Int J Gynecol Obstet 2019; 145 (Suppl. 1): 1–33 wileyonlinelibrary.com/journal/ijgo  | 1© 2019 International Federation of Gynecology and Obstetrics

DOI: 10.1002/ijgo.12802

S U P P L E M E N T A R T I C L E

The International Federation of Gynecology and Obstetrics (FIGO) initiative on pre- eclampsia: A pragmatic guide for first- trimester screening and prevention

Liona C. Poon1,* | Andrew Shennan2 | Jonathan A. Hyett3 | Anil Kapur4 |  Eran Hadar5 | Hema Divakar6 | Fionnuala McAuliffe7 | Fabricio da Silva Costa8 |  Peter von Dadelszen2 | Harold David McIntyre9 | Anne B. Kihara10 |  Gian Carlo Di Renzo11 | Roberto Romero12 | Mary D'Alton13 |  Vincenzo Berghella14 | Kypros H. Nicolaides15 | Moshe Hod5

1DepartmentofObstetricsandGynecology,PrinceofWalesHospital,TheChineseUniversityofHongKong,Shatin,HongKongSAR2Department of Women and Children's Health, School of Life Course Sciences, Faculty of Life Sciences and Medicine, King's College London, London, UK3DisciplineofObstetrics,GynecologyandNeonatology,FacultyofMedicine,UniversityofSydney,Sydney,NSW,Australia4WorldDiabetesFoundation,Bagsværd,Denmark5HelenSchneiderHospitalforWomen,RabinMedicalCenter,PetahTikva,andSacklerFacultyofMedicine,TelAvivUniversity,TelAviv,Israel6DivakarsSpecialityHospital,Bangalore,India7DepartmentofObstetricsandGynecology,TheNationalMaternityHospital,Dublin,Ireland8DepartmentofGynecologyandObstetrics,RibeirãoPretoMedicalSchool,UniversityofSãoPaulo,SãoPaulo,Brazil9UniversityofQueenslandMaterClinicalSchool,Brisbane,QLD,Australia10AfricanFederationofObstetricsandGynaecology,Khartoum,Sudan11CenterofPerinatalandReproductiveMedicine,DepartmentofObstetricsandGynecology,UniversityofPerugia,Perugia,Italy12PerinatologyResearchBranch,DivisionofObstetricsandMaternal-FetalMedicine,DivisionofIntramuralResearch,Eunice Kennedy ShriverNationalInstituteofChildHealthandHumanDevelopment,NationalInstitutesofHealth,USDepartmentofHealthandHumanServices,Bethesda,MD,andDetroit,MI,USA13ObstetricianandGynecologistin-Chief,ColumbiaUniversityIrvingMedicalCenter,NewYork-Presbyterian14DivisionofMaternal-FetalMedicine,DepartmentofObstetricsandGynecology,SidneyKimmelMedicalCollegeofThomasJeffersonUniversity,Philadelphia,PA,USA15FetalMedicineResearchInstitute,King‘sCollegeHospitalLondon,London,UK

*CorrespondenceLionaC.Poon,DepartmentofObstetricsandGynecology,PrinceofWalesHospital,TheChineseUniversityofHongKong,Shatin,HongKongSAR.Email:liona.poon@cuhk.edu.hk

Funding InformationPerkinElmerInc

[Correctionaddedon18June2019,afterfirstonlinepublication:theaffiliationforMaryD’Altonhasbeencorrectedinthisversion.]

www.wileyonlinelibrary.com/journal/ijgomailto:liona.poon@cuhk.edu.hkhttp://crossmark.crossref.org/dialog/?doi=10.1002%2Fijgo.12802&domain=pdf&date_stamp=2019-05-20

2  |     Poon ET AL.

ACKNOWLEDGMENTS

ThisprojectwasfundedbyanunrestrictededucationalgrantfromPerkinElmerInc,Waltham,USA,representedbyYvonneParker,VPMarketDevelopment,ReproductiveHealth,PerkinElmerInc.

CONFLICTS OF INTEREST

Theauthorshavenoconflictsofinteresttodeclare.

CONTRIBUTORS

Inadditiontotheauthors,thefollowingpeopleprovidedimportantcontributionsduringthecreationofthedocument.Thanksgotothefollow-inginternationalexperts:ThePregnancyandNCDCommitteemembers:MosheHod,HemaDivakar,MarkHanson,AnilKapur,HaroldDavidMcIntyre,MaryD‘Alton,AnneB.Kihara,RonaldMa,FionnualaMcAuliffe,GianCarloDiRenzo,andCNPurandare.Special thanks forFIGOguidanceandcoordinationgotoPresidentCarlosFuchtner,Past-PresidentCNPurandare,Past-ChairoftheSMNHCommitteeGerardVisser,alongwithJessicaMorrisandRachelGooden(FIGOProjectManagerandCoordinator).InadditiontotheFIGOExecutiveBoard,variousFIGOCommitteesandWorkingGroupscontributedtoandsupportedthedocument.

     |  3Poon ET AL.

CONTENTS

1. Executivesummary

2. TargetaudienceoftheFIGOinitiativeonpre-eclampsia

3. Assessmentofqualityofevidenceandgradingofrecommendations

4. Pre-eclampsia:background,definition,riskfactors,maternalandperinatalmorbidityandmortalityassociatedwithpre-eclampsia

5. Firsttrimesterpredictionofpre-eclampsia

6. Firsttrimesterpreventionofpre-eclampsia

7. Resource-basedapproachtoscreening

8. Cost-effectivenessofpre-eclampsiascreening

Supporting Information

AdditionalsupportinginformationmaybefoundonlineintheSupportingInformationsectionattheendofthearticle.

Appendix S1. Formulas for calculation of multiple of the median (MoM) values at 11–13weeks of gestation. Algorithm for prediction ofpre-eclampsia.

4  |     Poon ET AL.

LIST OF ABBREVIATIONS/ACRONYMS

ACOG AmericanCollegeofObstetricsandGynecologyAPS AntiphospholipidsyndromeART AssistedreproductivetechnologiesASPRE Combinedmultimarkerscreeningandrandomizedpatienttreatmentwithaspirinforevidence-basedpre-eclampsiapreventionAUC AreaunderreceiveroperatingcharacteristiccurveBMI BodymassindexCI ConfidenceintervalCRL Crown–rumplengthdBP DiastolicbloodpressureDR DetectionrateFGR FetalgrowthrestrictionFIGO InternationalFederationofGynecologyandObstetricsGI GastrointestinalGWAS Genome-wideassociationstudiesHELLP Hemolysis,elevatedliverenzyme,lowplateletHR HazardratioIAD InterarmdifferenceICSI IntracytoplasmicsperminjectionIGF Insulin-likegrowthfactorIQ IntelligencequotientISSHP InternationalSocietyfortheStudyofHypertensioninPregnancyIUFD Intrauterine fetal deathIUI IntrauterineinseminationIVF InvitrofertilizationMAP MeanarterialpressureMoM MultipleofmedianNCDs Noncommunicable diseasesNHFA NationalHeartFoundationofAustraliaNICE NationalInstituteforHealthandCareExcellenceNICU Neonatal intensive care unitOR OddsratioPAPP-A Pregnancy-associatedplasmaproteinAPE Pre-eclampsiaPLGF Placental growth factorPOC Point of careRR RelativerisksBP SystolicbloodpressureSD StandarddeviationSGA SmallforgestationalageSLE Systemic lupus erythematosusUA UmbilicalarteryUTPI UterinearterypulsatilityindexVEGF-R1Vascularendothelialgrowthfactorreceptor1VEGF Vascularendothelialgrowthfactor

     |  5Poon ET AL.

1  | EXECUTIVE SUMMARY

Pre-eclampsia (PE) is a multisystem disorder that typically affects2%–5% of pregnant women and is one of the leading causes of mater-nal and perinatal morbidity and mortality, especially when the condi-tionisofearlyonset.Globally,76000womenand500000babiesdieeach year from this disorder. Furthermore,women in low-resourcecountriesareatahigherriskofdevelopingPEcomparedwiththoseinhigh-resourcecountries.

Although a complete understanding of the pathogenesis of PEremainsunclear,thecurrenttheorysuggestsatwo-stageprocess.Thefirststageiscausedbyshallowinvasionofthetrophoblast,resultingininadequateremodelingofthespiralarteries.Thisispresumedtoleadto the second stage, which involves the maternal response to endo-thelialdysfunctionandimbalancebetweenangiogenicandantiangio-genicfactors,resultingintheclinicalfeaturesofthedisorder.

Accurate prediction and uniform prevention continue to eludeus.ThequesttoeffectivelypredictPE inthefirsttrimesterofpreg-nancyisfueledbythedesiretoidentifywomenwhoareathighriskofdevelopingPE, so thatnecessarymeasurescanbe initiatedearlyenoughto improveplacentationandthuspreventorat leastreducethefrequencyofitsoccurrence.Furthermore,identificationofan“atrisk”groupwillallowtailoredprenatalsurveillancetoanticipateandrecognizetheonsetoftheclinicalsyndromeandmanageitpromptly.

PE has been previously defined as the onset of hypertensionaccompanied by significant proteinuria after 20weeks of gestation.Recently,thedefinitionofPEhasbeenbroadened.Nowtheinterna-tionallyagreeddefinitionofPEistheoneproposedbytheInternationalSociety for the Study of Hypertension in Pregnancy (ISSHP).

Accordingtothe ISSHP,PE isdefinedassystolicbloodpressureat≥140mmHgand/ordiastolicbloodpressureat≥90mmHgonatleast two occasions measured 4 hours apart in previously normoten-sivewomenandisaccompaniedbyoneormoreofthefollowingnew-onsetconditionsatorafter20weeksofgestation:

1. Proteinuria (i.e. ≥30mg/mol protein:creatinine ratio;≥300mg/24hour; or ≥2+dipstick);

2. Evidenceofothermaternalorgandysfunction,including:acutekid-ney injury (creatinine ≥90μmol/L; 1 mg/dL); liver involvement (elevated transaminases, e.g. alanine aminotransferase or aspartate aminotransferase>40IU/L)withorwithoutrightupperquadrantor epigastric abdominal pain; neurological complications (e.g.eclampsia,alteredmentalstatus,blindness,stroke,clonus,severeheadaches, and persistent visual scotomata); or hematological com-plications (thrombocytopenia–platelet count

6  |     Poon ET AL.

Itprovidesboth thebestand themostpragmatic recommenda-tions according to the level of acceptability, feasibility, and ease ofimplementationthathavethepotential toproducethemostsignifi-cant impact indifferent resource settings. Suggestionsareprovidedforavarietyofdifferentregionalandresourcesettingsbasedontheirfinancial,human,andinfrastructureresources,aswellasforresearchprioritiestobridgethecurrentknowledgeandevidencegap.

To deal with the issue of PE, FIGO recommends the following:Public health focus:Thereshouldbegreater internationalatten-

tiongiventoPEandtothelinksbetweenmaternalhealthandnon-communicable diseases (NCDs) on the Sustainable Developmental Goals agenda. Public health measures to increase awareness, access, affordability, andacceptanceofpreconceptioncounselling, andpre-natal and postnatal services forwomen of reproductive age shouldbeprioritized.Greatereffortsarerequiredtoraiseawarenessofthebenefitsofearlyprenatalvisitstargetedatreproductive-agedwomen,particularlyinlow-resourcecountries.

Universal screening:Allpregnantwomenshouldbescreenedforpreterm PE during early pregnancy by the first-trimester combinedtestwithmaternal risk factorsandbiomarkersasaone-stepproce-dure. The risk calculator is available free of charge at https://fetal-medicine.org/research/assess/preeclampsia. FIGO encourages all countriesanditsmemberassociationstoadoptandpromotestrate-gies to ensure this. The best combined test is one that includes mater-nalriskfactors,measurementsofmeanarterialpressure(MAP),serum

placental growth factor (PLGF), and uterine artery pulsatility index(UTPI). Where it is not possible to measure PLGF and/or UTPI, the baselinescreeningtestshouldbeacombinationofmaternalriskfac-torswithMAP,andnotmaternalriskfactorsalone.Ifmaternalserumpregnancy-associatedplasmaproteinA(PAPP-A)ismeasuredforrou-tinefirst-trimesterscreeningforfetalaneuploidies,theresultcanbeincludedforPEriskassessment.Variationstothefullcombinedtestwouldleadtoareductionintheperformancescreening.Awomanisconsideredhighriskwhentheriskis1in100ormorebasedonthefirst-trimestercombinedtestwithmaternal risk factors,MAP,PLGF,and UTPI.

Contingent screening: Where resources are limited, routinescreeningforpretermPEbymaternalfactorsandMAPinallpregnan-cies and reserving measurements of PLGF and UTPI for a subgroup of thepopulation(selectedonthebasisoftheriskderivedfromscreen-ingbymaternalfactorsandMAP)canbeconsidered.

Prophylactic measures: Following first-trimester screening forpretermPE,womenidentifiedathighriskshouldreceiveaspirinpro-phylaxis commencing at 11–14+6weeks of gestation at a dose of~150mgtobetakeneverynightuntil36weeksofgestation,whendelivery occurs, orwhen PE is diagnosed. Low-dose aspirin shouldnot be prescribed to all pregnant women. In women with low calcium intake(

     |  7Poon ET AL.

2  | TARGET AUDIENCE OF THE FIGO INITIATIVE ON PRE- ECLAMPSIA

This document is directed at multiple stakeholders with the inten-tionofbringingattentiontoPE,whichisapreventablebutcommonandpotentiallylife-threateningcomplicationofpregnancywithgraveconsequencesforboththemothersandtheoffspring.ThisdocumentproposestocreateaglobalframeworkforactionforearlyscreeningandpreventionofPE.

The intended target audience includes:

• Healthcare providers:Allthosewhoarequalifiedtocareforpreg-nantwomenandtheirnewborns,butinparticularthoseresponsi-bleforscreeningforhigh-riskwomen(obstetricians,maternal-fetalmedicine specialists, internists, pediatricians, neonatologists, gen-eral practitioners/family physicians, midwives, nurses, advancepractice clinicians, nutritionists, pharmacists, community healthworkers,laboratorytechnicians,etc.)

• Healthcare delivery organizations and providers: governments, federalandstatelegislators,healthcaremanagementorganizations,health insurance organizations, international development agen-cies,andnongovernmentalorganizations.

• Professional organizations: international, regional, and nationalprofessional organizations of obstetricians and gynecologists,internists, family practitioners, pediatricians, neonatologists, andworldwidenationalorganizationsdedicatedtothecareofpregnantwomen with PE.

3  | QUALITY ASSESSMENT OF EVIDENCE AND GRADING OF STRENGTH OF RECOMMENDATIONS

Inassessing thequalityofevidenceandgradingof strengthof rec-ommendations, thedocument follows the terminologyproposedbythe Grading of Recommendations, Assessment, Development andEvaluation(GRADE)workinggroup(http://www.gradeworkinggroup.org/).Thissystemusesconsistentlanguageandgraphicaldescriptionsforthestrengthandqualityoftherecommendationsandtheevidenceonwhichtheyarebased.Strongrecommendationsarenumberedas1andconditional(weak)recommendationsarenumbered2.Forthequalityofevidence,cross-filledcirclesareused:⊕OOO denotes very low-qualityevidence;⊕⊕OO lowquality;⊕⊕⊕Omoderatequality;and ⊕⊕⊕⊕high-qualityevidence(Tables1and2).

T A B L E 1   Interpretationofstrongandconditional(weak)recommendationsaccordingtoGRADE.a,b

Implications1=Strong recommendation phrased as “we recommend” 2=Conditional (weak) recommendation phrased as “we

suggest”

Forpatients Nearlyallpatientsinthissituationwouldaccepttherecommendedcourseofaction.Formaldecisionaidsarenotneededtohelppatientsmakedecisionsconsistentwiththeir values and preferences

Mostpatientsinthissituationwouldacceptthesuggestedcourseofaction

For clinicians Accordingtotheguidelinesperformanceoftherecom-mendedactioncouldbeusedasaqualitycriterionorperformanceindicator,unlessthepatientrefuses

Decisionaidsmayhelppatientsmakeamanagementdecision consistent with their values and preferences

Forpolicymakers Therecommendationcanbeadaptedaspolicyinmostsituations

Stakeholdersneedtodiscussthesuggestion

aReprintedwithpermissionoftheAmericanThoracicSociety.©2019AmericanThoracicSociety.SchunemannHJ,JaeschkeR,CookDJ,etal.AnofficialATSstatement:gradingthequalityofevidenceandstrengthof recommendations inATSguidelinesandrecommendations.Am J Respir Crit Care Med 2006;174:605–614.TheAmericanJournalofRespiratoryandCriticalCareMedicineisanofficialjournaloftheAmericanThoracicSociety.bBothcaregiversandcarerecipientsneedtobeinvolvedinthedecision-makingprocessbeforeadoptingrecommendations.

T A B L E 2   InterpretationofqualityofevidencelevelsaccordingtoGRADE.a

Level of evidence Definition

High ⊕⊕⊕⊕

Weareveryconfidentthatthetrueeffectcorrespondstothatoftheestimatedeffect

Moderate ⊕⊕⊕O

Wearemoderatelyconfidentintheestimatedeffect.Thetrueeffectisgenerallyclosetotheestimatedeffect,butitmaybeslightlydifferent

Low ⊕⊕OO

Ourconfidenceintheestimatedeffectislimited.Thetrueeffectcouldbesubstantiallydifferentfromtheestimatedeffect

Verylow ⊕OOO

Wehaveverylittleconfidenceintheestimatedeffect.Thetrueeffectislikelytobesubstantiallydifferentfromtheestimatedeffect

aAdaptedwithpermissionfromBalshemetal.GRADEguidelines:3.Ratingthequalityofevidence.J Clin Epidemiol2011;64:401–6.©Elsevier(2011).

http://www.gradeworkinggroup.org/http://www.gradeworkinggroup.org/

8  |     Poon ET AL.

4  | PRE- ECLAMPSIA: BACKGROUND, DEFINITION, RISK FACTORS, MATERNAL AND PERINATAL MORBIDITY AND MORTALITY ASSOCIATED WITH PRE- ECLAMPSIA

4.1 | Introduction

Pre-eclampsia(PE) isamultisystemdisorderofpregnancypreviouslydefinedbytheonsetofhypertensionaccompaniedbysignificantpro-teinuriaafter20weeksofgestation.Recently,thedefinitionofPEhasbeen broadened.2–5 PE typically affects2%–5%ofpregnantwomenand is one of the leading causes of maternal and perinatal morbidity and mortality,especiallywhentheconditionisofearlyonset.6,7 Globally, 76 000 women and 500 000 babies die each year from this disorder.8 Furthermore,womeninlow-resourcecountriesareatahigherriskofdevelopingPEcomparedwiththoseinhigh-resourcecountries.

PEcanbesubclassifiedinto:

1. Early-onset PE (with delivery at

     |  9Poon ET AL.

40 years.19Onestudyhasevaluatedthematernalageassociatedriskaccording to the severity of PE. Using multivariate logistic regres-sionanalysis,adjustingforconfounders,theriskforlate-onsetPEhasbeenshowntoincreaseby4%withevery1-yearincreaseinmaternalage above 32 years.23 However, maternal age is not associated with increasedriskofearly-onsetPE.23

4.3.2 | Parity

Innulliparouswomen,the increasedriskofdevelopingPEhasbeenwidelyreported.OnesystematicreviewreportedthattheriskofPEis increased three-fold in nulliparouswomen.24 Another systematicreviewthat included26studiesreportedthatthis increasedriskforPEpersistsevenafteradjustingforotherriskfactors,suchasmaternalage,race,andbodymassindex(BMI)andthesummaryadjustedoddsratio (OR)was 2.71 (95%CI, 1.96–3.74).25 Parous women without priorhistoryofPEhavereducedriskofPE;however,thisprotectiveeffectislostwhentheconceptionpartnerisdifferent.26

4.3.3 | Previous history of PE

Alargepopulation-basedstudyincluding763795nulliparouswomenwithafirstdeliverybetween1987and2004showedthattheriskofPEwas4.1%inthefirstpregnancyand1.7%inlaterpregnanciesoverall.However,theriskwas14.7%inthesecondpregnancyforwomenwithahistoryofPEintheirfirstpregnancyand31.9%forwomenwhohadPEintheprevioustwopregnancies.TheriskofPEforparouswomenwithoutahistoryofPEwas1.1%.TheseobservationssuggestthattheriskofPEisgreaterinnulliparousthanparouswomenwithoutapriorhistoryofPE.Amongparouswomen,theriskofPEinsubsequentpreg-nancies depends on a prior history of PE.27Thisrelativeriskforsubse-quentPErangesfrom7to10timeshigherinasecondpregnancy.28–30

Astudy focusingonPEaccording to severityofdiseaseshowedthat a history of PE doubled the risk of developing early-onset PE(

10  |     Poon ET AL.

not associatedwith an increased risk of PE.44 High estrogen levels during implantationmay lead to impairedplacentationand reduceduteroplacentalcirculationaswellasdecreasednumberofuterinespi-ral arteries with vascular invasion.44–46 Women conceiving by intrau-terineinsemination,inparticularbydonorsperm,areatagreaterriskof developing PE.47–51 Those who have undergone donor ovum in vitrofertilization(IVF)appeartohaveahigherriskofPEthanthosewhohavehadautologousovumIVF.52Evidencefrom IVFpregnan-cieswithovumdonationsuggeststhattherearealteredextravilloustrophoblast and immunological changes in decidua basalis, which may impedethemodificationofthespiralarteries.53

4.3.6 | Family history of PE

AlthoughmostcasesofPEaresporadic,afamilialsusceptibilitytoPEhas been documented. Daughters or sisters of women with PE are 3–4timesmorelikelytodeveloptheconditionthanwomenwithouta family history.54–56Themodeofinheritanceseemstobecomplex,includingnumerousvariants,whichindividuallyhavesmalleffects,butcollectively contribute to an individual's susceptibility to the disor-der.Genome-wideassociationstudies(GWAS)usingsib-pairanalysishaveidentifiedplausible,yetconflicting,positionalcandidatemater-nal susceptibility genes forPE.GWASofPE affected families havedemonstratedsignificantlinkagetochromosomes2p,2q,4p,7p,9p,10q,11q,and22q.57 However, no other study has reproduced these significantorsuggestiveloci.

4.3.7 | Obesity

Thereissubstantialevidencetoshowthatobesity(BMI≥30kg/m2) con-fersa2to4-foldhigherriskforPE.58–64Theexactmechanismslinkingoverweight/obesityandPEremainunclear.Obesityisknownasastateofchronic,low-gradeinflammation,alsocalled“meta-inflammation”.65,66 Low-gradeinflammationcaninduceendothelialdysfunctionandplacen-tal ischemia by immune mediated mechanisms, which in turn lead to pro-ductionofinflammatorymediatorsresultinginanexaggeratedmaternalinflammatoryresponseanddevelopmentofPE.67

4.3.8 | Race and ethnicity

There is extensive evidence in the literature demonstrating theassociation between race and ethnicity and PE. Large populationstudies suggest that the risk of PE in Afro-Caribbean women isincreased by 20%–50%.68–72TheriskofPEisalsohigherinwomenofSouthAsianoriginthan inthoseofnon-Hispanicwhitewomen(adjustedOR1.3;95%CI,1.2–1.4).73 IncreasedriskofPEreflectsthemetabolic profiles of nonpregnantwomen associatedwith anincreased susceptibility to cardiovascular disease.74–76 BothAfro-CaribbeanandSouthAsianwomenaremoresusceptibletodevel-oping chronic hypertension, diabetes mellitus, and cardiovascular disease.Inalargeprospectiveobservationalcohortstudyofmorethan 79 000 singleton pregnancies recruited in London, UK, the risk of PE was significantly higher in women of Afro-Caribbean

andSouthAsian racialorigincomparedwithwhitewomen.77 The increasedriskremainssignificantevenafteradjustingforothercon-foundingriskfactorsforPE.

4.3.9 | Comorbidities

Therearecertainmedicalconditionsthatpredisposewomentodevel-oping PE. These include hyperglycemia in pregnancy (pre-pregnancytype 1 and type 2 diabetes mellitus, overt diabetes in pregnancy, and gestational diabetes requiring insulin treatment), pre-existing chronichypertension, renal disease, and autoimmune diseases such as sys-temiclupuserythematosus(SLE)andantiphospholipidsyndrome(APS).Recently,asystematicreviewandmeta-analysisevaluatedclinicalriskfactorsat less thanorequal to16weeksofgestation in25356655pregnant women in 27 countries.78PatientswithahistoryofchronichypertensionhaveahigherriskofdevelopingPEthanthosewithoutthiscondition(relativerisk[RR]5.4;95%CI,4.0–6.5).Pre-existingdia-betesmellitus,APS,SLE,andchronickidneydiseasearealsoassociatedwithanincreasedriskofdevelopingPE(RR3.7;95%CI,3.1–4.3;RR2.8;95%CI,1.8–4.3;RR2.5;95%CI,1.0–6.3;andRR1.8;95%CI,1.5–2.1,respectively).78

Interestingly,pre-existingdiabetesmellitusandPEsharemanyriskfactors including advanced maternal age, nulliparity, pre-pregnancyobesity, nonwhite racial propensity, and multiple pregnancy.79,80 Several common pathological pathways are present in both condi-tions.Theseincludeendothelialdysfunction(e.g.lowerflow-mediateddilatation),81,82 imbalance of angiogenic factors,81,83 increasedoxida-tivestress(e.g.lowtotalantioxidantstatus,highfreeradicals),84 and dyslipidemia (e.g. increased triglycerides).85,86 PE is a risk factor forfuture type 2 diabetes.87–90Thisrelationshipisstillevidentevenwhenwomenwho have PEwith gestational diabetes are excluded. Bothconditionsareassociatedwithinsulinresistance91–97 and women with PEhaveanincreasedriskofmetabolicsyndromeafterdelivery.98–100

4.4 | Maternal and perinatal morbidity and mortality associated with pre- eclampsia

4.4.1 | Maternal morbidity and mortality

The most common cause of death in women with PE is intracranial hemorrhage.101Otherseriouscomplicationsincludeplacentalabrup-tion,HELLPsyndrome,acutepulmonaryedema,respiratorydistresssyndrome, and acute renal failure.102

Chesley et al.103werethefirsttoproposetheconceptthatpreg-nancyisastresstest,basedontheobservationthatpregnantwomenwho have never developed PE have a lower risk of cardiovasculardisease than the general female population; whereas women witheclampsiahaveasimilarriskofcardiovasculardiseasein later lifeas

FIGOacknowledgesthemanymaternalriskfactorsthatareasso-ciated with the development of pre-eclampsia, which must betakenintoconsiderationforscreeningpractices.

     |  11Poon ET AL.

appropriately matched women with unknown pregnancy history.Therefore, while PE may not directly cause cardiovascular disease in later life, pregnancy itself acts as a challenge test to reveal under-lying metabolic risk factors for atherosclerosis and cardiovasculardisease.103 Evidence in support of this hypothesis is that PE and car-diovasculardiseasesharemanyrisksfactors,includingobesity,insulinresistance, diabetes mellitus, underlying hypertension, and dyslipid-emia.103–106A recentmeta-analysis demonstrated thatwomenwithpreviousPEhaveaRRof3.13(95%CI,2.51–3.89)forfuturedevelop-mentofchronichypertension,anORof2.28(95%CI,1.87–2.78)forfuturecardiovasculardisease,andanORof1.8(95%CI,1.43–2.21)for cardiovascular accident.107 It has been observed that the earlier the onsetofPE,themoreseveretheconditionandthehighertheriskofdevelopingsubsequentcardiovasculardisease.108

Compared with normotensive women, women with PE are also more likely to havemicroalbuminuria (amarker of renal damage) at3–5years after delivery.109 PEmay adversely impact future kidneyfunction sinceglomerular endotheliosis—a typical renal lesion inPEthatwas previously thought to resolve soon after delivery—can beobservedlongafterpregnancyinsomewomenwhohadPE.110Apro-spectivecohortstudyreportedanassociationbetweenPEandsubse-quentend-stagerenaldisease(RR4.7;95%CI,3.6–6.1).111PatientswithahistoryofPEshouldbeawareofthe increasedriskoffuturecardiovascular disease,107,108 metabolic syndrome,112,113 and chronic or end-stage renal disease.111 It remains to be determined whether lifestylemodificationsaswellasclosemonitoringforsignsandsymp-tomsofmetabolicsyndromesafterdeliveryamongpatientswithPEcanreducetheserisks.114

4.4.2 | Perinatal morbidity and mortality

PEisassociatedwithanumberofshort-andlong-termperinatalandneonatal complications, including death (Table3). These aremostlyrelatedtobirthweightandgestationalageatdeliveryandarethere-foremainlyattributedtoearly-onsetPE.

PE is commonly associated with placental lesions. The underly-ingvascularmanifestationsandthepresenceofoxidativestressandendothelial damage can lead to fetal growth restriction (FGR)withunderlyinghypoxiaandacidosis.Amulticenterprospectivestudyof30 639 unselected women with singleton pregnancies demonstrated

that in 614 (2%) women who developed PE there was an inverse signif-icantassociationbetweenthegestationalageatdeliveryandpreva-lenceofsmallforgestationalage(SGA)(r=−0.99,P

12  |     Poon ET AL.

fromBritishColumbiareportedprovincialoutcomesforinfantsbornbetween22and25weeksover17years (n=341).124 Some 20% of survivors hadmoderate disability (defined as cerebral palsy, or IQ2–3 SDs below the mean, or sensorineural hearing loss corrected with aids or visual impairment worse than 20/70), whereas 10% of survivors had severe disability (defined as nonambulatory cerebralpalsy, IQ more than three SDs below the mean, hearing loss not cor-rected, or legal blindness).

Regarding impact in adult life, the publication byOsmond andBarker125 suggested that the in-uteroenvironmentcould influenceadult health and disease state. Their hypothesis states that sub-optimal in-uteronutrient supply, as seen inplacental insufficiency,through metabolic and hormonal adaptations and altered organmorphology, leads to increased risksof insulin resistance,diabetes

mellitus, coronary artery disease, and hypertension. Thus, both the short-termandlong-termconsequencesofPE,intermsofimpactonindividualhealth, thefinancial costs inproviding theneededacuteintensivecare,andthelong-termconsequencestohealthandhumancapital,justifyeffortstofindeffectiveearlypredictionandpreventivestrategies.

FIGOrecommendsandsupportsthecallforgreaterattentionandfocus on the links between maternal health and NCDs in theSustainable Developmental Goals agenda, including efforts toensure early screening for all pregnant women for pre-existingNCDsortheirriskfactors.

     |  13Poon ET AL.

5  | FIRST TRIMESTER PREDICTION OF PRE- ECLAMPSIA

5.1 | Problems with existing methods of screening

ThecurrentapproachtoscreeningforPEistoidentifyriskfactorsfrommaternal demographic characteristics and medical history (maternalrisk factors).2–4,15,16,126,127 There are two key recommendations thathaveevolvedovertime.AccordingtotheNationalInstituteforHealthandClinicalExcellence(NICE)intheUK,womenshouldbeconsideredtobeathighriskofdevelopingPEiftheyhaveanyonehigh-riskfac-tor (hypertensive disease in previous pregnancy, chronic hypertension, chronic renal disease, diabetes mellitus, or autoimmune disease) or anytwomoderate-riskfactors(nulliparity,age≥40years,BMI≥35kg/m2, family history of PE, or interpregnancy interval >10 years).16 In the USA, the American College of Obstetricians and Gynecologists(ACOG)issuedtheHypertensioninPregnancyTaskForceReportrec-ommendingdailylow-doseaspirinbeginninginthelatefirsttrimesterforwomenwithahistoryofearly-onsetPEandpretermdeliveryatlessthan34weeksofgestation,orforwomenwithmorethanonepriorpregnancy complicated by PE.128 The US Preventive Services TaskForcepublishedasimilarguideline,althoughthelistofindicationsforlow-doseaspirinusewasmoreexpansive.129AnupdatedversionoftheUSPreventiveServicesTaskForceguidelinehasnowbeenendorsedbyACOG,theSocietyforMaternal-FetalMedicine,andtheAmericanDiabetes Association.130 Low-dose aspirin prophylaxis at 81mg/dfrom12and28weeksofgestation(optimallyat30kg/m2, family history of PE, sociodemo-graphiccharacteristics,andpersonalhistoryfactors).128 The approach recommendedbyNICEandACOGessentiallytreatseachriskfactorasaseparatescreeningtestwithadditivedetectionrateandscreen-positiverate.Althoughrecognitionofmaternal risk factorsmightbeuseful inidentifyingat-riskwomeninclinicalpractice, it isnotasuf-ficienttoolfortheeffectivepredictionofPE.131 In screening with use ofNICEguidelines,thedetectionrateis39%forpretermPEand34%fortermPEat10.3%false-positiverate.TherespectivedetectionratesinscreeningwithuseoftheUSPreventiveServicesTaskForcerecom-mendationswere90%and89%,at64.3%false-positiverate.132

5.2 | Screening with biomarkers

AnalternativeapproachtoscreeningforPE,whichallowsestimationofindividualpatient-specificrisksofPErequiringdeliverybeforeaspeci-fiedgestation,istouseBayestheoremtocombinetheaprioririskfrommaternalcharacteristicsandmedicalhistorywiththeresultsofvariouscombinationsofbiophysicalandbiochemicalmeasurements.Extensiveresearchinthelastdecadehasledtotheidentificationoffourpotentiallyusefulbiomarkersat11–13weeksofgestation:meanarterialpressure(MAP),uterinearterypulsatilityindex(UTPI),serumpregnancy-associated

plasmaproteinA(PAPP-A),andserumplacentalgrowthfactor(PLGF).The algorithm was originally developed from a study of 58 884 single-tonpregnanciesat11–13weeksofgestation,including1426(2.4%)thatsubsequentlydevelopedPE.TheestimateddetectionratesofpretermPEandallcasesofPE,atfixedfalse-positiverateof10%,were77%and54%.133Subsequently,datafromprospectivescreeningin35948single-tonpregnancies,including1058pregnancies(2.9%)thatexperiencedPE,wereusedtoupdatetheoriginalalgorithm.Thedetectionratesofpre-termPEandtermPEwere75%and47%,respectively,atfalse-positiverate of 10%.134Thepredictiveperformanceofthisalgorithmwasexam-inedinaprospectivemulticenterstudyof8775pregnancies,including239(2.7%)casesthatdevelopedPE.ThedetectionratesofpretermPEandtermPEwere75%and43%,respectively,atfalse-positiverateof10%.135InthescreenedpopulationintheASPREtrial,involving26941singletonpregnanciesfrom13maternityhospitalsinsixcountries(UK,Spain,Italy,Belgium,Greece,Israel),thedetectionratesofpretermPEandtermPE,afteradjustmentfortheeffectofaspirin,were77%and43%,respectively,atfalse-positiverateof9.2%.136

In the latest National Institute for Health Research (UK) com-missionedprospectivevalidationstudyoftheBayes-basedmodel in16 747 pregnancies, including 473 (2.8%) women who developed PE, thescreen-positiveratebytheNICEmethodwas10.3%,thedetectionrateforallPEwas30%,andforpretermPEitwas41%.Thedetectionrateofthemini-combinedtest(maternalfactors,MAP,andPAPP-A)for all PE was 43%, which was superior to that of the NICE method by 12.1% (95% CI, 7.9–16.2). In screening for preterm PE by a combina-tionofmaternalfactors,MAP,UTPI,andPLGF,thedetectionratewas82%, which was higher than that of the NICE method by 41.6% (95% CI, 33.2–49.9).137TheadditionofPAPP-Atothiscombinedmodeldidnot improve the overall screening performance.

Datafromthreereportedprospectivenoninterventionscreeningstudies at11–13weeksof gestation in a combined total of61174singleton pregnancies, including 1770 (2.9%) that developed PE, have demonstrated that screeningbyacombinationofmaternal risk fac-tors,MAP,PLGF, andUTPI andusing a risk cut-offof 1 in 100 forpretermPE inwhitewomen, the screen-positive ratewas10%anddetectionratesforearly-onset,preterm,andtermPEwere88%,69%,and40%,respectively.Withthesamemethodofscreeningandriskcut-offinwomenofAfro-Caribbeanracialorigin,thescreen-positiveratewas34%anddetectionratesforearly-onset,preterm,andtermPEwere100%,92%,and75%,respectively.138

AsecondaryanalysisofdatafromtheASPREtrialofatotalof34573womenwithsingletonpregnancieswhounderwentprospectivescreen-ing for preterm PE, including 239 (0.7%) cases of preterm PE, has shown thatinACOGorNICEscreen-positivewomenwhoarescreennegativebytheBayes-basedmethod,theriskofpretermPEisreducedtowithinorbelowbackgroundlevels.ThestudydemonstratedthatatleastoneoftheACOGcriteriawasfulfilledin22287(64.5%)pregnanciesandtheincidence of preterm PE was 0.97% (95% CI, 0.85–1.11). In the subgroup thatwasBayes-methodscreenpositive,theincidencewas4.80%(95%CI,4.14–5.55); inthosethatwerescreennegative itwas0.25%(95%CI,0.18–0.33%),andtherelativeincidenceinBayes-methodnegativetoBayes-methodpositivewas0.051(95%CI,0.037–0.071).In1392(4.0%)

14  |     Poon ET AL.

pregnanciesatleastoneoftheNICEhigh-riskcriteriawasfulfilledandinthis group the incidence of preterm PE was 5.17% (95% CI, 4.13–6.46). InthesubgroupsofscreenpositiveandscreennegativebytheBayesmethod the incidence of preterm PE was 8.71% (95% CI 6.93–10.89%) and0.65%(95%CI,0.25–1.67),respectively,andtherelativeincidencewas 0.075 (95% CI, 0.028–0.205). In 2360 (6.8%) pregnancies with at leasttwooftheNICEmoderate-riskcriteria,theincidenceofpretermPEwas1.74%(95%CI,1.28–2.35).InthesubgroupsofscreenpositiveandscreennegativebytheBayesmethodtheincidencewas4.91%(95%CI,3.54–6.79)and0.42%(95%CI,0.20–0.86),respectively,andtherelativeincidence was 0.085 (95% CI, 0.038–0.192).139 These results provide furtherevidencetosupportrisk-basedscreeningusingbiomarkers.

Thereisnowasubstantialbodyofevidencetosupportrisk-basedscreeningforpretermPEusingvariousbiomarkers.ThisapproachtoscreeninghasalsobeenvalidatedprospectivelyincountriesotherthanEurope.140–143Achecklist-basedmethodofscreeningusinginforma-tionfrommaternalhistorydoesnotperformaswellandcannolongerbeconsideredsufficientforpredictingpretermPEeffectively.

5.3 | Recommendations

FIGOrecommendsthefollowingfirst-trimesterscreeningproceduresfor singleton pregnancies.

5.3.1 | Maternal characteristics and medical history

Best practice recommendation: Maternal characteristics, medicalhistory and obstetric history (as shown inBox2)must be recordedaccurately.

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

Evidencetosupporttheinclusionoftheabove-listedmaternalriskfactorsinamultivariateregressionalgorithmoriginatesfromascreeningstudyof 120492 singletonpregnancies at 11–13weeks of gestation,including 2704 (2.2%) pregnancies that experienced PE.A competingriskmodelhasbeenutilizedtoproducerisksforPE,basedonacontin-uousmodel for thegestationalageatdeliverywithPE, treatingbirthsfromcausesotherthanPEascensoredobservations.144 This approach assumesthat,ifthepregnancyweretocontinueindefinitely,allwomenwouldexperiencePEandthatwhethertheydosoornotbeforeaspeci-fiedgestationalagedependsoncompetitionbetweendeliverybeforeorafterdevelopmentofPE.Theeffectofvariablesfrommaternalcharacter-isticsandhistoryistomodifythedistributionofgestationalageatdeliv-erywithPEsothatinpregnanciesatlowriskforPEthegestationalagedistributionisshiftedtotherightwiththeimplicationthat,inmostpreg-nancies, delivery will actually occur before development of PE (Fig. 1). In high-riskpregnanciesthedistributionisshiftedtotheleftandthesmallerthemeangestationalagethenthehighertheriskforPE(Fig.1).

Inthisriskfactor-basedmodel,increasedriskforPE,withaconse-quent shift in the Gaussian distribution of the gestational age at

delivery with PE to the left, is related to advancing maternal age,increasingweight,Afro-CaribbeanandSouthAsianorigin,medicalhis-toryofchronichypertension,diabetesmellitusandSLEorAPS,familyhistoryandpersonalhistoryofPE,andconceptionbyIVF.TheriskforPE decreases with increasing maternal height and in parous women withnopreviousPE;inthelatter,theprotectiveeffect,whichisrelatedinverselytotheinterpregnancyinterval,persistsbeyond15years.Atascreen-positiverateof11%,asdefinedbyNICE,thenewmodelpre-dicted40%and48%ofcasesofallPEandpretermPE,respectively.144 Theriskfactor-basedmodelhasbeenfurtherimprovedwiththeinclu-sionofgestationalageatdeliveryinthepreviouspregnancy.136

5.3.2 | Measurement of blood pressure

FIGOsupportstheuseofrisk-basedscreeningusingbiomarkersfor first-trimester prediction of pre-eclampsia over screeningmethodsthatusematernaldemographiccharacteristicsandmedi-calhistory(maternalriskfactors)only.

MAP is calculated from systolic (sBP) and diastolic blood pressure (dBP) readings. The measured sBP and dBP will be automatically converted to MAP by the risk calculator.

MAP=dBP+ (sBP−dBP)∕3

Box 2 Maternal characteristics, medical history, and obstetric history for pre- eclampsia screening in the first trimester.

Maternal age, y

Maternalweight,kg

Maternal height, cm

Maternalethnicity:white,Afro-Caribbean,SouthAsian,EastAsian,Mixed

Past obstetric history: nulliparous, parous without prior PE, parous with prior PE

Interpregnancy interval in years between the birth of the last child

Gestationalageatdelivery(weeks)andbirthweightofpreviouspregnancybeyond24wk

Family history of PE (mother)

Methodofconception:spontaneous,ovulationinduction,invitrofertilization

Smokinghabit

History of chronic hypertension

Historyofdiabetesmellitus:type1,type2,insulinintake

Historyofsystemiclupuserythematosusorantiphospholipidsyndrome

Abbreviation:PE,pre-eclampsia.

     |  15Poon ET AL.

Best practice recommendation: MAP should be measured aspart of the risk assessment for PE and it should bemeasured byvalidated automated and semiautomated devices (http://www.dableducational.org/sphygmomanometers/devices_1_clinical.html#ClinTable).

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

Best practice recommendation:Womenshouldbe inasittingposi-tion,with their armswell supported at the level of their heart andanappropriate-sizedadult cuff (small

16  |     Poon ET AL.

history of PE in the previous pregnancy, interpregnancy interval, chronic hypertension, and diabetes mellitus. Consequently, themeasurementofMAPisconvertedtoamultipleofmedian(MoM),adjusting for these associated maternal characteristics and gesta-tional age146(AppendixS1).

Poon et al.147firstreportedthevalueofMAPmeasuredbyvalidatedautomatedbloodpressuredevicesaccordingtoastandardizedproto-colat11–13weeksofgestationforthepredictionofPE.148 Maternal blood pressure was determined in 5590 singleton pregnant women by automated devices and appropriately trained doctors. For MAPaloneand incombinationwithmaternalhistory, thedetectionratesforPE,at10%false-positive rate,were38%and63%, respectively.A follow-up study ofmore than 9000 pregnancies at 11–13weeksof gestation compared the screening performance of sBP, dBP, andMAP.149AlthoughsBP,dBP,andMAPwereallfoundtoberaisedinwomenwhosubsequentlydevelopedPE,MAPperformedbestasamarker,withadetectionrateforearly-onsetPE,increasingfrom47%(basedonmaternalfactorsalone)to76%(basedonacombinationofmaternalfactorsandMAP)atafalse-positiverateof10%.149

Methodologically, based on the protocol of theNationalHeartFoundation of Australia (NHFA),148 blood pressure is measured in botharmsandaminimumoftworecordingsaremadeatone-minuteintervalsuntilvariationsbetweenconsecutivereadingsfalltowithin10mmHg in sBPand6mmHg indBP inbotharms.148 When this point of stability is achieved, the average of the last two stable mea-surements of the left and right arms is calculated and the higherof these two measurements from the two arms is used. However, in order to achieve the necessary point of blood pressure stability according to theNHFAprotocol, it has been shown that it is nec-essary to perform two measurements in both arms in about 50% of cases, three measurements in 25% of cases, and four measurements or more in 25%.145 In addition,whether blood pressure should betaken on the left or right arm remains controversial.The evidencesupportingsimultaneousmeasurementofbotharmsisderivedfromthe study published by Poon et al.150 In this study, the prevalence of bloodpressure interarmdifference (IAD),definedas IADofgreaterthan 10mmHg of sBP and dBPwas determined in 5435womenduringthefirsttrimesterofpregnancy.TheIADofsBPanddBPwasfoundin8.3%and2.3%ofnormalpregnantwomen,respectively.150 Asimplifiedprotocolforbloodpressuremeasurement(asdescribedabove) has been developed through a study of 25 505 singleton preg-nancies where blood pressure measurements were made using a val-idatedautomaticdeviceat11–13weeksofgestation.145 The results demonstrated that performance of screening for PE by taking theaverage of two measurements from both arms is comparable with theNHFAprotocol.

5.3.3 | Measurement of biochemical markers

Best practice recommendation:Infirst-trimesterscreening,thebestbiochemical marker is PLGF. PAPP-A is useful if measurements ofPLGF and UTPI are not available.

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

MaternalserumconcentrationsofPLGFandPAPP-Aaremeasuredby one of three commercially available automated devices. Quality control shouldbeappliedtoachieveconsistencyofmeasurementofbiomarkers.

5.3.3.1 | Placental growth factorPLGFisaglycosylateddimericglycoproteinsecretedbytrophoblasticcells and is part of the angiogenic vascular endothelial growth factor (VEGF)family.ItbindstoVEGFreceptor1(VEGFR-1),whichhasbeenshowntoincreaseduringpregnancy.PLGFissynthesizedinvillousandextravillous cytotrophoblasts, and has both vasculogenic and angio-genetic functions. Its angiogenetic abilities havebeen speculated toplay a role in normal pregnancy, and changes in the levels of PLGF or its inhibitory receptors have been implicated in the development of PE.151–153SeveralstudieshaveshownthatwomenwhosubsequentlydevelopPEhavesignificantlylowermaternalPLGFconcentrationsinthefirsttrimesterthanthosewithnormalpregnancies.154–157 This bio-markeralonehasadetectionrateof55%and33%,respectively,at10%false-positiverate,fortheidentificationofbothearly-andlate-onsetPE.158Asystematicreviewandmeta-analysisdemonstratedthatPLGFis superior to theotherbiomarkers forpredictingPE.159Specifically,maternalPLGFconcentrationsaloneachieveadetectionrateof56%at9%false-positiverateforthepredictionofearly-onsetPE.159

Several factorsaffect thevaluesofPLGF inpregnantwomen.Acohort study of more than 42 000 pregnancies, including 33 147 mea-suredbytheDELFIAXpresssystem(PerkinElmerLifeandAnalyticalSciences,Waltham,MA, USA), 7065measured by the Cobas e411system (Roche Diagnostics, Risch-Rotkreuz, Switzerland), and 2143measured by the B·R·A·H·M·S KRYPTOR compact PLUS (ThermoFisherScientific,Waltham,MA,USA),wasconductedtoevaluatetherelationship of PLGFwith analyzers andmaternal characteristics.138 Significant independent contributions to PLGF values are providedbythethreeanalyzersas listedabove,aswellasbygestationalage,maternalage,weight,racialorigin,cigarettesmoking,ahistoryofPEinthepreviouspregnancy,diabetesmellitus,andIVF.

5.3.3.2 | Pregnancy- associated plasma protein APAPP-Aisametalloproteinaseinsulin-likegrowthfactor(IGF)bindingprotein secretedby the syncytiotrophoblast thatplays an importantrole in placental growth and development. It enhances the mitogenic functionoftheIGFs.PEhasbeenshowntobeassociatedwithalowlevelofcirculatingPAPP-A,which ispresumablyduetothereducedavailabilityofunboundIGFstofulfiltheirfunctionalroleonacellularlevel.PAPP-Aisawell-establishedbiochemicalmarkerinthescreeningof trisomies21,18,and13. Ineuploidpregnancies,aPAPP-AMoMvalueatlessthanthe5thpercentile(0.4MoM)ispresentin8%–23%ofwomenwithPE.Therefore,asasinglemarkeritisnotanaccuratepredictive test for PE.160–162 A recent systematic review andmeta-analysis, including eight studies involving 132 076 pregnant women in

     |  17Poon ET AL.

thefirsttrimester,demonstratedthatthematernalPAPP-Aconcentra-tionoflessthanthe5thpercentileisassociatedwiththeriskofdevel-opingPEwithanORof1.94(95%CI,1.63–2.30). Ithasadetectionrateof16%(9%–28%)at8%false-positiveratetopredictPE.163

In a cohort study of more than 94 000 pregnancies, the rela-tionship between PAPP-A,measured by theDELFIA Xpress system(PerkinElmerLifeandAnalyticalSciences),andmaternalcharacteristicswas evaluated.164 Significant independent contributions to PAPP-Aareprovidedbygestationalage,maternalweight,height,racialorigin,cigarettesmoking,diabetesmellitus,methodofconception,previouspregnancywithorwithoutPE,andbirthweightZ-scoreoftheneonatein thepreviouspregnancy.ThemeasurementsofPLGFandPAPP-AshouldbeconvertedtoMoMs,adjustingfortheseassociatedmaternalcharacteristics,analyzers,andgestationalage138(AppendixS1).

5.3.4 | Measurement of uterine artery pulsatility index

Best practice recommendation: Where feasible UTPI should be measured.Atransabdominalultrasoundscanshouldbedoneat11+0 to 13+6weeks of gestation (corresponding to fetal crown–rumplength (CRL) of 42–84mm). Gestational age must be determinedfromthemeasurementofthefetalCRL.Thesamescanisutilizedforthemeasurementoffetaltranslucencythicknessanddiagnosisofanymajorfetaldefects.ForthemeasurementofUTPI,asagittalsectionof the uterus is obtained and the cervical canal and internal cervical osareidentified.Subsequently,keepingthetransducerinthemidlineitisgentlytiltedtothesideandcolorflowmappingisusedtoidentifyeachuterinearteryalongthesideofthecervixanduterusattheleveloftheinternalos(Fig.3).Pulsed-waveDopplerisusedwiththesam-plinggatesetat2mmtocoverthewholevesselandcareistakentoensurethattheangleofinsonationislessthan30°.Whenthreesimi-larconsecutivewaveformsareobtained(Fig.3),theUTPIismeasuredandthemeanUTPIoftheleftandrightarteriesiscalculated.165 The measurement of UTPI must be carried out by sonographers who have received the appropriate certificate of competence from the FetalMedicineFoundation(FMF)(www.fetalmedicine.org).

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

TheDopplerultrasoundassessingtheresistancetobloodflowinthe uterine arteries correlates with both histological studies and clin-ical severityofPE.Thisbiophysicalmarkerprovidesausefulnonin-vasivemethod for the assessment of theuteroplacental circulation.Studies have shown that a significant decrease of resistance in thespiral arteries occurswith advancing gestation,which is in keepingwith physiological changes throughout pregnancy.166,167 Persistent high impedance to flow in the uterine arteries is evidence of poorplacentation thatmanifests itself in the formof abnormal uteropla-centalflowvelocitywaveforms.HistologicalexaminationofplacentalbedbiopsiesofpregnanciesaffectedbyPEhasshownthatabsenceofphysiological changes of the spiral arteries is found more commonly in cases with high UTPI.168

Methodologically, the measurement of UTPI at the level of the internalosduringthefirsttrimesterismorereproduciblethanthoseobtained at the level of external iliac vessels crossover.169 In addi-tion,UTPIcanbeachievedattheleveloftheinternalcervicalosinagreaterproportionofwomenthanatthelevelofexternaliliacvesselcrossover.169

SeveralfactorscanaffectthevaluesofUTPIinpregnantwomen.Acohortstudyofmorethan83000pregnancieswasconductedtoevaluatetherelationshipbetweenUTPIandmaternalcharacteris-tics.138SignificantindependentcontributionstoUTPIareprovidedbygestationalage,maternalage,weight,racialorigin,ahistoryofPE in the previous pregnancy, and type 1 diabetes. Hence, before comparing the values between affected and unaffected groups,the UTPI value needs to be adjusted for these associated mater-nal characteristicsandgestational agebyconverting it toaMoM(AppendixS1).

Alargemeta-analysisoffirst-trimesterUTPImeasurementfortheprediction of PE included eight studies for the prediction of early-onsetPE(n=41692women)andelevenstudiesforthepredictionofPEofanygestation(n=39179women).170Thefirst-trimesterabnor-malUTPI isdefinedasgreaterthanthe90thpercentile,achievingadetectionrateof48%,at8%false-positiverate,fortheidentificationof early-onset PE. The detection rate for predicting late-onset PEreducesto26%ata7%false-positiverate.

[Correctionaddedon18June2019,afterfirstonlinepublication:Secondlastsentence‘Thefirst-trimesterabnormalUTPIisdefinedasgreaterthanthe90thpercentile’hasbeencorrectedforaccuracy inthisversion.]

F I G U R E 3   Identification of the uterine artery at the level of the internal os (left) and typical waveforms of the uterine artery Doppler in the firsttrimesterofpregnancy.CourtesyoftheFetalMedicineFoundation.[Colourfigurecanbeviewedatwileyonlinelibrary.com]

http://www.fetalmedicine.orgwww.wileyonlinelibrary.com

18  |     Poon ET AL.

The International Society of Ultrasound in Obstetrics andGynecology(ISUOG)hasrecentlypublisheditspracticeguidelineontheroleofultrasoundinscreeningforandfollow-upofPE.165

FIGOacknowledgesandendorsestheguidancefromISUOGwithregard to UTPI measurement methodology.

5.3.5 | Combined risk assessment

Best practice recommendation: Published algorithms should be used forconvertingthemeasuredvaluesofMAP,PLGF,andUTPI,withorwithoutPAPP-A, intoMoMsasdetailed above.Patient-specific riskforpretermPEiscalculatedusingtheBayes-basedmethod.Theriskcalculatorisavailablefreeofchargeonthewebpagehttps://fetalmedi-cine.org/research/assess/preeclampsia and on the FMF mobile app. It isalsoavailableonmedicalrecordssoftware.Awomanisconsideredhighriskwhentheriskisgreaterthanorequalto1in100basedonthefirst-trimestercombinedtestwithmaternal riskfactors,MAP,PLGF,and UTPI.1,136,171

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

Best practice recommendation:Basedonexistingevidence,thefirst-trimestercombinedtestismostpredictiveofpretermPEbutnottermPE. Thebestmodel is theone that combinesmaternal risk factors

withMAP,PLGF,andUTPI.Theperformanceofscreeningforpre-term PE of various combinations of the first-trimester test, basedondatafromthreepreviouslyreportedprospectivenoninterventionscreening studies, including a combined total of 61 174 singleton pregnancies, including 1770 (2.9%) that developed PE, is illustrated in Table 4.

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

Pragmatic practice recommendation: Where it is not possible to measurethebiochemicalmarkersand/orUTPI,thebaselinescreeningtestshouldbeacombinationofmaternalriskfactorswithMAP,andnotmaternalriskfactorsalone.PAPP-AisusefulifmeasurementsofPLGFandUTPI arenot availableThesevariations to the combinedtestwouldleadtoareductionintheperformancescreening.

Quality of evidenceStrength of recommendation

Moderate ⊕⊕⊕O Conditional

Asdemonstratedabove,biomarkersarebestusedinthecombina-tionstrategyforthepredictionofPE.Arecentsystematicreviewhasbeen conducted to evaluate theperformanceof simple riskmodels

T A B L E 4  Detectionrates,atscreen-positiverateof10%,ofpretermPEandtermPEbymaternalfactors,biomarkers,andtheircombination.a

Method of screeningRisk cut- off for PE

     |  19Poon ET AL.

(maternalcharacteristicsonly)versusspecializedmodelsthatincludespecialized tests such as the measurement of MAP, UTPI, and/ormaternalbiochemicalmarkersforthepredictionofPE.Seventymod-elsfrom29studieshavebeenidentified(17modelstopredictPE,31modelstopredictearly-onsetPE,and22modelstopredictlate-onsetPE).Amongthem,22weresimplemodelswhile48wereclassifiedasspecializedmodels.Comparingthesimpleandspecializedmodels,thelatterperformedbetterthanthesimplemodelsinpredictingearly-andlate-onsetPE,achievinganadditional18%(95%CI,0–56)indetectionrate for the predictionof PE at a fixed false-positive rate of 5%or10%.172Therefore,acombinationofvarioustestsratherthanasingletestisrecommendedforthepredictionofPE.

5.3.6 | Contingent screening

Pragmatic practice recommendation: Where resources are limited, routinescreeningforpretermPEbymaternal factorsandMAPinallpregnancies and reserving measurements of PLGF and UTPI for a sub-groupofthepopulation,selectedonthebasisoftheriskderivedfromscreeningbymaternalfactorsandMAPalonecanbeconsidered(Fig.4).

Quality of evidenceStrength of recommendation

Moderate ⊕⊕⊕O Conditional

In a prospective screening study including more than 120000singleton pregnancies, the performance of screening for preterm PE by this two-stagestrategywasexamined.Atafixedscreen-positiverateof10%,adetectionrateof71%wasachievedbythistwo-stagescreening,withscreeningbymaternalfactorsandMAP,basedontheabove-described combined algorithm, at 11+0–13+6weeks of gesta-tioninthefirststageandreservingmeasurementofPLGFandUTPIforthesecondstageandfor30%ofthepopulation.173

5.3.7 | Multiple pregnancies

Pragmatic practice recommendation: Thesamefirst-trimestercom-bined test for PE in singleton pregnancies can be adapted for screen-ingintwinpregnancies.ItleadstothedetectionofnearlyallaffectedcasesofPEbutatahighscreen-positiverate.

Quality of evidenceStrength of recommendation

Moderate ⊕⊕⊕O Conditional

InaprospectivescreeningstudyforPEin2219twinpregnanciesundergoingroutinefirst-trimestercombinedscreeningforaneuploidyand subsequently delivering two phenotypically normal live or still-bornbabiesatgreaterthanorequalto24weeksofgestation,theinci-dence of PE in dichorionic and monochorionic twin pregnancies was shown to be increased by four-fold and three-fold, respectively.174 IntwinpregnanciesthatdevelopedPE,thevaluesofMAPandUTPIwereincreasedandthevaluesofPLGFandPAPP-Aweredecreased.Thedistributionsoflog10MoMvaluesofbiomarkerswithgestationalage at delivery were similar to those that were previously reported in singleton pregnancies and it was therefore decided that the same first-trimestercombinedtestforsingletonpregnanciescouldbeappli-cabletotwinpregnancies.Inamixedpopulationofsingletonandtwinpregnancies,combinedscreeningbymaternalfactors,MAP,UTPI,andPLGFandatariskcut-offof1in75forpretermPE,thedetectionratesof preterm PE and all PE in singleton pregnancies were 77% and 57%, respectively, at a screen-positive rate of 13%; the respective ratesfortwinpregnancieswere99%and97%,atascreen-positiverateof75%.174TheadditionofPAPP-Adidnotimprovetheperformanceofscreening.

F I G U R E 4  Two-stagescreeningstrategyforpretermPEinwhichthewholepopulationundergoesfirst-stagescreeningbymaternalfactorsandMAPandaselectedproportionofthoseconsideredtobeatintermediateriskundergosecond-stagescreeningbyPLGFandUTPI.AdaptedwithpermissiongrantedbyWiley,fromWrightetal.173

Total population

First-stage screening

Screen positive

Screen positive

Second-stage screening

Screen negative

Screen negative

20  |     Poon ET AL.

6  | FIRST TRIMESTER PREVENTION OF PRETERM PRE- ECLAMPSIA

The current approach to prevention of PE is to commence low-dose aspirin at 75mg or 81mg daily in high-riskwomen as locallydefined.2–4,16,128Low-doseaspirintreatmentinpregnancyisthoughttoprevent thedevelopmentofPEby inhibiting thebiosynthesisofplacentalthromboxaneA2withminimaleffectsonvascularprostacy-clin levels.175TheenzymecyclooxygenaseplaysapivotalroleintheproductionofbothprostacyclinandthromboxaneA2.Aspirin inhib-its endothelial cyclooxygenase176 and this process is irreversible in platelets,where theenzyme is inhibited for their entire lifespan. Incontrast,whentheenzyme is resynthesized inendothelialcells, theprostacyclinproductionisre-establishedrelativelyrapidly.Thisselec-tive inhibitionof cyclooxygenaseand the resultingalteration in theprostacyclintothromboxaneA2ratiointheplacentaformsthebasisof using aspirin to prevent or delay the onset of PE.

Crandon and Isherwood177 demonstrated that nulliparous women whohadtakenaspirinmorethanonceafortnightthroughoutpreg-nancyhad a lower risk of PE than thosewhohadno reportedhis-toryofaspirinconsumption.In1985,anopen-labelrandomizedtrialshowedthatamongwomenatriskforPEorFGR,basedonobstetrichistory, pregnancies in women who received 300 mg of dipyridamole and150mgofaspirinbeginningat12weeksofgestationuntildeliv-erywerenotcomplicatedbyPE,fetalloss,andsevereFGRcomparedwiththoseinthenoninterventiongroup.178Alandmarkmeta-analysis,including 31 randomized trials of PE prevention, including 32217pregnancies, showed that patientswho received antiplatelet agentsespeciallyaspirinforthepreventionofPE,hada10%reductionofPE(RR0.90;95%CI,0.84–0.97),pretermbirthbefore34weeksofgesta-tion,andseriousadversepregnancyoutcomes(PE,delivery

     |  21Poon ET AL.

This recommendation is supported by the evidence from theASPRE trial.182 The choice for recommending night-time consump-tionof low-doseaspirin isbasedon the results froma randomized,double-blind,placebo-controlled,chronotherapytrialon350high-riskwomen.Thistrialdemonstratedthatwomentakinglow-doseaspirinat100mg, comparedwithplacebo,hada significantly lowerhazardratio (HR)ofseriousadverseoutcomes,acompositeofPE,pretermbirth,FGR,andIUFD(HR0.35;95%CI,0.22–0.56)andthattheeventrate of serious adverse outcomeswas significantly lower inwomentakinglow-doseaspirinintheevening,comparedwithinthemorningandafternoon(HR0.19;95%CI,0.10–0.39).184

The latest systematic reviewandmeta-analysis,which included16 randomized controlled trial studies for a totalof18907partic-ipants, demonstrated that administration of aspirinwas associatedwithareduction intheriskofpretermPE(RR0.62;95%CI,0.45–0.87).However,therewasnosignificanteffectontermPE(RR0.92;95% CI, 0.70–1.21). Only the subgroup in which aspirin was started at less thanorequal to16weeksofgestationatadoseofgreaterthan or equal to 100mg/dwas associatedwith a reduction in thefrequency of pretermPE (RR 0.33; 95%CI, 0.19–0.57;P=0.0001). Aspirinstartedafter16weeksoradministeredinadailydoseoflessthan100mgwasnotassociatedwithasignificantreductioninratesof preterm or term PE.185 Suggested aspirin dosages are provided in Table 5.

Low-doseaspirin isdefinedasdosageof less than300mg/d. In1979, Masotti etal.186 demonstrated that aspirin 2.5–3.5mg/kg isthe required dosage to produce a consistent inhibition of plateletaggregation with slight inhibition of prostaglandin production. Therecommendeddosageof150mg/dwouldbesufficientforanaveragewomanwithaweightof65kgatbooking.

Pragmatic practice recommendation: Where it is not possible to source the above suggested aspirin regime, the minimum dosage of aspirintobeprescribedtohigh-riskwomenshouldbe100mg/d.

Quality of evidenceStrength of recommendation

Moderate ⊕⊕⊕O Conditional

Best practice recommendation:High-riskwomenmustbe informedand counselled about the importance of treatment adherence and assessed for compliance at each prenatal visit.

Quality of evidenceStrength of recommendation

High ⊕⊕⊕⊕ Strong

InasecondaryanalysisofdatafromtheASPREtrial,theinflu-enceofadherenceonthebeneficialeffectofaspirininpreventionofpretermPEwasevaluated.Thechoiceofcut-offforgoodcom-pliancewasredefinedasgreaterthan90%,whichwasbasedonanexploratoryanalysisof the treatmenteffect.PretermPEoccurredin 5/555 (0.9%) participants in the aspirin groupwith adherence

greaterthanorequalto90%,in8/243(3.3%)ofparticipantsintheaspirin group with adherence less than 90%, in 22/588 (3.7%) of participants in theplacebo groupwith adherence greater thanorequalto90%,andin13/234(5.6%)ofparticipantsintheplacebogroupwithadherence lessthan90%.TheORintheaspiringroupfor preterm PE was 0.24 (95% CI, 0.09–0.65) for adherence more than 90% and 0.59 (95% CI, 0.23–1.53) for adherence less than 90%.The beneficial effect of aspirin in preventing preterm PE isdependent on adherence.187Inaddition,therewasnoevidenceofheterogeneity intheaspirineffect insubgroupsdefinedaccordingtomaternal characteristics and obstetric history,with the excep-tionofchronichypertension.Inwomenwithchronichypertension,preterm PE occurred in 10.2% (5/49) in the aspirin group and in 8.2%(5/61)intheplacebogroup(adjustedOR1.29;95%CI,0.33–5.12);therespectivevaluesinthosewithoutchronichypertensionwere 1.1% (8/749) in the aspirin group and 3.9% (30/761) in the placebogroup(adjustedOR0.27;95%CI,0.12–0.60).Prophylacticaspirin may not be as effective in lowering preterm PE risk inwomenwith chronic hypertension comparedwith other high-riskgroups. Further, in women with adherence of more than 90% the adjustedORintheaspiringroupwas0.24(95%CI,0.09–0.65), inthe subgroup with chronic hypertension it was 2.06 (95% CI, 0.40–10.71), and in those without chronic hypertension it was 0.05 (95% CI, 0.01–0.41).

Pragmatic practice recommendation: If vaginal spotting occurs itmust be duly assessed but does not necessitate stopping aspirin prophylaxis.

Quality of EVIDENCEStrength of recommendation

High ⊕⊕⊕⊕ Strong

Several systematic reviews of randomized controlled trials havedemonstrated that the use of low-dose aspirin during pregnancy isnot associatedwith hemorrhagic complications.179,188,189 The study bytheUSPreventiveServicesTaskForce,whichincludedmorethan23000 pregnantwomen, showed that the risk of placental abrup-tion (RR1.17;95%CI,0.93–1.48) andpostpartumhemorrhage (RR1.02;95%CI,0.96–1.09)didnotsignificantly increasewith theuse

T A B L E 5  Proposedaspirinregimeforpretermpre-eclampsiaprevention.

Maternal weight, kg

Daily required dosage, mg Administration, mg

22  |     Poon ET AL.

of aspirin.188Inaddition,womenwhowereexposedtolow-doseaspi-rin during pregnancy had similar mean blood loss to those who were notexposedtolow-doseaspirin.188Arecentmeta-analysisinvolving12 585 pregnant women showed that the use of aspirin at less than 100mg/dorgreaterthanorequalto100mg/d,regardlessof initia-tiontime (≤16or>16weeks),wasnotassociatedwithan increasedriskofplacentalabruptionorprepartumhemorrhage.190IntheASPREtrial,womenexposedtoaspirindidnothaveincreasedriskofbleedingadverse events.136Duringthetrial,womenwithvaginalspottingwerenotadvisedtostopthetrialmedication.

Best practice recommendation: Inwomenwith low calcium intake(

     |  23Poon ET AL.

F I G U R E 5  Pathwayofpretermpre-eclampsiascreeningandprevention.

24  |     Poon ET AL.

7  | RESOURCE- BASED APPROACH TO SCREENING

Implementation of guidelines is a constant challenge. The reality isthat most low-resource countries around the world are unable toimplementafirst-trimesterscreeningprogramforpretermPEbasedon the combined test.

Recommendationsthatarerigidandimpracticalinreal-lifesettingsare unlikely to be implemented andhencemayproduce little or noimpact.Ontheotherhand,pragmaticbutlessthanidealrecommen-dations may produce significant impact owing to morewidespreadimplementation.

The FIGO approach is three pronged: (1) to promote, encourage, and advocate ideal evidence-based guidance; (2) to offer pragmatic options for resource-constrained situations based on local experience backed by less than optimal evidence; and (3) to promote research aimed at improving the evidence base in both well-resourced and resource-constrained contexts.

FIGOrecommendationsarebasedonavailableresourcesatcoun-trylevelandevidenceoflocalpractice.Countriesworldwidefallintofour resource categories. There are also variations seenwithin anycountry.Anaffluentcountrymayhavepocketsofpoorlyfundedcareand,conversely,alow-ormiddle-resourcecountrymayhave“stateoftheart”careintheprivatesectorforaselectedfew.

High-resource countries:IncludescountriesorregionslikeNorthAmerica,WesternEurope,Japan,SouthKorea,Australia,etc.

Upper/middle-resource countries: Includes countries like Brazil,China,Colombia,Hungary,Malaysia,Mexico,Romania,SouthAfrica,Turkey,etc.

Low/middle-resource countries: Includes countries like India,Indonesia,Pakistan,Nigeria,Egypt,Vietnam,thePhilippines,etc.

Low-resource countries: Includes countries like Bangladesh,Nepal, Cambodia, Kenya, Tanzania, Uganda, Ethiopia, Congo, SierraLeone, etc.

Inlow-andmiddle-incomecountrieswhereresourcesarelimited,variationsofthefirst-trimestercombinedtestcanbeconsideredbutthebaselinetestisonethatcombinesmaternalriskfactorswithMAP.Intheabsenceofoneortwoofthebiomarker(s),riskcalculationcanstillbedonebutthedetectionratesforpretermPEwillbereduced,inturnleadingtoareductioninthetreatmenteffectsizebyaspirinprophylaxisbutmaymakethetreatmenteconomicallyfeasible.

As99%ofseriousmorbidityoccursinlow-andmiddle-incomesettings,anypredictionandpreventionstrategiesneedtobeappliedin these settings to impact on the global burden. An estimated70000 women die each year from PE. Low- and middle-incomecountries are disproportionately affected by avoidable maternaldeaths,with84%ofmortalitiesoccurringinSub-SaharanAfricaandSouthern Asia.200 Regional variation in hypertension disorders inpregnancy as a cause of maternal mortality is also seen, with the highest burden in LatinAmerica and the Caribbean,where thesedisorders are responsible for 22% of deaths compared with 14% globally.Foreverywomanthatdies,another20sufferlife-alteringmorbidity.

PredictionandpreventionofPE isan importantgoal inreducingavoidable mortality and morbidity on a global scale but enabling this in diversehealthsystemsremainsachallenge.Predictionmodelsbasedonearlymaternalcharacteristicstoimproveriskstratificationrequirecontactwith healthcare services in the first trimester.A systematicanalysis of early prenatal care visits between 1990 and 2013 showed that although worldwide coverage increased by 40%, in 2013 only 24% ofvisitswere in low-income countries comparedwith 82% in high-income countries.201Althoughprogressincoverageofprenatalcareisimprovingovertime,itremainsfarfromuniversal.Furtherunderstand-ing of contextual factors influencing barriers to prenatal care suchasacceptability,affordability,andgeographicalaccessibility iscrucialwhenconsideringapproachestoimprovequalityandcoverageofcare.

Communityoutreachstrategiessuchasmobileclinicsareanexam-pleofattemptstoincreasecoverage;however,qualityofcareinthesesettingshasrarelybeenevaluated.Across-sectionalstudycomparingqualityofprenatalcarebetweenfixedandmobileclinicsinHaitifoundlow referral rates, with 95% of women found to be hypertensive not being referred to a higher level of care to screen for proteinuria.202 This highlightstheneedtofocusonprovidereducation,consistentadher-ence to clinical guidelines, and improvement in referral pathways.

A health systems approach must be advocated for, particularlyin low- and middle-income countries where systems are weakest.Concentrating solely ondeliveryof health servicesor new technol-ogiesaimedatriskpredictionisunlikelytoworkalonetoreducetheburden of mortality and morbidity. However, it remains an import-ant goal and must be accompanied by necessary pathways to deliver effectiveprophylaxis,suchaswithaspirin.Althoughaffordable,edu-cationandprescribing/deliverypathwaysmustbeestablished if thisstrategy is tobeeffective.Workforce,availabilityofessentialdrugs,informationsystems,governance, andfinancing thereforemustalsobe addressed.

Prioritiesforunder-resourcedsettingsmaybefocusedaroundtheavailabilityofaccurate,functionalbloodpressuremonitoringdevices;facilityforassessmentofrisksuchasproteinuria;trainingofstaff(par-ticularlyifultrasoundistobeusedforscreening),includingforappro-priateescalationandmanagement;ensuringreliablesupplychainsfortheprovisionofaspirin,antihypertensives,andmagnesiumsulphate;and the availability of laboratory services. This will need to be in paral-lelwithanystrategytoimplementprevention.

Cost-effectiveness of an early PE prediction and preventionapproach has been evaluated in high-income countries and hasshown substantial cost saving.203 However, in low- and middle-income countries,multiple system-level barriers exist that impactimplementation,evaluation,andsustainabilityofsuchapproaches.Economic analysis of basic screening devices such as blood pres-suremonitors andurinedipsticks foruse in low-resource settingshasshownthatsimpledevicesaremostcost-effective.204 The more sophisticated tools for first-trimester screening must be evalu-atedinthiscontext.Moreworkisrequiredtoevaluatethebalancebetweenbetterdetectionversuscostofscreeningwithcurrentsuc-cessfulpredictionalgorithms,andtreatmentpathwayswhenappliedtoalow-incomesetting.

     |  25Poon ET AL.

A key barrier to early prediction and prevention of PE in low-resource settings is the delayedfirst prenatalvisit or even contactwithahealthcareworker.Further, inmanyplacesbloodpressure isnot beingmeasured at all.More efforts to raise awareness of thebenefitsof an early prenatalvisit (targeted atwomenof reproduc-tive age, primary healthcare workers, women's self-help groups,etc.), coupled with skills development of primary healthcare pro-viders on risk assessment, accurate blood pressure measurement,counsellingskills,andensuringaspirinavailabilityandadherencetotreatment and follow-upwill have a far greater impact onPE out-comesthanmakingmoreadvancedtestingtechnologyandprotocolsavailable.IntegratingPEriskassessmentasanintegralpartofbasic

first-trimester evaluation protocol (measuring weight, blood pres-sure, hemoglobin, blood sugar, etc.) will go a long way to improve implementation.

• All countries have an obligation to implement the best pre-eclampsiatestingandmanagementpracticestheycan.

• FIGOacknowledgesthatforglobalprogresstobemade,India,China,Nigeria,Pakistan,thePhilippines,Indonesia,Bangladesh,Brazil,andMexicomustbekeytargetsforfocusedpre-eclamp-siaattention.

26  |     Poon ET AL.

8  | COST- EFFECTIVENESS OF PRE- ECLAMPSIA SCREENING

Itiswellestablishedthatsignificanthealthcareresourceshavetobeinvested to prevent morbidity and mortality related to PE and that, as a consequence, bothmaternal and neonatal costs are inflatedcompared with an uncomplicated pregnancy. Given the relativefrequencyofPE,thesecostsaresignificantforthehealthcaresys-temandthewidespreadimplementationofaprediction/preventionstrategywouldhelpeasethisburden.Asthebesttestsforpredic-tion involve a multivariate model that includes several investiga-tivetools,manyconsiderscreeningtobecomplexandexpensive.ItisthereforeimportanttofirstrecognizethecurrentcostsofPE,and the potential benefit in spending a fraction of the sum thatwouldberecoupedthrougheffectivepreventiononcomprehensivescreening.

The reported costs of PE do vary depending on the jurisdic-tion.ExamplesofcostsincludethosereportedwithintheUSAandIrish healthcare systems.A reviewof billing data collected by theCalifornianMedi-Calhealthcareprogramestimated thecostofanuncomplicated vaginal delivery to be US $4500 in 2011 (US $4900 based on Consumer Price Index to 2017).205 The average incre-mental cost for a pregnancy complicated by hypertensive disease was US$8200, with an estimated total incremental cost for allCalifornian births of more than US $200 million. Costs were high-est forwomenwhohad severedisease requiringdelivery at early(

     |  27Poon ET AL.

inprevalenceofdiseaseandaCAD$14.4millioncostsavingtotheCanadian healthcare system. This saving was demonstrated despite theconservativenatureofsomeofthecostings.Thecostofamother/infant being delivered at less than 34weeks of gestationwas onlycostedatCAD$13268.21,thereforethecostsavingsfrompreventionofearly-onsetPEmaybeunderestimated.Thecostoffirst-trimesterscreeningwasestimatedatCAD$668.84;however,incircumstanceswherefirst-trimesteraneuploidyscreeningisperformedthis is likelylower(CAD$100pertest),whichwouldleadtoacostreduction(andfurthersavingtothehealthsystem)ofanadditionalCAD$220millionper year.

Afourthcost-effectivenessanalysis focusedoncalciumprophy-laxisandusedadecisionanalyticmodeltoexaminetheimpactofthistreatment ifprescribedtoallpregnantwomen, towomen identifiedasbeinghighriskforPE,ortowomenwithlowdietaryintakeofcal-cium.213Thesethreemodels ledtocorrespondingreductionsofdis-easeprevalenceof25%,8%,and13%,respectively—alldemonstratingcost savings ranging from €2 to 4.6 million per 100 000 pregnancies. Onceagain,thelowcostoftheinterventionmakesuniversalprophy-laxisappealing.

None of these models have adequately considered the long-term health effects of PE.Whilematernal cerebrovascular eventsare rare, Pourat et al.205estimateda lifetimecostofUS$659156

for such an event in a 25-year-oldwoman. Similarly,womenwhohadPEhavehigherrisksofothercardiovascularpathologiesinmid-dleagethatmaybeavoidedthrougheffectivefirst-trimesterscreen-ingandprophylaxis.Preterminfantshavesignificantriskofcerebralpalsyandneurodevelopmentaldelay—disabilitiesthathaveanasso-ciated estimated cost of US$38250 per year.205 These children/youngadultsalsohavehigherrisksofhypertension,type2diabetes,andmetabolicsyndrome—allassociatedwiththeirownburdensandhealthcare costs.

Further cost-effective analyses are needed to demonstrate thevalueoffirst-trimesterscreeningindifferentpopulations,withdiffer-entdiseaseprevalenceanddifferentmodels/costsofmedicalcare.Tothis point, allmodels have suggested that the introduction of first-trimester prediction and prevention dominates current screeningstrategies. This is largely driven by the cost savings associated with reductionofpretermdelivery.

• FIGO considers early screening to be a measure that would mostlikelyincreasesavingstothehealthsystem.

• FIGOcallsformorecost-effectiveanalysestobeconductedtoshowthisbenefittopolicymakers.

28  |     Poon ET AL.

REFERENCES

1. FIGOWorkingGrouponGoodClinicalPractice inMaternal–FetalMedicine. Good clinical practice advice: First trimester screeningandpreventionofpre-eclampsiainsingletonpregnancy.Int J Gynecol Obstet. 2019;144:325–329.

2. Tranquilli AL, Dekker G, Magee L, et al. The classification, diag-nosis and management of the hypertensive disorders of preg-nancy:A revised statement from the ISSHP.Pregnancy Hypertens. 2014;4:97–104.

3. Magee LA, PelsA, HelewaM, Rey E, vonDadelszen P; CanadianHypertensive Disorders of Pregnancy (HDP) Working Group.Diagnosis,evaluation,andmanagementof thehypertensivedisor-ders of pregnancy. Pregnancy Hypertens. 2014;4:105–145.

4. LoweSA,BowyerL,LustK,etal.SOMANZguidelinesfortheman-agement of hypertensive disorders of pregnancy 2014. Aust N Z J Obstet Gynaecol. 2015;55:e1–e29.

5. Brown MA, Magee LA, Kenny LC, et al. The hypertensive disor-ders of pregnancy: ISSHP classification, diagnosis &managementrecommendations for international practice. Pregnancy Hypertens. 2018;13:291–310.

6. VillarJ,SayL,GulmezogluAM,etal.Eclampsiaandpre-eclampsia:Ahealthproblemfor2000years.In:CritchlyH,MacLeanA,PostL,WalkJ,eds.Pre-Eclampsia.London:RCOGPress;2003:189–207.

7. RonsmansC,GrahamWJ.Maternalmortality:Who,when,where,and why. Lancet. 2006;368:1189–1200.

8. Kuklina EV, Ayala C, CallaghanWM. Hypertensive disorders andsevere obstetric morbidity in the United States. Obstet Gynecol. 2009;113:1299–1306.

9. LisonkovaS,JosephKS.Incidenceofpreeclampsia:Riskfactorsandoutcomes associated with early- versus late-onset disease. Am J Obstet Gynecol. 2013;209:544.

10. LisonkovaS,SabrY,MayerC,YoungC,SkollA,JosephKS.Maternalmorbidityassociatedwithearly-onsetandlate-onsetpreeclampsia.Obstet Gynecol. 2014;124:771–781.

11. Redman CW, Sargent IL. Latest advances in understanding pre-eclampsia. Science. 2005;308:1592–1594.

12. JimB,KarumanchiSA.Preeclampsia:Pathogenesis,prevention,andlong-termcomplications.Semin Nephrol. 2017;37:386–397.

13. Chaiworapongsa T, Chaemsaithong P, Yeo L, Romero R. Pre-eclampsia part 1: Current understanding of its pathophysiology. Nat Rev Nephrol. 2014;10:466–480.

14. Thilaganathan B. Pre-eclampsia and the cardiovascular-placentalaxis.Ultrasound Obstet Gynecol. 2018;51:714–717.

15. WorldHealthOrganization.WHO Recommendations for Prevention and Treatment of pre-Eclampsia and Eclampsia. Geneva: WHO; 2011.

16. National Collaborating Centre for women's and Children's Health(UK). Hypertension in Pregnancy: The Management of Hypertensive Disorders During Pregnancy.London:RCOGPress;2010.

17. AmericanCollegeofObstetriciansandGynecologists;TaskForceonHypertensioninPregnancy.Hypertensioninpregnancy.Reportof the American College of Obstetricians and Gynecologists’Task Force on Hypertension in Pregnancy. Obstet Gynecol. 2013;122:1122–1131.

18. LoweSA,BrownMA,DekkerGA,etal.Guidelinesforthemanage-ment of hypertensive disorders of pregnancy 2008. Aust N Z J Obstet Gynaecol. 2009;49:242–246.

19. KhalilA,SyngelakiA,MaizN,ZinevichY,NicolaidesKH.Maternalage and adverse pregnancy outcome: A cohort study.Ultrasound Obstet Gynecol. 2013;42:634–643.

20. Lamminpaa R, Vehvilainen-Julkunen K, Gissler M, Heinonen S.Preeclampsia complicated by advanced maternal age: A registry-based study on primiparous women in Finland 1997-2008. BMC Pregnancy Childbirth. 2012;12:47.

21. YogevY,MelamedN,BardinR,Tenenbaum-GavishK,Ben-ShitritG,Ben-HaroushA.Pregnancyoutcomeatextremelyadvancedmater-nal age. Am J Obstet Gynecol. 2010;203:558.e1–558.e7.

22. Balasch J, Gratacos E. Delayed childbearing: Effects on fertil-ity and the outcome of pregnancy. Curr Opin Obstet Gynecol. 2012;24:187–193.

23. PoonLC,KametasNA,ChelemenT,LealA,NicolaidesKH.Maternalriskfactorsforhypertensivedisordersinpregnancy:Amultivariateapproach. J Hum Hypertens. 2010;24:104–110.

24. Duckitt K, Harrington D. Risk factors for pre-eclampsia at ante-natal booking: Systematic review of controlled studies. BMJ. 2005;330:565.

25. LuoZ-C,AnN,XuH-R,LaranteA,AudibertF,FraserWD.Theeffectsandmechanismsofprimiparityontheriskofpre-eclampsia:Asys-tematicreview.Paediatr Perinat Epidemiol. 2007;21(Suppl.1):36–45.

26. Robillard PY, Hulsey TC, Alexander GR, Keenan A, de Caunes F,PapiernikE.Paternitypatternsandriskofpreeclampsia inthe lastpregnancyinmultiparae.J Reprod Immunol. 1993;24:1–12.

27. Hernandez-Diaz S, Toh S, Cnattingius S. Risk of pre-eclampsia infirst and subsequent pregnancies: Prospective cohort study.BMJ. 2009;338:b2255.

28. CampbellDM,MacGillivray I,Carr-HillR.Pre-eclampsia in secondpregnancy. Br J Obstet Gynaecol. 1985;92:131–140.

29. Sibai BM, el-Nazer A, Gonzalez-Ruiz A. Severe preeclampsia-eclampsia in young primigravid women: Subsequent preg-nancy outcome and remote prognosis. Am J Obstet Gynecol. 1986;155:1011–1016.

30. LieRT,RasmussenS,BrunborgH,GjessingHK,Lie-NielsenE,IrgensLM. Fetal and maternal contributions to risk of pre-eclampsia:Populationbasedstudy.BMJ. 1998;316:1343–1347.

31. Odegard RA, Vatten LJ, Nilsen ST, Salvesen KA, Austgulen R.Risk factors and clinical manifestations of pre-eclampsia. BJOG. 2000;107:1410–1416.

32. vanRijnBB,HoeksLB,BotsML,FranxA,BruinseHW.Outcomesofsubsequentpregnancyafterfirstpregnancywithearly-onsetpre-eclampsia. Am J Obstet Gynecol. 2006;195:723–728.

33. LangenveldJ,JansenS,vanderPostJ,WolfH,MolBW,GanzevoortW.Recurrenceriskofadeliverybefore34weeksofpregnancyduetoanearlyonsethypertensivedisorder:Asystematicreview.Am J Perinatol. 2010;27:565–571.

34. Rousso D, Panidis D, Gkoutzioulis F, Kourtis A, Mavromatidis G,KalahanisI.Effectoftheintervalbetweenpregnanciesonthehealthof mother and child. Eur J Obstet Gynecol Reprod Biol. 2002;105:4–6.

35. King JC.The risk ofmaternal nutritional depletion and poor out-comes increases in early or closely spaced pregnancies. J Nutr. 2003;133:1732S–1736S.

36. Conde-AgudeloA,Rosas-BermudezA,Kafury-GoetaAC.Effectsofbirthspacingonmaternalhealth:Asystematicreview.Am J Obstet Gynecol. 2007;196:297–308.

37. MigniniLE,CarroliG,BetranAP,etal. InterpregnancyintervalandperinataloutcomesacrossLatinAmericafrom1990to2009:Alargemulti-countrystudy.BJOG. 2016;123:730–737.

38. WinikoffB.Theeffectsofbirthspacingonchildandmaternalhealth.Stud Fam Plann. 1983;14:231–245.

39. KlebanoffMA.Theintervalbetweenpregnanciesandtheoutcomeofsubsequentbirths.N Engl J Med. 1999;340:643–644.

40. JacksonRA,GibsonKA,WuYW,CroughanMS.Perinataloutcomesinsingletonsfollowinginvitrofertilization:Ameta-analysis.Obstet Gynecol. 2004;103:551–563.

41. TrogstadL,MagnusP,MoffettA,StoltenbergC.Theeffectofrecur-rentmiscarriageand infertilityontheriskofpre-eclampsia.BJOG. 2009;116:108–113.

42. Pandey S, ShettyA, HamiltonM, Bhattacharya S, Maheshwari A.Obstetricandperinataloutcomesinsingletonpregnanciesresulting

     |  29Poon ET AL.

fromIVF/ICSI:Asystematicreviewandmeta-analysis.Hum Reprod Update. 2012;18:485–503.

43. Thomopoulos C, Tsioufis C, Michalopoulou H, Makris T,PapademetriouV, Stefanadis C. Assisted reproductive technologyand pregnancy-related hypertensive complications: A systematicreview. J Hum Hypertens. 2013;27:148–157.

44. MartinAS,MonsourM,KawwassJF,BouletSL,KissinDM,JamiesonDJ. Risk of preeclampsia in pregnancies after assisted reproduc-tive technology and ovarian stimulation. Matern Child Health J. 2016;20:2050–2056.

45. AlbrechtED,BonaguraTW,BurleighDW,EndersAC,AberdeenGW,PepeGJ.Suppressionofextravilloustrophoblastinvasionofuterinespiral arteries by estrogen during early baboon pregnancy. Placenta. 2006;27:483–490.

46. Imudia AN, Awonuga AO, Doyle JO, et al. Peak serum estradiollevel during controlled ovarian hyperstimulation is associatedwith increased risk of small for gestational age and preeclamp-sia in singleton pregnancies after in vitro fertilization. Fertil Steril. 2012;97:1374–1379.

47. Fisher SJ. Why is placentation abnormal in preeclampsia? Am J Obstet Gynecol. 2015;213:S115–S122.

48. SmithGN,WalkerM,TessierJL,MillarKG. Increased incidenceofpreeclampsiainwomenconceivingbyintrauterineinseminationwithdonorversuspartnerspermfortreatmentofprimaryinfertility.Am J Obstet Gynecol. 1997;177:455–458.

49. HoyJ,VennA,HallidayJ,KovacsG,WaalwykK.Perinatalandobstet-ricoutcomesofdonor inseminationusingcryopreservedsemen inVictoria,Australia.Hum Reprod. 1999;14:1760–1764.

50. SalhaO, SharmaV,DadaT, et al.The influence of donated gam-etes on the incidence of hypertensive disorders of pregnancy. Hum Reprod. 1999;14:2268–2273.

51. NeedJA,BellB,MeffinE,JonesWR.Pre-eclampsiainpregnanciesfromdonorinseminations.J Reprod Immunol. 1983;5:329–338.

52. Simeone S, Serena C, Rambaldi MP, Marchi L, Mello G, MecacciF. Risk of preeclampsia and obstetric outcome in donor oocyteand autologous in vitro fertilization pregnancies.Minerva Ginecol. 2016;68:9–14.

53. NakabayashiY,NakashimaA,YoshinoO, et al. Impairment of theaccumulationofdecidualTcells,NKcells,andmonocytes,andthepoor vascular remodeling of spiral arteries, were observed in oocyte donationcases,regardlessofthepresenceorabsenceofpreeclamp-sia. J Reprod Immunol. 2016;114:65–74.

54. Arngrimsson R, Bjornsson S, Geirsson RT, Bjornsson H, WalkerJJ, Snaedal G. Genetic and familial predisposition to eclampsiaand pre-eclampsia in a defined population. Br J Obstet Gynaecol. 1990;97:762–769.

55. Cincotta RB, Brennecke SP. Family history of pre-eclampsia as apredictor forpre-eclampsia inprimigravidas. Int J Gynaecol Obstet. 1998;60:23–27.

56. WilliamsPJ,BroughtonPF.Thegeneticsofpre-eclampsiaandotherhypertensive disorders of pregnancy. Best Pract Res Clin Obstet Gynaecol. 2011;25:405–417.

57. Zintzaras E, Kitsios G, Harrison GA, et al. Heterogeneity-basedgenome search meta-analysis for preeclampsia. Hum Genet. 2006;120:360–370.

58. CnattingiusS,BergstromR,LipworthL,KramerMS.Prepregnancyweightandtheriskofadversepregnancyoutcomes.N Engl J Med. 1998;338:147–152.

59. WeissJL,MaloneFD,EmigD,etal.Obesity,obstetriccomplicationsandcesareandeliveryrate–apopulation-basedscreeningstudy.Am J Obstet Gynecol. 2004;190:1091–1097.

60. LeungTY,LeungTN,SahotaDS,etal.TrendsinmaternalobesityandassociatedrisksofadversepregnancyoutcomesinapopulationofChinese women. BJOG. 2008;115:1529–1537.

61. SyngelakiA, Bredaki FE,Vaikousi E,MaizN,NicolaidesKH.Bodymassindexat11-13weeks’gestationandpregnancycomplications.Fetal Diagn Ther. 2011;30:250–265.

62. LiuL,HongZ,ZhangL.Associationsofprepregnancybodymassindex and gestational weight gain with pregnancy outcomesin nulliparous women delivering single live babies. Sci Rep. 2015;5:12863.

63. RahmanMM,AbeSK,KandaM, et al.Maternal bodymass indexandriskofbirthandmaternalhealthoutcomesinlow-andmiddle-incomecountries:Asystematicreviewandmeta-analysis.Obes Rev. 2015;16:758–770.

64. WeiY-M,YangH-X,ZhuW-W,etal.Riskofadversepregnancyout-comesstratifiedforpre-pregnancybodymassindex.J Matern Fetal Neonatal Med. 2016;29:2205–2209.

65. Catalano PM. Obesity, insulin resistance, and pregnancy outcome. Reproduction. 2010;140:365–371.

66. GregorMF,Hotamisligil GS. Inflammatorymechanisms in obesity.Annu Rev Immunol. 2011;29:415–445.

67. SpradleyFT,PaleiAC,GrangerJP.Immunemechanismslinkingobe-sity and preeclampsia. Biomolecules. 2015;5:3142–3176.

68. Mittendorf R, Lain KY,WilliamsMA,Walker CK. Preeclampsia. Anested, case-control study of risk factors and their interactions. J Reprod Med. 1996;41:491–496.

69. KnuistM,BonselGJ,ZondervanHA,TreffersPE.Riskfactorsforpre-eclampsiainnulliparouswomenindistinctethnicgroups:Aprospec-tivecohortstudy.Obstet Gynecol. 1998;92:174–178.

70. MostelloD,CatlinTK,RomanL,HolcombWLJ,LeetT.Preeclampsiain the parous woman: Who is at risk? Am J Obstet Gynecol. 2002;187:425–429.

71. CaugheyAB,StotlandNE,WashingtonAE,EscobarGJ.Maternaleth-nicity, paternal ethnicity, and parental ethnic discordance: Predictors of preeclampsia. Obstet Gynecol. 2005;106:156–161.

72. GhoshG,Grewal J,MannistoT, et al. Racial/ethnic differences inpregnancy-relatedhypertensivediseaseinnulliparouswomen.Ethn Dis. 2014;24:283–289.

73. RussellRB,GreenNS,SteinerCA,etal.Costofhospitalizationforpreterm and low birth weight infants in the United States. Pediatrics. 2007;120:e1–e9.

74. BurtVL,WheltonP,RoccellaEJ,etal.PrevalenceofhypertensionintheUSadultpopulation.ResultsfromtheThirdNationalHealthand Nutrition Examination Survey, 1988-1991. Hypertension. 1995;25:305–313.

75. KestenbaumB,SeligerSL,EasterlingTR, et al.Cardiovascular andthromboembolic events following hypertensive pregnancy. Am J Kidney Dis. 2003;42:982–989.

76. BellamyL,CasasJ-P,HingoraniAD,WilliamsDJ.Pre-eclampsiaandrisk of cardiovascular disease and cancer in later life: Systematicreviewandmeta-analysis.BMJ. 2007;335:974.

77. KhalilA,RezendeJ,AkolekarR,SyngelakiA,NicolaidesKH.Maternalracial origin and adverse pregnancy outcome: A cohort study.Ultrasound Obstet Gynecol. 2013;41:278–285.

78. BartschE,MedcalfKE,ParkAL,RayJG.Clinicalriskfactorsforpre-eclampsia determined in early pregnancy: Systematic review andmeta-analysisoflargecohortstudies.BMJ. 2016;353:i1753.

79. MuddLM,OweKM,MottolaMF,PivarnikJM.Healthbenefitsofphysical activity during pregnancy: An international perspective.Med Sci Sports Exerc. 2013;45:268–277.

80. SchneiderS,FreerksenN,RohrigS,HoeftB,MaulH.Gestationaldia-betesandpreeclampsia–similarriskfactorprofiles?Early Hum Dev. 2012;88:179–184.

81. Conti E, Zezza L, Ralli E, et al. Growth factors in preeclampsia:A vascular disease model. A failed vasodilation and angiogenicchallenge from pregnancy onwards?. Cytokine Growth Factor Rev. 2013;24:411–425.

30  |     Poon ET AL.

82. Guimaraes MF, Brandao AHF, Rezende CA, et al. Assessment ofendothelial function in pregnant women with preeclampsia andgestationaldiabetesmellitusbyflow-mediateddilationofbrachialartery. Arch Gynecol Obstet. 2014;290:441–447.

83. Kane SC, Costa Fda S, Brennecke