Andersons pediatric cardiology 287

CriticalAppraisal
Moststudiesthatreportestimatesofprevalencearespecificallydesignedtodo
soandhaveitastheprimaryaimofthestudy.Beforeacceptingtheresultsof
thesestudies,itisimportantcriticallytoappraisethemethodsandreporting,to
determineifthefindingsarevalid,reliable,andrelevant.Criticalappraisalis
alsorequiredwhenonewishestocompareestimatesarisingfromdifferent
studies,suchascomparisonsoftrendsovertimeoracrossdifferentpopulations.
Considerationmustbegiventoalloftheaspectsbywhichthedenominatorand
numeratoraredefinedandderivedandtotheadequacyofreportingofthe
associatedmethodsandresults.InBox13.1,questionsforwhichtheanswers
shouldbeevidentwhencriticallyappraisingareportregardingprevalenceare
noted.


Box13.1

QuestionsforCriticallyAppraisingaReport
AboutPrevalence
▪Whatwasthestatedpurposeoraimofthestudy?
▪Howaccurateandvalidwastheestimatereported?
▪Howaccurateandvalidwasthedefinitionandcompletenessofthe
numerator?
▪Whatformsofcongenitalheartdiseasewereincludedor
excluded?
▪Whatnomenclatureandsystemofclassificationwasusedto
describeandgroupcases?
▪Whatwerethemethodsbywhichcaseswereascertained
andreported?
▪Whatwastheoveralldesignofthestudyanddata
collection?
▪Howwerecasesdetectedandreportedfromthepopulation

studied?
▪Whatfactors,particularlyrelatingtothesystemforhealth


care,mayhaveinfluencedtheascertainmentofcases?
Specifically,dideveryoneinthepopulationstudiedhave
equalaccesstohealthcareandthereferralcenter,andwas
screeningandverificationinfluencedbydifferencesin
qualityandavailabilityofexpertiseortechnology?
▪Howwerediagnosesconfirmedorverified?
▪Howlongwasthefollow-up,andwasitsufficientlylongto
capturecaseswithlaterclinicalmanifestation?
▪Ifauniversalscreeningwasapplied,wasitappliedequally
totheentirepopulationatanearlyenoughtimepoint,was
itsufficientlysensitiveandspecific,andwasanassessment
ofverificationperformed?
▪Howaccurateandvalidwasthedefinitionandcompletenessofthe
denominator?
▪Whatwasthedefinitionofthepopulationstudied?
▪Whatwerethesourcesofdatausedtoderivethe
denominator,andweretheyvalidandreliable?
▪Howaccuratelyandcompletelydoesthepopulationstudied
reflectthetargetpopulationorpopulationatlarge?
▪Arethecasescompletedderivedfromthepopulation
studiedasdefinedinthedenominator?Werethecasesand
populationcharacterizedaspartofthesamestudywiththe
samemethodology,asinacohortstudy?
▪Howreliableistheestimateofprevalenceestimate?Areconfidence
intervalsprovided?
▪Howdoestheestimatefromthestudycomparewiththosereportedfrom

otherstudieswithcomparablemethodology?
▪Istheestimateapplicabletoyourownclinicalpopulation?Isthepopulation
studiedsimilartoyourownclinicalpopulationintermsofsetting,time,
geography,anddemographiccharacteristics?Istheestimateofprevalence
relevanttoyourownclinicalorresearchquestion?


FactorsInfluencingEstimatesof
Prevalence
Ideally,factorsthatmighthaveatruecausalrelationshiptothedevelopmentof
congenitalcardiacdiseaseandhenceinfluencethetrueincidenceshouldbe
identified.Identificationofsuchfactorsmayallowforthepreventionof
congenitalcardiacdiseaseorprovidenewknowledgeastoetiologyand
development.Theidentificationoftheseassociationscomesmainlyfrom
observationalstudies.

EnvironmentalFactors
Themostwidelyheldbeliefisthatcongenitalcardiacdiseaseistheproductof
aninteractionbetweengeneticandenvironmentalfactors.However,evidence
hasyettoprogresssufficientlyfartoprovethisnotion,althoughadvancesinthe
understandingofgeneticunderpinningscontinuetoincrease.Forexample,
recentstudieshavesuggestedthepresenceofaninteractionbetweenmaternal
genesandmaternalexposurestoaromatichydrocarbonsandcongenitalheart
disease.7Numerousenvironmentalfactorshavebeenlinkedtosuch
development,independentofanyalreadyknowngeneticinfluenceor
predisposition.4
AscientificstatementfromtheAmericanHeartAssociationreviewedstudies
ofprevalenceandsummarizedstudiesofenvironmentalfactors.8InTable13.3,
positiveassociationsformaternalillnessesandmaternalexposuresare
highlighted.

Table13.3
FetalExposuresandIncreasedRiskofCongenitalHeartDefectsa
OddsRatio
MATERNALILLNESS
Phenylketonuria52–55
Pregestationaldiabetes56–59
Febrileillness56,60–62
Influenza62,63
Maternalrubella64

>6
3.1–18
1.8–2.9
2.1
b