primarynetlife
netlife 时间:2021-03-17 阅读:()
RESEARCHCosteffectivenessofpneumococcalvaccinationamongDutchinfants:aneconomicanalysisofthesevenvalentpneumococcalconjugatedvaccineandforecastforthe10valentand13valentvaccinesMarkHRozenbaum,healtheconomistandmodellerofinfectiousdiseases,1ElisabethAMSanders,professorinpaediatricimmunologyandinfectiousdiseases,2AlbertJanvanHoek,healtheconomistandmodellerofinfectiousdiseases,1AngeliqueGSCJansen,researchfellow,2,3ArievanderEnde,associateprofessormedicalmicrobiology,4GermievandenDobbelsteen,seniorresearchscientist,5GerwinDRodenburg,researchfellow,2EelkoHak,associateprofessorinclinicalepidemiologyofinfectiousdiseases,1,2,3,6MaartenJPostma,professorinpharmacoeconomics1,6ABSTRACTObjectivesToupdatecosteffectivenessestimatesforthefourdose(3+1)scheduleofthesevenvalentpneumococcalconjugatedvaccine(PCV-7)intheNetherlandsandtoexploretheimpactoncosteffectivenessofreduceddoseschedulesandimplementationof10valentand13valentpneumococcalvaccines(PCV-10andPCV-13).
DesignEconomicevaluationcomparingPCV-7,PCV-10,andPCV-13withnovaccinationusingadecisiontreeanalyticmodelbuiltfromdatainpreviousstudies.
SettingTheNetherlands.
PopulationAcohortof180000newbornsfolloweduntil5yearsofage.
MainoutcomemeasuresCosts;gainsinlifeyearsandqualityadjustedlifeyears(QALYs);andincrementalcosteffectivenessratios.
ResultsUnderbasecaseassumptions—thatis,assumingafiveyearprotectiveperiodofthevaccineandnoassumednetindirecteffects(herdprotectionminusserotypereplacement)amongchildrenagedover5years—vaccinationwithPVC-7inafourdose(3+1)schedulewasestimatedtoprevent71and5778casesofinvasiveandnon-invasivepneumococcaldisease,respectively,inchildrenagedupto5years.
Thiscorrespondswithatotalnetgainof173lifeyearsor277QALYs.
TheincrementalcosteffectivenessratioofPCV-7wasestimatedat113891(98300;$145000)perQALY,wellovertheratioof50000perQALYrequiredforPCV-7toberegardedaspotentiallycosteffective.
Athreedose(2+1)scheduleofPCV-7reducedtheincrementalcosteffectivenessratioto82975perQALY.
Forvariousassumptionsandincluding10%ofthemaximumnetindirecteffectsamongindividualsaged5yearsandover,PCV-10andPCV-13hadincrementalcosteffectivenessratiosrangingfrom31250to52947perQALY.
ConclusionsThecurrentDutchinfantvaccinationprogrammeoffourdosesofPCV-7isnotcosteffectivebecauseofincreasesininvasivediseasecausedbynon-vaccineserotypes,whichreducestheoveralldirecteffectsofvaccinationandoffsetspotentialpositiveherdprotectionbenefitsinunvaccinatedindividuals.
The10valentand13valentpneumococcalvaccinescouldhavebetternethealthbenefitsthanPCV-7throughlessreplacementdiseaseandincreasedherdprotection.
Boththeseeffectscouldsubstantiallyreducetheincrementalcosteffectivenessratiotopossiblyacceptablelevels,iftotalprogrammecostscanbeloweredbyreducedschedules,reductionsinvaccineprices,orboth.
INTRODUCTIONGiventhemultitudeofnewvaccinesavailableforintroductionintonationalimmunisationprogrammes,healtheconomicmodellingofvariousimmunisationplansisbecomingincreasinglyimportantininformingdecisionsonhealthpolicy.
Thedecisiontointroducethesevenvalentpneumococcalconjugatedvaccine(PCV-7)intotheDutchnationalimmunisationpro-grammeforinfants,forexample,hasinpartbeendri-venbycosteffectivenessconsiderations.
1TheDutchHealthCouncilestimatedtheincrementalcosteffec-tivenessratioofvaccinationwithPCV-7comparedwithnovaccinationat70000(60300;$89200)andlessthan20000perqualityadjustedlifeyear(QALY)in2001and2005,respectively.
1Crucialfactorsresponsibleforthechangefromapotentiallyunfavour-ablecosteffectivenessratioin2001,exceeding50000perQALY,toafavourableratioin2005weretheinclu-sionofdataonobservedherdprotectioneffectsinadultsafternationwideimplementationofPCV-7intheUSAin2000andlimiteddiseasedevelopmentcausedbypneumococcalserotypesnotpresentinthe1UnitofPharmacoEpidemiologyandPharmacoEconomics,DepartmentofPharmacy,UniversityofGroningen,Groningen,Netherlands2WilhelminaChildren'sHospital,UniversityMedicalCenterUtrecht,Utrecht,Netherlands3JuliusCenterforHealthSciencesandPrimaryCare,UniversityofUtrecht,Utrecht,Netherlands4CenterforInfectionandImmunityAmsterdam,DepartmentofMedicalMicrobiologyandtheNetherlandsReferenceLaboratoryforBacterialMeningitis,AcademicMedicalCenterAmsterdam,Amsterdam,Netherlands5NetherlandsVaccineInstitute,Bilthoven,Netherlands6DepartmentofEpidemiology,UniversityMedicalCenterGroningen,UniversityofGroningen,Groningen,NetherlandsCorrespondenceto:MHRozenbaumm.
h.
rozenbaum@rug.
nlCitethisas:BMJ2010;340:c2509doi:10.
1136/bmj.
c2509BMJ|ONLINEFIRST|bmj.
compage1of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-7replacingpneumococcalserotypeseliminatedbythevaccine(replacementdisease).
2-4Nexttodirecteffectsoninvasivediseaseinvacci-nees,expectedsavingsfromherdprotectionwerealsopartofhealtheconomicstudiesinotherEuropeancountriesthatintroducedPCV-7intotheirnationalimmunisationprogrammes.
5-10Boththefourdose(3+1)vaccinescheduleandthereducedthreedose(2+1)schedule,asimplementedinNorwayandtheUK,1112arehighlyeffectiveagainstinvasivepneumococcaldiseasecausedbyvaccineserotypes.
However,thenetoverallbenefitofnationalimmu-nisationprogrammesinmanyEuropeancountrieshasbeenreducedbyincreasesininvasivediseasecausedbynon-vaccineserotypes.
12-15Importantly,inthefirst18-30monthsaftertheintroductionofPCV-7intheNetherlands,France,andtheUK,nooverallreductionininvasivediseaseinnon-vaccineeswasobserved.
121315Giventhatbothincreasesininvasivediseasecausedbynon-vaccineserotypesandabsenceofherdprotec-tionmayconsiderablyaffectthecosteffectivenessofthecurrentDutchvaccinationprogramme,wesetouttoupdatecosteffectivenessestimatesforthecurrentfourdosescheduleofPCV-7byusingrecentdataonepidemiologyandresourceuse.
Also,weinvestigatethecosteffectivenessofreduceddoseschedulesandvaccinepricereductionscombinedwiththeimplementationof10valentand13valentpneumo-coccalvaccines(PCV-10andPCV-13).
METHODSModelWedesignedadecisiontreeanalyticmodelstructurethatbuildsonourpreviouslyreportedmodel.
616Var-iousdatasourceswereusedtopopulateourmodel;theseincludedclinicaltrialsandobservationalstudiesforeffectivenessofpneumococcalvaccines,laboratorydataforincidenceandserotypedistributionsofpneu-mococcaldisease,andregistrationsforresourceuseandcosts.
Figure1showsthediseasemodelforthehealtheffectsofpneumococcalvaccination,includingthepos-sibilityofsubsequentpneumococcaldiseasesuchasnon-invasivepneumonia,otitismedia,andinvasivepneumococcaldisease.
Assumptionsregardingbothcostsandqualityoflifearesummarisedintable1andaremorethoroughlydiscussedinwebextra1.
Intheanalyses,acohortof180000newborns,repre-sentingtheDutchbirthcohort,wasrunthroughthedecisiontreetwice(basecaseanalysis):onceasamainlyvaccinatedcohort(PCV-7/PCV-10/PCV-13);andonceasanunvaccinatedcohort.
Theanalytictimeframeofthestudywasfiveyearsbecausevaccineeffectivenesscouldnotbeassumedbeyondfiveyears.
However,longtermeffectsofinvasivepneumococcaldiseasewereextrapolatedoverthefulllifetimeoftheindivi-dualsinthecohort(thatis,untildeathor100years).
SimpleComplexTympanostomyAcuteotitismedia(4)Pneumonia(9)VaccinationofallhealthyinfantsPolicyonpneumococcalconjugatedvaccineNovaccination[+]Meningitis(1,15)NotympanostomyNoinfectionCochlearimplant(12)Nocochlearimplant(13)TreatedbyaGP(3)Death(8)RecoveryHearingproblemsDeath(6)RecoveryDeath(7)RecoveryDisability(11)Admittedtohospital(2)Bacteraemiawithfocus(1,14)Non-invasivepneumococcaldiseaseInvasivepneumococcaldiseaseBacteraemiawithoutfocus(1,14)Invasivepneumonia(1,14)Death(5)Specialeducation(10)RecoverywithoutsequelaeFig1|Decisiontreeusedinconjunctionwiththecohortof180000newborns.
Numbersbetweenbracketscorrespondtodatashownintable1.
Theboxesrepresentdecisionnodes,withgreencolourindicatingprobabilisticstatesandredcolourindicatingendstates.
The"Novaccination"armisacloneofthe"Vaccinationofallhealthyinfants"arm(asrepresentedbythe+sign;risksdifferbetweenbotharms)RESEARCHpage2of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromTable1|ParametersusedintheeconomicmodelMeanorrangeDistributionCorrespondingbranchesinfig1ReferencesCase-fatalityrate(birthcohort)Meningitis9%Beta(3,32)517Pneumonia0%N/AN/A17Bacteraemiawithfocus0%N/AN/A17Bacteraemiawithoutfocus9%Beta(2,21)617Mortality(non-invasivepneumoniaandacuteotitismedia)0%N/AN/AAssumedCase-fatalityrate(agefiveyearsorolder)Meningitis9-92%Beta(agedependent)517Pneumonia0-29%Beta(agedependent)717Bacteraemiawithfocus0-33%Beta(agedependent)817Bacteraemiawithoutfocus9-67%Beta(agedependent)617Respiratoryinfections0%N/AN/AAssumedVaccineefficacyInvasivepneumococcaldisease(allvaccineserotypes)97.
4%Lognormal(SE0.
044)119Non-invasivepneumonia(admittedtohospital)11.
1%Lognormal(SE0.
082)223Non-invasivepneumonia(seenbygeneralpractitioner)6.
0%Lognormal(SE0.
032)323Acuteotitismedia7.
0%Lognormal(SE0.
011)422Directcosts()Costofhospitaladmission*Invasivepneumococcaldisease(agedependent)1091-27318Triangular(agedependent)117,39Non-invasivepneumonia26-2614Triangular(severitydependent)916,39,40Acuteotitismedia17-381Triangular(severitydependent)416,39,40Specialeducation(annualcosts)9798-16962Triangular(agedependent)1016Institutionalcare(annualcosts)39583Triangular(29,687;39,583;49,478)1139Cochlearimplantation56633Triangular(0;0.
004;0.
01)1241Indirectcosts()Invasivepneumococcaldisease0-974Triangular(severitydependent)117,39Non-invasivepneumonia(admittedtohospital)0-2529Triangular(severitydependent)117,39Non-invasivepneumonia(seenbygeneralpractitioner)115-315Triangular(severitydependent)916,39Acuteotitismedia58-23Triangular(severitydependent)416,39Totaldropinqualityoflife(QALYs)Disability§0.
53Beta(estimated)1142Bilateralhearingloss(firstyear)§0.
45Beta(estimated)128,43Bilateralhearingloss:cochleardevice§0.
18Beta(estimated)128,43Allotherhearingloss§0.
09Beta(estimated)1342Hospitaladmissionforbacteraemia**0.
0079Beta(estimated)148,44Hospitaladmissionformeningitis0.
0232Beta(estimated)158,44Hospitaladmissionfornon-invasivepneumonia0.
006Triangular(0.
001,0.
006,0.
01)28Non-invasivepneumoniatreatedbyageneralpractitioner0.
004Triangular(0,0.
004,0.
01)38Acuteotitismedia0.
005Triangular(0,0.
005,0.
01)48OtherparametersIncreaseinnon-vaccineserotypeofinvasivepneumococcaldisease100%Triangular(50%,100%,150%)N/A12,13NetindirecteffectforPCV-10andPCV-1310%Triangular(0%,10%,30%)N/AAssumedDiscountratehealtheffects1.
5%N/AN/A27Discountratecosts4%N/AN/A27*Basedontheaveragedurationofhospitalstay(bothintensivecareandgeneralward)andcorrespondingunitcosts.
39SeealsowebtableBforagespecificcostsofhospitaladmission.
Indirectcostscausedbyabsenceatworkofparentstakingcareoftheirchildren.
Indirectcostscausedbyabsenceatworkofpatientduetohospitaladmission.
§Peryear.
Percase.
**SameQALYdecrementwasassumedforinvasivepneumonia,bacteraemiawithfocus,andbacteraemiawithoutfocus.
Seealsowebextra2.
Indirecteffectsintheanalysedbirthcohort.
Seealsowebextra3.
Indirecteffectsforthoseaged5yearsandolder.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine.
RESEARCHBMJ|ONLINEFIRST|bmj.
compage3of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromBaselinediseaserisksSurveillancedataontheincidenceandserotypedistri-butionofinvasivepneumococcaldiseasebeforenationalimplementationofPCV-7wereavailablefortheperiod2004-2006,includingdataonage,primaryfocusofinfection,resourceuse,hospitaladmission,andoutcome.
1317Thecase-fatalityrateformeningitisandbacteraemiawithoutfocusinchildrenwasesti-matedtobe9%(table1),17whichisinlinewiththeinternationalliterature.
5818Invasivepneumoniaandbacteraemiawithfocuswereassumednottoresultindeathinchildren.
17Inourmodel,severementalandphysicalhandicapresultingfrommeningitiswasassumedtooccurin13%ofcasesofpneumococcalmeningitisinchildren,ofwhich50%wouldrequirespecialeducationand25%intensive"roundtheclock"institutionalcare.
6Jansenetalfoundthathear-ingproblemsoccurredin32%ofcasesofmeningitis,ofwhich50%wereseriousenoughtorequireacochlearhearingdevice.
17Baselinerisksfornon-invasivepneu-moniarequiringhospitaladmissionandfornon-inva-sivepneumoniaandacuteotitismediatreatedingeneralpractitionersurgerieswereestimatedfromnationalhospitalandgeneralpractitionerrecords,respectively(seewebtableA).
VaccineefficaciesVaccineefficacyagainstinvasivepneumococcaldiseasewasassumedat97.
4%aftertwodosesforallsevenser-otypesofpneumococcaldiseasecoveredbyPCV-7(table1).
19Thisvalueseemstobeaconservativeesti-mateifonetakesintoaccountthefactthatonlyonevac-cinefailurehasbeenreportedintheNetherlandsinthefirsttwoyearsafterintroductionofroutineinfantvacci-nationinJune2006.
Routinevaccinationforinfantsina2+1doseschedulewasintroducedinNorwayin2006,andsimilarlynovaccinefailureshadoccurreduptoJune2008.
11Protectionagainstinvasivediseasewasthusesti-matedtolastforfiveyearsinthebasecaseanalysis.
20Furthermore,inrandomisedcontrolledsettings,thevac-cinewasshowntobeeffectiveagainstnon-invasivepneumoniaandotitismediainchildren.
21-23Fornon-invasivepneumonia,efficacyofpneumococcalvaccina-tionseemstoincreasewithdiagnosticcertainty.
23Inourmodel,weappliedtheefficacyestimateof11.
1%for"clinicalpneumoniaandperihilarfindings"tochildrenadmittedtohospitalwiththediagnosisofpneumoniaintheNetherlands.
23ThisdefinitionofpneumoniaseemstobestfitthetypesofpneumoniastreatedinDutchhospitals.
Anefficacyof6.
0%wasassumedforpatientswhovisitedageneralpractitionerandwerediagnosedwithpneumonia.
23Intworando-misedstudies,PCV-7wasfoundtoprevent6.
4%to7.
0%ofallcasesofacuteotitismedia.
221924Theinter-pretationofthesestudiesfortheDutchsettingisham-peredbyseveralfactors,includingthefactthatthecausalmicro-organismisnotrecordedincasesofotitismediaintheNetherlands.
Inourmodel,weusedanoverallefficacyestimateof7.
0%forotitismediaonthebasisofthemostrecentdatafromtheKaiserPer-manentetrial.
22Givenevidenceforthedurationofpro-tectionagainstnon-invasivepneumoniaandrecentUSsurveillancedata,weassumedthatvaccinatedchildrenwereprotectedagainstnon-invasivepneumoniaandotitismediauptotheirsecondyearoflife,startingaftertheseconddoseofthevaccine.
212526Avaccineefficacyof97.
4%againstallserotypesincludedwasassumedforPCV-10andPCV-13,simi-lartotheassumedefficacyofPCV-7.
IntheabsenceofclinicaldataontheefficacyofPCV-10andPCV-13againstnon-invasivepneumoniaandacuteotitismedia,theefficacyofthesetwovaccineswasassumedtoincreaseproportionallywiththeincreaseinserotypecoverageforinvasivepneumococcaldisease.
IndirecteffectsAswellasestimationsofthedirecteffects,wealsoesti-matedindirecteffectsofvaccinationinourmodel.
Weincludedinourbasecaseanalysisherdprotectionagainstinvasivepneumococcaldiseaseforchildreninthebirthcohortnotyetfullyprotectedbythevaccineandfornon-vaccinatedchildren,assumingthisprotec-tionwouldbeaseffectiveasvaccination(table2).
1213Wealsoincreasedtheincidenceofinvasivepneumo-coccaldiseasecausedbynon-vaccineserotypestoTable2|Basecaseserotypecoverageandefficacyfordirecteffectsandassumptionsonindirecteffectsfortheanalysedbirthcohortandtheremainingpopulation(thoseaged5yearsorolder)forPCV-7,PCV-10,andPCV-13PCV-7PCV-10PCV-13Serotypescovered4,6B,9V,14,18C,19F,23F4,6B,9V,14,18C,19F,23F,1,5,7F4,6B,9V,14,18C,19F,23F,3,6A,19A,1,5,7FIncreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypesintheanalysedbirthcohort(serotypereplacement)100%100%100%Efficacyandlevelofherdprotectionagainstvaccineserotypesofinvasivepneumococcaldiseaseintheanalysedbirthcohort*97.
4%97.
4%97.
4%Netindirecteffectintheremainingpopulation0%10%10%*Herdprotectionwasassumedfortheentirebirthcohort,includingthosenotyet(fully)protectedbythevaccine(eitherinfantstooyoungtobevaccinatedorthosewhoreceivedonlyasingledoseofthevaccine)andnon-vaccinatedchildren(5%ofabirthcohortfortheNetherlands),assumingaprotectioneffectof97.
4%againstvaccineserotypes,similartotheefficacyofthevaccine.
Netindirectbenefitsaredefinedasthebenefitsresultingfromprotectionagainstinvasivepneumococcaldiseasecausedbyvaccineserotypesminustheincreaseofinvasivepneumococcaldiseasecausedbynon-vaccineserotypes.
Thepotentialmaximumwasdefinedasfullreductionininvasivepneumococcaldiseasecasescausedbyvaccineserotypesintheabsenceofanyreplacementdisease.
Lowerpercentagescanbedefinedasacombinationofadecreaseinvaccineserotypeinvasivepneumococcaldiseaseandanincreaseindiseasefrominnon-vaccineserotypes.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine.
RESEARCHpage4of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom100%fortheanalysedbirthcohort(thatis,wedoubledtheincidenceofinvasivepneumococcaldiseasecausedbynon-vaccineserotypes)onthebasisofsurveillancedatafromearlyafternationalintroductionofPCV-7intheNetherlandsandtheUK.
1213Seewebextra2foramoreindepthdescriptionoftheassumptionsforourestimationofindirecteffectsinthebirthcohort.
Noserotypeinformationforacuteotitismediaandnon-invasivepneumoniaisavailableintheNether-lands,andserotypereplacementforthesediseasesmaybeassumedtobealreadyincludedinthevaccineefficacyestimatesinthefirstefficacystudies.
21-23There-fore,wedidnotincludeanadditionalincreaseofnon-vaccineserotypediseasebutalsoleftoutpotentialherdeffectsforotitismediaandnon-invasivepneumonia(seewebextra2).
WeassumedinourbasecaseanalysisforPCV-7thatnonetindirecteffectwouldexistforindividualsout-sidethemodelledcohort.
ThisassumptionwasmadebecausenoreductionintheincidenceofinvasivepneumococcaldiseasehasbeenobservedaftertheintroductionofroutinevaccinationwithPCV-7forindividuals5yearsofageorolderandbecausetheobservedherdprotectioneffectintheUKinthethirdyearafterintroducingroutinevaccinationwascompletelycounteredbyariseininvasivepneumococ-caldiseasecausedbynon-vaccineserotypes.
12Inthisrespect,netindirecteffectsaredefinedascasesofinva-sivediseaseavertedbyherdprotectionminusinvasivecasesofreplacementdisease.
Netindirecteffectsmayoccurinthefuture,espe-ciallyifserotypecoverageisextendedbyachangefromsevenserotypevaccinestovaccineswithbroaderserotypecoverage.
1317Therefore,inthebasecaseana-lysisforPCV-10andPCV-13,anetindirecteffectforinvasivediseaseat10%ofthepotentialmaximumwasappliedforthoseaged5yearsorolder(seewebextra3).
Inparticular,thepotentialmaximumwasdefinedaspreventionofallcasesofinvasivediseasecausedbyserotypesinthevaccineandabsenceofanyreplace-mentdisease.
Netprotectiveindirectbenefitsagainstotitismediaandnon-invasivepneumoniawerenotincludedinanyoftheanalyses.
25Giventhatthereismuchuncertaintyaboutthedevel-opmentofindirecteffects,theseassumptionswerevar-iedoverawiderangeinthesensitivityanalyses.
OutcomemeasuresandcosteffectivenessanalysisThesimulationmodeltracksallthespecificdiseasecasesandthedeaths,costs,changesinQALYsandlifeyears,andindirecteffects(herdprotectionandser-otypereplacement).
WewereabletodeterminethenetcostsandnetlifeyearsandQALYsgainedbysummingallthecosts,lifeyears,andQALYsandcalculatingthedifferencesfortheevaluationswithandwithoutvacci-nation.
TheincrementalcosteffectivenessratiowascalculatedbydividingthenetcostsbyeitherlifeyearsorQALYs.
Healtheffectsandcostweredis-countedat1.
5%and4%fortimepreference,respec-tively,accordingtotheDutchguidelinesforcosteffectivenessresearch.
27Incrementalcosteffectivenessratiosforroutinevac-cinationwerecalculatedbycomparingdifferentvacci-nationschedulesagainstnovaccination.
FollowingrecentlypublishedevidenceontheefficacyofPCV-7inreduceddoseschedules,2829weinvestigatedtheeffectofathreedoseschedule(thatis,2+1)totesttheeffectoflowertotalvaccinationcosts(seewebextra4).
WealsoforecastedtheincrementalcosteffectivenessofpotentialshiftsfromPCV-7topneumococcalvac-cinesthatincludeadditionalserotypes(thatis,PCV-10andPCV-13).
ForPCV-7,theestimatedcurrentcostof50perdosewithintheDutchnationalimmunisationpro-grammewasused.
616ForPCV-13,theofficiallylistedpriceof68.
56wasapplied,withadministrationcostsof5.
95beingadded(totalcostperdose74.
51).
16ForPCV-10,noofficiallylistedpriceisavailableintheNetherlands.
GiventhatweknowthepricingofPCV-10inothercountriesisconservativecomparedwithPCV-13,weassumedthetotalcostperdoseofPCV-10atthemidpointbetweenPCV-7andPCV-13(thatis,62.
25).
30Table3|BasecaseanalysisresultsfortheanalysedDutchbirthcohortAcuteotitismediaNon-invasivepneumoniaInvasivepneumococ-caldiseaseInvasivepneumococcaldiseaserelatedtonetindirecteffectsforindividualsaged5yearsorolder*TotalCases(undiscounted)Novaccination170788193851882410NAPCV-716541618979117210NAPCV-1016466418922802260NAPCV-1316391218865382229NACasesavertedPCV-75372406710NAPCV-106124463108150NAPCV-136876520150181NAQALYsgained(years)PCV-72722480277PCV-10302361314707PCV-13342470384891Lifeyearsgained(years)PCV-7001730173PCV-1000255312566PCV-1300336381717Directsavings(1000s),excludingvaccinationcostsPCV-7126375172502226PCV-10144427245413984422PCV-13161479318116965518Indirectsavings(1000s;directeffects)relatedtoproductionlossesPCV-732074460440PCV-103658467161677PCV-134109493202799*Onlynetindirecteffectsagainstinvasivepneumococcaldiseasewereincludedinthemodelforindividualsaged5yearsorolder.
ForPCV-7,nonetindirecteffectswereincludedintothemodelforindividualsaged5yearsorolderinthebasecaseanalysis.
NA,notapplicable;PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine;QALY,qualityadjustedlifeyears.
RESEARCHBMJ|ONLINEFIRST|bmj.
compage5of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromScenarioandsensitivityanalysesWeperformedunivariate,threshold,scenario,andprobabilisticsensitivityanalyses.
Intheunivariatesen-sitivityanalyses,allrelevantparameterswerevariedby25%toexploretheimpactofeachparameterrelativetoeachother.
Onespecificthresholdanalysiswasper-formedinwhichtheeffectoftheparameterontheincrementalcosteffectivenessratiowasinvestigatedbyvaryingthenetindirecteffectsonindividualsaged5yearsorolderoverarangeof0%to30%.
Fortheprobabilisticsensitivityanalyses,parametersweregen-eratedusingMonteCarlosampling,withoutcomevaluesgeneratedbyrunningthemodel5000times.
Lognormal,beta,andtriangulardistributionswereusedexceptformultinomialprobabilities,whereDirichletdistributionswereassumed(seetable1forspecificdistributions).
RESULTSCosteffectivenessofPCV-7Inthebasecaseanalysis,theestimatedburdenofpneu-mococcalinfectionforabirthcohortfollowedforfiveyearswas170788casesofacuteotitismediaand19385casesofnon-invasivepneumonia,ofwhich2645caseswouldresultinhospitaladmission(table3).
Applyingthebasecaseassumptions,5372(31%)casesofacuteotitismediaand406(21%)casesofnon-invasivepneumoniawouldbepreventedbyvaccina-tionwithPCV-7,correspondingtogainsof27and2QALYs,respectively.
Additionally,188casesofinvasivepneumococcaldiseaseayearwereestimatedinchildrenunder5yearsofage:65casesofmeningitis;45casesofinva-sivepneumococcaldisease;38casesofbacteraemiawithfocus;and40casesofbacteraemiawithoutfocus.
Intotal,71(38%)casesofinvasivediseasewouldbepreventedbyvaccinationwithPCV-7,corre-spondingtoatotalgainof173lifeyearsor248QALYs.
Inadditiontothehealthgains,vaccinationwithPCV-7wouldalsopreventapproximately2.
2millionofdirectcostsand0.
4millionofindirectcosts.
Assumingafourdoseschedule,theannualcostofvac-cinationisestimatedat34.
2million.
Dividingtheincrementalcostsbytheincrementalhealthbenefitsresultsinanincrementalcosteffectivenessratioof113891perQALYgainedforPCV-7.
Anincremen-talcosteffectivenessratiooflessthan50000perQALYwouldberequiredforPCV-7toberegardedaspotentiallycosteffective.
Shiftingfroma3+1dosescheduletoa2+1regimencouldimprovecosteffec-tivenessofPCV-7to82975perQALY(table4).
CosteffectivenessofPCV-10andPCV-13Comparedwithnovaccination,vaccinationwithPCV-10wouldprevent6124casesofotitismedia,463casesofnon-invasivepneumonia,and258casesofinvasivepneumococcaldisease,ofwhich150wouldbeavertedbynetindirecteffectsinindividualsaged5yearsandolder.
Overallthesehealthbenefitswouldresultinagainof707QALYs.
VaccinationwithPCV-13wouldprevent6876casesofotitismedia,520casesofnon-invasivepneumonia,and331casesofinvasivepneumococcaldisease,resultinginatotalgainof891QALYs.
Dividingtheincrementalcostsbytheincrementalhealthbenefitsforthe10valentand13valentvaccinesproducedincrementalcosteffectivenessratiosof52947and50042perQALYforPCV-10andPCV-13,respectively.
A2+1doseschedulecouldreducetheseincrementalcosteffectivenessratiosto37891forPCV-10andto35743forPCV-13(table4).
A25%reductioninthevaccinepriceofPCV-10andPCV-13(to50perdose,thecostofTable4|Incrementalcosteffectivenessratiosinthebasecaseanalysis,sensitivityanalysis,andseveralscenarioanalysesPCV-7(/QALY)PCV-10(/QALY)PCV-13(/QALY)3+1dosescheduleWithoutnetpositiveindirecteffectsforindividualsaged5yearsorolder*1138919915191705With10%netpositiveindirecteffectsforindividualsaged5yearsorolder*599375294750042With20%netpositiveindirecteffectsforindividualsaged5yearsorolder*3969835146334792+1dosescheduleWithoutnetpositiveindirecteffectsforindividualsaged5yearsorolder*829757208366572With10%netpositiveindirecteffectsforindividualsaged5yearsorolder*430703789135743With20%netpositiveindirecteffectsforindividualsaged5yearsorolder*281012471823488Reductioninthecostofthevaccine(50perdose)NA4110631250Excludingherdeffectsintheanalysedbirthcohortforinvasivepneumococcaldisease1290695777055055Includingherdeffectsintheanalysedbirthcohortfornon-invasivepneumococcaldisease1111535221149407Higherutilitylosses§675814013638664Exclusionofproductivitylosses(analysisfromahealthcareperspective)1154815390450938Efficacyagainstacuteotitismedia785274304841457*Inclusionofnetpositiveindirecteffects(herdprotectionagainstvaccineserotypediseaseminusincreasesinnon-vaccineserotypepneumococcaldisease).
Seealsowebextra2.
Basecasescenario.
Scenarioswerecalculatedholdingallotherassumptions,similartothebasecaseanalysis(thatis,withnonetindirectbenefitsforPCV-7and10%netindirectbenefitsforPCV-10andPCV-13).
§UtilitiesreportedbyProsseretalwereusedforchildrenagedupto5yearsold.
40Efficacyagainstacuteotitismediawasassumedtobe33.
6%,aswasshownfortheprecursorvaccineofPCV-10byPrymulaetal.
41NA,notapplicable;PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine;QALY,qualityadjustedlifeyear.
RESEARCHpage6of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-7)wouldreducethecosteffectivenessratiosto41106and31250,respectively.
Assumingbothadose(tothreedoses)andapricereduction(to50perdose),thecosteffectivenessratiosforPCV-10andPCV-13wouldbeaslowas29013and21654perQALY,respectivelyScenarioandsensitivityanalysesFigure2showstheparametersthatproducedthelar-gestvariationinthecosteffectivenessratioforPCV-7whenvariedby25%.
Apartfromvaccineefficacyagainstinvasivepneumococcaldisease,themostimportantdeterminantsofthecosteffectivenessofPCVwerethetotalvaccinationcosts,theincreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypes,andthecasefatalityrateformeningitis.
InunivariatesensitivityanalysesforPCV-10andPCV-13,generallysimilarbutsmallerchangesintheincre-mentalcosteffectivenessratiowereobserved.
Thechangesweresmallerbecauseoftherelativeimpor-tanceofindirectbenefitsintheunvaccinatedpopula-tionforPCV-10andPCV-13.
Figure3showstheimpactofvaryingthelevelofnetindirecteffectsofvaccinationinindividualsaged5yearsorover.
Atleast14%oftheestimatednetindir-ecteffectwouldbeneededinordertomakePCV-7costeffective(thatis,lessthan50000perQALY).
Severalscenarioanalysesaredisplayedintable4,whichagainshowthelargeimpactofindirecteffectsandreduceddoseschedulesonthecosteffectivenessofpneumococcalvaccination.
Finally,figure4showscosteffectivenessacceptabil-itycurvesforsixdifferentscenarios.
ThisfigureclearlyshowsthatadministeringPCV-7ina3+1dosesche-dulecannotbeconsideredascosteffectivecomparedwithnovaccination.
Theincrementalcosteffective-nessratiosofPCV-10andPCV-13arelikelytobemorefavourablethanthatforPCV-7,yetstillthetotalcostsofvaccinationshouldbereducedinordertounambiguouslyconsidervaccinationcosteffective.
DISCUSSIONOureconomicanalysisindicatesthatthecurrentnationalvaccinationprogrammewithPCV-7intheNetherlandsisnotcosteffective.
Asseveralpaperssug-gestthatloweringthenumberofdosesfromfourtothreewillnotaffectthevaccineefficacyforthepneu-mococcalvaccine,11282931weinvestigatedthepotentialimpactofsuchreduced-doseschedules.
Althougha2+1reduceddoseschedulecouldlowerthetotalcostofvaccinationand,therefore,reducetheincrementalcosteffectivenessratiobyapproximately30%,itisunlikelythatuniversalvaccinationwithPCV-7willbecomeacceptableonthegroundsofcosteffectiveness.
MorefavourableincrementalcosteffectivenessratioswereshownforPCV-10andPCV-13,aslongasnetpositiveindirecteffectsforindividualsaged5yearsorolderwereincludedintheanalyses.
Inpar-ticular,scenariosthatusedreducedtotalvaccinationcostsbyusinga2+1dosescheduleshowedthatincre-mentalcosteffectivenessratioswoulddecreasedownto37891and35743perQALYforPCV-10andPCV-13,respectively.
Theseratiosarelikelytobecon-sideredascosteffectivegivenvariouscountryspecificthresholds.
StrengthsandweaknessesThisisthefirsteconomicevaluationofnationalvacci-nationagainstpneumococcaldiseasethathasincludedserotypereplacementfortheanalysedbirthcohortbyusingpost-vaccinationdata.
1213Weestimatedthenumberofcasesofinvasivepneumococcaldiseaseavertedbyvaccinationandtheincreaseininvasivepneumococcaldiseasecausedbynon-vaccinesero-typesonthebasisofthemostrecentdataavailable.
17Giventherelativelysmallnumberofcasesreportedduringthesurveillanceperiodoftwoyears,ourpredic-tionsregardingtheincreaseofdiseasecausedbynon-vaccineserotypesmayhavelimitedprecision;how-ever,theyarebasedonthebestdatacurrentlyavail-able.
Inparticular,theestimatedincreaseof100%forinvasivediseasecausedbyserotypesnotcoveredbyPCV-7wasbasedonnationalobservationalstudiesfromtheNetherlandsandtheUK.
121317Ontheonehand,thisspecificassumptionmaybetoopessimistic.
Ontheotherhand,datafromtheUKshowanongoingincreaseinthecasesofinvasivepneumococcaldiseasecausedbynon-vaccineserotypesandnoplateauhasyetbeenreachedinthethirdyearafterPCV-7intro-duction,suggestingthattheeventualincreaseindis-easecausedbynon-vaccineserotypesmightevenbehigher.
12Thereare,however,someimportantdiffer-encesbetweentheNetherlandsandtheUK.
IncontrasttotheNetherlands,theUKusesareduceddosesche-duleofPCV-7at2,4,and13months.
Also,theintro-ductionofPCV-7intheUKwasfollowedbyacatch-upCosteffectivenessratio(/QALY)80000100000120000140000160000180000Increaseinnon-vaccineserotypeinvasivepneumococcaldiseaseVaccineefficacyagainstnon-vaccineserotypeinvasivepneumococcaldiseaseChangeinQALYwithcochlearimplantChangeinQALYacuteotitismediaChangeinQALYwithdisabilityTotalvaccinationcostsCasefatalityrateformeningitis(%)Casefatalityrateforbacteraemiawithoutfocus(%)VaccineefficacyagainstacuteotitismediaFig2|Sensitivityanalysisonthebasecasecosteffectivenessratioforthesevenvalentpneumococcalconjugatedvaccine.
Theparameterswerevariedby25%.
Darkbarsshowtheincrementalcosteffectivenessratioaftera25%decreaseintheparameter,whereaslightbarsshowtheincrementalcosteffectivenessratioaftera25%increase(notethatitwasnotpossibletoincreasevaccineefficacy).
QALY,qualityadjustedlifeyearRESEARCHBMJ|ONLINEFIRST|bmj.
compage7of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfromprogrammeforallchildrenagedlessthan2years.
IntheNetherlands,bycontrast,vaccinationwasimple-mentedwithoutacatch-upprogramme.
Severalalter-nativescenariosregardingserotypereplacementwereexploredinthesensitivityanalyses,whichshowedthatourconclusionsregardingtheincrementalcosteffec-tivenessratiosforallthreevaccineswerequiterobust.
InourbasecaseanalysisforPCV-7,weassumedthattherewasnonetindirecteffectofvaccinationforindi-vidualsoutsidethemodelledbirthcohortbecausenooverallreductionininvasivepneumococcaldiseaseinnon-vaccineeshasbeenobservedinanyEuropeancountry,incontrasttotheUS.
121315ThedifferencebetweenresultsobtainedintheUSandthoserecordedinEuropemaybepartlyexplainedbythe60%to70%coverageofthesevenvaccineserotypesinEurope,comparedwiththemorethan80%coverageintheUS.
32ThisdisparityleavesmoreroomforreplacementdiseaseinEurope.
Countryspecificdifferencesinthecirculatingserotypescausingdisease(inclusiveofsecu-larchangesintime)couldalsocontributetotheloweroverallreductionofinvasivepneumococcaldiseaseinEuropecomparedwiththeUS.
33Furthermore,intheNetherlands,asinmostpartsofEurope,thebaselineincidenceratesofinvasivepneumococcaldiseaseinchildrenaresubstantiallylowerthanintheUSandalmostexclusivelybasedoncultureconfirmedcasesofchildrenadmittedtohospital.
1734Anotherpoten-tiallyrelevantdifferenceintheintroductionofPCV-7intheNetherlandscomparedwiththeUSisthehighvaccineuptake(>95%)amongallnewbornsintheNetherlandsforallfourdosesofthevaccination,whichcouldpotentiallyleadtomorerapiddevelop-mentofreplacementdisease.
34Potentialnetindirecteffectsinnon-vaccineesweremodelledusingstraightforwardcalculus.
Ideally,theimpactofpneumococcalvaccinationshouldhavebeenmodelledusingasocalleddynamictransmissionmodel,inwhichthetransmissionandcarriageofStrep-tococcuspneumoniaeistakenexplicitlyintoaccount.
However,becausethetransitiondynamicsofSpneu-moniaearecomplexandserotypedependent,anddetaileddataregardingthesetransmissiondynamicsarealsoquitelimited,dynamicallymodellingallrele-vantserotypesofSpneumoniaewouldbeverycompli-cated.
ForPCV-10andPCV-13,anetindirecteffectof10%wasincludedinthebasecaseanalysis.
Thisesti-mateofindirectbenefitmaybeconservativeifcom-paredwiththemuchhighernetindirectprotectivebenefitsobservedintheUSafterimplementationofroutinevaccinationwithsimilarorlowervaccineser-otypecoverage.
2-4Furthermore,wedidnotincludethebenefitsarisingfromthepreventionofantibioticresistanceinourmodelbecausetheimpactofthisinclusionisexpectedtobesmallgiventhatpenicillinresistanceislessthan0.
4%intheNetherlands.
17Finally,similartoalmostallpreviouscosteffectivenessanalysesforpneumococcalvaccination,ouranalytictimeframewasequaltotheassumedprotectionperiod,afterwhichweassumedthathealtheffectsandcostswouldbesimilarinthevaccinatedandunvaccinatedgroup.
ComparisonwithotherstudiesThecosteffectivenessofPCV-7isworsethanthatcal-culatedinourpreviousstudiesandinotherrecenthealtheconomicstudies.
5-10161835Thisdisparityismostlybecauseoftheexclusionofherdprotectioneffectsandtheinclusionofserotypereplacementinourstudy.
Otherfactorscontributingtotheworseincrementalcosteffectivenessratioweretheuseofalowerdeathrateforinvasivepneumococcaldiseaseandlowerindirectcoststhaninourpreviousstudies.
616Severalrecentlypublishedcosteffectivenessstudiesincludednetvaccinebenefitsforunvaccinatedadultsandelderlypeopleintheirbasecaseanalysis.
5-7935Thesestudiesreportedvaccinationtobecostsaving59oratleastcosteffective.
6735ThethreestudiesthatexcludedherdprotectioninthebasecaseanalysisreportedrelativelyunfavourablecosteffectivenessratiosforPCV-7comparedwithotherrecommendedinfantvaccinations.
81618Whenweexcludedtheincreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypesbutleftallotherassumptionsthesameasinthebasecaseanalysis,ourresultsweresimilartothoseofthesethreestudies—thatis,wefoundanunfavourablecosteffectivenessratio.
81618OurcosteffectivenessresultsshowthatthecurrentvaccinationscheduleforPCV-7mightbefarmoreexpensiveperQALYgainedcomparedwithotherrou-tineinfantvaccinationprogrammesrecentlyimple-mented,suchasforhumanpapillomavirus36,orwithothervaccinesthathavenotyetbeenimplementedinanationalprogrammeintheNetherlands,suchashepa-titisB37andvaricella.
38ImplicationsandfutureresearchAdministrationofPCV-7at2,3,4,and11monthswasintroducedtotheNetherlandsaspartofthenationalimmunisationprogrammein2006partiallyonthebasisoffavourablecosteffectivenessdata.
ThecurrentanalysisshowsunfavourablecosteffectivenessoftheNetindirectbenefits(%)040000600008000010000012000020000051015202530PCV-10PCV-13PCV-7Costeffectivenessratio(/QALY)Fig3|Theeffectoncosteffectivenessratiosofvaryingthelevelofnetindirecteffectofvaccinationforindividualsaged5yearsorolder.
Thehorizontaldashedlineshowsthethresholdat50000perQALY.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine(3+1doseschedule);QALY,qualityadjustedlifeyearRESEARCHpage8of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-73+1doseschedulebecauseofincreasesininva-sivediseasecausedbynon-vaccineserotypes,whichoffsettheherdprotectivebenefitsinindividualsout-sidetheanalysedbirthcohort.
Althoughthecosteffec-tivenessofPCV-7isunfavourablefromahealtheconomicspointofview,itisfavourablefromapublichealthpointofview—asignificantdecreaseincasesofpneumococcaldiseasehasoccurredintheNetherlandsoverthepasttwoyears.
13Switchingtothe10valentor13valentvaccinewouldextendtheserotypecoveragetoahigherlevelthanthatcurrentlyachievedwithPCV-7,whichmightreducethepotentialfordiseasecausedbynon-vaccineserotypesandincreasetheoverallbenefitsinvaccinatedchildren.
Herdprotectiveeffectsaremorelikelytooccurwithbroadvaccinecoverage,renderingvaccinationpoten-tiallycosteffective.
Vaccinationwouldbeparticularlycosteffectiveifamorevalentvaccineisusedincombi-nationwithdosereductions,pricereductions,orboth.
OurpapershouldhelpguidefuturedecisionstopotentiallyreducedosesofpneumococcalvaccineortoshiftfromPCV-7tovaccinesthatcoveradditionalserotypes.
Furtherresearchshouldbedirectedtobuildingadynamicmodeltoentangleandexplicitlypredicttheindirecteffectsofdiseasereplacementandherdprotectiononvaccineefficacyandthusfurtherenhancethevalidityofcosteffectivenessapproachesappliedtopneumococcalvaccination.
Contributors:MJP,EH,andGvdDdesignedthestudy.
MHR,AJvH,andMJPdesignedthecomputermodelandcarriedoutthecomputersimulationsandanalysis.
DataanalyseswereperformedbyAGSCJ,GDR,andMHRundersupervisionofEAMS,MJP,andAvdE.
MHR,MJP,AJvH,andEAMSdraftedthemanuscript.
Allauthorscommentedondraftsandcontributedtothefinalversion.
MHRandMJParetheguarantorsofthestudy.
Funding:MHRwasfundedbyanunrestrictedgrantfromWyethHoofddorp.
AJvHwasfinancedbytheNetherlandsVaccineInstitute,Bilthoven.
ThisworkhasbeenpreviouslypresentedataworkshoponpneumococcalvaccinesattheEuropeanPublicHealthAssociationconferenceinLisbon,Portugal,whichwassupportedbyaresearchgrantfromGlaxoSmithKlineNetherlands.
Theauthors'workwasindependentofthefunders,whohadnoroleinthestudydesign,analysisofdata,writingofthemanuscript,ordecisiontosubmitforpublication.
Competinginterests:AllauthorshavecompletedtheUnifiedCompetingInterestformatwww.
icmje.
org/coi_disclosure.
pdf(availableonrequestfromthecorrespondingauthor)anddeclare:(1)MHRwasfundedbyanunrestrictedgrantfromWyethHoofddorp;andAJvHwasfinancedbytheNetherlandsVaccineInstitute,Bilthoven;(2)MJPhasreceivedtravelgrantsfromGlaxoSmithKlineandWyethtoattendexpertmeetingsinReykjavik,Iceland,andIstanbul,Turkey;EAMShasreceivedunrestrictedgrantsfromWyethandBaxterforresearch,consultingfeesfromWyethandGlaxoSmithKline,lecturingfeesfromWyeth,andgrantsupportfromWyethandGlaxoSmithKlineforvaccinestudies;andAvdEhasreceivedunrestrictedgrantsfromWyethandNovartis;(3)Nospouses,partners,orchildrenwithrelationshipswithcommercialentitiesthatmighthaveaninterestinthesubmittedwork;(4)Nonon-financialintereststhatmayberelevanttothesubmittedwork.
Datasharing:Noadditionaldataavailable.
1HealthCounciloftheNetherlands.
Vaccinationofinfantsagainstpneumococcalinfections[inDutch].
2005.
www.
gezondheidsraad.
nl/sites/default/files/05@13N.
pdf.
2HsuHE,ShuttKA,MooreMR,BeallBW,BennettNM,CraigAS,etal.
Effectofpneumococcalconjugatevaccineonpneumococcalmeningitis.
NEnglJMed2009;360:244-56.
3LexauCA,LynfieldR,DanilaR,PilishviliT,FacklamR,FarleyMM,etal.
Changingepidemiologyofinvasivepneumococcaldiseaseamongolderadultsintheeraofpediatricpneumococcalconjugatevaccine.
JAMA2005;294:2043-51.
4WhitneyCG,FarleyMM,HadlerJ,HarrisonLH,BennettNM,LynfieldR,etal.
Declineininvasivepneumococcaldiseaseaftertheintroductionofprotein-polysaccharideconjugatevaccine.
NEnglJMed2003;348:1737-46.
5ClaesC,ReinertRR,vonderSchulenburgJM.
CosteffectivenessanalysisofheptavalentpneumococcalconjugatevaccineinGermanyconsideringherdimmunityeffects.
EurJHealthEcon2009;10:25-38.
6HubbenGA,BosJM,GlynnDM,vanderEndeA,vanAlphenL,PostmaMJ.
Enhanceddecisionsupportforpolicymakersusingawebinterfacetohealth-economicmodels—illustratedwithacost-effectivenessanalysisofnation-wideinfantvaccinationwiththe7-valentpneumococcalconjugatevaccineintheNetherlands.
Vaccine2007;25:3669-78.
7McIntoshED,ConwayP,WillinghamJ,HollingsworthR,LloydA.
PneumococcalpneumoniaintheUK—howherdimmunityaffectsthecost-effectivenessof7-valentpneumococcalconjugatevaccine(PCV).
Vaccine2005;23:1739-45.
8MelegaroA,EdmundsWJ.
Cost-effectivenessanalysisofpneumococcalconjugatevaccinationinEnglandandWales.
Vaccine2004;22:4203-14.
9SilfverdalSA,BergS,HemlinC,JokinenI.
Thecost-burdenofpaediatricpneumococcaldiseaseinSwedenandthepotentialcost-effectivenessofpreventionusing7-valentpneumococcalvaccine.
Vaccine2009;27:1601-8.
10TilsonL,UsherC,ButlerK,FitzsimonsJ,O'HareF,CotterS,etal.
EconomicevaluationofauniversalchildhoodpneumococcalconjugatevaccinationstrategyinIreland.
ValueHealth2008;11:898-903.
WHATISALREADYKNOWNONTHISTOPICRecentpneumococcalsurveillancestudiesshowasignificantincreaseinnon-vaccineserotypedisease,whichreducestheoverallhealtheffectsofvaccinationandoffsetspotentialpositiveherdprotectionbenefitsinunvaccinatedindividualsExistingeconomicanalysesofpneumococcalvaccinationdidnotincorporatetheincreaseindiseasecausedbynon-vaccineserotypes,andmostincludedtoooptimisticherdprotectiveeffectsWHATTHISSTUDYADDSThecurrentfourdose(3+1)scheduleofthesevenvalentpneumococcalconjugatedvaccineusedintheNetherlandsisnotcosteffectiveVaccinationwith10valentor13valentvaccinescouldsubstantiallyreducetheincrementalcosteffectivenessofpneumococcalvaccinationtoapotentiallyacceptablelevelReducingthetotalprogrammecostforpneumococcalvaccinationbyreducingdoseschedules,reducingvaccineprices,orbothisnecessarytounambiguouslyrenderroutineinfantvaccinationcosteffectiveintheNetherlandsPCV-7basecase(3+1doseschedule)PCV-725%reductioninvaccinationcostsPCV-10basecase(3+1doseschedule)PCV-1025%reductioninvaccinationcostsPCV-13basecase(3+1doseschedule)PCV-1325%reductioninvaccinationcostsThreshold(¤/QALY)Proportionbelowthreshold(%)04000080000120000160000200000020406080100Fig4|CosteffectivenessacceptabilitycurvesforbasecasevaccinationschedulesandforalternativescenariosforPCV-7,PCV-10,andPCV-13RESEARCHBMJ|ONLINEFIRST|bmj.
compage9of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom11VestrheimDF,LovollO,AabergeIS,CaugnantDA,HolbyEA,BakkeH,etal.
Effectivenessofa2+1doseschedulepneumococcalconjugatevaccinationprogrammeoninvasivepneumococcaldiseaseamongchildreninNorway.
Vaccine2008;26:3277-81.
12HealthProtectionAgency.
CumulativeweeklynumberofreportsofinvasivepneumococcaldiseaseinEnglandandWalesbyepidemiologicalyearJuly-June.
2009.
www.
hpa.
org.
uk.
13RodenburgGD,deGreeffSC,JansenAG,deMelkerHE,SchoulsLM,HakE,etal.
Effectsofpneumococcalconjugatevaccine2yearsafteritsintroduction,theNetherlands.
EmergInfectDis2010;16:816-23.
14Munoz-AlmagroC,JordanI,GeneA,LatorreC,Garcia-GarciaJJ,PallaresR.
Emergenceofinvasivepneumococcaldiseasecausedbynonvaccineserotypesintheeraof7-valentconjugatevaccine.
ClinInfectDis2008;46:174-82.
15LepoutreA,VaronE,GeorgesS,GutmannL,Levy-BruhlD.
ImpactofinfantpneumococcalvaccinationoninvasivepneumococcaldiseasesinFrance,2001-2006.
EuroSurveill2008;13:ii,18962.
16BosJM,RumkeH,WelteR,PostmaMJ.
Epidemiologicimpactandcost-effectivenessofuniversalinfantvaccinationwitha7-valentconjugatedpneumococcalvaccineintheNetherlands.
ClinTher2003;25:2614-30.
17JansenAGSC,RodenburgGD,GreefSC,HakE,VeenhovenRH,SpanjaardL,etal.
InvasivepneumococcaldiseaseintheNetherlands:syndromes,outcomeandpotentialvaccinebenefits.
Vaccine2009;27:2394-401.
18Giorgi-RossiP,MeritoM,BorgiaP.
Cost-effectivenessofintroducingtheconjugatedpneumococcalvaccinetoroutinefreeimmunizationsforinfantsinLazio,Italy.
HealthPolicy2009;89:225-38.
19BlackS,ShinefieldH,FiremanB,LewisE,RayP,HansenJR,etal.
Efficacy,safetyandimmunogenicityofheptavalentpneumococcalconjugatevaccineinchildren.
NorthernCaliforniaKaiserPermanenteVaccineStudyCenterGroup.
PediatrInfectDisJ2000;19:187-95.
20MadhiSA,AdrianP,KuwandaL,JassatW,JonesS,LittleT,etal.
Long-termimmunogenicityandefficacyofa9-valentconjugatepneumococcalvaccineinhumanimmunodeficientvirusinfectedandnon-infectedchildrenintheabsenceofaboosterdoseofvaccine.
Vaccine2007;25:2451-7.
21HansenJ,BlackS,ShinefieldH,CherianT,BensonJ,FiremanB,etal.
Effectivenessofheptavalentpneumococcalconjugatevaccineinchildrenyoungerthan5yearsofageforpreventionofpneumonia:updatedanalysisusingWorldHealthOrganizationstandardizedinterpretationofchestradiographs.
PediatrInfectDisJ2006;25:779-81.
22FiremanB,BlackSB,ShinefieldHR,LeeJ,LewisE,RayP.
Impactofthepneumococcalconjugatevaccineonotitismedia.
PediatrInfectDisJ2003;22:10-6.
23BlackSB,ShinefieldHR,LingS,HansenJ,FiremanB,SpringD,etal.
Effectivenessofheptavalentpneumococcalconjugatevaccineinchildrenyoungerthanfiveyearsofageforpreventionofpneumonia.
PediatrInfectDisJ2002;21:810-5.
24EskolaJ,KilpiT,PalmuA,JokinenJ,HaapakoskiJ,HervaE,etal.
Efficacyofapneumococcalconjugatevaccineagainstacuteotitismedia.
NEnglJMed2001;344:403-9.
25NelsonJC,JacksonM,YuO,WhitneyCG,BoundsL,BittnerR,etal.
Impactoftheintroductionofpneumococcalconjugatevaccineonratesofcommunityacquiredpneumoniainchildrenandadults.
Vaccine2008;26:4947-54.
26CentersforDiseaseControlandPrevention.
Pneumoniahospitalizationsamongyoungchildrenbeforeandafterintroductionofpneumococcalconjugatevaccine—UnitedStates,1997-2006.
MMWRMorbMortalWklyRep2009;58:1-4.
27BrouwerWB,NiessenLW,PostmaMJ,RuttenFF.
Needfordifferentialdiscountingofcostsandhealtheffectsincosteffectivenessanalyses.
BMJ2005;331:446-8.
28GoldblattD,SouthernJ,AshtonL,RichmondP,BurbidgeP,TasevskaJ,etal.
Immunogenicityandboostingafterareducednumberofdosesofapneumococcalconjugatevaccineininfantsandtoddlers.
PediatrInfectDisJ2006;25:312-9.
29WhitneyCG,PilishviliT,FarleyMM,SchaffnerW,CraigAS,LynfieldR,etal.
Effectivenessofseven-valentpneumococcalconjugatevaccineagainstinvasivepneumococcaldisease:amatchedcase-controlstudy.
Lancet2006;368:1495-502.
30VespaG,ConstenlaDO,PepeC,SafadiMA,BerezinE,CassiodeMoraesJ,etal.
Estimatingthecost-effectivenessofpneumococcalconjugatevaccinationinBrazil.
RevPanamSaludPublica2009;26:518-28.
31KayhtyH,AhmanH,ErikssonK,SorbergM,NilssonL.
Immunogenicityandtolerabilityofaheptavalentpneumococcalconjugatevaccineadministeredat3,5and12monthsofage.
PediatrInfectDisJ2005;24:108-14.
32HausdorffWP,BryantJ,ParadisoPR,SiberGR.
Whichpneumococcalserogroupscausethemostinvasivedisease:implicationsforconjugatevaccineformulationanduse,partI.
ClinInfectDis2000;30:100-21.
33FenollA,GranizoJJ,AguilarL,GimenezMJ,Aragoneses-FenollL,HanquetG,etal.
TemporaltrendsofinvasiveStreptococcuspneumoniaeserotypesandantimicrobialresistancepatternsinSpainfrom1979to2007.
JClinMicrobiol2009;47:1012-20.
34CentersforDiseaseControlandPrevention.
Invasivepneumococcaldiseaseinchildren5yearsafterconjugatevaccineintroduction—eightstates,1998-2005.
MMWRMorbMortalWklyRep2008;57:144-8.
35RayGT,WhitneyCG,FiremanBH,CiurylaV,BlackSB.
Cost-effectivenessofpneumococcalconjugatevaccine:evidencefromthefirst5yearsofuseintheUnitedStatesincorporatingherdeffects.
PediatrInfectDisJ2006;25:494-501.
36JitM,ChoiYH,EdmundsWJ.
EconomicevaluationofhumanpapillomavirusvaccinationintheUnitedKingdom.
BMJ2008;337:a769.
37TilsonL,ThorntonL,O'FlanaganD,JohnsonH,BarryM.
CosteffectivenessofhepatitisBvaccinationstrategiesinIreland:aneconomicevaluation.
EurJPublicHealth2008;18:275-82.
38RozenbaumMH,vanHoekAJ,VegterS,PostmaMJ.
Cost-effectivenessofvaricellavaccinationprograms:anupdateoftheliterature.
ExpertRevVaccines2008;7:753-82.
39OostenbrinkJB,BouwmansCAM,KoopmanschapMA,RuttenFHH.
Guidelinesforcostingresearch,methodsandstandardizedpricesforeconomicevaluationsinhealthcare.
DutchHealthCareInsuranceBoard,2004.
40ProsserLA,RayGT,O'BrienM,KleinmanK,SantoliJ,LieuTA.
Preferencesandwillingnesstopayforhealthstatespreventedbypneumococcalconjugatevaccine.
Pediatrics2004;113:283-90.
41PrymulaR,PeetersP,ChrobokV,KrizP,NovakovaE,KaliskovaE,etal.
PneumococcalcapsularpolysaccharidesconjugatedtoproteinDforpreventionofacuteotitismediacausedbybothStreptococcuspneumoniaeandnon-typableHaemophilusinfluenzae:arandomiseddouble-blindefficacystudy.
Lancet2006;367:740-8.
Accepted:16February2010RESEARCHpage10of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom
DesignEconomicevaluationcomparingPCV-7,PCV-10,andPCV-13withnovaccinationusingadecisiontreeanalyticmodelbuiltfromdatainpreviousstudies.
SettingTheNetherlands.
PopulationAcohortof180000newbornsfolloweduntil5yearsofage.
MainoutcomemeasuresCosts;gainsinlifeyearsandqualityadjustedlifeyears(QALYs);andincrementalcosteffectivenessratios.
ResultsUnderbasecaseassumptions—thatis,assumingafiveyearprotectiveperiodofthevaccineandnoassumednetindirecteffects(herdprotectionminusserotypereplacement)amongchildrenagedover5years—vaccinationwithPVC-7inafourdose(3+1)schedulewasestimatedtoprevent71and5778casesofinvasiveandnon-invasivepneumococcaldisease,respectively,inchildrenagedupto5years.
Thiscorrespondswithatotalnetgainof173lifeyearsor277QALYs.
TheincrementalcosteffectivenessratioofPCV-7wasestimatedat113891(98300;$145000)perQALY,wellovertheratioof50000perQALYrequiredforPCV-7toberegardedaspotentiallycosteffective.
Athreedose(2+1)scheduleofPCV-7reducedtheincrementalcosteffectivenessratioto82975perQALY.
Forvariousassumptionsandincluding10%ofthemaximumnetindirecteffectsamongindividualsaged5yearsandover,PCV-10andPCV-13hadincrementalcosteffectivenessratiosrangingfrom31250to52947perQALY.
ConclusionsThecurrentDutchinfantvaccinationprogrammeoffourdosesofPCV-7isnotcosteffectivebecauseofincreasesininvasivediseasecausedbynon-vaccineserotypes,whichreducestheoveralldirecteffectsofvaccinationandoffsetspotentialpositiveherdprotectionbenefitsinunvaccinatedindividuals.
The10valentand13valentpneumococcalvaccinescouldhavebetternethealthbenefitsthanPCV-7throughlessreplacementdiseaseandincreasedherdprotection.
Boththeseeffectscouldsubstantiallyreducetheincrementalcosteffectivenessratiotopossiblyacceptablelevels,iftotalprogrammecostscanbeloweredbyreducedschedules,reductionsinvaccineprices,orboth.
INTRODUCTIONGiventhemultitudeofnewvaccinesavailableforintroductionintonationalimmunisationprogrammes,healtheconomicmodellingofvariousimmunisationplansisbecomingincreasinglyimportantininformingdecisionsonhealthpolicy.
Thedecisiontointroducethesevenvalentpneumococcalconjugatedvaccine(PCV-7)intotheDutchnationalimmunisationpro-grammeforinfants,forexample,hasinpartbeendri-venbycosteffectivenessconsiderations.
1TheDutchHealthCouncilestimatedtheincrementalcosteffec-tivenessratioofvaccinationwithPCV-7comparedwithnovaccinationat70000(60300;$89200)andlessthan20000perqualityadjustedlifeyear(QALY)in2001and2005,respectively.
1Crucialfactorsresponsibleforthechangefromapotentiallyunfavour-ablecosteffectivenessratioin2001,exceeding50000perQALY,toafavourableratioin2005weretheinclu-sionofdataonobservedherdprotectioneffectsinadultsafternationwideimplementationofPCV-7intheUSAin2000andlimiteddiseasedevelopmentcausedbypneumococcalserotypesnotpresentinthe1UnitofPharmacoEpidemiologyandPharmacoEconomics,DepartmentofPharmacy,UniversityofGroningen,Groningen,Netherlands2WilhelminaChildren'sHospital,UniversityMedicalCenterUtrecht,Utrecht,Netherlands3JuliusCenterforHealthSciencesandPrimaryCare,UniversityofUtrecht,Utrecht,Netherlands4CenterforInfectionandImmunityAmsterdam,DepartmentofMedicalMicrobiologyandtheNetherlandsReferenceLaboratoryforBacterialMeningitis,AcademicMedicalCenterAmsterdam,Amsterdam,Netherlands5NetherlandsVaccineInstitute,Bilthoven,Netherlands6DepartmentofEpidemiology,UniversityMedicalCenterGroningen,UniversityofGroningen,Groningen,NetherlandsCorrespondenceto:MHRozenbaumm.
h.
rozenbaum@rug.
nlCitethisas:BMJ2010;340:c2509doi:10.
1136/bmj.
c2509BMJ|ONLINEFIRST|bmj.
compage1of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-7replacingpneumococcalserotypeseliminatedbythevaccine(replacementdisease).
2-4Nexttodirecteffectsoninvasivediseaseinvacci-nees,expectedsavingsfromherdprotectionwerealsopartofhealtheconomicstudiesinotherEuropeancountriesthatintroducedPCV-7intotheirnationalimmunisationprogrammes.
5-10Boththefourdose(3+1)vaccinescheduleandthereducedthreedose(2+1)schedule,asimplementedinNorwayandtheUK,1112arehighlyeffectiveagainstinvasivepneumococcaldiseasecausedbyvaccineserotypes.
However,thenetoverallbenefitofnationalimmu-nisationprogrammesinmanyEuropeancountrieshasbeenreducedbyincreasesininvasivediseasecausedbynon-vaccineserotypes.
12-15Importantly,inthefirst18-30monthsaftertheintroductionofPCV-7intheNetherlands,France,andtheUK,nooverallreductionininvasivediseaseinnon-vaccineeswasobserved.
121315Giventhatbothincreasesininvasivediseasecausedbynon-vaccineserotypesandabsenceofherdprotec-tionmayconsiderablyaffectthecosteffectivenessofthecurrentDutchvaccinationprogramme,wesetouttoupdatecosteffectivenessestimatesforthecurrentfourdosescheduleofPCV-7byusingrecentdataonepidemiologyandresourceuse.
Also,weinvestigatethecosteffectivenessofreduceddoseschedulesandvaccinepricereductionscombinedwiththeimplementationof10valentand13valentpneumo-coccalvaccines(PCV-10andPCV-13).
METHODSModelWedesignedadecisiontreeanalyticmodelstructurethatbuildsonourpreviouslyreportedmodel.
616Var-iousdatasourceswereusedtopopulateourmodel;theseincludedclinicaltrialsandobservationalstudiesforeffectivenessofpneumococcalvaccines,laboratorydataforincidenceandserotypedistributionsofpneu-mococcaldisease,andregistrationsforresourceuseandcosts.
Figure1showsthediseasemodelforthehealtheffectsofpneumococcalvaccination,includingthepos-sibilityofsubsequentpneumococcaldiseasesuchasnon-invasivepneumonia,otitismedia,andinvasivepneumococcaldisease.
Assumptionsregardingbothcostsandqualityoflifearesummarisedintable1andaremorethoroughlydiscussedinwebextra1.
Intheanalyses,acohortof180000newborns,repre-sentingtheDutchbirthcohort,wasrunthroughthedecisiontreetwice(basecaseanalysis):onceasamainlyvaccinatedcohort(PCV-7/PCV-10/PCV-13);andonceasanunvaccinatedcohort.
Theanalytictimeframeofthestudywasfiveyearsbecausevaccineeffectivenesscouldnotbeassumedbeyondfiveyears.
However,longtermeffectsofinvasivepneumococcaldiseasewereextrapolatedoverthefulllifetimeoftheindivi-dualsinthecohort(thatis,untildeathor100years).
SimpleComplexTympanostomyAcuteotitismedia(4)Pneumonia(9)VaccinationofallhealthyinfantsPolicyonpneumococcalconjugatedvaccineNovaccination[+]Meningitis(1,15)NotympanostomyNoinfectionCochlearimplant(12)Nocochlearimplant(13)TreatedbyaGP(3)Death(8)RecoveryHearingproblemsDeath(6)RecoveryDeath(7)RecoveryDisability(11)Admittedtohospital(2)Bacteraemiawithfocus(1,14)Non-invasivepneumococcaldiseaseInvasivepneumococcaldiseaseBacteraemiawithoutfocus(1,14)Invasivepneumonia(1,14)Death(5)Specialeducation(10)RecoverywithoutsequelaeFig1|Decisiontreeusedinconjunctionwiththecohortof180000newborns.
Numbersbetweenbracketscorrespondtodatashownintable1.
Theboxesrepresentdecisionnodes,withgreencolourindicatingprobabilisticstatesandredcolourindicatingendstates.
The"Novaccination"armisacloneofthe"Vaccinationofallhealthyinfants"arm(asrepresentedbythe+sign;risksdifferbetweenbotharms)RESEARCHpage2of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromTable1|ParametersusedintheeconomicmodelMeanorrangeDistributionCorrespondingbranchesinfig1ReferencesCase-fatalityrate(birthcohort)Meningitis9%Beta(3,32)517Pneumonia0%N/AN/A17Bacteraemiawithfocus0%N/AN/A17Bacteraemiawithoutfocus9%Beta(2,21)617Mortality(non-invasivepneumoniaandacuteotitismedia)0%N/AN/AAssumedCase-fatalityrate(agefiveyearsorolder)Meningitis9-92%Beta(agedependent)517Pneumonia0-29%Beta(agedependent)717Bacteraemiawithfocus0-33%Beta(agedependent)817Bacteraemiawithoutfocus9-67%Beta(agedependent)617Respiratoryinfections0%N/AN/AAssumedVaccineefficacyInvasivepneumococcaldisease(allvaccineserotypes)97.
4%Lognormal(SE0.
044)119Non-invasivepneumonia(admittedtohospital)11.
1%Lognormal(SE0.
082)223Non-invasivepneumonia(seenbygeneralpractitioner)6.
0%Lognormal(SE0.
032)323Acuteotitismedia7.
0%Lognormal(SE0.
011)422Directcosts()Costofhospitaladmission*Invasivepneumococcaldisease(agedependent)1091-27318Triangular(agedependent)117,39Non-invasivepneumonia26-2614Triangular(severitydependent)916,39,40Acuteotitismedia17-381Triangular(severitydependent)416,39,40Specialeducation(annualcosts)9798-16962Triangular(agedependent)1016Institutionalcare(annualcosts)39583Triangular(29,687;39,583;49,478)1139Cochlearimplantation56633Triangular(0;0.
004;0.
01)1241Indirectcosts()Invasivepneumococcaldisease0-974Triangular(severitydependent)117,39Non-invasivepneumonia(admittedtohospital)0-2529Triangular(severitydependent)117,39Non-invasivepneumonia(seenbygeneralpractitioner)115-315Triangular(severitydependent)916,39Acuteotitismedia58-23Triangular(severitydependent)416,39Totaldropinqualityoflife(QALYs)Disability§0.
53Beta(estimated)1142Bilateralhearingloss(firstyear)§0.
45Beta(estimated)128,43Bilateralhearingloss:cochleardevice§0.
18Beta(estimated)128,43Allotherhearingloss§0.
09Beta(estimated)1342Hospitaladmissionforbacteraemia**0.
0079Beta(estimated)148,44Hospitaladmissionformeningitis0.
0232Beta(estimated)158,44Hospitaladmissionfornon-invasivepneumonia0.
006Triangular(0.
001,0.
006,0.
01)28Non-invasivepneumoniatreatedbyageneralpractitioner0.
004Triangular(0,0.
004,0.
01)38Acuteotitismedia0.
005Triangular(0,0.
005,0.
01)48OtherparametersIncreaseinnon-vaccineserotypeofinvasivepneumococcaldisease100%Triangular(50%,100%,150%)N/A12,13NetindirecteffectforPCV-10andPCV-1310%Triangular(0%,10%,30%)N/AAssumedDiscountratehealtheffects1.
5%N/AN/A27Discountratecosts4%N/AN/A27*Basedontheaveragedurationofhospitalstay(bothintensivecareandgeneralward)andcorrespondingunitcosts.
39SeealsowebtableBforagespecificcostsofhospitaladmission.
Indirectcostscausedbyabsenceatworkofparentstakingcareoftheirchildren.
Indirectcostscausedbyabsenceatworkofpatientduetohospitaladmission.
§Peryear.
Percase.
**SameQALYdecrementwasassumedforinvasivepneumonia,bacteraemiawithfocus,andbacteraemiawithoutfocus.
Seealsowebextra2.
Indirecteffectsintheanalysedbirthcohort.
Seealsowebextra3.
Indirecteffectsforthoseaged5yearsandolder.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine.
RESEARCHBMJ|ONLINEFIRST|bmj.
compage3of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromBaselinediseaserisksSurveillancedataontheincidenceandserotypedistri-butionofinvasivepneumococcaldiseasebeforenationalimplementationofPCV-7wereavailablefortheperiod2004-2006,includingdataonage,primaryfocusofinfection,resourceuse,hospitaladmission,andoutcome.
1317Thecase-fatalityrateformeningitisandbacteraemiawithoutfocusinchildrenwasesti-matedtobe9%(table1),17whichisinlinewiththeinternationalliterature.
5818Invasivepneumoniaandbacteraemiawithfocuswereassumednottoresultindeathinchildren.
17Inourmodel,severementalandphysicalhandicapresultingfrommeningitiswasassumedtooccurin13%ofcasesofpneumococcalmeningitisinchildren,ofwhich50%wouldrequirespecialeducationand25%intensive"roundtheclock"institutionalcare.
6Jansenetalfoundthathear-ingproblemsoccurredin32%ofcasesofmeningitis,ofwhich50%wereseriousenoughtorequireacochlearhearingdevice.
17Baselinerisksfornon-invasivepneu-moniarequiringhospitaladmissionandfornon-inva-sivepneumoniaandacuteotitismediatreatedingeneralpractitionersurgerieswereestimatedfromnationalhospitalandgeneralpractitionerrecords,respectively(seewebtableA).
VaccineefficaciesVaccineefficacyagainstinvasivepneumococcaldiseasewasassumedat97.
4%aftertwodosesforallsevenser-otypesofpneumococcaldiseasecoveredbyPCV-7(table1).
19Thisvalueseemstobeaconservativeesti-mateifonetakesintoaccountthefactthatonlyonevac-cinefailurehasbeenreportedintheNetherlandsinthefirsttwoyearsafterintroductionofroutineinfantvacci-nationinJune2006.
Routinevaccinationforinfantsina2+1doseschedulewasintroducedinNorwayin2006,andsimilarlynovaccinefailureshadoccurreduptoJune2008.
11Protectionagainstinvasivediseasewasthusesti-matedtolastforfiveyearsinthebasecaseanalysis.
20Furthermore,inrandomisedcontrolledsettings,thevac-cinewasshowntobeeffectiveagainstnon-invasivepneumoniaandotitismediainchildren.
21-23Fornon-invasivepneumonia,efficacyofpneumococcalvaccina-tionseemstoincreasewithdiagnosticcertainty.
23Inourmodel,weappliedtheefficacyestimateof11.
1%for"clinicalpneumoniaandperihilarfindings"tochildrenadmittedtohospitalwiththediagnosisofpneumoniaintheNetherlands.
23ThisdefinitionofpneumoniaseemstobestfitthetypesofpneumoniastreatedinDutchhospitals.
Anefficacyof6.
0%wasassumedforpatientswhovisitedageneralpractitionerandwerediagnosedwithpneumonia.
23Intworando-misedstudies,PCV-7wasfoundtoprevent6.
4%to7.
0%ofallcasesofacuteotitismedia.
221924Theinter-pretationofthesestudiesfortheDutchsettingisham-peredbyseveralfactors,includingthefactthatthecausalmicro-organismisnotrecordedincasesofotitismediaintheNetherlands.
Inourmodel,weusedanoverallefficacyestimateof7.
0%forotitismediaonthebasisofthemostrecentdatafromtheKaiserPer-manentetrial.
22Givenevidenceforthedurationofpro-tectionagainstnon-invasivepneumoniaandrecentUSsurveillancedata,weassumedthatvaccinatedchildrenwereprotectedagainstnon-invasivepneumoniaandotitismediauptotheirsecondyearoflife,startingaftertheseconddoseofthevaccine.
212526Avaccineefficacyof97.
4%againstallserotypesincludedwasassumedforPCV-10andPCV-13,simi-lartotheassumedefficacyofPCV-7.
IntheabsenceofclinicaldataontheefficacyofPCV-10andPCV-13againstnon-invasivepneumoniaandacuteotitismedia,theefficacyofthesetwovaccineswasassumedtoincreaseproportionallywiththeincreaseinserotypecoverageforinvasivepneumococcaldisease.
IndirecteffectsAswellasestimationsofthedirecteffects,wealsoesti-matedindirecteffectsofvaccinationinourmodel.
Weincludedinourbasecaseanalysisherdprotectionagainstinvasivepneumococcaldiseaseforchildreninthebirthcohortnotyetfullyprotectedbythevaccineandfornon-vaccinatedchildren,assumingthisprotec-tionwouldbeaseffectiveasvaccination(table2).
1213Wealsoincreasedtheincidenceofinvasivepneumo-coccaldiseasecausedbynon-vaccineserotypestoTable2|Basecaseserotypecoverageandefficacyfordirecteffectsandassumptionsonindirecteffectsfortheanalysedbirthcohortandtheremainingpopulation(thoseaged5yearsorolder)forPCV-7,PCV-10,andPCV-13PCV-7PCV-10PCV-13Serotypescovered4,6B,9V,14,18C,19F,23F4,6B,9V,14,18C,19F,23F,1,5,7F4,6B,9V,14,18C,19F,23F,3,6A,19A,1,5,7FIncreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypesintheanalysedbirthcohort(serotypereplacement)100%100%100%Efficacyandlevelofherdprotectionagainstvaccineserotypesofinvasivepneumococcaldiseaseintheanalysedbirthcohort*97.
4%97.
4%97.
4%Netindirecteffectintheremainingpopulation0%10%10%*Herdprotectionwasassumedfortheentirebirthcohort,includingthosenotyet(fully)protectedbythevaccine(eitherinfantstooyoungtobevaccinatedorthosewhoreceivedonlyasingledoseofthevaccine)andnon-vaccinatedchildren(5%ofabirthcohortfortheNetherlands),assumingaprotectioneffectof97.
4%againstvaccineserotypes,similartotheefficacyofthevaccine.
Netindirectbenefitsaredefinedasthebenefitsresultingfromprotectionagainstinvasivepneumococcaldiseasecausedbyvaccineserotypesminustheincreaseofinvasivepneumococcaldiseasecausedbynon-vaccineserotypes.
Thepotentialmaximumwasdefinedasfullreductionininvasivepneumococcaldiseasecasescausedbyvaccineserotypesintheabsenceofanyreplacementdisease.
Lowerpercentagescanbedefinedasacombinationofadecreaseinvaccineserotypeinvasivepneumococcaldiseaseandanincreaseindiseasefrominnon-vaccineserotypes.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine.
RESEARCHpage4of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom100%fortheanalysedbirthcohort(thatis,wedoubledtheincidenceofinvasivepneumococcaldiseasecausedbynon-vaccineserotypes)onthebasisofsurveillancedatafromearlyafternationalintroductionofPCV-7intheNetherlandsandtheUK.
1213Seewebextra2foramoreindepthdescriptionoftheassumptionsforourestimationofindirecteffectsinthebirthcohort.
Noserotypeinformationforacuteotitismediaandnon-invasivepneumoniaisavailableintheNether-lands,andserotypereplacementforthesediseasesmaybeassumedtobealreadyincludedinthevaccineefficacyestimatesinthefirstefficacystudies.
21-23There-fore,wedidnotincludeanadditionalincreaseofnon-vaccineserotypediseasebutalsoleftoutpotentialherdeffectsforotitismediaandnon-invasivepneumonia(seewebextra2).
WeassumedinourbasecaseanalysisforPCV-7thatnonetindirecteffectwouldexistforindividualsout-sidethemodelledcohort.
ThisassumptionwasmadebecausenoreductionintheincidenceofinvasivepneumococcaldiseasehasbeenobservedaftertheintroductionofroutinevaccinationwithPCV-7forindividuals5yearsofageorolderandbecausetheobservedherdprotectioneffectintheUKinthethirdyearafterintroducingroutinevaccinationwascompletelycounteredbyariseininvasivepneumococ-caldiseasecausedbynon-vaccineserotypes.
12Inthisrespect,netindirecteffectsaredefinedascasesofinva-sivediseaseavertedbyherdprotectionminusinvasivecasesofreplacementdisease.
Netindirecteffectsmayoccurinthefuture,espe-ciallyifserotypecoverageisextendedbyachangefromsevenserotypevaccinestovaccineswithbroaderserotypecoverage.
1317Therefore,inthebasecaseana-lysisforPCV-10andPCV-13,anetindirecteffectforinvasivediseaseat10%ofthepotentialmaximumwasappliedforthoseaged5yearsorolder(seewebextra3).
Inparticular,thepotentialmaximumwasdefinedaspreventionofallcasesofinvasivediseasecausedbyserotypesinthevaccineandabsenceofanyreplace-mentdisease.
Netprotectiveindirectbenefitsagainstotitismediaandnon-invasivepneumoniawerenotincludedinanyoftheanalyses.
25Giventhatthereismuchuncertaintyaboutthedevel-opmentofindirecteffects,theseassumptionswerevar-iedoverawiderangeinthesensitivityanalyses.
OutcomemeasuresandcosteffectivenessanalysisThesimulationmodeltracksallthespecificdiseasecasesandthedeaths,costs,changesinQALYsandlifeyears,andindirecteffects(herdprotectionandser-otypereplacement).
WewereabletodeterminethenetcostsandnetlifeyearsandQALYsgainedbysummingallthecosts,lifeyears,andQALYsandcalculatingthedifferencesfortheevaluationswithandwithoutvacci-nation.
TheincrementalcosteffectivenessratiowascalculatedbydividingthenetcostsbyeitherlifeyearsorQALYs.
Healtheffectsandcostweredis-countedat1.
5%and4%fortimepreference,respec-tively,accordingtotheDutchguidelinesforcosteffectivenessresearch.
27Incrementalcosteffectivenessratiosforroutinevac-cinationwerecalculatedbycomparingdifferentvacci-nationschedulesagainstnovaccination.
FollowingrecentlypublishedevidenceontheefficacyofPCV-7inreduceddoseschedules,2829weinvestigatedtheeffectofathreedoseschedule(thatis,2+1)totesttheeffectoflowertotalvaccinationcosts(seewebextra4).
WealsoforecastedtheincrementalcosteffectivenessofpotentialshiftsfromPCV-7topneumococcalvac-cinesthatincludeadditionalserotypes(thatis,PCV-10andPCV-13).
ForPCV-7,theestimatedcurrentcostof50perdosewithintheDutchnationalimmunisationpro-grammewasused.
616ForPCV-13,theofficiallylistedpriceof68.
56wasapplied,withadministrationcostsof5.
95beingadded(totalcostperdose74.
51).
16ForPCV-10,noofficiallylistedpriceisavailableintheNetherlands.
GiventhatweknowthepricingofPCV-10inothercountriesisconservativecomparedwithPCV-13,weassumedthetotalcostperdoseofPCV-10atthemidpointbetweenPCV-7andPCV-13(thatis,62.
25).
30Table3|BasecaseanalysisresultsfortheanalysedDutchbirthcohortAcuteotitismediaNon-invasivepneumoniaInvasivepneumococ-caldiseaseInvasivepneumococcaldiseaserelatedtonetindirecteffectsforindividualsaged5yearsorolder*TotalCases(undiscounted)Novaccination170788193851882410NAPCV-716541618979117210NAPCV-1016466418922802260NAPCV-1316391218865382229NACasesavertedPCV-75372406710NAPCV-106124463108150NAPCV-136876520150181NAQALYsgained(years)PCV-72722480277PCV-10302361314707PCV-13342470384891Lifeyearsgained(years)PCV-7001730173PCV-1000255312566PCV-1300336381717Directsavings(1000s),excludingvaccinationcostsPCV-7126375172502226PCV-10144427245413984422PCV-13161479318116965518Indirectsavings(1000s;directeffects)relatedtoproductionlossesPCV-732074460440PCV-103658467161677PCV-134109493202799*Onlynetindirecteffectsagainstinvasivepneumococcaldiseasewereincludedinthemodelforindividualsaged5yearsorolder.
ForPCV-7,nonetindirecteffectswereincludedintothemodelforindividualsaged5yearsorolderinthebasecaseanalysis.
NA,notapplicable;PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine;QALY,qualityadjustedlifeyears.
RESEARCHBMJ|ONLINEFIRST|bmj.
compage5of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromScenarioandsensitivityanalysesWeperformedunivariate,threshold,scenario,andprobabilisticsensitivityanalyses.
Intheunivariatesen-sitivityanalyses,allrelevantparameterswerevariedby25%toexploretheimpactofeachparameterrelativetoeachother.
Onespecificthresholdanalysiswasper-formedinwhichtheeffectoftheparameterontheincrementalcosteffectivenessratiowasinvestigatedbyvaryingthenetindirecteffectsonindividualsaged5yearsorolderoverarangeof0%to30%.
Fortheprobabilisticsensitivityanalyses,parametersweregen-eratedusingMonteCarlosampling,withoutcomevaluesgeneratedbyrunningthemodel5000times.
Lognormal,beta,andtriangulardistributionswereusedexceptformultinomialprobabilities,whereDirichletdistributionswereassumed(seetable1forspecificdistributions).
RESULTSCosteffectivenessofPCV-7Inthebasecaseanalysis,theestimatedburdenofpneu-mococcalinfectionforabirthcohortfollowedforfiveyearswas170788casesofacuteotitismediaand19385casesofnon-invasivepneumonia,ofwhich2645caseswouldresultinhospitaladmission(table3).
Applyingthebasecaseassumptions,5372(31%)casesofacuteotitismediaand406(21%)casesofnon-invasivepneumoniawouldbepreventedbyvaccina-tionwithPCV-7,correspondingtogainsof27and2QALYs,respectively.
Additionally,188casesofinvasivepneumococcaldiseaseayearwereestimatedinchildrenunder5yearsofage:65casesofmeningitis;45casesofinva-sivepneumococcaldisease;38casesofbacteraemiawithfocus;and40casesofbacteraemiawithoutfocus.
Intotal,71(38%)casesofinvasivediseasewouldbepreventedbyvaccinationwithPCV-7,corre-spondingtoatotalgainof173lifeyearsor248QALYs.
Inadditiontothehealthgains,vaccinationwithPCV-7wouldalsopreventapproximately2.
2millionofdirectcostsand0.
4millionofindirectcosts.
Assumingafourdoseschedule,theannualcostofvac-cinationisestimatedat34.
2million.
Dividingtheincrementalcostsbytheincrementalhealthbenefitsresultsinanincrementalcosteffectivenessratioof113891perQALYgainedforPCV-7.
Anincremen-talcosteffectivenessratiooflessthan50000perQALYwouldberequiredforPCV-7toberegardedaspotentiallycosteffective.
Shiftingfroma3+1dosescheduletoa2+1regimencouldimprovecosteffec-tivenessofPCV-7to82975perQALY(table4).
CosteffectivenessofPCV-10andPCV-13Comparedwithnovaccination,vaccinationwithPCV-10wouldprevent6124casesofotitismedia,463casesofnon-invasivepneumonia,and258casesofinvasivepneumococcaldisease,ofwhich150wouldbeavertedbynetindirecteffectsinindividualsaged5yearsandolder.
Overallthesehealthbenefitswouldresultinagainof707QALYs.
VaccinationwithPCV-13wouldprevent6876casesofotitismedia,520casesofnon-invasivepneumonia,and331casesofinvasivepneumococcaldisease,resultinginatotalgainof891QALYs.
Dividingtheincrementalcostsbytheincrementalhealthbenefitsforthe10valentand13valentvaccinesproducedincrementalcosteffectivenessratiosof52947and50042perQALYforPCV-10andPCV-13,respectively.
A2+1doseschedulecouldreducetheseincrementalcosteffectivenessratiosto37891forPCV-10andto35743forPCV-13(table4).
A25%reductioninthevaccinepriceofPCV-10andPCV-13(to50perdose,thecostofTable4|Incrementalcosteffectivenessratiosinthebasecaseanalysis,sensitivityanalysis,andseveralscenarioanalysesPCV-7(/QALY)PCV-10(/QALY)PCV-13(/QALY)3+1dosescheduleWithoutnetpositiveindirecteffectsforindividualsaged5yearsorolder*1138919915191705With10%netpositiveindirecteffectsforindividualsaged5yearsorolder*599375294750042With20%netpositiveindirecteffectsforindividualsaged5yearsorolder*3969835146334792+1dosescheduleWithoutnetpositiveindirecteffectsforindividualsaged5yearsorolder*829757208366572With10%netpositiveindirecteffectsforindividualsaged5yearsorolder*430703789135743With20%netpositiveindirecteffectsforindividualsaged5yearsorolder*281012471823488Reductioninthecostofthevaccine(50perdose)NA4110631250Excludingherdeffectsintheanalysedbirthcohortforinvasivepneumococcaldisease1290695777055055Includingherdeffectsintheanalysedbirthcohortfornon-invasivepneumococcaldisease1111535221149407Higherutilitylosses§675814013638664Exclusionofproductivitylosses(analysisfromahealthcareperspective)1154815390450938Efficacyagainstacuteotitismedia785274304841457*Inclusionofnetpositiveindirecteffects(herdprotectionagainstvaccineserotypediseaseminusincreasesinnon-vaccineserotypepneumococcaldisease).
Seealsowebextra2.
Basecasescenario.
Scenarioswerecalculatedholdingallotherassumptions,similartothebasecaseanalysis(thatis,withnonetindirectbenefitsforPCV-7and10%netindirectbenefitsforPCV-10andPCV-13).
§UtilitiesreportedbyProsseretalwereusedforchildrenagedupto5yearsold.
40Efficacyagainstacuteotitismediawasassumedtobe33.
6%,aswasshownfortheprecursorvaccineofPCV-10byPrymulaetal.
41NA,notapplicable;PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine;QALY,qualityadjustedlifeyear.
RESEARCHpage6of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-7)wouldreducethecosteffectivenessratiosto41106and31250,respectively.
Assumingbothadose(tothreedoses)andapricereduction(to50perdose),thecosteffectivenessratiosforPCV-10andPCV-13wouldbeaslowas29013and21654perQALY,respectivelyScenarioandsensitivityanalysesFigure2showstheparametersthatproducedthelar-gestvariationinthecosteffectivenessratioforPCV-7whenvariedby25%.
Apartfromvaccineefficacyagainstinvasivepneumococcaldisease,themostimportantdeterminantsofthecosteffectivenessofPCVwerethetotalvaccinationcosts,theincreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypes,andthecasefatalityrateformeningitis.
InunivariatesensitivityanalysesforPCV-10andPCV-13,generallysimilarbutsmallerchangesintheincre-mentalcosteffectivenessratiowereobserved.
Thechangesweresmallerbecauseoftherelativeimpor-tanceofindirectbenefitsintheunvaccinatedpopula-tionforPCV-10andPCV-13.
Figure3showstheimpactofvaryingthelevelofnetindirecteffectsofvaccinationinindividualsaged5yearsorover.
Atleast14%oftheestimatednetindir-ecteffectwouldbeneededinordertomakePCV-7costeffective(thatis,lessthan50000perQALY).
Severalscenarioanalysesaredisplayedintable4,whichagainshowthelargeimpactofindirecteffectsandreduceddoseschedulesonthecosteffectivenessofpneumococcalvaccination.
Finally,figure4showscosteffectivenessacceptabil-itycurvesforsixdifferentscenarios.
ThisfigureclearlyshowsthatadministeringPCV-7ina3+1dosesche-dulecannotbeconsideredascosteffectivecomparedwithnovaccination.
Theincrementalcosteffective-nessratiosofPCV-10andPCV-13arelikelytobemorefavourablethanthatforPCV-7,yetstillthetotalcostsofvaccinationshouldbereducedinordertounambiguouslyconsidervaccinationcosteffective.
DISCUSSIONOureconomicanalysisindicatesthatthecurrentnationalvaccinationprogrammewithPCV-7intheNetherlandsisnotcosteffective.
Asseveralpaperssug-gestthatloweringthenumberofdosesfromfourtothreewillnotaffectthevaccineefficacyforthepneu-mococcalvaccine,11282931weinvestigatedthepotentialimpactofsuchreduced-doseschedules.
Althougha2+1reduceddoseschedulecouldlowerthetotalcostofvaccinationand,therefore,reducetheincrementalcosteffectivenessratiobyapproximately30%,itisunlikelythatuniversalvaccinationwithPCV-7willbecomeacceptableonthegroundsofcosteffectiveness.
MorefavourableincrementalcosteffectivenessratioswereshownforPCV-10andPCV-13,aslongasnetpositiveindirecteffectsforindividualsaged5yearsorolderwereincludedintheanalyses.
Inpar-ticular,scenariosthatusedreducedtotalvaccinationcostsbyusinga2+1dosescheduleshowedthatincre-mentalcosteffectivenessratioswoulddecreasedownto37891and35743perQALYforPCV-10andPCV-13,respectively.
Theseratiosarelikelytobecon-sideredascosteffectivegivenvariouscountryspecificthresholds.
StrengthsandweaknessesThisisthefirsteconomicevaluationofnationalvacci-nationagainstpneumococcaldiseasethathasincludedserotypereplacementfortheanalysedbirthcohortbyusingpost-vaccinationdata.
1213Weestimatedthenumberofcasesofinvasivepneumococcaldiseaseavertedbyvaccinationandtheincreaseininvasivepneumococcaldiseasecausedbynon-vaccinesero-typesonthebasisofthemostrecentdataavailable.
17Giventherelativelysmallnumberofcasesreportedduringthesurveillanceperiodoftwoyears,ourpredic-tionsregardingtheincreaseofdiseasecausedbynon-vaccineserotypesmayhavelimitedprecision;how-ever,theyarebasedonthebestdatacurrentlyavail-able.
Inparticular,theestimatedincreaseof100%forinvasivediseasecausedbyserotypesnotcoveredbyPCV-7wasbasedonnationalobservationalstudiesfromtheNetherlandsandtheUK.
121317Ontheonehand,thisspecificassumptionmaybetoopessimistic.
Ontheotherhand,datafromtheUKshowanongoingincreaseinthecasesofinvasivepneumococcaldiseasecausedbynon-vaccineserotypesandnoplateauhasyetbeenreachedinthethirdyearafterPCV-7intro-duction,suggestingthattheeventualincreaseindis-easecausedbynon-vaccineserotypesmightevenbehigher.
12Thereare,however,someimportantdiffer-encesbetweentheNetherlandsandtheUK.
IncontrasttotheNetherlands,theUKusesareduceddosesche-duleofPCV-7at2,4,and13months.
Also,theintro-ductionofPCV-7intheUKwasfollowedbyacatch-upCosteffectivenessratio(/QALY)80000100000120000140000160000180000Increaseinnon-vaccineserotypeinvasivepneumococcaldiseaseVaccineefficacyagainstnon-vaccineserotypeinvasivepneumococcaldiseaseChangeinQALYwithcochlearimplantChangeinQALYacuteotitismediaChangeinQALYwithdisabilityTotalvaccinationcostsCasefatalityrateformeningitis(%)Casefatalityrateforbacteraemiawithoutfocus(%)VaccineefficacyagainstacuteotitismediaFig2|Sensitivityanalysisonthebasecasecosteffectivenessratioforthesevenvalentpneumococcalconjugatedvaccine.
Theparameterswerevariedby25%.
Darkbarsshowtheincrementalcosteffectivenessratioaftera25%decreaseintheparameter,whereaslightbarsshowtheincrementalcosteffectivenessratioaftera25%increase(notethatitwasnotpossibletoincreasevaccineefficacy).
QALY,qualityadjustedlifeyearRESEARCHBMJ|ONLINEFIRST|bmj.
compage7of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfromprogrammeforallchildrenagedlessthan2years.
IntheNetherlands,bycontrast,vaccinationwasimple-mentedwithoutacatch-upprogramme.
Severalalter-nativescenariosregardingserotypereplacementwereexploredinthesensitivityanalyses,whichshowedthatourconclusionsregardingtheincrementalcosteffec-tivenessratiosforallthreevaccineswerequiterobust.
InourbasecaseanalysisforPCV-7,weassumedthattherewasnonetindirecteffectofvaccinationforindi-vidualsoutsidethemodelledbirthcohortbecausenooverallreductionininvasivepneumococcaldiseaseinnon-vaccineeshasbeenobservedinanyEuropeancountry,incontrasttotheUS.
121315ThedifferencebetweenresultsobtainedintheUSandthoserecordedinEuropemaybepartlyexplainedbythe60%to70%coverageofthesevenvaccineserotypesinEurope,comparedwiththemorethan80%coverageintheUS.
32ThisdisparityleavesmoreroomforreplacementdiseaseinEurope.
Countryspecificdifferencesinthecirculatingserotypescausingdisease(inclusiveofsecu-larchangesintime)couldalsocontributetotheloweroverallreductionofinvasivepneumococcaldiseaseinEuropecomparedwiththeUS.
33Furthermore,intheNetherlands,asinmostpartsofEurope,thebaselineincidenceratesofinvasivepneumococcaldiseaseinchildrenaresubstantiallylowerthanintheUSandalmostexclusivelybasedoncultureconfirmedcasesofchildrenadmittedtohospital.
1734Anotherpoten-tiallyrelevantdifferenceintheintroductionofPCV-7intheNetherlandscomparedwiththeUSisthehighvaccineuptake(>95%)amongallnewbornsintheNetherlandsforallfourdosesofthevaccination,whichcouldpotentiallyleadtomorerapiddevelop-mentofreplacementdisease.
34Potentialnetindirecteffectsinnon-vaccineesweremodelledusingstraightforwardcalculus.
Ideally,theimpactofpneumococcalvaccinationshouldhavebeenmodelledusingasocalleddynamictransmissionmodel,inwhichthetransmissionandcarriageofStrep-tococcuspneumoniaeistakenexplicitlyintoaccount.
However,becausethetransitiondynamicsofSpneu-moniaearecomplexandserotypedependent,anddetaileddataregardingthesetransmissiondynamicsarealsoquitelimited,dynamicallymodellingallrele-vantserotypesofSpneumoniaewouldbeverycompli-cated.
ForPCV-10andPCV-13,anetindirecteffectof10%wasincludedinthebasecaseanalysis.
Thisesti-mateofindirectbenefitmaybeconservativeifcom-paredwiththemuchhighernetindirectprotectivebenefitsobservedintheUSafterimplementationofroutinevaccinationwithsimilarorlowervaccineser-otypecoverage.
2-4Furthermore,wedidnotincludethebenefitsarisingfromthepreventionofantibioticresistanceinourmodelbecausetheimpactofthisinclusionisexpectedtobesmallgiventhatpenicillinresistanceislessthan0.
4%intheNetherlands.
17Finally,similartoalmostallpreviouscosteffectivenessanalysesforpneumococcalvaccination,ouranalytictimeframewasequaltotheassumedprotectionperiod,afterwhichweassumedthathealtheffectsandcostswouldbesimilarinthevaccinatedandunvaccinatedgroup.
ComparisonwithotherstudiesThecosteffectivenessofPCV-7isworsethanthatcal-culatedinourpreviousstudiesandinotherrecenthealtheconomicstudies.
5-10161835Thisdisparityismostlybecauseoftheexclusionofherdprotectioneffectsandtheinclusionofserotypereplacementinourstudy.
Otherfactorscontributingtotheworseincrementalcosteffectivenessratioweretheuseofalowerdeathrateforinvasivepneumococcaldiseaseandlowerindirectcoststhaninourpreviousstudies.
616Severalrecentlypublishedcosteffectivenessstudiesincludednetvaccinebenefitsforunvaccinatedadultsandelderlypeopleintheirbasecaseanalysis.
5-7935Thesestudiesreportedvaccinationtobecostsaving59oratleastcosteffective.
6735ThethreestudiesthatexcludedherdprotectioninthebasecaseanalysisreportedrelativelyunfavourablecosteffectivenessratiosforPCV-7comparedwithotherrecommendedinfantvaccinations.
81618Whenweexcludedtheincreaseininvasivepneumococcaldiseasecausedbynon-vaccineserotypesbutleftallotherassumptionsthesameasinthebasecaseanalysis,ourresultsweresimilartothoseofthesethreestudies—thatis,wefoundanunfavourablecosteffectivenessratio.
81618OurcosteffectivenessresultsshowthatthecurrentvaccinationscheduleforPCV-7mightbefarmoreexpensiveperQALYgainedcomparedwithotherrou-tineinfantvaccinationprogrammesrecentlyimple-mented,suchasforhumanpapillomavirus36,orwithothervaccinesthathavenotyetbeenimplementedinanationalprogrammeintheNetherlands,suchashepa-titisB37andvaricella.
38ImplicationsandfutureresearchAdministrationofPCV-7at2,3,4,and11monthswasintroducedtotheNetherlandsaspartofthenationalimmunisationprogrammein2006partiallyonthebasisoffavourablecosteffectivenessdata.
ThecurrentanalysisshowsunfavourablecosteffectivenessoftheNetindirectbenefits(%)040000600008000010000012000020000051015202530PCV-10PCV-13PCV-7Costeffectivenessratio(/QALY)Fig3|Theeffectoncosteffectivenessratiosofvaryingthelevelofnetindirecteffectofvaccinationforindividualsaged5yearsorolder.
Thehorizontaldashedlineshowsthethresholdat50000perQALY.
PCV-7/10/13,seven/10/13valentpneumococcalconjugatedvaccine(3+1doseschedule);QALY,qualityadjustedlifeyearRESEARCHpage8of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
DownloadedfromPCV-73+1doseschedulebecauseofincreasesininva-sivediseasecausedbynon-vaccineserotypes,whichoffsettheherdprotectivebenefitsinindividualsout-sidetheanalysedbirthcohort.
Althoughthecosteffec-tivenessofPCV-7isunfavourablefromahealtheconomicspointofview,itisfavourablefromapublichealthpointofview—asignificantdecreaseincasesofpneumococcaldiseasehasoccurredintheNetherlandsoverthepasttwoyears.
13Switchingtothe10valentor13valentvaccinewouldextendtheserotypecoveragetoahigherlevelthanthatcurrentlyachievedwithPCV-7,whichmightreducethepotentialfordiseasecausedbynon-vaccineserotypesandincreasetheoverallbenefitsinvaccinatedchildren.
Herdprotectiveeffectsaremorelikelytooccurwithbroadvaccinecoverage,renderingvaccinationpoten-tiallycosteffective.
Vaccinationwouldbeparticularlycosteffectiveifamorevalentvaccineisusedincombi-nationwithdosereductions,pricereductions,orboth.
OurpapershouldhelpguidefuturedecisionstopotentiallyreducedosesofpneumococcalvaccineortoshiftfromPCV-7tovaccinesthatcoveradditionalserotypes.
Furtherresearchshouldbedirectedtobuildingadynamicmodeltoentangleandexplicitlypredicttheindirecteffectsofdiseasereplacementandherdprotectiononvaccineefficacyandthusfurtherenhancethevalidityofcosteffectivenessapproachesappliedtopneumococcalvaccination.
Contributors:MJP,EH,andGvdDdesignedthestudy.
MHR,AJvH,andMJPdesignedthecomputermodelandcarriedoutthecomputersimulationsandanalysis.
DataanalyseswereperformedbyAGSCJ,GDR,andMHRundersupervisionofEAMS,MJP,andAvdE.
MHR,MJP,AJvH,andEAMSdraftedthemanuscript.
Allauthorscommentedondraftsandcontributedtothefinalversion.
MHRandMJParetheguarantorsofthestudy.
Funding:MHRwasfundedbyanunrestrictedgrantfromWyethHoofddorp.
AJvHwasfinancedbytheNetherlandsVaccineInstitute,Bilthoven.
ThisworkhasbeenpreviouslypresentedataworkshoponpneumococcalvaccinesattheEuropeanPublicHealthAssociationconferenceinLisbon,Portugal,whichwassupportedbyaresearchgrantfromGlaxoSmithKlineNetherlands.
Theauthors'workwasindependentofthefunders,whohadnoroleinthestudydesign,analysisofdata,writingofthemanuscript,ordecisiontosubmitforpublication.
Competinginterests:AllauthorshavecompletedtheUnifiedCompetingInterestformatwww.
icmje.
org/coi_disclosure.
pdf(availableonrequestfromthecorrespondingauthor)anddeclare:(1)MHRwasfundedbyanunrestrictedgrantfromWyethHoofddorp;andAJvHwasfinancedbytheNetherlandsVaccineInstitute,Bilthoven;(2)MJPhasreceivedtravelgrantsfromGlaxoSmithKlineandWyethtoattendexpertmeetingsinReykjavik,Iceland,andIstanbul,Turkey;EAMShasreceivedunrestrictedgrantsfromWyethandBaxterforresearch,consultingfeesfromWyethandGlaxoSmithKline,lecturingfeesfromWyeth,andgrantsupportfromWyethandGlaxoSmithKlineforvaccinestudies;andAvdEhasreceivedunrestrictedgrantsfromWyethandNovartis;(3)Nospouses,partners,orchildrenwithrelationshipswithcommercialentitiesthatmighthaveaninterestinthesubmittedwork;(4)Nonon-financialintereststhatmayberelevanttothesubmittedwork.
Datasharing:Noadditionaldataavailable.
1HealthCounciloftheNetherlands.
Vaccinationofinfantsagainstpneumococcalinfections[inDutch].
2005.
www.
gezondheidsraad.
nl/sites/default/files/05@13N.
pdf.
2HsuHE,ShuttKA,MooreMR,BeallBW,BennettNM,CraigAS,etal.
Effectofpneumococcalconjugatevaccineonpneumococcalmeningitis.
NEnglJMed2009;360:244-56.
3LexauCA,LynfieldR,DanilaR,PilishviliT,FacklamR,FarleyMM,etal.
Changingepidemiologyofinvasivepneumococcaldiseaseamongolderadultsintheeraofpediatricpneumococcalconjugatevaccine.
JAMA2005;294:2043-51.
4WhitneyCG,FarleyMM,HadlerJ,HarrisonLH,BennettNM,LynfieldR,etal.
Declineininvasivepneumococcaldiseaseaftertheintroductionofprotein-polysaccharideconjugatevaccine.
NEnglJMed2003;348:1737-46.
5ClaesC,ReinertRR,vonderSchulenburgJM.
CosteffectivenessanalysisofheptavalentpneumococcalconjugatevaccineinGermanyconsideringherdimmunityeffects.
EurJHealthEcon2009;10:25-38.
6HubbenGA,BosJM,GlynnDM,vanderEndeA,vanAlphenL,PostmaMJ.
Enhanceddecisionsupportforpolicymakersusingawebinterfacetohealth-economicmodels—illustratedwithacost-effectivenessanalysisofnation-wideinfantvaccinationwiththe7-valentpneumococcalconjugatevaccineintheNetherlands.
Vaccine2007;25:3669-78.
7McIntoshED,ConwayP,WillinghamJ,HollingsworthR,LloydA.
PneumococcalpneumoniaintheUK—howherdimmunityaffectsthecost-effectivenessof7-valentpneumococcalconjugatevaccine(PCV).
Vaccine2005;23:1739-45.
8MelegaroA,EdmundsWJ.
Cost-effectivenessanalysisofpneumococcalconjugatevaccinationinEnglandandWales.
Vaccine2004;22:4203-14.
9SilfverdalSA,BergS,HemlinC,JokinenI.
Thecost-burdenofpaediatricpneumococcaldiseaseinSwedenandthepotentialcost-effectivenessofpreventionusing7-valentpneumococcalvaccine.
Vaccine2009;27:1601-8.
10TilsonL,UsherC,ButlerK,FitzsimonsJ,O'HareF,CotterS,etal.
EconomicevaluationofauniversalchildhoodpneumococcalconjugatevaccinationstrategyinIreland.
ValueHealth2008;11:898-903.
WHATISALREADYKNOWNONTHISTOPICRecentpneumococcalsurveillancestudiesshowasignificantincreaseinnon-vaccineserotypedisease,whichreducestheoverallhealtheffectsofvaccinationandoffsetspotentialpositiveherdprotectionbenefitsinunvaccinatedindividualsExistingeconomicanalysesofpneumococcalvaccinationdidnotincorporatetheincreaseindiseasecausedbynon-vaccineserotypes,andmostincludedtoooptimisticherdprotectiveeffectsWHATTHISSTUDYADDSThecurrentfourdose(3+1)scheduleofthesevenvalentpneumococcalconjugatedvaccineusedintheNetherlandsisnotcosteffectiveVaccinationwith10valentor13valentvaccinescouldsubstantiallyreducetheincrementalcosteffectivenessofpneumococcalvaccinationtoapotentiallyacceptablelevelReducingthetotalprogrammecostforpneumococcalvaccinationbyreducingdoseschedules,reducingvaccineprices,orbothisnecessarytounambiguouslyrenderroutineinfantvaccinationcosteffectiveintheNetherlandsPCV-7basecase(3+1doseschedule)PCV-725%reductioninvaccinationcostsPCV-10basecase(3+1doseschedule)PCV-1025%reductioninvaccinationcostsPCV-13basecase(3+1doseschedule)PCV-1325%reductioninvaccinationcostsThreshold(¤/QALY)Proportionbelowthreshold(%)04000080000120000160000200000020406080100Fig4|CosteffectivenessacceptabilitycurvesforbasecasevaccinationschedulesandforalternativescenariosforPCV-7,PCV-10,andPCV-13RESEARCHBMJ|ONLINEFIRST|bmj.
compage9of10on9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom11VestrheimDF,LovollO,AabergeIS,CaugnantDA,HolbyEA,BakkeH,etal.
Effectivenessofa2+1doseschedulepneumococcalconjugatevaccinationprogrammeoninvasivepneumococcaldiseaseamongchildreninNorway.
Vaccine2008;26:3277-81.
12HealthProtectionAgency.
CumulativeweeklynumberofreportsofinvasivepneumococcaldiseaseinEnglandandWalesbyepidemiologicalyearJuly-June.
2009.
www.
hpa.
org.
uk.
13RodenburgGD,deGreeffSC,JansenAG,deMelkerHE,SchoulsLM,HakE,etal.
Effectsofpneumococcalconjugatevaccine2yearsafteritsintroduction,theNetherlands.
EmergInfectDis2010;16:816-23.
14Munoz-AlmagroC,JordanI,GeneA,LatorreC,Garcia-GarciaJJ,PallaresR.
Emergenceofinvasivepneumococcaldiseasecausedbynonvaccineserotypesintheeraof7-valentconjugatevaccine.
ClinInfectDis2008;46:174-82.
15LepoutreA,VaronE,GeorgesS,GutmannL,Levy-BruhlD.
ImpactofinfantpneumococcalvaccinationoninvasivepneumococcaldiseasesinFrance,2001-2006.
EuroSurveill2008;13:ii,18962.
16BosJM,RumkeH,WelteR,PostmaMJ.
Epidemiologicimpactandcost-effectivenessofuniversalinfantvaccinationwitha7-valentconjugatedpneumococcalvaccineintheNetherlands.
ClinTher2003;25:2614-30.
17JansenAGSC,RodenburgGD,GreefSC,HakE,VeenhovenRH,SpanjaardL,etal.
InvasivepneumococcaldiseaseintheNetherlands:syndromes,outcomeandpotentialvaccinebenefits.
Vaccine2009;27:2394-401.
18Giorgi-RossiP,MeritoM,BorgiaP.
Cost-effectivenessofintroducingtheconjugatedpneumococcalvaccinetoroutinefreeimmunizationsforinfantsinLazio,Italy.
HealthPolicy2009;89:225-38.
19BlackS,ShinefieldH,FiremanB,LewisE,RayP,HansenJR,etal.
Efficacy,safetyandimmunogenicityofheptavalentpneumococcalconjugatevaccineinchildren.
NorthernCaliforniaKaiserPermanenteVaccineStudyCenterGroup.
PediatrInfectDisJ2000;19:187-95.
20MadhiSA,AdrianP,KuwandaL,JassatW,JonesS,LittleT,etal.
Long-termimmunogenicityandefficacyofa9-valentconjugatepneumococcalvaccineinhumanimmunodeficientvirusinfectedandnon-infectedchildrenintheabsenceofaboosterdoseofvaccine.
Vaccine2007;25:2451-7.
21HansenJ,BlackS,ShinefieldH,CherianT,BensonJ,FiremanB,etal.
Effectivenessofheptavalentpneumococcalconjugatevaccineinchildrenyoungerthan5yearsofageforpreventionofpneumonia:updatedanalysisusingWorldHealthOrganizationstandardizedinterpretationofchestradiographs.
PediatrInfectDisJ2006;25:779-81.
22FiremanB,BlackSB,ShinefieldHR,LeeJ,LewisE,RayP.
Impactofthepneumococcalconjugatevaccineonotitismedia.
PediatrInfectDisJ2003;22:10-6.
23BlackSB,ShinefieldHR,LingS,HansenJ,FiremanB,SpringD,etal.
Effectivenessofheptavalentpneumococcalconjugatevaccineinchildrenyoungerthanfiveyearsofageforpreventionofpneumonia.
PediatrInfectDisJ2002;21:810-5.
24EskolaJ,KilpiT,PalmuA,JokinenJ,HaapakoskiJ,HervaE,etal.
Efficacyofapneumococcalconjugatevaccineagainstacuteotitismedia.
NEnglJMed2001;344:403-9.
25NelsonJC,JacksonM,YuO,WhitneyCG,BoundsL,BittnerR,etal.
Impactoftheintroductionofpneumococcalconjugatevaccineonratesofcommunityacquiredpneumoniainchildrenandadults.
Vaccine2008;26:4947-54.
26CentersforDiseaseControlandPrevention.
Pneumoniahospitalizationsamongyoungchildrenbeforeandafterintroductionofpneumococcalconjugatevaccine—UnitedStates,1997-2006.
MMWRMorbMortalWklyRep2009;58:1-4.
27BrouwerWB,NiessenLW,PostmaMJ,RuttenFF.
Needfordifferentialdiscountingofcostsandhealtheffectsincosteffectivenessanalyses.
BMJ2005;331:446-8.
28GoldblattD,SouthernJ,AshtonL,RichmondP,BurbidgeP,TasevskaJ,etal.
Immunogenicityandboostingafterareducednumberofdosesofapneumococcalconjugatevaccineininfantsandtoddlers.
PediatrInfectDisJ2006;25:312-9.
29WhitneyCG,PilishviliT,FarleyMM,SchaffnerW,CraigAS,LynfieldR,etal.
Effectivenessofseven-valentpneumococcalconjugatevaccineagainstinvasivepneumococcaldisease:amatchedcase-controlstudy.
Lancet2006;368:1495-502.
30VespaG,ConstenlaDO,PepeC,SafadiMA,BerezinE,CassiodeMoraesJ,etal.
Estimatingthecost-effectivenessofpneumococcalconjugatevaccinationinBrazil.
RevPanamSaludPublica2009;26:518-28.
31KayhtyH,AhmanH,ErikssonK,SorbergM,NilssonL.
Immunogenicityandtolerabilityofaheptavalentpneumococcalconjugatevaccineadministeredat3,5and12monthsofage.
PediatrInfectDisJ2005;24:108-14.
32HausdorffWP,BryantJ,ParadisoPR,SiberGR.
Whichpneumococcalserogroupscausethemostinvasivedisease:implicationsforconjugatevaccineformulationanduse,partI.
ClinInfectDis2000;30:100-21.
33FenollA,GranizoJJ,AguilarL,GimenezMJ,Aragoneses-FenollL,HanquetG,etal.
TemporaltrendsofinvasiveStreptococcuspneumoniaeserotypesandantimicrobialresistancepatternsinSpainfrom1979to2007.
JClinMicrobiol2009;47:1012-20.
34CentersforDiseaseControlandPrevention.
Invasivepneumococcaldiseaseinchildren5yearsafterconjugatevaccineintroduction—eightstates,1998-2005.
MMWRMorbMortalWklyRep2008;57:144-8.
35RayGT,WhitneyCG,FiremanBH,CiurylaV,BlackSB.
Cost-effectivenessofpneumococcalconjugatevaccine:evidencefromthefirst5yearsofuseintheUnitedStatesincorporatingherdeffects.
PediatrInfectDisJ2006;25:494-501.
36JitM,ChoiYH,EdmundsWJ.
EconomicevaluationofhumanpapillomavirusvaccinationintheUnitedKingdom.
BMJ2008;337:a769.
37TilsonL,ThorntonL,O'FlanaganD,JohnsonH,BarryM.
CosteffectivenessofhepatitisBvaccinationstrategiesinIreland:aneconomicevaluation.
EurJPublicHealth2008;18:275-82.
38RozenbaumMH,vanHoekAJ,VegterS,PostmaMJ.
Cost-effectivenessofvaricellavaccinationprograms:anupdateoftheliterature.
ExpertRevVaccines2008;7:753-82.
39OostenbrinkJB,BouwmansCAM,KoopmanschapMA,RuttenFHH.
Guidelinesforcostingresearch,methodsandstandardizedpricesforeconomicevaluationsinhealthcare.
DutchHealthCareInsuranceBoard,2004.
40ProsserLA,RayGT,O'BrienM,KleinmanK,SantoliJ,LieuTA.
Preferencesandwillingnesstopayforhealthstatespreventedbypneumococcalconjugatevaccine.
Pediatrics2004;113:283-90.
41PrymulaR,PeetersP,ChrobokV,KrizP,NovakovaE,KaliskovaE,etal.
PneumococcalcapsularpolysaccharidesconjugatedtoproteinDforpreventionofacuteotitismediacausedbybothStreptococcuspneumoniaeandnon-typableHaemophilusinfluenzae:arandomiseddouble-blindefficacystudy.
Lancet2006;367:740-8.
Accepted:16February2010RESEARCHpage10of10BMJ|ONLINEFIRST|bmj.
comon9March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
c2509on2June2010.
Downloadedfrom
- primarynetlife相关文档
- solvenetlife
- necessarilynetlife
- Simonnetlife
- Kooperationnetlife
- propensitynetlife
- modellingnetlife
hypervmart：英国/荷兰vps,2核/3GB内存/25GB NVMe空间/不限流量/1Gbps端口/Hyper-V,$10.97/季
hypervmart怎么样?hypervmart是一家国外主机商，成立于2011年，提供虚拟主机、VPS等，vps基于Hyper-V 2012 R2，宣称不超售，支持linux和windows，有荷兰和英国2个数据中心，特色是1Gbps带宽、不限流量。现在配置提高，价格不变，性价比提高了很多。(数据中心不太清楚，按以前的记录，应该是欧洲)，支持Paypal付款。点击进入：hypervmart官方网...
Digital-VM：服务器,$80/月;挪威/丹麦英国/Digital-VM:日本/新加坡/digital-vm：日本VPS仅$2.4/月
digital-vm怎么样？digital-vm在今年1月份就新增了日本、新加坡独立服务器业务，但是不知为何，期间终止了销售日本服务器和新加坡服务器，今天无意中在webhostingtalk论坛看到Digital-VM在发日本和新加坡独立服务器销售信息。服务器硬件是 Supermicro、采用最新一代 Intel CPU、DDR4 RAM 和 Enterprise Samsung SSD内存，默认...
江苏云服务器 2H2G 20M 79元/月 大宽带159元/月 高性能挂机宝6元/月 香港CN2 GIA、美国200G防御 CN2 GIA 折后18元/月 御速云
介绍：御速云成立于2021年的国人商家，深圳市御速信息技术有限公司旗下品牌，为您提供安全可靠的弹性计算服务，随着业务需求的变化，您可以实时扩展或缩减计算资源，使用弹性云计算可以极大降低您的软硬件采购成本，简化IT运维工作。主要从事VPS、虚拟主机、CDN等云计算产品业务，适合建站、新手上车的值得选择，拥有华东江苏、华东山东等国内优质云产品；香港三网直连（电信CN2GIA联通移动CN2直连）；美国高...
netlife为你推荐
-
金评媒朱江请问朱江恺撒堡KX系列的钢琴怎么样？12306崩溃亲们，为什么12306手机订票系统打不开，显示网络异常，留学生认证留学生服务中心认证内容和范围？psbc.com怎样登录wap.psbc.com曹谷兰曹谷兰事件 有吧友知道吗www.bbb551.com广州欢乐在线551要收费吗？www.hyyan.comDOTA6.51新手选什么英雄为好，请详细讲述出装备顺序，加点顺序，以及注意事项。谢谢www.zhiboba.com上什么网看哪个电视台直播NBAdadi.tv电视机如何从iptv转换成tv？关键词分析如何进行关键词指数分析