/
Age Adjustment Using the  Projected U Age Adjustment Using the  Projected U

Age Adjustment Using the Projected U - PDF document

mitsue-stanley
mitsue-stanley . @mitsue-stanley
Follow
478 views
Uploaded On 2015-04-23

Age Adjustment Using the Projected U - PPT Presentation

S Population Richard J Klein MPH and Charlotte A Schoenborn MPH Introduction Age adjustment using the direct method is the application of observed agespeci64257c rates to a standard age distribution to eliminate differences in crude rates ID: 54127

Population Richard Klein

Share:

Link:

Embed:

Download Presentation from below link

Download Pdf The PPT/PDF document "Age Adjustment Using the Projected U" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

AgeAdjustmentUsingthe2000ProjectedU.S.PopulationRichardJ.Klein,M.P.H.,andCharlotteA.Schoenborn,M.P.H.Ageadjustment,usingthedirectmethod,istheapplicationofobservedage-speciÞcratestoastandardagedistributiontoeliminatedifferencesincruderatesinpopulationsofinterestthatresultfromdifferencesinthepopulationsÕagedistributions.Thisadjustmentisusuallydonewhencomparingtwoormorepopulationsatonepointintimeoronepopulationattwoormorepointsintime.Ageadjustmentisparticularlyrelevantwhenpopulationsbeingcomparedhavedifferentagestructures,forexample,theU.S.whiteandHispanicpopulations.Theclassicliteratureonageadjusting,aswellasmorerecentNationalCenterforHealthStatistics(NCHS)publications,hasfocusedonadjustingdeathratesandprovidescomprehensivediscussionsofage-adjustmenttechniques.However,ageadjustmentcanbeappliedtoanypopulation-basedevent.Thisreportdescribesseveralsetsofage-adjustmentweights,basedontheyear2000projectedU.S.population,thatwereusedtogenerateage-adjustedbaselinedataforanumberofHealthyPeople2010objectives.ThisworkbuildsonthefoundationsetbytheDepartmentofHealthandHumanServices(DHHS),whichestablishedtheyear2000projectedU.S.populationasthestandardpopulationforageadjustingmortalitystatistics.ThesetsofweightsprovidedinthispaperusetheagegroupingsmostcommonlyfoundinexistingDHHSpublications.Age-adjustmentweights,aswellasexamplesofcomputercodesappropriateforcalculatingage-adjustedratesforcomplexsamplesurveysusingtheSUDAANsoftwarepackage,areprovidedinthisreport.TheweightsandproceduresshownarenotintendedasÞxedrulesforageadjustment,butratherasguidelinestopromoteandfacilitateconsistencyandcomparabilityinage-adjustmentproceduresamongusersofhealth-relateddata.HistoricalPerspectiveSeveralyearsago,NCHScoordinatedaneffortthatresultedinanagreementamongcertainFederalandStateagenciestoageadjustmortalitydatausingtheyear2000projectedU.S.population.InAugust1998,theSecretaryoftheDepartmentofHealthandHumanServices(DHHS) TheauthorsgratefullyacknowledgetheeffortsofVirginiaM.Freid(DivisionofHealthandUtilizationAnalysis)whowasamajorcontributortotheplanningofthisreportandDavidA.Fessler(asummerinternwiththeDivisionofHealthPromotionStatistics)whocontributedsubstantiallytotheproductionofthedatatablesforthisreport. Number20January2001 issuedapolicystatementdirectingallDHHSagenciestousetheyear2000projectedU.S.population(hereafterreferredtoasthe2000StandardPopulation)forageadjustingdeathratesbeginningnolaterthandatayear1999.agreementresolvedlong-standingproblemsassociatedwiththeuseofdifferentstandardpopulationsforageadjustingdeathrates(U.S.1940population,U.S.1970population,U.S.1980population,etc.)byvariousagenciespublishingmortalitystatistics,andtheresultantdifficultiesexperiencedbydatausersincomparingtheserates.Thedecisiontoageadjustmortalitydatausingthe2000StandardPopulationfocusedonmortalitydatafromtheNationalVitalStatisticsSystemand,therefore,wasindependentofotherNCHSdatasystems(e.g.,NationalHealthInterviewSurvey,NationalHealthandNutritionExaminationSurvey,NationalHospitalDischargeSurvey,NationalNursingHomeSurvey,NationalAmbulatoryMedicalCareSurvey,NationalHospitalMedicalCareSurvey,andNationalHospiceandHomeCareSurvey).Dataforthosesystemshavebeenageadjustedusingavarietyofstandardpopulationssuchas1970U.S.civiliannoninstitutionalizedpopulationand1980U.S.residentpopulation,whichcanbefoundinappendixIIofUnitedStates,1999.Toimprovecomparabilityofage-adjustedratesamongdatasystems,thesedataforNCHSsurveysarebeingrecalculatedusingthe2000StandardPopulationinHealth,UnitedStates,2000UnitedStates,2001.Inaddition,beginningwith1999datafortheChronicDiseaseIndicators,CDCwillbeproducingage-adjusteddataforStatesfromtheBehavioralRiskFactorSurveillanceSystemusingthe2000StandardPopulation.CalculationofAge-AdjustmentWeights:Background,Process,andTheprocessofderivingage-adjustmentweightsforsurveyandotherpopulation-baseddatabeganbyidentifyingagecategoriestypicallyusedforageadjustmentandtryingtoensurethatallthevariousgroupingscouldbeconstructedfromasingleMasterListtable1ThesourceofthepopulationguresintheMasterLististheofficialBureauoftheCensusprojectionsfortheyear2000publishedinCurrentPopulationReports,seriesP-25,no.1130,table2.Itisthesamepublicationonwhichthestandardpopulationformortalitydataisbased.TheCensusprojectionsaredisaggregatedbysexandrace.However,onlythetotalsforbothsexesandallracescombinedwereusedtodeveloptheMasterList.Thecomplete2000populationprojectionsareavailablefromtheBureauoftheCensuswebsiteat:.census.gov/prod/1/pop/p25-TheMasterList,consistingof24agegroups,wasdevelopedtoreducediscrepanciesthatcanoccurwhenextractingandcombiningnumbersfromthepublishedCensusprojections.Theseprojectionsforthe2000population,bothforsingleyearsofageandfor5-yearagegroups,areroundedtothousands.However,populationcountsforthe5-yearagegroupswerecalculatedbysummingtheunroundedcountsforsingleyears(unpublished)andthenroundingthatsum.Asaconsequence,publishedsumsofthesingleyearsofagedonotnecessarilyaddtothepublished5-yeartotals,andweightsbasedonthepublishedCensusprojectionsmaybeslightlydifferentdependingonwhethersingle-yearor5-yearagegroupsarecombinedtoproducetheweights.ByningtheMasterListasthestartingpointforderivingcpopulationestimatesandage-adjustmentweights,suchdiscrepanciesareminimized.OncetheMasterListwasestablished,weightswerecomputedforanumberofcommonlyusedagedistributions.Theweightforagivenagegroupisthesizeofthecorrespondingagegroupinthestandardpopulationdividedbythetotalstandardpopulationforallagesofinterest.Table2shows22differentpopulationagedistributionsusedincalculatingbaselinedatafromvariousmajordatasystemsfortheHealthyPeople2010objectivesthatuseageadjustment.MajordatasystemsarethosethatareresponsiblefortrackingveormoreHealthyPeople2010objectives.TableAliststhedistributionsthatwereusedbythemajordatasystems.AgeDistributions#4and#18eachappearintwodifferentdatasystemgroups.Thespecicgroupingsuseddependontheagegroupscoveredbytheobjectiveandonthenumberofeventsincludedinthemeasureusedfortheobjective.Thesegroupingscanbeusedasageneral Table1.Masterlist:2000U.S.projectedpopulationandage-adjustmentweightsinthousandsAllages274,6341.000000Under1year3,7950.0138181year3,7590.0136872–4years.................11,4330.0416305years3,8960.0141866–8years.................11,8000.0429669years4,2240.01538010–11years8,2580.03006912–14years................11,7990.04296315–17years................11,8190.04303518–19years8,0010.02913320–24years18,2570.06647825–29years17,7220.06453030–34years19,5110.07104435–39years22,1800.08076240–44years22,4790.08185145–49years19,8060.07211850–54years17,2240.06271655–59years13,3070.04845460–64years10,6540.03879365–69years9,4100.03426470–74years8,7260.03177375–79years7,4150.02700080–84years4,9000.01784285yearsandover4,2590.015508 NOTE:WhenusingtheNCHSMasterListtoobtainadditionalagegroupsnotshownintable2,theage-adjustmentweightsshouldberecalculatedusingtheappropriatedenominatorandmustaddto1(see‘‘CalculationofAge-AdjustmentWeights:Background,Process,andProcedures’’). Table2:Selectedagedistributionsandage-adjustmentweightsbasedonthe2000projectedU.S.populationinthousandsweightAgeinthousandsDistribution#1Distribution#2Allages274,6341.000000Allages274,6341.000000Under1year3,7950.013818Under12years47,1650.1717384years15,1920.0553171219years31,6190.11513114years39,9770.1455652029years35,9790.13100724years38,0770.1386463039years41,6910.15180634years37,2330.1355734049years42,2850.15396844years44,6590.1626135059years30,5310.11117054years37,0300.1348346069years20,0640.07305764years23,9610.0872477079years16,1410.05877374years18,1360.06603780yearsandover9,1590.03335084years12,3150.04484285yearsandover4,2590.015508Distribution#3Distribution#4Allages274,6341.000000Allages274,6341.000000Under18years70,7830.257736Under18years70,7830.25773644years108,1500.3937971844years108,1500.39379754years37,0300.1348344564years60,9910.22208164years23,9610.0872476574years18,1360.06603774years18,1360.06603775yearsandover16,5740.06034975yearsandover16,5740.060349Distribution#5Distribution#62yearsandover267,0801.0000002yearsandover267,0801.0000005years15,3290.057395217years63,2290.23674211years24,2820.0909171844years108,1500.40493519years31,6190.1183884554years37,0300.13864729years35,9790.1347125564years23,9610.08971539years41,6910.1560996574years18,1360.06790549years42,2850.15832375yearsandover16,5740.06205659years30,5310.11431469years20,0640.07512479years16,1410.06043580yearsandover9,1590.034293Distribution#7Distribution#812yearsandover227,4691.00000018yearsandover203,8511.00000019years31,6190.1390041824years26,2580.12881029years35,9790.1581712544years81,8920.40172539years41,6910.1832824564years60,9910.29919449years42,2850.18589365yearsandover34,7100.17027159years30,5310.13422169years20,0640.08820579years16,1410.07095980yearsandover9,1590.040265Distribution#9Distribution#1018yearsandover203,8511.00000018yearsandover203,8511.00000024years26,2580.1288101829years43,9800.21574634years37,2330.1826483039years41,6910.20451744years44,6590.2190774049years42,2850.20743164years60,9910.2991945059years30,5310.14977165yearsandover34,7100.1702716069years20,0640.09842579years16,1410.07918080yearsandover9,1590.044930Distribution#11Distribution#1220yearsandover195,8501.00000020yearsandover195,8501.00000029years35,9790.1837072039years77,6700.39657939years41,6910.2128724059years72,8160.37179549years42,2850.21590560yearsandover45,3640.23162659years30,5310.15589069years20,0640.10244679years16,1410.08241580yearsandover9,1590.046765 guideforresearchersdevelopingmeasurestotrackhealthvariables.Theactualchoiceofagegroupingstobeusedforageadjusting,however,isananalyticdecisionandis,ofcourse,notlimitedtotheagedistributionsshownintable2Whencreatingthe22agedistributionsshownintable2theage-specicpopulationcountswereconstructedtobeconsistentwiththepublishedCensusprojections.Insomecases,minoradjustmentstotheage-adjustmentweightswerenecessaryfortheweightstosumtoone.WiththeexceptionofAgeDistribution#1(describedbelow),age-speciccountsfortheagedistributionsshownintable2werederiveddirectlyfromtheMasterList(table1).Thetotalpopulationforeachdistributionwasobtainedbysummingthecomponentage-specicpopulationcounts.Age-adjustmentweightsforeachoftheagedistributionswereobtainedbydividingtheage-speciccountsbythetotalpopulationoverallagesinthatagedistribution,roundingtosixdecimalplaces,andmakinganynecessaryadjustments(wheretheimpactwouldbetheleast)sothattheweightssummedto1(roundedtosixdecimalsplaces:1.000000.)Notethatregardlessofwhethertheagedistributioncoversallagesorasubsetofallages(e.g.,adults,children,elderly),theweightsarerenormalizedtoalwayssumto1.AgeDistribution#1,shownintable2,consistsofagegroupsandage-adjustmentweightspreviouslypublishedbytheNCHSDivisionofVitalStatisticsandusedforageadjustingmortalitystatistics.Anerratumsheetwaspublishedthatshowstheweightforthe7584year-oldgroupas0.044842,insteadof0.044841initiallyprinted.Thisminordifferenceisanexampleoftheadjustmentsthatmustbemadetotheactualweightfortheweightstosumto1.Theactualcalculatedweightwas0.044841498,whichwhenroundedtosixdecimalplaces,was0.044841.However,usingthisweight,thesumofalloftheweightsfellshortof1.Aftermakingaminormodicationthatincreasedtheweightto0.044842,theage-adjustmentweightssummedto1.Asimilarprocedurewasusedtoobtaintheweightsforveagedistributionspresentedthatrequiredsmalladjustmentssothattheweightssummedto1(seetableBInallcases,themodication(eitherincreasingordecreasingthesixthdecimalplacesothattheweightswouldsumto1)wasmadetotheagegroupwheretheimpactontheage-adjustmentcalculationwouldbetheleast.Nocationswererequiredforanyotheragegroups.Itshouldbenotedthatthenumberofagegroupsusedforageadjustingmayhaveasmallimpactontheleveloftheage-adjustedrate.Forexample,AgeDistributions#8and#9intable2arebothforadults18yearsandolder.However,AgeDistribution#9usesveagegroups(providingmoredetailinthe2544year-oldagegroup)whileAgeDistribution#8usesfouragegroups.Ageadjustmentusingoneofthesedistributionswouldyieldslightlydifferentestimatesfromtheother.Inchoosingthenumberofagegroupsforaspecianalysis,twocompetinginterestsneedtobeaddressed.First, Table2:Selectedagedistributionsandage-adjustmentweightsbasedonthe2000projectedU.S.population—Con.inthousandsweightAgeinthousandsDistribution#13Distribution#1420yearsandover195,8501.00000025yearsandover177,5931.00000044years100,1490.5113562534years37,2330.20965464years60,9910.3114173544years44,6590.25146865yearsandover34,7100.1772274564years60,9910.34343165yearsandover34,7100.195447Distribution#15Distribution#1640yearsandover...........118,1801.00000045yearsandover95,7011.00000049years42,2850.3578024549years19,8060.20695764years41,1850.3484945064years41,1850.43035165yearsandover34,7100.29370465yearsandover34,7100.362692Distribution#17Distribution#1850yearsandover75,8951.00000065yearsandover34,7101.00000064years41,1850.5426586574years18,1360.52250165yearsandover34,7100.45734275yearsandover16,5740.477499Distribution#19Distribution#20Under18years70,7831.000000Under65years239,9241.000000Under5years18,9870.268242Under18years70,7830.29502211years28,1780.3980901844years108,1500.45076817years23,6180.3336684564years60,9910.254210Distribution#21Distribution#225–64years220,9371.00000018–64years169,1411.00000017years51,7960.2344381824years26,2580.15524344years108,1500.4895062534years37,2330.22013064years60,9910.2760563544years44,6590.26403464years60,9910.360593 4 themoreagegroupsutilized,thetighterthecontroloftheeffectofthedifferencesintheagedistributionamonggroupsortimeperiodsbeingcompared.Ontheotherhand,whentoomanyagegroupsareusedandthedatabecomesparseforcertainagegroups,theresultantage-specicratesusedinthecomputationoftheage-adjustedratemayhavelargervariances.Itshouldalsobenotedthatthegreaterthedifferencebetweentheagedistributionofthestandardpopulationusedtoageadjust(inthiscase,the2000StandardPopulation)andtheagedistributionofthestudypopulation,thegreaterthedifferencebetweenthecruderateandtheage-adjustedrateforthestudypopulation.Whentheagedistributionofthestandardpopulationandtheagedistributionofthestudypopulationarethesame,thecruderateisequaltotheage-adjustedrate.These22agedistributionscannotcoverallpossiblesituationsthatwouldrequireageadjustment.AnalystsdesiringagebreakoutsnotspeciedherecanusetheMasterListtogenerateotheragedistributionsandtheirassociatedage-adjustmentweights.Insomecases,theMasterListmaynotcontainalloftheagegroupsdesired.Additionaldetail,includingage-sex-speciccountsforuseinage-sexadjustment,isavailablefromtheoriginalCensuspopulationWhenextractingadditionalsubpopulationsfromtheCensusprojections,thenewstandardpopulationcountsshouldbeasconsistentaspossiblewiththeMasterList,andtheweightsofthenewlyconstructedagedistributionsmustsumto1.Age-AdjustedversusCrudeRates:DeÞnitionsandExamplesSupposethatfrequenciesoftheoccurrenceofacharacteristic(e.g.,beingacurrentcigarettesmoker)havebeendeterminedineachofseveralagegroupsformembersofastudypopulation(e.g.,1997U.S.population,ages18yearsandolder).Theseareage-specicfrequenciesofthecharacteristic(cigarettesmoker).Theage-specicratesofthecharacteristicaretheratiosoftheage-specifrequenciestotherespectiveage-specicsizesofthestudyThecruderateofthecharacteristicisthetotalnumberofoccurrencesofthecharacteristic(cigarettesmoker)acrossallagegroupsdividedbythetotalstudypopulation(allpersonsages18yearsandolder).Age-specicandcruderatesareoftenexpressedasratesper100population(percents)orasratesper1,000or100,000population.Supposethatpopulationsizesforthesameagegroupshavebeenobtainedforaselectedstandardpopulation.Thentheage-adjustedrateofthecharacteristicissimplyaweightedaverageoftheage-specicrates.Theweightforagivenagegroupisthesizeofthecorrespondingagegroupinthestandardpopulationdividedbythetotalstandardpopulationforallagesofinterest.Theweightsthusaddto1.Theage-adjustedratehastheadvantage(overthecruderate)ofbeingcontrolledforagewhencomparedtoage-adjustedratesforotherstudypopulationsthathavedifferentagestructures.(Age-adjustedratesbeingcomparedallbebasedonthesamestandardpopulation.)Toillustrate,supposeonewantstoexaminesmokingratesbylevelofeducationforpeople18yearsandolder.Becausetheagedistributionsofvariouseducationgroupsdiffer,itisdesirabletoageadjusttheratestoensurethatanyeducationdifferencesthatareobservedarenotconfoundedbydifferencesintheagestructureofthevariouseducationgroups.Inthisexample,eacheducationgroup(e.g.,lessthanhighschool,highschoolgraduate)isconsideredtobeaseparatestudypopulation,andthesmokingrateforeachisageadjusted.Examplesofthecalculationofage-adjustedsmokingprevalenceratesbylevelofeducationareshownintablesC.Thesetablesillustratetwoequivalent(exceptfordifferencescausedbyrounding)methodsforcalculatingage-adjustedrates.TableCillustratesthesecalculationsusingtheactualage-specicpopulationfrequenciesfromthe2000StandardPopulationAgeDistribution#8asweights.TableDillustratesthesecalculationsusingage-adjustmentweightscomputedastheratiooftheage-specicpopulationfrequenciesintheStandardPopulationAgeDistribution#8tothetotalstandardpopulationintheagerangecovered.Theage-adjustmentweightsintableDwereroundedtosixdecimalplaces,butcouldhavebeenroundedtoanynumberwiththepossibilityofslightlydifferentresults.Noticethatthetwomethodsyieldidenticalresultsinpercentsshowntotwodecimalplaces.ApplicationtoHealthyPeopleObjectivesAgeadjustmentwasusedformostoftheHealthyPeople2000objectivesusingmortalitydata,butonlyforafewselectedotherobjectives.InHealthyPeople2010,age TableA.AgedistributionsusedbymajorHealthyPeople2010datasystemsDatasystemsAgedistributionsusedforHealthyPeople2010objectivesNHIS,BRFSS,MEPS........#3,6,8,9,14,15,16,17,18,19,22NHANES,CSFII...........#2,5,7,10,11,12NAMCS,NHAMCS..........#4,13..................#4,18,20,21................#1 BRFSSisBehavioralRiskFactorSurveillanceSystem.CSFIIisContinuingSurveyofFoodIntakesbyIndividuals.MEPSisMedicalExpenditurePanelSurvey.NAMCSisNationalAmbulatoryMedicalCareSurvey.NHAMCSisNationalHospitalAmbulatoryMedicalCareSurvey.NHANESisNationalHealthandNutritionExaminationSurvey.NHDSisNationalHospitalDischargeSurvey.NHISisNationalHealthInterviewSurvey.NVSS-MisNationalVitalStatisticsSystem-Mortality. TableB.ModiÞcationstoage-adjustmentweights#2.............4049years0.1539690.153968#6.............4554years0.1386480.138647#7.............5059years0.1342200.134221#15............65yearsandolder0.2937050.293704#20............Under18years0.2950230.295022 5 adjustmentwasalsousedformostoftheobjectivesusingmortalitydata.However,theuseofageadjustmentwasexpandedtoincludemanyotherHealthyPeople2010objectivesthatmeasurehealthoutcomesandriskfactors.InHealthyPeople2010,age-adjusteddataareshownforobjectivesandpopulationsubgroupsthatincludeeitherallagesorawiderangeofages.Objectivesorpopulationsubgroupsthatincludegroupswithrelativelynarrowageranges(generallylessthan40years)arenotadjusted.Forsomepopulationgroups,age-adjustedratesareconsiderablydifferentfromcruderates.Thisoccursbecausethepopulationdistributionofthegroupisquitedifferent TableC.Exampleofcalculationofage-adjusted1997smokingprevalenceamongU.S.adults,bylevelofeducation,using2000StandardEducationlevelandage1997NHISNumberofNumberof(ageadjusted)ColumnAColumnBColumnCColumnDColumnE(columnB)(columnD)ColumnF(columnEsum/columnDsum)Alleducationlevels,18yearsandover193,844,14424.747,879,504203,851,00050,004,16224.5324years24,819,36128.77,123,15726,258,0007,536,04644years82,800,70228.623,681,00181,892,00023,421,11264years54,473,36524.413,291,50160,991,00014,881,80465yearsandover31,750,71612.03,810,08634,710,0004,165,200Alleducationlevels,18yearsandover192,588,94824.547,655,050203,851,00050,029,82224.5424years24,738,88928.87,119,73126,258,0007,556,92444years82,341,23228.623,531,47081,892,00023,403,08864years54,211,57624.413,228,26860,991,00014,882,52865yearsandover31,297,25112.13,775,58134,710,0004,187,282NohighschooldiplomaorGED,18yearsandover37,647,64930.511,482,533203,851,00069,188,13533.9424years5,556,52236.82,044,80026,258,0009,662,94444years...............................11,211,30341.14,607,84681,892,00033,657,61264years9,640,95634.93,364,69460,991,00021,285,85965yearsandover...........................11,238,86813.21,483,53134,710,0004,581,720GED,18yearsandover5,058,51048.82,468,553203,851,00091,893,71845.0824years685,62866.7457,31426,258,00017,514,08644years2,681,17553.61,437,11081,892,00043,894,11264years1,266,06639.0493,76660,991,00023,786,49065yearsandover425,64119.382,14934,710,0006,699,030Highschoolgraduate,18yearsandover53,155,20328.715,255,543203,851,00059,055,70928.9724years6,826,55231.42,143,53726,258,0008,245,01244years21,463,97335.57,619,71081,892,00029,071,66064years14,984,20328.74,300,46660,991,00017,504,41765yearsandover9,880,47512.21,205,41834,710,0004,234,620Somecollegeorhigher,18yearsandover96,727,58619.018,378,241203,851,00037,256,53518.2824years...............................11,670,18721.22,474,08026,258,0005,566,69644years46,984,78121.09,866,80481,892,00017,197,32064years28,320,35117.95,069,34360,991,00010,917,38965yearsandover9,752,26710.31,004,48434,710,0003,575,130 2000StandardPopulation:Distribution#8(inthousands) Adjustmentweight Ages18yearsandover203,8511.00000024years26,2580.12881044years81,8920.40172564years60,9910.29919465yearsandover34,7100.170271 NOTE:Forpurposesofageadjusting,eacheducationlevelcanbeconsideredastudypopulation.Excludespersonswithunknownsmokingstatus.Percentsinthistableshouldbeenteredintoformulasasproportions,e.g.,24.7%=0.247.Multiplytherateforeachage-education-speciÞcgroup(excluding18yearsandover)incolumnBbythestandardpopulationfortheappropriateagegroup(columnD)togettheage-adjustedfrequencyforeachagegroup(columnE).The18yearsandoverfrequenciesarethesumsoftheage-specifcfrequencies.Foreacheducationlevel,sumthefourage-adjustedfrequencies(columnE)togetthetotalage-adjustedfrequencyforages18yearsandover.Dividethatsumbythetotalstandardpopulation18yearsandovertogettheage-adjustedrateforthateducationlevelandmultiplyby100toexpresstherateasapercent(columnF).Includespersonswithunknowneducationstatus.Excludespersonswithunknowneducationstatus. TableD.Exampleofcalculationofage-adjusted1997smokingprevalenceamongU.S.adults,bylevelofeducation,usingage-adjustmentweightsbasedonthe2000StandardPopulationEducationlevelandage1997NHISNumberof(2000StandardColumnAColumnBColumnC(columnA)(columnB)ColumnDColumnE(columnB)(columnD)(columnEsumexpressedasapercent)Alleducationlevels,18yearsandover193,844,14424.747,879,5041.0000000.24529724.5324years24,819,36128.77,123,1570.1288100.03696844years82,800,70228.623,681,0010.4017250.11489364years54,473,36524.413,291,5010.2991940.07300365yearsandover31,750,71612.03,810,0860.1702710.020433Alleducationlevels,18yearsandover192,588,94824.547,655,0501.0000000.24542424.5424years24,738,88928.87,119,7310.1288100.03707144years82,341,23228.623,531,4700.4017250.11480564years54,211,57624.413,228,2680.2991940.07300765yearsandover31,297,25112.13,775,5810.1702710.020541NohighschooldiplomaorGED,18yearsandover37,647,64930.511,482,5331.0000000.33940633.9424years5,556,52236.82,044,8000.1288100.04740244years...............................11,211,30341.14,607,8460.4017250.16510964years9,640,95634.93,364,6940.2991940.10441965yearsandover...........................11,238,86813.21,483,5310.1702710.022476GED,18yearsandover5,058,51048.82,468,5531.0000000.45078945.0824years685,62866.7457,3140.1288100.08591644years2,681,17553.61,437,1100.4017250.21532564years1,266,06639.0493,7660.2991940.11668665yearsandover425,64119.382,1490.1702710.032862Highschoolgraduate,18yearsandover53,155,20328.715,255,5431.0000000.28970028.9724years6,826,55231.42,143,5370.1288100.04044644years21,463,97335.57,619,7100.4017250.14261264years14,984,20328.74,300,4660.2991940.08586965yearsandover9,880,47512.21,205,4180.1702710.020773Somecollegeorhigher,18yearsandover96,727,58619.018,378,2411.0000000.18276418.2824years...............................11,670,18721.22,474,0800.1288100.02730844years46,984,78121.09,866,8040.4017250.08436264years28,320,35117.95,069,3430.2991940.05355665yearsandover9,752,26710.31,004,4840.1702710.017538 2000StandardPopulation:Distribution#8(inthousands) Adjustmentweight Ages18yearsandover203,8511.00000024years26,2580.12881044years81,8920.40172564years60,9910.29919465yearsandover34,7100.170271 NOTE:Forpurposesofageadjusting,eacheducationlevelcanbeconsideredastudypopulation.Excludespersonswithunknownsmokingstatus.Percentsinthistableshouldbeenteredintoformulasasproportions,e.g.,24.7%=.247.Multiplytherateforeachage-education-specicgroup(excluding18yearsandover)incolumnBbytheage-adjustmentweightfortheappropriatestandardagegroup(columnD)togettheadjustmentfactorforeachagegroup(columnE).The18yearsandoverageadjustmentfactorsarethesumsoftheage-specicadjustmentfactors.Foreacheducationlevel,sumthefourage-specicadjustmentfactors(columnE)togetthetotalage-adjustedrateforthateducationlevelandmultiplyby100toexpresstherateasapercent(columnF).Includespersonswithunknowneducationstatus.Excludespersonswithunknowneducationstatus. fromthedistributionofthestandardpopulation,which,forHealthyPeople2010,isbasedontheprojectedyear2000populationfortheentireUnitedStates.Forexample,age-adjustedratesintheHispanicpopulation(especiallyMexicanAmericans)foroutcomesandbehaviorsthataregenerallymorefrequentamongtheolderpopulationareconsiderablyhigherthanthecorrespondingcruderates.ThisisaresultoftheHispanicpopulationsmuchyoungeragedistributionthanthestandardpopulation.Age-adjustedbaselinedataarenotedinPeople2010andinTrackingHealthyPeople2010PartB:OperationalDeTrackingHealthyPeoplealsoincludesadetaileddiscussionoftheage-adjustmentproceduresandweightsusedforHealthyPeople2010.Allage-adjusteddatafortheHealthyPeople2010objectivesareadjustedtothe2000StandardPopulation,althoughanumberofdifferentagedistributionswereusedfordifferentobjectivesbasedonthedataproducedbydifferentdatasystems.TheagedistributionsusedforHealthyPeople2010datawerederivedfromasingleMasterListtable1)andareshownintable2listingoftheagedistributionsusedtocalculateage-adjustedratesfordatafromthemajorHealthyPeople2010datasystemsisshownintableAAgeAdjustmentUsingSUDAANSoftwareWhenanalyzingdatafromcomplexsamplesurveys(e.g.,NationalHealthInterviewSurvey,NationalHealthandNutritionExaminationSurvey),thesampledesignmustbetakenintoaccount.SUDAAN,softwareforthestatisticalanalysisofcorrelateddata,isoftenusedtocalculatevarianceestimatesforNCHSandothersurveydata.SUDAANisspecicallydesignedforanalysisofcluster-correlateddatafromstudiesinvolvingrecurrentevents,longitudinaldata,repeatedmeasures,multivariateoutcomes,multistagesampledesigns,stratieddesigns,unequallyweighteddata,andwithout-replacementsamples.tsmarginalorpopulation-averagedmodelsusinggeneralizedestimatingequations(GEE).Robustvarianceestimatesarecomputedthatfullyaccountforintraclustercorrelation,unequalweighting,stratication,andwithout-replacementsamples.ThissectionprovidesSUDAANcodethatcanbeusedtoageadjustsurveydata.MoreinformationonSUDAANcanbefoundathttp://www.rti.org/patents/sudaan/sudaan.htmlEitherage-specicpopulationfrequenciesorthecweightsshownintable2(orderivedfromtheMasterList)maybeusedasweightstoproduceage-adjustedratesusingSUDAANsPROCDESCRIPTprocedure.AllofthecalculationsshownintablesCcanbeaccomplishedwithjusttwolinesofSUDAANcodeaddedtotheexistingPROCDESCRIPTcode.Figure1showsPROCDESCRIPTcodeusingstandardpopulationAgeDistribution#8fromtable2.Theexamplevariable(AGEGRP8)isarecodebasedonallagesofNationalHeathInterviewSurveysampleadultsandconsistsofthefollowingfouragecategories:1824years,25years,4564years,and65yearsandover.ThenumbersshownafterthewordSTDWGTrepresentthenumberofpersonsineachoftheseagecategoriesinthe2000StandardPopulation,roundedtothenearestthousand.TheagegroupingsofAGEGRP8andtheagegroupingsofthestandardpopulation(AgeDistribution#8)mustmatch.SUDAANusestheage-specicpopulationfrequenciestocalculatetheage-adjustmentweightsandthenproducesanage-adjustedrateforthetotaladultpopulation,ages18andolder.Figure2showsPROCDESCRIPTcodeusingthesamevariableasabove(AGEGRP8)andthesamefouragecategories:1824years,2544years,4564years,and65yearsandolder.However,insteadoftheage-specistandardpopulationfrequenciesshownintheSTDVARstatementingure1,thisversionusestheage-adjustmentweights.Theseage-adjustmentweightsaretheproportioneachgrouprepresentsofthetotalstandardpopulationages18andolder(roundedtosixdecimals)andcorrespondtothecategoriesoftheAGEGRP8variableonthedatale(i.e.,ages1824(0.128810),ages2544years(0.401725),ages4564(0.299194),and65yearsandolder(0.170271).Thestandardpopulationagegroupingsmustmatchthoseofthestudypopulation.Forsomesurveydata,Figure1.SUDAANcodetoproduceage-adjustedratesusing2000StandardPopulationgures(inthousands) PROCDESCRIPT;VARCATLEVELSTDVARAGEGRP8;STDWGT26258818926099134710;TABLES Figure2.SUDAANcodetoproduceage-adjustedratesusing2000StandardPopulationage-adjustmentweights PROCDESCRIPT;VARCATLEVELSTDVARAGEGRP8;STDWGT0.1288100.4017250.2991940.170271;TABLES 8 adjustmenttothetotalpopulation(i.e.,allages)isappropriate.Inothercases,suchastheexampleusedhere,subsetsoftheall-agespopulationaretheappropriatestandard(e.g.,persons18yearsandolderfortheNHISSampleAdultquestionnaire).Itshouldbekeptinmindthatinsomecases,the2000StandardPopulationmaynotbethemostappropriatestandard,suchasinstudiesofpopulationswithverydifferentagestructuresthanthetotalU.S.population.However,totheextentthatanalysesemploythesamestandardpopulationforageadjustment,itwillbepossibletocompareandinterpretresultsacrossdatasystemsandacrossdatayears.Usingage-adjustmentstandardsotherthanthe2000StandardPopulationshouldbeappropriately1.ShryockHS,SiegelJS,etal.Themethodsandmaterialsofdemography.AcademicPress.NewYork.1976.2.FeinleibM,ZarateAO(eds).Reconsideringageadjustmentprocedures:WorkshopProceedings.VitalHealthStat4(29).3.AndersonRN,RosenbergHM.ReportoftheSecondWorkshoponAgeAdjustment.NationalCenterforHealthStatistics.VitalHealthStat4(30).1998.4.AndersonRN,RosenbergHM.Agestandardizationofdeathrates:Implementationoftheyear2000standard.Nationalvitalstatisticsreports;vol47no3.Hyattsville,Maryland:NationalCenterforHealthStatistics.1998.5.CurtinLR,KleinRJ.Directstandardization(age-adjusteddeathrates).HealthPeopleStatisticalNotesNo.6(revised).March1995.http://www.rti.org/patents/sudaan/sudaan.html8.ShalalaDE.HHSpolicyforchangingthepopulationstandardforageadjustingdeathrates.MemorandumfromtheSecretary.August26,1998.9.NationalCenterforHealthStatistics.Health,UnitedStates,1999,withhealthandagingchartbook.Hyattsville,Maryland.10.DayJC.PopulationprojectionsoftheUnitedStatesbyage,sex,race,andHispanicorigin:1995to2050,U.S.BureauoftheCensus,CurrentPopulationReports,P251130,U.S.GovernmentPrintingOffice,Washington.1996..census.gov/prod/1/pop/p25-11.DepartmentofHealthandHumanServices(DHHS).TrackingHealthyPeople2010.Washington:U.S.GovernmentPrintingOffice.2000. PublishedissuesofHealthyPeopleStatisticalNotesNumberTitleDateofIssueHealthyPeople20001HealthStatusIndicatorsfortheYear2000Fall19912InfantMortalityWinter19913HealthStatusIndicators:DenitionsandNationalDataSpring19924IssuesRelatedtoMonitoringtheYear2000ObjectivesSummer19935RevisionstoHealthyPeople2000BaselinesJuly19936DirectStandardization(Age-AdjustedDeathRatesMarch19957YearsofHealthyLifeApril19958EvaluatingPublicHealthDataSystems:APracticalApproachJune19959MonitoringAirQualityinHealthyPeople2000September199510HealthStatusIndicators:DifferentialsbyRaceandHispanicOriginSeptember199511OperationalDenitionsforYear2000Objectives:PriorityArea20,ImmunizationandInfectiousDiseasesFebruary199712OperationalDenitionsforYear2000Objectives:PriorityArea13,OralHealthMay199713HealthyPeople2000MidcourseRevisions:ACompendiumAugust199714OperationalDenitionsforYear2000Objectives:PriorityArea14,MaternalandInfantHealthDecember199715PriorityDataNeeds:SourcesofNational,State,andLocal-levelDataandDataCollectionSystemsDecember199716OperationalDenitionsforYear2000Objectives:PriorityArea6,MentalHealthandMentalDisordersFebruary199817OperationalDenitionsforYear2000Objectives:PriorityArea21,ClinicalPreventiveServicesDecember199818OperationalDenitionsforYear2000Objectives:PriorityArea1,PhysicalActivityandFitnessDecember199819HealthyPeople2000:AnAssessmentBasedontheHealthStatusIndicatorsfortheUnitedStatesandEachStateNovember2000HealthyPeople201020AgeAdjustmentUsingthe2000ProjectedU.S.PopulationJanuary2001 10 DEPARTMENTOFHEALTH&HUMANSERVICESCentersforDiseaseControlandPreventionNationalCenterforHealthStatistics6525BelcrestRoadHyattsville,Maryland20782-2003 OFFICIALBUSINESSPENALTYFORPRIVATEUSE,$300 Toreceivethispublicationregularly,contacttheNationalCenterforHealthStatisticsbycalling301-458-4636www.cdc.gov/nchs/ SuggestedcitationKleinRJ,SchoenbornCA.Ageadjustmentusingthe2000projectedU.S.population.HealthyPeopleStatisticalNotes,no.20.Hyattsville,Maryland:NationalCenterforHealthStatistics.January2001.DHHSPublicationNo.(PHS)2001-12371-0015(1/01) FIRSTCLASSMAILPOSTAGE&FEESPAIDPERMITNO.G-284