Andersons pediatric cardiology 285

withfunctional
abnormalitiesonly
insomecases

resolvewithoutbecomingclinicallymanifest(e.g.,aorticvalvewithtwoleaflets,the
prolapsingmitralvalve,silentpersistentpatencyofthearterialductorsmallseptaldefects,
includedpatencyoftheovalforamen).

Ratios,Rates,andProportions
Confusionandmisuseregardingthetermsincidenceandprevalencearerooted
inmisunderstandingregardingwhatconstitutesaratio,arate,andaproportion.
ThesedefinitionsaresummarizedinTable13.2.
Table13.2
StatisticalandMethodologicDefinitions,SomeConsiderations,andExamplesUsedinIdentifying
theFrequencyofCongenitalCardiacDisease
Term
Ratio

Definition,SomeConsiderations,andExamples
Arelationshipbetweentwoquantitieswherethenumeratorandadenominatoraremutually
exclusive(e.g.,theratioofmalestofemaleswithtransposedarterialtrunks).
Proportion
Thefrequencyofaconditionorcategorywithinthepopulationspecifiedinthedenominator
independentoftime(e.g.,theproportionofmuscularseptaldefectsamongallventricularseptal
defects).
Rate
Thefrequencyofoccurrenceofaconditionthattakesintoaccountthetimeperiodoverwhichthe
numeratorwasaccumulated(e.g.,therateoftheoccurrenceofathromboticepisodeinpatients
duringtheirfirstyearsubsequenttotheFontanoperation).
Kaplan-Meier Thecumulativefrequencyofoccurrenceofaconditionthatreflectsinstantaneouschangesinrates
estimates

andriskovertimeand,unlikerate,doesnotassumethattheconditionoccursuniformlyoverthe
specifiedperiodoftime(e.g.,freedomfromofathromboticepisodeinpatientssubsequenttothe
Fontanoperationover10-yearperiod,showingmaximaloccurrenceinfirstpostoperativeyear).
Incidence
Therateofnewoccurrencesofaconditionoveraspecifiedperiodoftimeinapopulationatrisk
(e.g.,newcasesperyearofKawasakidiseaseper100,000children<5years).
Prevalence
Theproportionofpreexistingandnewoccurrencesofaconditionidentifiedinapopulationat
risk,eitheratasinglepointintimeoroveraspecifiedperiod(e.g.,prevalenceofcongenitalheart
diseaseatbirthamongstalllivebornindividuals).
Denominator Representsthepopulationofinterestoratriskoveradefinedtimeperiodandshouldbethefirst
considerationwhenestimatingtheprevalence.
Numerator
Representsthenumberofcasesofcongenitalheartdiseasethatwereidentifiedfromapopulation
atriskoveradefinedperiodoftime.
Ascertainment Thecompletenessofidentificationofallthecasesfromthepopulationatrisk.Thisisakeyaspect
ofcases
ofappraisalofthenumeratorofprevalenceestimates(e.g.,activesurveillanceisbetterthan
nonmandatoryreporting).
Verificationof Theaccuracyofdefinitionorclassificationofallcases.Themethodofverificationcanaffectthe
cases
prevalence(e.g.,clinicalidentificationversusautopsy,angiography,andechocardiography).33
Definitionof Thespecificationanduseofclearandstandarddefinitionsofcongenitalheartlesionsdescribed
cases
withacceptednomenclature.Itisalsoimportanttoconsiderwhichlesionswereincludedand
excluded.

IncidenceandPrevalence

UsingthedefinitionsinTable13.2,whatmanymistakenlyrefertoasthe
incidenceofcongenitalcardiacdiseaseisinrealitytheprevalence,consistingof
thenumberofnewbornswhoaresubsequentlyconfirmedtohavecongenitally
malformedheartsasobservedwithinadefinedpopulationofliveborn
individualsoveraspecifiedtime.However,incidenceandratesareveryrelevant
inthestudyofprognosisandthenaturalandmodifiednaturalhistory.
Theprevalenceatorforavariableperiodfollowinglivebirthisimportantin
definingthemaximalburdenofcongenitalcardiacdiseaseinthepopulation.The
naturalandmodifiednaturalhistoryspecifictocertaindefectsandstrategiesfor
theirmanagementinfluencethechangingprevalenceinthepopulationovertime,
withthenumeratoranddenominatorbothdecreasingduetodeaths,andthe
numeratorperhapsdecreasingduetospontaneousresolutionofsomelesions.
Theabsolutenumberandcharacteristicsofpatientsatanygiventimepointare
importantfordefiningtheburdenposedbydisease,whichisakeydeterminant
indefiningtherequirementsforresources.Withtheincreasingsurvivalof
patientsintoadulthood,andtheirtransitionintothesystemofhealthcare
providingforadults,thisnumberhastakenonincreasingimportance,butits
accurateestimationisfraughtwithnumerousmethodologicchallenges.

Denominator
Thedefinitionofthedenominatorisakeypieceofinformationbecauseit
characterizesthepopulationtowhichtheestimateofincidenceorprevalence
maybeapplied.Knowledgeofthepopulationcharacteristicsisalsonecessaryto
evaluatesimilarpopulationstowhichtheestimatemightbeextrapolated.The
mostvaliddenominatorsareonesthatareenumeratedduringthetimeof
ascertainmentofthecases,whichisbestachievedbyusingaprospectivecohort
studydesign.However,suchanapproachisintensiveintermsofbothtimeand
resources.Mostestimatesofprevalenceuseastheirdenominatorthetotal
numberoflivebirthsoccurringoveraspecifiedperiodoftimederivedfroma
geographicallydefinedpopulation.

Aconvenientandreadilyavailabledenominatorthatisusedfrequentlyisthe
numberofbirthsreportedtoagovernmentaloradministrativesystemof
registrationorvitalstatistics.Ifused,informationregardingthecompletenessof
ascertainment/reportingofbirths,usuallythroughauditsofthedata,shouldbe
sought.Someregistrieswillincludestillbirthsandfetaldeaths,andthis
informationshouldbeclearlyindicated.Registriesofbirthsusuallycollect


additionalinformation,specificallydemographicsofthebirthparents,butmay
alsoincludedataregardingclinicaldiagnosesandcharacteristicspresentor
evidentatbirth.Someestimatesofprevalencewilladditionallyrelyonthis
informationforascertainmentofcases.Ifso,thenthevalidityandreliabilityof
thatdataneedtobecarefullyscrutinized.

Numerator
Toevaluatethenumeratorofaprevalenceestimate,onemustknowhowthe
caseswereidentifiedandverifiedfromthedatasource,whattypesoflesions
wereincludedandexcluded,andwhatnomenclatureandclassificationscheme
wasused.

AscertainmentofCases
Acomprehensiveandactiveprospectivesurveillanceofallsourcesofcasesis
likelytoyieldthemostcompleteascertainmentbecausecaseascertainmentisa
specificandplannedendeavor.Moststudiesofthisnaturerelyonclinical
presentationorevaluationofalivingsubjectastheinitialentrypoint.The
durationoffollow-upforcaseascertainmentmustbesufficientlylongthatall
casesareidentified,especiallybecausesomeimportantcongenitalheartdisease
maynotmanifestveryearlyinlife.

VerificationofCases

Giventheprecedingdiscussionthatmostprevalencestudiesrelyonaclinical
diagnosisforthemajorityofcaseidentification,verificationorconfirmationof
thediagnosisisimportant.Earlierstudieshaveincludedcasesverifiedby
autopsy,surgery,orcardiaccatheterization,whichmayhavelimited
ascertainmentbutassuredverification.Theseprevalenceestimateshavebeen
somewhatlowandskewedtowardmoreseriouscongenitalheartlesions.
Echocardiographyhasbecomethestandardfortheinitialverificationofcases
andhasproventohaveahighsensitivityandspecificity.4

SourcesofData
Datasourcesregardingcasesvarywidelyinthedegreetowhichascertainment
waspassiveoractiveandthedegreetowhichtheyaccuratelyrelatetoadefined