rentperiodontitis.From1693FDPsana-lyzed,onlynineFDPswerelostduetoperiodontitis.Theannualfailurerateswere0.07forconventionaland0.1forcantileverFDPstranslatinginto5-yearratesoflossofFDPsduetoperiodontitisof0.4%forconventionalFDPsand0.5%forcantileverFDPs(Table15).Onestudy(Fayyad&al-Rafee1996a,1996b),however,reportedrecurrentperio-dontitistoaffectabutmentsof12.8%oftheFDPsafteronly5.1years.Astheinformationprovideddidnotdifferentiatebetweenperiodontitisreportedasacompli-cationorperiodontitisleadingtothelossoftheFDP,thestudywasexcludedfromtheanalysis.Implant-supportedreconstructionsPeri-implantmucosallesionswerereportedtoinvariouswaysbythedifferentauthors.Severalstudiesprovidedinformationonsofttissuecomplicationsandperiimplanti-tis,whileotherstudiesreportedsignsofinßammation(pain,redness,swellingand Table13.ComplicationratesandsuccessofcantileverFDPsStudyYearofTotalofFDPsNumberofTotalFDPsrate(per100FDPyears)successafter5years(%)mmerleetal.2000115103910353.7782.8Decocketal.19961686417415.5375.8Carlson&Yontchev1996129.5101059.5262.1Palmquist&Swartz19933418Ð2317n.a.n.a.n.a.Budtz-Jorgensen&Isidor199041581964.0881.5Reichen-Graden&Lang1989216.241303.0885.7Totalsummaryestimate(95%CI)39111922074.62(3.54Ð6.04)79.4(73.9Ð83.8%)BasedonstandardPoissonÕsregression,testforheterogeneityn.a.,notavailable;FDP,Þxeddentalprosthesis. Table14.Complicationratesandsuccessofimplant-supportedFDPsStudyYearofpublicationTotalnumberofpatientsNumberofTotalrate(per100patientsyears)successafter5years(%)Jemtetal.20024252219511.2856.9ggeretal.200133571634.2980.7rtrop&Jemt19995853028110.6858.6Wennerberg&Jemt199913356360810.3659.6Totalsummaryestimate(95%CI)26612212479.78(8.07Ð11.86)61.3(55.3Ð66.8%)BasedonstandardPoissonÕsregression,testforheterogeneityFDP,Þxeddentalprosthesis.Pjeturssonetal.SurvivalofFDPsandsinglecrowns106|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard TechnicalcomplicationsTooth-supportedreconstructionsThemostfrequenttechnicalcomplicationbytooth-supportedreconstructionswaslossofretention(fractureofthelutingcement).From1801FDPsanalyzed,thiscomplicationoccurredin121reconstruc-tions.Theannualcomplicationrateswere0.66forconventionalFDPs.However,forcantileverFDPsitwassigniÞcantly0.019)higher(1.75).Thistranslatesinto5-yearratesoflossofretentionof3.3%forconventionalFDPsand8.4%forcantileverFDPs(Table15).Twostudies(Karlsson1986),(Gustavsen&Silness1986)reportedhighincidenceoflossofretention.Thiscorrelatedtotheincreasedincidenceofcariesreportedinoneofthetwostudies(Karlsson1986).Thehighestannualcomplicationrate(5.7)wasreportedforthe12-unitFDPsinthemandiblesupportedbytwocaninesonly.Thesecondmostcommontechnicalcomplicationwasfractureofmaterials.Theseincludedfracturesoftheframework,theveneersorthecorebuild-ups.Fifty-nineoutof2287FDPsexaminedexperi-encedsomekindofmaterialfractures.Theannualfailurerateswere0.32forconven-tionaland0.61forcantileverFDPstrans-latinginto5-yearratesofmaterialfracturesof1.6%forconventionalFDPsand3%forcantileverFDPs(Table15).Whenceramicfractureorceramiccippingwereanalyzedseparately,the5-yearcomplicationratesforconventionalFDPsincreasedto2.9%,andforcantileverFDPstheincidencewentupto3.5%(Table15).FractureofabutmentteethwasreportedontheabutmentandtheFDPlevels.ForcantileverFDPstheannualfailurerateofabutmenttoothfracturewas0.30translat-inginto5-yearratesofabutmenttoothfracturesof2.9%(95%CI:1.7Ð5%).TheincidenceoffracturesofanabutmenttoothleadingtothelossofthewholeFDPswasreportedfor1411FDPs.Theannualfailurerateswere0.2forconventionaland0.24forcantileverFDPstranslatinginto5-yearratesofFDPslostduetoabutmenttoothfractureof1%forconventionalFDPsand1.2%forcantileverFDPs(Table15).Implant-supportedreconstructionsThemostcommontechnicalcomplicationbyimplant-supportedreconstructionswas Table16.Summaryofcomplicationsbyimplant-supportedreconstructionsComplicationImplant-supportedFDPsCombinedtooth-implant-supportedFDPsImplant-supportedsinglecrownsNumberofimplantsoreventrates(95%CI)(95%CI)(%)Numberofimplantsorrates(95%CI)rates(95%CI)Numberofimplantsoreventrates(95%CI)rates(95%CI)Estimatedrateofsofttissuecomplications7511.79(1.05Ð3.03)8.6(5.1Ð14.1%)1841.44(0.35Ð5.96)7(1.7Ð25.8%)2672.03(1.05Ð3.95)9.7Estimatedrateofboneloss2mmn.a.n.a.n.a.n.a.5091.31(0.61Ð2.79)6.3Estimatedrateofestheticcomplicationsn.a.n.a.n.a.n.a.4181.82(0.64Ð5.12)8.7Estimatedrateofabutmenttoothintrusionn.a.n.a.5061.07(0.4Ð2.87)5.2(2Ð13.3%)n.a.n.a.Estimatedrateofimplantfracture25590.11(0.05Ð0.23)0.5(0.3Ð1.1%)5300.2(0.08Ð0.34)0.8(0.4Ð1.7%)13120.03(0.006Ð0.13)0.14(0.03Ð0.64%)Estimatedrateofabutmentorscrewfracture25900.3(0.16Ð0.57)1.5(0.8Ð2.8%)5110.11(0.06Ð0.22)0.6(0.3Ð1.1%)5100.07(0.018Ð0.28)0.35Estimatedrateoflooseabutmentsorscrews24531.15(0.76Ð1.74)5.6(3.7Ð8.3%)2961.44(0.95Ð2.17)6.9(4.7Ð10.3%)7522.72(1.17Ð6.3)12.7Estimatedrateoflostaccessholerestorations1691.778.9n.a.n.a.n.a.n.a.Estimatedrateoflossofretention931.18(0.6Ð2.34)5.7(3Ð11%)2861.53(1.09Ð2.13)7.3(5.3Ð10.1%)3741.13(0.44Ð2.91)5.5Estimatedrateofveneerfracture9482.53(1.6Ð4.02)11.9(7.7Ð18.2%)1251.51(0.98Ð2.3)7.2(4.8Ð10.9%)5080.92(0.48Ð1.75)4.5Estimatedrateofceramiccippingorfracture5211.84(1.03Ð3.27)8.8(5Ð15.1%)1251.51(0.98Ð2.3)7.2(4.8Ð10.9%)4020.71(0.34Ð1.46)3.5Estimatedrateofframework6230.13(0.06Ð0.32)0.7(0.3Ð1.6%)1200.33(0.12Ð0.91)1.6(0.6Ð4.4%)3480.61(0.22Ð1.73)3BasedonstandardPoissonÕsregression.Basedonrandom-effectsPoissonÕsregression.n.a.,notavailable;FDP,Þxeddentalprosthesis.Pjeturssonetal.SurvivalofFDPsandsinglecrowns108|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard Theannualfailureratesfordifferentrecon-structionsrangedfrom0.48to3.54.Fromthissingleclinic,thehighestfailuresratesafter10yearswereseenforcantileverFDPsandcombinedtooth-implant-supportedFDPsandcorrespondedwelltothoseofthemeta-analysisofthedentalprostheticliterature.Analyzingtheproportionofpatientsfreeofallcomplications,theconventionaltooth-supportedFDPshadthelowestan-nualcomplicationsrate(3.41),followedbycantileverFDPs(4.62)andsolelyimplant-supportedFDPs(9.78).The5-yearsuproportionfortooth-supportedFDPswas84.3%signiÞcantly(0.008)higherthanfortheimplant-supportedFDPsForconventionaltooth-supportedFDPs,themostfrequentcomplicationswerebio-logicalcomplicationslikecariesandlossofpulpvitality.ForcantileverFDP,thein-cidenceofbiologicalcomplicationswassimilartothatofconventionalFDPs.Technicalcomplicationslikelossofreten-tionandmaterialfractureswasmorefrequent.Forimplant-supportedreconstructions,theincidenceofbiologicalcomplicationswassimilarforallthreegroups.However,theincidenceoftechnicalcomplicationswassigniÞcantlyhigherforimplant-sup-portedthanfortooth-supportedreconstruc-tions.Fractureoftheveneermaterial(ceramicfractureorchipping),abutmentorscrewlooseningandlossofretentionwerethemostfrequentlyencounteredtechnicalcomplications.Fractureoftheveneermaterialwasmorefrequentinstudiesreportingongold-acrylreconstruc-tions.ThehighfailurerateoftheFDPswithagoldframeworkandacrylicveneersisthoughttobeduetohighnumbersofveneerfracturesandestheticcomplica-tions.Forabutmentorscrewloosening,onestudy(Henryetal.1996),reportingontheÞrstgenerationofSCsonBrane-markimplants,wasaclearoutlier.Thisgroupsreportedthatwhenthetitaniumscrewswerereplacedwithnewgoldabut-mentscrewsandwhennewabutmentreplacedtheolderone,thisincidenceofthiscomplicationwasdramaticallyre-duced.Fractureofcomponentslikeframe-work,implants,abutmentorscrewswereararecomplicationwithonestudyoncrownsmadeonprefabricatedceramiccaps(Schelleretal.1998)reportinganunusuallyhighincidenceofcorefractures.ThissystematicreviewaddressingthesurvivalandcomplicationratesofFDPofdifferentdesignswasbasedonaseriesofÞvesystematicreviewsthatallusedthesamesearchstrategy,thesameinclusion/exclusioncriteriaandthesamestatisticalmethodology.Insteadofperformingaformalqualityassessmentoftheincludedstudiesandsensitivityanalysis(Junietal.1999),thisreviewusedstringentinclusioncriteria.Forexample,onlystudieswithaclinicalfollowupexaminationwereincludedtoavoidthepotentialinaccuraciesineventdescriptioninstudiesthatbasedtheiranalysisonsubjectivepatientself-reports.Onelimitationofthisreviewisthatitwasmainlybasedonstudiesthatwereconductedinaninstitutionalenvironment,suchasuniversityorspecializedimplantclinics.Therefore,thelong-termoutcomesobserved,cannotbegeneralizedtodentalservicesprovidedinprivatepractice.Afurtherlimitationisthatthepublishedinformationdidnotallowestimatingan-nualfailureratesseparatelyfordifferenttimeperiodsoryearsafterinsertionofthereconstruction.Thus,itwasnotpossibletoassesswhetherannualfailureratesin-creasedovertime.Moreover,twoofthethreeliteraturesearchesonlyincludedEnglish-languagepublications.Thiscouldbeproblematicfortworeasons:(a)theprecisionofsum-maryestimatesisreducedifasubstantialnumberofadditionalstudiespublishedinotherlanguagesexist;(b)biasmaybeintroducediftheresultsofstudiespub-lishedinEnglishdiffersystematicallyfromthosepublishedinotherlanguages.However,inthethirdsearch,inwhichGerman-andFrench-languageliteraturewasalsoincluded,itwasseenthatall26includedstudieswerepublishedinEnglish.ThisisinagreementwithanempiricalstudyfromEggeretal.(2003)thatfoundlittleeffectoftheinclusion/exclusionoftrialspublishedinlanguagesotherthanEnglishoncombinedeffectestimatesinmeta-analysesofRCTs.ResearchimplicationsItwasevidentfromthesearchoftheentiredentalliteratureofFDPthatthereisaneedforlongitudinalstudieswith10ormoreyearsofobservation.Thisisespeciallyevidentforimplant-supportedreconstruc-tionsandthereisadeÞnitiveneedformorelongitudinalstudiesaddressingsuchrecon-structions.ThepresentsystematicreviewrevealedseveralshortcomingsintheconductandreportingofclinicalstudiesofFDPresult-inginthefollowingrecommendations:Long-termcohortstudiesonreconstruc-tionsshouldhavecompletefollow-upin-formationpreferentiallywithsimilarlengthoffollow-upforallpatients.Thismeansthatdataonwell-deÞnedtimeper-iodsshouldbereportedfortheentireco-hort.OwingtovariousdeÞnitionsofsuccess,authorsshouldreportdataonsurvivalincombinationwithincidenceofcomplications.Survivalandsuccess(freeofallcomplications)ofthesuprastructuresshouldbereported.Well-deÞnedcriteriashouldbeusedfortheassessmentofthebiologicalandtechnicalcomplications.Datafromclinicalandradiographicassess-mentsshouldbedescribedusingfrequencydistributions.Collaborativeeffortstocon-ductapooledindividualpatientdataana-lysisofthepatientsandimplantsinthevariousstudieswouldallowdevelopmentandusecommondeÞnitionsofcomplica-tionsandtoobtainaclearerpictureofthelong-termsurvival.BiologicalcomplicationsdeÞnedby(1)thethresholdlevelofpocket-probingdepth(PPD),(2)thepresence/absenceofbleedingonprobing(BOP)/suppurationas-sessedatanyexaminationintervaland(3)crestalbonelossovertimeshouldbede-scribedforimplantsandneighboringteeth.Technicalcomplicationsshouldbedi-videdinto(1)major;suchas,implantfracture,lossofsuprastructures,(2)med-ium;suchas,abutmentorabutmentfrac-ture,veneerorframeworkfractures,estheticandphoneticcomplicationsand(3)minor;suchasabutmentandscrewloosening,lossofretention,lossofscrewholesealing,veneerchipping(tobepo-lished)andocclusaladjustments.Thetypeandnumberofeventsoftechnicalcomplicationspertimeintervalaswellastime/costrequiredshouldalsobereported.ClinicalimplicationsBasedontheresultsofthepresentsystematicreview,planningofprostheticPjeturssonetal.SurvivalofFDPsandsinglecrowns110|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard toothimplants.ClinicalOralImplantsResearch:38Ð43.¬mmerle,C.H.,Ungerer,M.C.,Fantoni,P.C.,Bra¬gger,U.,Burgin,W.&Lang,N.P.(2000)Long-termanalysisofbiologicandtechnicalas-pectsofÞxedpartialdentureswithcantilevers.InternationalJournalofProsthodontics409Ð415.Henry,P.J.,Laney,W.R.,Jemt,T.,Harris,D.,Krogh,P.H.,Polizzi,G.,Zarb,G.A.&Herrmann,I.(1996)Osseointegratedimplantsforsingle-toothreplacement:aprospective5-yearmulticenterstudy.InternationalJournalofOral&Maxillo-facialImplants:450Ð455.Hochman,N.,Ginio,I.&Ehrlich,J.(1987)ThecantileverÞxedpartialdenture:a10-yearfollow-up.JournalofProstheticDentistry542Ð545.Hochman,N.,Mitelman,L.,Hadani,P.E.&Zalk-ind,M.(2003)Aclinicalandradiographicevalua-tionofÞxedpartialdentures(FDPs)preparedbydentalschoolstudents:aretrospectivestudy.JournalofOralRehabilitation:165ÐHochman,N.,Yaffe,A.&Ehrlich,J.(1992)Splint-ing:aretrospective17-yearfollow-upstudy.Jour-nalofProstheticDentistry:600Ð602.Hosny,M.,Duyck,J.,vanSteenberghe,D.&Naert,I.(2000)Within-subjectcomparisonbetweencon-nectedandnonconnectedtooth-to-implantÞxedpartialprostheses:upto14-yearfollow-upstudy.InternationalJournalofProsthodontic340Ð346.Isidor,F.&Budtz-Jrgensen,E.(1990)Periodontalconditionsfollowingtreatmentwithdistallyex-tendingcantileverbridgesorremovablepartialdenturesinelderlypatients.A5-yearstudy.Jour-nalofPeriodontology:21Ð26.Jemt,T.,Bergendal,B.,Arvidson,K.,Bergendal,T.,Karlsson,L.D.,Linden,B.,Rundcrantz,T.&Wendelhag,I.(2002)Implant-supportedweldedtitaniumframeworksintheedentulousmaxilla:a5-yearprospectivemulticenterstudy.Interna-tionalJournalofProsthodontics:415Ð421.Jemt,T.&Lekholm,U.(2005)Singleimplantsandbuccalbonegraftsintheanteriormaxilla:mea-surementsofbuccalcrestalcontoursina6-yearprospectiveclinicalstudy.ClinicalImplantDen-tistry&RelatedResearch:127Ð35.Jemt,T.,Lekholm,U.&Adell,R.(1989)Osseoin-tegratedimplantsinthetreatmentofpartiallyedentulouspatients:apreliminarystudyon876consecutivelyplacedÞxtures.InternationalJour-nalofOral&MaxillofacialImplants:211Ð217.Jokstad,A.&Mjor,I.A.(1996)TenyearsÕclinicalevaluationofthreelutingcements.JournalofDentistry:309Ð315.Karlsson,S.(1986)AclinicalevaluationofÞxedbridges,10yearsfollowinginsertion.JournalofOralRehabilitation:423Ð432.Karlsson,S.(1989)FailuresandlengthofserviceinÞxedprosthodonticsafterlong-termfunction.Alongitudinalclinicalstudy.SwedishDentalJour-:185Ð192.Kindberg,H.,Gunne,J.&Kronstrom,M.(2001)Tooth-andimplant-supportedprostheses:aretro-spectiveclinicalfollow-upupto8years.Interna-tionalJournalofProsthodontic:575Ð581.Koth,D.L.,McKinney,R.V.,Steßik,D.E.&Davis,Q.B.(1988)Clinicalandstatisticalanalysiesofhumanclinicaltrialswiththesinglecrystalalu-miniumoxideendostealdentalimplant:Þve-yearresults.TheJournalofProsthetikDentistryLaurell,L.,Lundgren,D.,Falk,H.&Hugoson,A.(1991)Long-termprognosisofextensivepolyunitcantileveredÞxedpartialdentures.JournalofProstheticDentistry:545Ð552.Lekholm,U.,Gunne,J.,Henry,P.,Higuchi,K.,Linden,U.,Bergstrom,C.&vanSteenberghe,D.(1999)SurvivaloftheBranemarkimplantinpartiallyedentulousjaws:a10-yearprospectivemulticenterstudy.InternationalJournalofOral&MaxilofacialImplants:639Ð645.Lekholm,U.,vanSteenberghe,D.,Herrmann,I.,Bolender,C.,Folmer,T.,Gunne,J.,Henry,P.,Higuchi,K.,Laney,W.R.&Linden,U.(1994)Osseointegratedimplantsinthetreatmentofpartiallyedentulousjaws:aprosopective5-yearmulticenterstudy.InternationalJournalofOralandMaxillofacialImplants:627Ð635.Levin,L.,Pathael,S.,Dolev,E.&Schwartz-Arad,D.(2005)Aestheticvs.surgicalsuccessofsingledentalimplants:1-to9-yearfollow-up.PracticalProceduresandAestheticDentistry:533Ð538.Libby,G.,Arcuri,M.R.,LaVelle,W.E.&Hebl,L.(1979)LongevityofÞxedpartialdentures.JournalofProstheticDentistry:127Ð131.Mengel,R.,Schroder,T.&Flores-de-Jacoby,L.(2001)Osseointergratedimplantsinpatientstrea-tedforgeneralizedchronicperiodontitisandgeneralizedaggressiveperiodontitis:3-and5-yearresultsofaprospectivelong-termstudy.JournalofPeriodontology:977Ð989.Mericske-Stern,R.,Grutter,L.,Rosch,R.&Mer-icske,E.(2001)Clinicalevaluationandprostheticcomplicationsofsingletoothreplacementsbynon-submergedimplants.ClinicalOralImplantsResearch:309Ð318.Naert,I.,Koutsikakis,G.,Duyck,J.,Quirynen,M.,Jacobs,R.&vanSteenberghe,D.(2002a)Biologicoutcomeofimplant-supportedrestorationsinthetreatmentofpartialedentulism.PartI:along-itudinalclinicalevaluation.ClinicalOralIm-plantResearch:381Ð389.Naert,I.,Koutsikakis,G.,Quirynen,M.,Duyck,J.,vanSteenberghe,D.&Jacobs,R.(2002b)Biologicoutcomeofimplant-supportedrestorationsinthetreatmentofpartialedentulism.Part2:along-itudinalradiographicevaluation.ClinicalOralImplantResearch:390Ð395.Naert,I.E.,Duyck,J.A.,Hosny,M.M.F.&vanSteenberghe,D.(2001)Freestandingandtooth-implantconnectedprosthesesinthetreatmentofpartiallyedentulouspatientsPartI:anupto15yearsclinicalevaluation.ClinicalOralImplantResearch:237Ð244.Napankangas,R.,Salonen,M.A.&Raustia,A.M.(1997)A10-yearfollow-upstudyofÞxedmetalceramicprosthodontics.JournalofOralRehabi-:713Ð717.Napankangas,R.,Salonen-Kemppi,M.A.&Raus-tia,A.M.(2002)LongevityofÞxedmetalceramicbridgeprostheses:aclinicalfollow-upstudy.Jour-nalofOralRehabilitation:140Ð145.Nickenig,H.-J.,Scha¬fer,C.&Spiekermann,H.(2006)Suvivalandcomplicationratesofcom-binedtooth-implant-supportedÞxedpartialden-tures.ClinicalOralImplantsResearch:506.Nyman,S.&Lindhe,J.(1979)Alongitudinalstudyofcombinedperiodontalandprosthetictreatmentofpatientswithadvancedperiodontaldisease.JournalofPeriodontology:163Ð169.dman,P.A.&Karlsson,S.(1988)Follow-upstudyofpatientswithbridgeconstructionsperformedbyprivatedentalsurgeonsandatauniversityclinic,8yearsfollowinginsertion.JournalofOralRe-habilitation:55Ð63.Olsson,M.,Gunne,J.,Astrand,P.&Borg,K.(1995)Bridgessupportedbyfree-standingimplantsvs.bridgessupportedbytoothandimplant.AÞve-yearprospectivestudy.ClinicalOralImplantResearch:114Ð121.rtorp,A.&Jemt,T.(1999)Clinicalexperiencesofimplant-supportedprostheseswithlaser-weldedtitaniumframeworksinthepartiallyedentulousjaw:a5-yearfollow-upstudy.ClinicalImplantDentistry&RelatedResearch:84Ð91.wall,B.E.,Almfeldt,I.&Helbo,M.(1991)Twenty-yearexperiencewith12-unitÞxedpartialdenturessupportedbytwoabutments.Interna-tionalJournalofProsthodontics:24Ð29.Palmer,R.M.,Palmer,P.J.&Smith,B.J.(2000)A5-yearprospectivestudyofAstrasingletoothim-plants.ClinicalOralImplantsResearch179Ð182.Palmqvist,S.&Swartz,B.(1993)ArtiÞcialcrownsandÞxedpartialdentures18to23yearsafterplacement.InternationalJournalofProsthodon-:279Ð285.Petersson,K.,Pamenius,M.,Eliasson,A.,Narby,B.,Holender,F.,Palmquist,S.&Hakansson,J.(2006)20-yearfollow-upofpatientsreceivinghigh-costdentalcarewithintheSwedishDentalinsurancesystem:1977Ð1978to1998Ð2000.SwedishDentalJournal:77Ð86.Polizzi,G.,Fabbro,S.,Furri,M.,Herrmann,I.&Squarzoni,S.(1999)Clinicalapplicationofnar-rowBranemarkSystemimplantsforsingle-toothrestorations.InternationalJournalofOral&MaxillofacialImplants:496Ð503.Preiskel,H.W.&Tsolka,P.(2004)Cement-andscrew-retainedimplant-supportedprostheses:upto10yearsoffollow-upofanewdesign.Interna-tionalJournalofOral&MaxillofacialImplants:87Ð91.Reichen-Graden,S.&Lang,N.P.(1989)Periodontalandpulpalconditionsofabutmentteeth.StatusafterfourtoeightyearsfollowingtheincorporationofÞxedreconstructions.SchweizerMonatsschriftfu¨rZahnmedizin1381Ð1385.Roberts,D.H.(1970a)Thefailureofretainersinbridgeprostheses.Ananalysisof2000retainers.BritishDentalJournal:117Ð124.Roberts,D.H.(1970b)Therelationshipbetweenageandthefailurerateofbridgeprostheses.BritishDentalJournal:175Ð177.Romeo,E.,Lops,D.,Margutti,E.,GhisolÞ,M.,Chiapasco,M.&Vogel,G.(2004)Long-termsurvivalandsuccessoforalimplantsinthetreat-mentoffullandpartialarches:a7-yearprospec-tivestudywiththeITIdentalimplantsystem.Pjeturssonetal.SurvivalofFDPsandsinglecrowns112|Clin.OralImpl.Res.(Suppl.3),2007/97Ð1132007TheAuthors.Journalcompilation2007BlackwellMunksgaard