THE INTRODUCTION OF UMBILICAL ARTERY LACTATE ...

THEINTRODUCTIONOFUMBILICALARTERYLACTATESAMPLINGASA

TOOLTOIMPROVEINTRAPARTUMCAREINSOUTHAFRICA

DrEmmaRoseAllanson

MBBS(Hons)

MPHTM

FRANZCOG

Athesissubmittedintotalfulfilmentofthe

requirementsofthedegreeof

DoctorofPhilosophy

DivisionofObstetricsandGynaecology,FacultyofHealthandMedicalSciences

TheUniversityofWesternAustralia

2019

ii

DEDICATION

ForBenjaminandTheodore,whoareeverything.

iii

INMEMORIAM

TeresaIka

1967-2010

Aboveall,forteachingmetoalwaysfightwithequalmeasuresofstrengthandgrace.

iv

ABSTRACT

Perinatal mortality remains one of the most pressing global health problems, and

improvementsinthequalityofcarearoundthetimeofchildbirthhavethepotentialto

save millions ofwomen’sand newborn lives, particularly in low and middle-income

countries (LMIC).One of the major contributors to perinatal mortality in LMIC is

intrapartum hypoxia and its sequelae, for which the lack of consistently available

qualityintrapartumcareisasignificantcontributor.

The ability to identify andmanage acute intrapartumhypoxia is a significant clinical

challenge,andthecapacitytoachievethisinaLMICformsthebasisofthisthesis.The

specific focus of this thesiswas a comprehensive series of clinical trial and data set

analysesbasedinoneLMIC,SouthAfrica,wheretheperinatalmortalityrate(PNMR)is

33 per 1000 births,with 5 deaths per 1000 births being attributable to intrapartum

asphyxia. Thus, the principle aim of this research was to use the concept of a

commonlyusedtoolinhigh-incomecountries(measurementoffetallactatetoassess

acidosis at birth from an umbilical cord sample), and apply it in a resource limited

settinginordertoimproveintrapartumcare.

Wehypothesisedthatintroducingumbilicalartery(UA)lactatesamplingasaqualityof

care feedback tool would modify clinicians’ behaviours and improve maternal and

neonataloutcomesovertime.Inordertotestthishypothesis,itwasnecessarytofirst

v

develop an understanding of perinatal mortality in South Africa, secondly to be

assuredofthediagnosticaccuracyofUAacidosistestingwithlactate,thirdlytoexplore

theimpactofqualityofcareaudit,andfinallytoimplementUAlactatetesting,assess

itsimpact,anditsfeasibilityoutsideofaresearchsetting.Theconceptsexploredand

analysesusedareoutlinedbelow:

1. Tounderstand the causesofperinataldeath in SouthAfricausingpopulation

baseddata. Frequency distributionswere performed for outcome summaries

and Fisher’s exact test was used to interrogate the relationship between

maternalconditionsandperinatalmortality,andthetimingofperinataldeath.

2. To synthesise the existing evidence on the diagnostic test accuracy of UA

lactate inmeasuring fetal acidosis and predicting poor neonatal outcome by

comparing lactate with other assessments of acidosis in the immediate

newborn and by assessing the capability of lactate to predict neonatal

outcomes. A meta-analysis of correlations was performed applying the

hierarchical summary receiveroperatingcharacteristicsmodelandabivariate

model.

3. Todeterminetheimpactofqualityofcareauditonperinatalmortality in163

facilities inSouthAfricawithat leastfiveyearsofcontinuousaudit.Temporal

trends in perinatal mortality rates were tested using the extended Mantel-

HaenszelM2statisticwithonedegreeoffreedom.

4. To investigate the utility of UA lactate measurements in a South African

hospital for assessing intrapartum care and predicting neonatal outcomes.

vi

Using receiver operator curve (ROC) analysis, the optimal lactate cut-off for

discriminatingbetweentherequirementforneonatal resuscitation,admission

tonurseryandApgar<7at5minuteswasdetermined.

5. To ascertain the clinical impact of introducing UA lactate sampling, in

conjunctionwitheducationoncardiotocograph (CTG) interpretationand fetal

andlactatephysiology.Usinglinearandlogisticregressionanalysis,theimpact

of introducing UA lactate sampling onmaternal and neonatal outcomeswas

determined.

6. To assess theUA lactate results, and intrapartum and neonatal outcomes of

mothers with and without human immunodeficiency virus (HIV) infection.

Univariate comparisonsweremade using independent t-tests for continuous

outcomesandChi-squareorFisherexacttestsforcategoricalcomparisons.

7. ToassesstheattitudesandbarriersofclinicalstafftotheimplementationofUA

lactate in South Africa. In addition to descriptive statistics, the relationship

between various attitudes and barriers before and after a training program

wereanalysedusingFisher’sexacttest,andordinalregressionmodelsusedto

interrogate the role of various factors (e.g. age) on the identification of

individualbarriers.

There are several original contributions arising from this thesis, including the first

systematicreviewandmeta-analysisofthediagnostictestaccuracyofumbilicallactate

formeasuringacidosisandpredictingneonataloutcome,acomprehensiveanalysisof

theimpactofqualityofcareaudit,andthelargestassessmentofaninterventionusing

vii

UA lactate measurement and its impact in a LMIC on intrapartum obstetric

management.Specifically,thekeyfindingsareasfollows:

• In a comprehensive data set representative of a total obstetric population in

South Africa, 49.5% and 74.2% of women having a late stillbirth or early

neonataldeath, respectively,werehealthy (nocomplicationsofpregnancyor

labour)andaquarteroftheperinataldeathsfollowedintrapartumasphyxia.

• Inthefirstsystematicreviewofthediagnostictestaccuracyofumbilicallactate

for measuring acidosis and predicting neonatal outcome, including 38,284

women, lactate was shown to correlate with pH (pooled ES -0.650 95% CI -

0.663to-0.637,p<0.01),BE(pooledES-0.71095%CI -0.721—0.699,p<0.01),

andApgarscoreat5minutes(pooledES0.395%CI0.193-0.407p<0.01).Using

thecommonthresholdoutcomeofHIE,lactatehasasensitivityof69.7%anda

specificityof93%.

• In the largest analysis of the impact of QoC audit on perinatalmortality,we

interrogatedthechangesinPNMRin163SouthAfricanfacilitieswithatleast5

yearsofcontinuousQoCauditdata.3,406,347birthsand85,728deathswere

included and 29% of facilities had a decrease in their PNMRwith 5 years of

continuousQoCaudit.Facilitieswithincreasingmortalityappearedlesslikelyto

haveacomprehensiveQoCauditprocess.

• Ina largecohortof946babieswithaUAlactatesample,wedeterminedthat

theoptimal cut-off for lactate for theoutcomeofneonatal resuscitationwas

5.46mmol/L (sensitivity 68%, specificity 72%). For neonatal admission the

optimal lactate cut-offwas 4.95mmol/L (sensitivity 61%, specificity 59%) and

viii

for an Apgar score of <7 at five minutes the optimal lactate cut-off was

5.65mmol/L(sensitivity64%,specificity69%).

• In a before and after trial assessing the impact of introducing UA lactate

sampling,wefoundnodifferenceinmeanlactate(4.6mmol/L{95%CI4.4-4.8}

comparedwith4.9mmol/L{95%CI4.7-5.1},p=0.089).Howeversuspectedfetal

compromisewasreducedinthepost-interventionperiod:30·2%vs.22·1%,aOR

0·71, 95% CI 0·52-0·96, p=0·027 without worsening neonatal outcomes.

Caesarean section rateswere significantly reduced in the univariate analysis:

pre- 40·3% vs. post-intervention 31·6% (p=0·007). This reduction remained

significantwhenadjustedforpreviouscaesareansection,primiparity,maternal

HIV infection and preterm birth (aOR 0·72, 95%CI 0·54-0·98, p=0·035).

Moreover,emergency caesareandeliveries in theunit asawhole (not just in

women recruited to the trial) decreased from692/2436 (28·4%) to399/2232

(17·0%,P<0·001), andneonatalnurseryadmissionsdecreased from450/2436

(18·5%)to349/2232(15·6%,P=0·010).

• In the 21.6% of women in our before and after trial that were HIV positive,

therewasnoimpactofeithertheirdiseaseortheuseofHAARTonthepreterm

delivery rate, mode of delivery, neonatal resuscitation rate, 1 or 5 minute

Apgarscorebelow7,ortherateofhavingaUAlactatelevel>5.45mmol/L.

• InthefinalstudyassessingtheattitudesandbarrierstointroducingUAlactate

sampling,themajorityofmidwivesanddoctorsprovidingintrapartumcareina

SouthAfricandistricthospitalwerepositiveaboutboththeroleofUAlactate

analysis as well as the potential benefits it provides. Training aided in

ix

overcoming some of the perceived barriers to implementing UA lactate

sampling.

Thethesisoutlinedbelowdemonstratesthattheuseofumbilicalarterylactateisboth

feasible andeffective in SouthAfrica, and can contribute to improvedmaternal and

perinataloutcomes.

SCHOLARSHIPACKNOWLEDGEMENTS

ThisresearchwassupportedbyanAustralianGovernmentResearchTrainingProgram

(RTP)Scholarship.ThisresearchwassupportedbyanAustralianPostGraduateAward,

anAthelstanandAmySawMedicalScholarship,andaWomenandInfants’Research

FoundationGordonKingDoctorofPhilosophyScholarship.

x

DECLARATION

I,EmmaRoseAllanson,certifythat:

Thisthesishasbeensubstantiallyaccomplishedduringenrolmentinthedegree.

This thesis does not containmaterial that has been accepted for the award of any

otherdegreeordiplomainmyname,inanyuniversityorothertertiaryinstitution.

Nopartofthisworkwill, inthefuture,beused inasubmission inmyname,forany

otherdegreeordiplomainanyuniversityorothertertiaryinstitutionwithouttheprior

approval of The University of Western Australia and where applicable, any partner

institutionresponsibleforthejoint-awardofthisdegree.

This thesisdoesnotcontainanymaterialpreviouslypublishedorwrittenbyanother

person,exceptwhereduereferencehasbeenmadeinthetext.

Thework(s)arenotinanywayaviolationorinfringementofanycopyright,trademark,

patent,orotherrightswhatsoeverofanyperson.

Theresearchinvolvinghumandatareportedinthisthesiswasassessedandapproved

by The University of Western Australia Human Research Ethics Committee on the

xi

26/02/2014 (Reference number RA/4/1/6581), the University of Pretoria, Pretoria,

SouthAfricaonthe10/02/2014(Referencenumber7/2014).

Writtenpatientconsenthasbeenreceivedandarchivedfortheresearch involving

patientdatareportedinthisthesis.

Thisthesiscontainspublishedworkand/orworkpreparedforpublication,someof

whichhasbeenco-authored.

Signature:

Date:6thMarch,2019

xii

AUTHORSHIPDECLARATION:CO-AUTHOREDPUBLICATIONS

In accordance with the University of Western Australia’s regulations regarding

Research Higher Degrees, this thesis is presented as a series of journal papers. The

contributionofthecandidateandco-author(s)forthepaperscomprisingchapters3,4,

5,6,7,8and9areherebysetforth.

Detailsofthework:

ThepaperpresentedinChapter4isfirstauthoredbythecandidateandco-authoredbyMariMullerandRobertCPattinsonandispublishedas:

Allanson ER,MullerM, Pattinson RC. Causes of perinatalmortality and associatedmaternal complications in a South African province: challenges in predicting pooroutcomes.BMCPregnancyChildbirth.2015;15(1):37.

Locationinthesis:

Chapter4

Studentcontributiontowork:

EA reviewed the perinatalmortality data from the PPIP database, analysed the dataandwrote thedraftandrevisionsandapproved the finalmanuscript.RPcontributedsignificantlytothestudydesign,reviewedthedraftandapprovedthefinalmanuscript.MMcoordinated all of thedata collection andquality controlled thedata collection,reviewedthedraftandapprovedthefinalmanuscript.

Co-authorsignaturesanddates:

MariMuller

RobertCPattinson

xiii

Detailsofthework:

ThepaperpresentedinChapter5isfirstauthoredbythecandidateandco-authoredby Talal Waqar, Christopher R H White, Ozge Tuncalp and Jan E Dickinson and ispublishedas:

Allanson ER, Waqar T, White CR, Tunçalp Ö, Dickinson JE. Umbilical lactate as ameasure of acidosis and predictor of neonatal risk: a systematic review. BJOG.2017;124(4):584-94.

Locationinthesis:

Chapter5


EAandJDdevelopedthereviewprotocol.EAandCWexecutedthesearchstrategy.EA,CW, TW, and OT completed data extraction and quality assessment. EA and OTundertook the meta-analyses. EA drafted the manuscript. All authors revised themanuscriptandapprovedthefinalversion.


TalalWaqar

HWhite

OzgeTuncalp

JanEDickinson

xiv

Detailsofthework:

ThepaperpresentedinChapter65isfirstauthoredbythecandidateandco-authoredbyRobertCPattinsonandispublishedas:

Allanson ER, Pattinson RC. Quality-of-care audits and perinatal mortality in SouthAfrica.BullWorldHealthOrgan.2015;93(6):424-8.

Locationinthesis:

Chapter6


EAreviewedtheperinatalmortalitydatafromthePPIPdatabase,analysedthedataandwrotethedraftandrevisionsandapprovedthefinalmanuscript.RPcontributedsignificantly to the study design, reviewed the draft and approved the finalmanuscript.


RobertCPattinson

Detailsofthework:

ThepaperpresentedinChapter7isfirstauthoredbythecandidateandco-authoredbyRobertCPattinson,ElizabethANathanandJanEDickinsonandispublishedas:

Allanson ER, Pattinson RC, Nathan EA, Dickinson JE. The introduction of umbilicalcord lactate measurement and associated neonatal outcomes in a South Africantertiaryhospitallaborward.JMaternFetalNeonatalMed.2018;31(10):1272-8.

Locationinthesis:

Chapter7


ERAdesignedthestudy,supervisedtheconductofthestudy,collectedthedata,wrotethe initial manuscript, and approved the final manuscript. EAN contributed to thedesignofthestudy,analysedthedata,andapprovedthefinalmanuscript.RCPandJEDcontributed to the design of the study, supported the conduct of the study, andapprovedthefinalmanuscript.

xv


RobertCPattinson

ElizabethANathan

JanEDickinson

Detailsofthework:

ThepaperpresentedinChapter8isfirstauthoredbythecandidateandco-authoredbyRobertCPattinson,ElizabethANathanandJanEDickinsonandunderreviewwithBMC Pregnancy Childbirth as:Allanson ER, Pattinson RC, Nathan EA, Dickinson JE Abeforeandafterstudyoftheimpactonobstetricandperinataloutcomesfollowingtheintroduction of an educational package of fetal heart rate monitoring educationcoupledwithumbilicalarterylactatesamplinginalowresourcesettinglabourwardinsouthafrica

Locationinthesis:

Chapter8



xvi


RobertCPattinson

ElizabethANathan

JanEDickinson

Detailsofthework:

ThepaperpresentedinChapter9isfirstauthoredbythecandidateandco-authoredbyRobertCPattinson,ElizabethANathanandJanEDickinsonandispublishedas:

Allanson ER, Pattinson RC, Nathan EA, Dickinson JE. Impact of maternal HIV onumbilical cord lactate measurement at delivery in a South African labor ward. Int JGynaecolObstet.2018;141(3):366-70.

Locationinthesis:

Chapter9



xvii


RobertCPattinson

ElizabethANathan

JanEDickinson

Detailsofthework:

ThepaperpresentedinChapter10isfirstauthoredbythecandidateandco-authoredbyKateGrobicki,RobertCPattinsonandJanEDickinsonandispublishedas:

Allanson ER, Grobicki K, Pattinson RC, Dickinson JE. Attitudes towards theimplementationofuniversalumbilicalarterylactateanalysisinaSouthAfricandistricthospital.BMCPregnancyChildbirth.2016;16(1):166.

Locationinthesis:

Chapter10


EAdesignedthestudy,supervisedtheconductofthestudy,analysedtheresultsandwrotetheinitialmanuscript.KGcontributedtothedesignofthestudy,conductedthetrainingand surveysandapproved the finalmanuscript.RCPand JEDcontributed tothedesignofthestudyandapprovedthefinalmanuscript.

xviii


KateGrobicki

RobertCPattinson

xix

Studentsignature:

Date:6thMarch,2019

I, Professor Jan Dickinson, certify that the student statements regarding theircontributiontoeachoftheworkslistedabovearecorrect

Coordinatingsupervisorsignature:

Date:6thMarch,2019

xx

ACKNOWLEDGEMENTS

ItisasomewhatoverwhelmingprocesstositandreflectupontheprocessofaPhDas

the end draws near, and to wonder where on earth one is to begin with

acknowledgements.TherearemanypeopleIhavebeenprivilegedtohavebymyside

throughoutthisentireprocess,andmanymorewhohavecontributedatstepsalong

theway.Iwillinnowaydojusticetotheprocessofthankingallofthemintheseshort

words;pleasejustknowthatmygratitudeisdeepandlong-lasting.

Firstly, to thewomenofKalafongandZithulelehospitals inSouthAfrica - this thesis

wouldneverhavecome to fruitionwithoutyou.Youareall strength inmotherhood

embodiedandtheimpressionyouleftisdeep.Tothemidwives,doctorsandresearch

staffatKalafong,thankyouforyourdedicationtothisprojectfromtheverybeginning,

andformakingmeawelcomepartofyourfamily.

ProfessorRobertPattinsonandDrJennyMakin,myloveforyoubothisgreat.In2014I

packedupmythingsandarrivedaloneinSouthAfrica,completelyunsureofwhatwas

tocome.Thatyearturnedouttobelifechanging,andlaunchedmanyofthepathsthat

I now find my career winds along. You are two extraordinary people whose

commitmenttoglobalhealthandrespectofandadvocacyforwomen issomething I

aspiretoeveryday.

xxi

I am very grateful to Dr Özge Tunçalp, who took on the role of supervisor when I

headeduptoSwitzerlandin2015.Thankyouforguidingandsupportingmethrough

the second year ofmy PhD, spendingmany hours reading papers, advisingme, and

becoming one of my fondest friends. To Metin Gulmezoglu, thank you for the

opportunity to join your group, for teaching me much about the world of global

maternalhealth,andshowingmethetruevalueofcreatingateamintheworkplace.

I am so lucky to have had extraordinary researcher Professor John Newnham as a

constant inmyentiremedical andPhD life as a supervisor,mentor, andgeneral life

guide. Many of my fondest memories of the PhD process are sitting in your office

debatingthemeaningoflife,thevalueofoverseasfellowshipsandthelessonslearnt

fromPhDs. Thank you for givingme the finalpush I needed to geton thatplane to

South Africa - for that I will be forever grateful. To Talal, Chris, Kate, Mari, and

Elizabethwhoallco-authoredpaperswithmethroughthisprocess;thankyouforyour

hardwork,adviceandsupport.AlsotothethirdmemberofmyPhDadvisoryboard,

ProfessorYeeLeung– thankyou for themanyyearsofmentorship,and forenticing

meovertothedarksideofsub-specialtytraining.

To Professor JanDickinson,what can I say? There is noway inwhichwords cando

justicetohowyouhaveshapedmylife.Isuspectitisveryraretofindsomeonewhois

truly amentor in theway youare.Anextraordinary clinician and researcher, I have

beenveryprivilegedtohaveyouasasupervisor.Youhavehadmybackeverystepof

xxii

my careerandhaveguidedme through thisPhDwithmorewisdom than I canever

hope tohave.Thankyou forbeingamentor, colleagueandwonderful friend tome,

andIwouldbethrilledifinmycareer,Icanbeafractionofthewomanyouare.

TomyfamilywhomIlovedearly-Joel,Tess,Dylan,WilliamandAlexandra,thankyou

forallofyoursupport,andsorryfortheimportanteventsImissedbeingaway.ToGina

andAndrew,mywonderfulparents;youcreatedaworldwhereIbelievedIcoulddo

anything and knew that youwould always be bymy side. Your support ofmy long-

windedcareerpathhasbeenunwavering.IthankyouandIloveyou.

TomybeautifulhusbandBen.Youareanextraordinaryman,andIamsoluckytobe

yourwife.Ithasbeenajoytonavigatelifewithyouthelast14years,andmygoodness

I am looking forward to the next many years of adventures. Thank you for the

immeasurable support youhavegivenmeduring thisPhD. I loveyou. Finally, tomy

son, Theodore. You have bought me nothing but happiness since the day you so

casually arrived in to this world. May you continue to be the life loving, sweet,

inquisitive littleman that youarenow forevermore. Iwill always loveyouand Iwill

alwayshaveyourback.

xxiii

PRESENTATIONSANDARTICLESARISINGFROMTHISTHESIS

PRESENTATIONS

South African Society of Obstetricians and Gynaecologists National Congress. May

19th, 2014. Cape Town, South Africa. “The role of audit in changing perinatal

mortality”

South African Society of Obstetricians and Gynaecologists Mini Symposium. August

23rd,2014.Pretoria,SouthAfrica.“Perinatalmortality:currentandfuturefocus.”

Priorities in Perinatal Care. March 17th, 2015. Drakensberg, South Africa. “Causes

ofperinatalmortality and associated maternal complications in a South African

province:Challengesinpredictingpooroutcomes”

Women and Infants Research Foundation Rising Stars Event. September 15th, 2016.

Perth,Western Australia. “Introducing umbilical artery lactate sampling as a tool to

improveintrapartumcareinamiddleincomecountrysetting”

xxiv

RoyalAustralianandNewZealandCollegeofObstetriciansandGynaecologistsAnnual

ScientificMeeting.October17th,2016.Perth,WesternAustralia.“FromAlpstoZebras

andEverythinginBetween”

ARTICLES

Allanson ER, Muller M, Pattinson RC. Causes of perinatal mortality and associated

maternal complications in a South African province: challenges in predicting poor

outcomes.BMCPregnancyChildbirth.2015;15:37.

Allanson ER, Pattinson RC. Quality-of-care audits and perinatal mortality in South

Africa.BullWorldHealthOrgan.2015;93(6):424-428.

Allanson ER, Waqar T, White CR, Tunçalp Ö, Dickinson JE. Umbilical lactate as a

measure of acidosis and predictor of neonatal risk: a systematic review. BJOG.

2017;124(4):584-594.

Allanson ER, Pattinson RC, Nathan EA, Dickinson JE. The benefit of introducing

umbilicalarterylactatesamplingintoalowresourcesettinglabourward:abeforeand

aftertrialinSouthAfrica.(UnderconsiderationbyBMCPregnancyChildbirth,2019)

xxv

Allanson ER, Grobicki K, Pattinson RC, Dickinson JE. Attitudes towards the

implementationofuniversalumbilicalarterylactateanalysisinaSouthAfricandistrict

hospital.BMCPregnancyChildbirth.2016;16(1):166.

AllansonER,PattinsonRC,NathanEA,DickinsonJE.Theintroductionofumbilicalcord

lactate measurement and associated neonatal outcomes in a South African tertiary

hospitallaborward.JMaternFetalNeonatalMed.2018;31(10):1272-1278.

Allanson ER, Pattinson RC, Nathan EA, Dickinson JE. Impact of maternal HIV on

umbilical cord lactatemeasurement at delivery in a South African labor ward. Int J

GynaecolObstet.2018;141(3):366-70.

BerghAM,AllansonE, PattinsonRC.What is needed for taking emergencyobstetric

and neonatal programmes to scale? Best Pract Res Clin Obstet Gynaecol.

2015;29(8):1017-1027.

Allanson ER, Tunçalp Ö, Vogel JP, Khan DN, Oladapo OT, Long Q, Gülmezoglu AM.

Implementationofeffectivepracticesinhealthfacilities:asystematicreviewofcluster

randomisedtrials.BMJGlobHealth.2017;2(2):e000266.

xxvi

AllansonE,TunçalpÖ,GardosiJ,PattinsonRC,ErwichJJ,FlenadyVJ,FrøenJF,Neilson

J,ChouD,MathaiM,SayL,GülmezogluM.Classifyingthecausesofperinataldeath.

BullWorldHealthOrgan.2016;94(2):79-79A.

AllansonER,TuncalpÖ,GardosiJ,PattinsonRC,VogelJP,ErwichJ,FlenadyVJ,Froen

JF,NeilsonJ,QuachA,FrancisA,ChouD,MathaiM,SayL,GülmezogluAM.Givinga

voice to millions: developing the WHO application of ICD-10 to deaths during the

perinatalperiod:ICD-PM.BJOG.2016;123(12):1986-1899.

AllansonER,TunçalpÖ,GardosiJ,PattinsonRC,FrancisA,VogelJP,ErwichJJ,Flenady

VJ,FrøenJF,NeilsonJ,QuachA,ChouD,MathaiM,SayL,GülmezogluAM.TheWHO

applicationofICD-10todeathsduringtheperinatalperiod(ICD-PM):resultsfrompilot

databasetestinginSouthAfricaandUnitedKingdom.BJOG.2016;123(12):2019-2028.

AllansonER,VogelJP,TunçalpӦ,GardosiJ,PattinsonRC,FrancisA,ErwichJJ,Flenady

VJ, Frøen JF,Neilson J,QuachA. Application of ICD-PM to preterm-relatedneonatal

deathsinSouthAfricaandUnitedKingdom.BJOG.2016;123(12):2029-2036.

xxvii

AllansonER,TunçalpӦ,GardosiJ,PattinsonRC,FrancisA,VogelJP,ErwichJJ,Flenady

VJ,FrøenJF,NeilsonJ,QuachA.OptimisingtheInternationalClassificationofDiseases

to identify the maternal condition in the case of perinatal death. BJOG.

2016;123(12):2037-2046.

LeisherSH,TeohZ,ReinebrantH,AllansonE,BlencoweH,ErwichJJ,FrøenJF,Gardosi

J,GordijnS,GülmezogluAM,HeazellAE,KortewegF,LawnJ,McClureEM,PattinsonR,

SmithGC,TunçalpÖ,WojcieszekAM,FlenadyV.Classification systems for causesof

stillbirth and neonatal death, 2009–2014: an assessment of alignment with

characteristicsforaneffectiveglobalsystem.BMCPregnancyChildbirth.2016;16:269.

LeisherSH,TeohZ,ReinebrantH,AllansonE,BlencoweH,ErwichJJ,FrøenJF,Gardosi

J,GordijnS,GülmezogluAM,HeazellAE,KortewegF,LawnJ,McClureEM,PattinsonR,

Smith GC, Tunçalp Ö, Wojcieszek AM, Flenady V. Seeking order amidst chaos: a

systematicreviewofclassificationsystemsforcausesofstillbirthandneonataldeath,

2009–2014.BMCPregnancyChildbirth.2016;16(1):295.

FrøenJF,FribergIK,LawnJE,BhuttaZA,PattinsonRC,AllansonER,FlenadyV,McClure

EM,FrancoL,GoldenbergRL,KinneyMV,LeisherSH,PittC,IslamM,KheraA,Dhaliwal

L, AggarwalN, PainaN, TemmermanM, Lancet Ending Preventable Stillbirths Series

study group. Stillbirths: progress and unfinished business. Lancet.

2016;387(10018):574-586.

xxviii

WojcieszekAM,ReinebrantHE, LeisherSH,AllansonE,CooryM,Erwich JJ, Frøen JF,

GardosiJ,GordijnS,GulmezogluM,HeazellAE,KortewegFJ,McClureE,PattinsonR,

Silver RM, Smith GC, Teoh Z, Tunçalp Ö, Flenady V. Characteristics of a global

classification system for perinatal deaths: aDelphi consensus study.BMCPregnancy

Childbirth.2016;16(1):223.

xxix

CONTENTS

Dedication.............................................................................................................................ii

InMemoriam........................................................................................................................iii

Abstract................................................................................................................................iv

Scholarshipacknowledgements............................................................................................ix

Declaration............................................................................................................................x

Authorshipdeclaration:Co-authoredpublications...............................................................xii

Acknowledgements.............................................................................................................xx

Presentationsandarticlesarisingfromthisthesis.............................................................xxiii

Presentations...............................................................................................................xxiii

Articles.........................................................................................................................xxiv

Abbreviations........................................................................................................................1

Tableoffigures......................................................................................................................6

Tableoftables.......................................................................................................................9

Tableofimages....................................................................................................................13

CHAPTERONE:THESISAIMANDOUTLINE............................................................................15

Thesisaim........................................................................................................................16

Thesisoutline..................................................................................................................19

References.......................................................................................................................27

CHAPTERTWO:LITERATUREREVIEW...................................................................................29

Chapteroutline................................................................................................................30

Introduction....................................................................................................................31

Definitionsofperinatalmortality.....................................................................................32

Perinatalmortalityinmiddleincomesettings..................................................................36

Perinatalmortalityinthemiddle-incomecountryliterature:areview.............................39

Perinatalmortalityinroutineobstetriccare....................................................................40

Facilitybaseddeliveryandperinatalmortalityinmiddle-incomecountries.....................42

xxx

Qualityofcare................................................................................................................44

Caesareansectioninmiddleincomecountries................................................................46

PerinatalmortalityinSouthAfrica..................................................................................48

CausesofperinatalmortalityinSouthafrica...................................................................49

PerinatalmortalityinSouthAfrica:areview...................................................................53

Theglobalclassificationofperinatalmortality.................................................................54

Fetalphysiologyandtheimpactofhypoxia.....................................................................58

Acidbasebalanceinthefetus.........................................................................................58

Intrapartumhypoxia.......................................................................................................59

Thefetalresponsetohypoxia.........................................................................................60

Thepretermfetalresponsetohypoxiaandasphyxia......................................................64

Lactateasameasurementofmetabolicacidosis..............................................................65

Theclinicalconsequencesofhypoxiaandacidaemia.......................................................66

Theconceptofasphyxia..................................................................................................66

Causesofacutehypoxiaandasphyxiaintheintrapartumperiod....................................67

Optionsformeasuringhypoxiaanditssequelae..............................................................70

pH...................................................................................................................................70

Baseexcess(BE)..............................................................................................................71

Apgarscore.....................................................................................................................71

Lactate............................................................................................................................72

Measuringumbilicallactate.............................................................................................73

Timingoflactatesampling..............................................................................................74

Methodofsampling........................................................................................................76

Cordarteriallactatelevels..............................................................................................78

Fetalheartratemonitoring..............................................................................................80

TheCTGandacidosis:areview.......................................................................................85

Theneonatewithacidosis...............................................................................................93

xxxi

Neonatalacidosisandadverseoutcomes........................................................................94

Thewellnewbornwithacidosis......................................................................................95

Qualityofcareaudits.......................................................................................................97

Definingqualityofcareandqualityofcareaudit............................................................97

Qualityofcareauditinmaternalandnewbornhealth....................................................99

Qualityofcareauditandnearmisscases......................................................................102

MaternalHIVInfection..................................................................................................103

Conclusion.....................................................................................................................104

References.....................................................................................................................105

CHAPTER THREE: THESIS OUTLINE AND METHODOLOGY OF THE RESEARCH…………………………………….....................................................................................................134

Thesisoutline................................................................................................................135

Outlineoftheresearchinchapterfour..........................................................................137

Aim................................................................................................................................137

Methodology.................................................................................................................138

Inclusioncriteria............................................................................................................138

Dataextracted...............................................................................................................138

Ethics.............................................................................................................................139

Outlineoftheresearchinchapterfive...........................................................................140

Objectivesofthereview................................................................................................140

Criteriaforconsideringstudiesinthereview.................................................................140

Generalinclusion/exclusioncriteria.............................................................................140

Typesofstudydesignsfordiagnostictestaccuracy........................................................141

Typesofparticipants.....................................................................................................141

Typesofoutcomes.........................................................................................................141

Searchstrategyforidentificationofstudies...................................................................142

Customisedsearchstrategy...........................................................................................143

Methodsofthereview..................................................................................................158

xxxii

Analysisplan..................................................................................................................164

Outlineoftheresearchinchaptersix.............................................................................166

Aim................................................................................................................................166

Methodology.................................................................................................................166

Ethics.............................................................................................................................167

Analysis.........................................................................................................................167

Outlineoftheresearchinchaptersseven,eightandnine..............................................168

Background...................................................................................................................168

Studyobjective.............................................................................................................168

Primary aims................................................................................................................169

Secondary aims............................................................................................................170

Methodology................................................................................................................170

Phase1..........................................................................................................................171

Studydesignandstudysetting......................................................................................171

Studypopulation..........................................................................................................171

Studyproceduresanddatacollected............................................................................172

Datacollected...............................................................................................................173

Samplesizecalculation.................................................................................................175

Phase two.....................................................................................................................176


Studypopulation..........................................................................................................176

Studyproceduresanddatacollected............................................................................177

Teachingintervention...................................................................................................177

Datacollected...............................................................................................................180

Ethics............................................................................................................................181

Outlineoftheresearchinchapterten...........................................................................182

Background...................................................................................................................182

xxxiii

Studyobjective.............................................................................................................183

Primaryaims.................................................................................................................183

Methodology................................................................................................................183


Informedconsent..........................................................................................................184

Pre-intervention...........................................................................................................184

Intervention..................................................................................................................185

Postintervention..........................................................................................................186

CHAPTER FOUR: CAUSES OF PERINATAL MORTALITY AND ASSOCIATED MATERNALCOMPLICATIONS IN A SOUTH AFRICAN PROVINCE: CHALLENGES IN PREDICTING POOROUTCOMES........................................................................................................................187

Chapteroverview..........................................................................................................188

Abstract........................................................................................................................189

Background...................................................................................................................189

Methods.......................................................................................................................190

Results..........................................................................................................................191

Discussion.....................................................................................................................191

Conclusion....................................................................................................................194

References....................................................................................................................195

CHAPTER FIVE: UMBILICAL LACTATE AS A MEASURE OF ACIDOSIS AND PREDICTOR OFNEONATALRISK:ASYSTEMATICREVIEW...........................................................................196


Abstract........................................................................................................................198

Introduction..................................................................................................................198

Methods.......................................................................................................................199

Results..........................................................................................................................200

Discussion.....................................................................................................................203

Conclusion....................................................................................................................205

References....................................................................................................................206

xxxiv

Referencesforchapterintroduction..............................................................................209

CHAPTERSIX:QUALITYOFCAREAUDITSANDPERINATALMORTALITYINSOUTHAFRICA..211


Abstract........................................................................................................................213

Introduction..................................................................................................................213

Approach......................................................................................................................213

Relevantchanges..........................................................................................................214

Lessonslearnt...............................................................................................................215

References....................................................................................................................217

CHAPTER SEVEN: THE INTRODUCTIONOFUMBILICAL CORD LACTATEMEASUREMENTANDASSOCIATED NEONATAL OUTCOMES IN A SOUTH AFRICAN TERTIARY HOSPITAL LABOURWARD................................................................................................................................218


Abstract........................................................................................................................220

Introduction..................................................................................................................220

Materialsandmethods.................................................................................................221

Results..........................................................................................................................221

Discussion.....................................................................................................................224

Conclusion....................................................................................................................225

References....................................................................................................................225

CHAPTER EIGHT: PREVENTING THE FIRST CAESAREAN: A BEFORE AND AFTER TRIALDEMONSTRATINGTHEBENEFITOFINTRODUCINGUMBILICALARTERYLACTATESAMPLINGINTOALOWRESOURCESETTINGLABOURWARD..................................................................227


Referenceforchapterintroduction................................................................................231

Abstract.........................................................................................................................235

Introduction..................................................................................................................238

Materialsandmethods..................................................................................................239

Statisticsandanalysis....................................................................................................241

xxxv

Ethics.............................................................................................................................243

Results...........................................................................................................................244

Maternalcharacteristics................................................................................................244

Pregnancyandlabour...................................................................................................245

Lactateresults...............................................................................................................247

Birthcharacteristics......................................................................................................247

Neonatalcharacteristics................................................................................................248

Modeofdeliveryandneonataloutcomesforalldeliveries...........................................251

Discussion......................................................................................................................251

Conclusion.....................................................................................................................256

References.....................................................................................................................260

CHAPTERNINE: IMPACTOFMATERNALHIVONUMBILICALCORDLACTATEMEASUREMENTATDELIVERYINASOUTHAFRICANLABOURWARD...........................................................263


Abstract........................................................................................................................265

Introduction..................................................................................................................265

Materialsandmethods.................................................................................................266

Results..........................................................................................................................266

Discussion.....................................................................................................................267

References....................................................................................................................269

CHAPTER TEN: ATTITUDES TOWARDS THE IMPLEMENTATION OF UNIVERSAL UMBILICALARTERYLACTATEANALYSISINASOUTHAFRICANDISTRICTHOSPITAL...............................270


Abstract........................................................................................................................272

Background...................................................................................................................272

Methods.......................................................................................................................273

Results..........................................................................................................................273

Discussion.....................................................................................................................275

xxxvi

Conclusion....................................................................................................................276

References....................................................................................................................277

CHAPTERELEVEN:DISCUSSION..........................................................................................278

Overview.......................................................................................................................279

Applicabilityoftheresearch..........................................................................................281

Themanagementofintrapartumasphyxia....................................................................281

TheuseandbenefitsofUAlactatesampling.................................................................282

Translatingresearchintoclinicalpractice......................................................................285

TheimplementationofQOCaudit.................................................................................286

Strengthsoftheresearch...............................................................................................288

Limitationsoftheresearch............................................................................................289

Futuredirectionsforresearch........................................................................................291

Conclusion.....................................................................................................................291

References.....................................................................................................................292

APPENDICES.......................................................................................................................295

AppendicesforChapter2...............................................................................................296

Appendix2.1:PapersbythethesisauthoronthedevelopmentoftheICD-PM.............296


Appendix 3.1: Patient information and consent form for the “introduction of universalcordlactatesamplingandcorrelationwithneonataloutcomes”...................................330

Appendix3.2:Datacollectionsheet..............................................................................334

Appendix3.3:PretrainingCTGtestfordoctors(questions)...........................................336

Appendix3.4:PretrainingCTGtestfordoctors(answers).............................................350

Appendix3.5:PosttrainingCTGtestfordoctors...........................................................351

Appendix3.6:PosttrainingCTGtestfordoctors(answers)...........................................371

Appendix3.7:PretrainingCTGtestformidwives(questions).......................................373

Appendix3.8:PretrainingCTGtestformidwives(answers)..........................................380

Appendix3.9:PosttrainingCTGtestformidwives........................................................381

xxxvii

Appendix3.10:PosttrainingCTGtestformidwives(answers)......................................391


Appendix5.1:Searchstrategy.......................................................................................393

Appendix5.2:Datafromeachincludedstudy,includingthethresholdsforeachoutcome.....................................................................................................................................394

Appendix5.3Characteristicsofincludedstudies(n=12)................................................398

AppendicesforChapter10.............................................................................................403

Appendix10.1Questionnaire:attitudesandbarrierstowardsimplementationofuniversalcordlactatesampling....................................................................................................404

Appendix10.2Patient/participant’sinformationleaflet&informedconsentform.....410

1

ABBREVIATIONS

ABE actualbaseexcess

aOR adjustedoddsratio

ACTH adrenocorticotropichormone

AFA advancedfetalassessment

ANS autonomicnervoussystem

APH antepartumhaemorrhage

ART anti-retroviraltherapy

ATP adenosinetriphosphate

AUC areaunderthecurve

AVD assistedvaginaldelivery

AVP argininevasopressin

BG bloodgas

BE baseexcess

BPM beatsperminute

CI confidenceintervals

CO2 carbondioxide

2

CRH corticotropin-releasinghormone

CS caesareansection

CTG cardiotocograph

CV coefficientofvariation

DHIS DistrictHealthInformationSystem

ENND earlyneonataldeath

ESMOE EssentialStepsintheManagementofObstetricEmergencies

FDIU fetaldeathinutero

FIGO InternationalFederationofGynecologyandObstetrics

FSB freshstillbirth

GR grams

HAART highlyactiveanti-retroviraltherapy

HIE hypoxicischaemicencephalopathy

HIV humanimmunodeficiencyvirus

HMD hyalinemembranedisease

HPA hypothalamic-pituitary-adrenal

HR heartrate

3

ICD-10 International Statistical Classification of Diseases and Related Health

Problems10thRevision

ICD-PM The WHO application of ICD-10 to perinatal deaths: ICD-perinatal

mortality

IQ interquartilerange

IUGR intrauterinegrowthrestriction

KEMH KingEdwardMemorialHospital

LBW lowbirthweight

LFD latefetaldeath

LMIC lowandmiddle-incomecountries

LND lateneonataldeath

MAP meanarterialpressure

MAS meconiumaspirationsyndrome

MeSH medicalsubjectheadings

MDG millenniumdevelopmentgoals

MIC middle-incomecountries

MMR maternalmortalityrate

MSB maceratedstillbirth

4

NEC necrotisingenterocolitis

NND neonataldeath

NNJ neonataljaundice

OR oddsratio

PCI perinatalcareindicators

PMTCT preventionofmothertochildtransmission

PNMR perinatalmortalityrate

POC pointofcare

PPIP PerinatalProblemsIdentificationprogram

PT provincialtertiaryhospital

QoC qualityofcare

QUADAS2 Qualityassessmentofdiagnostictestaccuracystudies2

RCT randomisedcontrolledtrial

RDS respiratorydistresssyndrome

ROC receiveroperatorcurve

SBE standardbaseexcess

SD standarddeviation

SDG sustainabledevelopmentgoals

5

TTN transienttachypnoeaofthenewborn

UA umbilicalartery

UV umbilicalvein

VBAC vaginalbirthaftercaesarean

WHO WorldHealthOrganization

WHOMCS WorldHealthOrganizationmulti-countrysurvey

6

TABLEOFFIGURES

Chapter2

Figure2.1 Thecalculationofperinatalmortality

Figure2.2: Distribution of perinatalmortality across three time periods in

SouthAfrica

Figure2.3 Aerobicandanaerobicmetabolism.

Figure2.4 Anexampleof thecolorimetric reactionusedbytheAccutrend

monitor

Figure2.5 CTG with the normal fetal baseline HR lying within the 110-

160bpmareadefinedbythegreyline

Figure2.6: CTGsshowingdecreased,normalandincreasedvariability

Figure2.7: CTGshowingaccelerationsabovethebaseline

Figure2.8: CTGshowingearlydecelerations

Figure2.9: CTGshowinglatedecelerations

Figure2.10: CTGshowingvariabledecelerations

Figure2.11: CTGshowingprolongeddecelerations

7

Chapter5

Figure1: PRISMAflowchart

Figure2: Forestplotofumbilicalarteriallactatesensitivityandspecificity

foracidosisdefinedbypH<7.2


foracidosisdefinedbyBE(>8)


forneonatalpoorneurologicaloutcomeincludingHIE

Figure5: Plotof fittedmodel formeta-analysis sensitivity and specificity

ofumbilicallactateforpoorneurologicaloutcomeincludingHIE

Figure6: Risk of bias and applicability concerns graph: review authors’

judgementsabouteachdomainpresentedaspercentagesacross

includedstudies

Figure7: Riskofbiasandapplicabilityconcernssummary:reviewauthors’

judgementabouteachdomainforeachincludedstudy

8

Chapter6

Figure1: Trendsinperinatalmortalityforthefirstfiveyearsofquality-of-

careauditsinhealth-carefacilities,SouthAfrica

Chapter7

Figure1: Receiver operating characteristics curves for UA lactate for

neonatal resuscitation in all babies (black solid line), preterm

babies (light grey solid line), and term babies (dark grey solid

line).Thedashedlineindicatesthediagonalreferenceline

Figure2: Receiver operating characteristics curves for UA lactate for

neonataladmissiontothenurseryinallbabies(blacksolidline),

pretermbabies(lightgreysolidline),andtermbabies(darkgrey

solidline).Thedashedlineindicatesthediagonalreferenceline

Figure3: ReceiveroperatingcharacteristicscurvesforApgarscore<7at5

minutesinallbabies(blacksolidline),pretermbabies(lightgrey

solidline),andtermbabies(darkgreysolidline).Thedashedline

indicatesthediagonalreferenceline

9

TABLEOFTABLES

Chapter2

Table2.1 Perinatalmortalityratesinmiddleincomecountries

Table2.2 PerinatalcareindicatorsatKalafongHospital,2013

Table2.3 Causes and frequency of clinical events that result in

intrapartumhypoxiaandasphyxiainLMIC

Table2.4: Cut-offvaluesforcohortsassessingacidosisusinglactatelevels

Table2.5: ComponentsoftheCTG.

Chapter4

Table1: Fetal/neonatal and maternal characteristics of macerated and

freshstillbirthsandearlyneonataldeaths

Table2: Comparisonofmaternalconditioninthelatestillbirthandearly

neonataldeaths

Table3: Obstetric causes of death in late stillbirth and early neonatal

death

10

Table4: Causesofneonataldeath

Chapter5

Table1: Meta-analysisofcorrelationcoefficientsof lactatecomparedto

threeothermeasuresofacidosis(pH,SBEandApgarscoreat5

minutes)

Chapter6

Table1: Obstetric causes of death in quality-of-care audits for health-

carefacilitieswithincreasinganddecreasingperinatalmortality,

SouthAfrica

Chapter7

Table1: Maternalandpregnancycharacteristics

11

Chapter8

Table8.1: Preandposttestscoresfordoctors

Table8.2: Preandposttestscoresformidwives

Table1: Maternalcharacteristicsinpre-andpost-interventiongroups

Table2: Intrapartumcomplicationsinpre-andpost-interventiongroups

Table3: Neonatalcharacteristicspre-andpost-interventiongroups

Table4: Neonatal outcomes stratified by birth weight <2500gr and ≥

2500gr

Chapter9

Table1: Maternalcharacteristics

Table2: Deliverycharacteristics

12

Chapter10

Table1: Characteristicsofrespondents

Table2: Perceived barriers to introduction of universal umbilical artery

lactateanalysispreandposttraining

Table3: Identification of barriers to implementing universal umbilical

artery lactate analysis pre and post training (the influence of

clinicalrole)

Chapter11

Table11.1 PerinatalcareindicatorsforKalafongHospital,2014

13

TABLEOFIMAGES

Chapter1

Image1.1 WomenintheantenatalwardatKalafonghospital

Image1.2 AnelephantatKrugerNationalPark,Mpumalanga

Image1.3 BedsoutforwashingatKalafongHospital

Image1.4 IntrapartumCTGatKalafongHospital

Image1.5 CTGpriortointrapartumdeathatKalafongHospital

Chapter2

Image2.1: Saving Babies. The most recent report from the Perinatal

ProblemsIdentificationProgram

Image2.2a

and2.2b: AccutrendPlusTM©hand-heldlactatemeter

Image2.3: Asegmentofumbilicalcordwiththelargerveinclearlyvisible

Image2.4: Acrosssectionofumbilicalcordshowing2arteriesand1vein

Image2.5: A1mlnon-heparinisedsyringeand26-gaugeneedle

14

Chapter3

Image3.1: KalafongHospitalbynight,2014

Image3.2: AKalafonghospitalobstetricsregistrartestingaUAsample

Chapter11

Image11.1 CrowdedantenatalwardatKalafongHospital

Image11.2 The primary author of this thesis and SisterMaria, Matron of

KalafongHospitallabourward

15

CHAPTERONE:THESISAIMANDOUTLINE

16

THESISAIM

Millionsofwomenandtheirbabiesdieeveryyeararoundthetimeofchildbirthand

improving this statistic remains a priority of the Sustainable Development Goals

(SDGs)(1). Thus, the basis of this dissertation centres on the concept of perinatal

mortalityinlowandmiddle-incomecountries(LMIC)andparticularlythecontribution

of intrapartum care to this. Lack of quality of intrapartum care around the time of

childbirth leads to adverse outcomes, frequently as a result of hypoxia and its

sequelae, which in LMIC is a significant contributor to both stillbirths and early

neonatal deaths(2). Intrapartum hypoxaemia can be considered the intermittent or

permanent fall in fetal blood oxygen levels during labour, which may lead to

intrapartum hypoxia, or a fall in oxygen levels in organs(3). With the influence of

maternal,fetal,andcarefactors,thiscanleadtoacidosis(increaseofhydrogenionsin

thetissues(4)),andfinallytointrapartumasphyxia,aconglomerateofclinicalfeatures

includinganeurologicallydepressedneonateatbirth,aswellas intrapartumorearly

neonataldeath(5).

Themiddle-incomecountryofSouthAfricahasaperinatalmortalityrate(PNMR)of33

per 1000 births, with only small improvement in this statistic in since 2000(6).

Intrapartum asphyxia causes at least 10% of all stillbirths in South Africa(7) and

additionally,at leastaquarterofearlyneonataldeathsareattributedtointrapartum

asphyxia(8),withanuntoldnumberofsurvivorssufferinglongtermdisability(9).

17

ReducingthePNMRinLMICrequiresinterventionsthatmeetthecontextualneedsof

the relevant populations. That is, any intervention needs to be easily applied,

inexpensive, easily taught and sustainable. Reducing intrapartum hypoxia requires

accurate recognition of the condition, followed by an appropriate clinical response.

Interventions toachieve this includeelectronic fetalheart ratemonitoring,access to

instrumental deliveries and caesarean section facilities, and effective neonatal

resuscitation, in addition to adequate staffing with access to teaching and case

reflection(9-11).

WhileelectronicfetalheartratemonitoringisfrequentlyusedinSouthAfricawiththe

aim of detecting fetal hypoxia and compromise(12), there are several limitations to

CTGusethatmaylimititsimpactonintrapartumoutcomes.Dependingonthesystem

usedforinterpretation,specificityforCTGandfetalacidosiscanbeaslowas18%,with

positivepredictivevaluesaround35%(13).However,thesensitivityofelectronicfetal

heartratemonitoringmaybeimprovedwithumbilicalcordbloodsamplingtoassess

acid-basebalanceimmediatelyafterdelivery.Umbilicalcordsamplingsupplementsthe

clinical picture provided by fetal heart rate patterns, as it objectively reflects fetal

tissue hypoxia and acidosis, which have morbid and mortal sequelae(14). Sampling

allows immediate case reflection with positive or negative reinforcement of

management decisions for eachdelivery to occur(15).Although this clearly does not

change the intrapartum management for the sampled case, doing so has several

potentialbenefits.Firstly,inthecaseofanormalresult, intrapartumasphyxiacanbe

18

excludedandhealthcareprovidersreassuredabouttheirclinicaldecisionmaking(14).

In the case of an abnormal result, staff can both reflect on intrapartum care and

potentiallytakestepstochangemanagementstrategies inthefuture(15). Itwasthis

latterconceptthatformedthebasisofthisthesis.

UAbloodgasmeasurementallowsanobjectiveassessmentofthemetabolicstateof

thefetuspriortodeliveryandprovidesabiochemicalrepresentationoffetalacid-base

statusintheintrapartumperiod.Itisconventionallyperformedusingameasurement

ofpH(14).WhilstUApHisthegoldstandardformeasurementoffetalhypoxiathathas

resulted inacidosis, thecostandmaintenanceof therequiredequipment (bloodgas

analyser)renderthisnotfeasibleforresourcepoorsettings,includingmostfacilitiesin

South Africa. An alternate way to measure acidosis is with UA cord blood lactate;

whichcanbereadilytestedonahandhelddeviceandcorrelateswellwithpHvaluesin

theassessmentoffetalhypoxia(16).Thus,theaimofthisresearchwastointroduceUA

lactatesamplinginaresourcelimitedsetting,asaqualityofcarefeedbacktool,which

it was hypothesised, would alter clinicians’ behaviours and improve maternal and

neonatal outcomes over time. This research was performed at Kalafong Hospital,

Pretoria,SouthAfrica(Image1.1),withapproximately6500deliveriesperyearina16

bedlabourward,usuallystaffedwith3midwivesand3doctorsatanyonetime.

19

THESISOUTLINE

In this thesis I have explored the causes of perinatal mortality in South Africa,

undertakena systematic reviewandmeta-analysisof thediagnostic test accuracyof

lactate,completedacomprehensiveanalysisoftheimpactofqualityofcareauditon

perinatal mortality, implemented UA lactate testing to ascertain its capacity in this

setting topredict short-termneonataloutcomes,assessed the impactof introducing

UA lactate sampling, investigated the role maternal HIV infection may play in the

resultsofthetest,aswellasitsfeasibilityoutsideofaresearchsetting.Chaptertwo

reviews the literature to provide anunderstandingof perinatalmortality in lowand

middleincomesettings(andinSouthAfricaspecifically),normalfetalphysiology,fetal

lactate,qualityofcareaudit,nearmisscases,andotherfactorsrelatedtointrapartum

care,includingmaternalHIVinfectionandcaesareansectionrates.

20

Image1.1WomenintheantenatalwardatKalafonghospital

Chapter four explores the causes of perinatal mortality and associated maternal

conditionsinSouthAfrica.ItusesdatafromMpumalanga,aprovinceineasternSouth

Africa bordering Swaziland and Mozambique (and well known for the presence of

KrugerPark),wheremorethan90%ofwomengivebirthinahealthcarefacilityandall

facilitiescaptureperinatalmortalityqualityofcareauditdatausingtheSouthAfrican

Perinatal Problems Identification Program (PPIP) database. A secondary analysis of

PPIP data for the Provincewas undertaken for the period October 2013 to January

2014, inclusive. Data for each individual late perinatal death were reviewed. The

frequencies of maternal and perinatal characteristics for late fetal deaths were

21

examined and the relationships between the maternal condition and perinatal

outcomesanalysed.

Image1.2AnelephantatKrugerNationalPark,Mpumalanga

Chapter five is a systematic review and diagnostic test accuracy of umbilical cord

lactatetomeasureacidosisandpredictneonataloutcome.Tosynthesisetheevidence

for the diagnostic test accuracy of umbilical cord lactate in measuring acidosis and

predictingpoorneonataloutcome,publishedandunpublishedstudiesbetween1990

and 2014 from PubMed/Medline, EMBASE, Cochrane Central Register of Controlled

Trials, and clinicaltrials.gov, extracting cross-sectional and randomised studies that

assessedfetalacidosis(usinglactateastheindextest)withorwithoutanassessment

of neonatal outcome, were reviewed. Correlations between index and reference

test(s) were recorded, as were the raw data, to classify the predictive ability of

22

umbilical lactateforneonataloutcomes.Meta-analysisofcorrelationwasperformed.

Estimatesofthestudies'observedsensitivitiesandspecificitieswereplottedonForest

plots with 95% confidence intervals (CI). Where possible, we combined data using

meta-analysis, applying the hierarchical summary receiver operating characteristics

modelandabivariatemodel.

Chaptersixexplores the impactsofqualityofcareauditonPNMR inSouthAfrica.A

review of several small studies showed a reduction in perinatal mortality after the

introductionofperinatal auditsbutno large scale researchexists(17). Therefore,we

analysedthechangesinPNMRin163facilitiesinSouthAfricawithatleastfiveyearsof

continuous quality of care audit.We tested for temporal trends in PNMR using the

extendedMantel-HaenszelM2statisticwithonedegreeoffreedom.

Chapters seven to nine originate from research undertaken at Kalafong Hospital in

2014.(Image1.3)ChapterseveninvestigatestheutilityofUAlactatemeasurementsat

Kalafong Hospital to assess intrapartum care and predict neonatal outcomes. From

March 3rd to 12th of November 2014, a prospective cohort study of UA lactatewas

conducted.Usingreceiveroperatorcurve (ROC)analysis,wedeterminedtheoptimal

lactatecut-offfordiscriminatingbetweentherequirementforneonatalresuscitation,

admissiontotheneonatalnurseryandApgarscorelessthansevenatfiveminutes.

23

Image1.3BedsoutforwashingatKalafongHospital

ChaptereightcentresontheconceptthattheintroductionofUAlactatesamplinginall

deliveries inotherunitsproducesprogressive improvements inneonatalbiochemical

markersand reducedadmission tohigh care facilitiesover time, independentof the

rate of obstetric intervention(15). Thus, from 3rdMarch - 12th November 2014, we

undertook a before and after trial to determine the impact of introducing UA cord

lactate sampling, using a hand-held lactatemeter, in conjunctionwith education on

CTGinterpretationandfetalandlactatephysiology.(Images1.4&1.5)Usinglinearand

logistic regression analysis, we determined the impact of introducing UA lactate

samplingonmaternalandneonataloutcomes.

24

\

Image1.4IntrapartumCTGatKalafongHospital

InChapternine,theUAlactatelevelsandperinataloutcomesamongwomenwithand

withoutHIV infectionwereassessed. InacohortofwomenwithUAlactatesamples,

maternalandneonataldatawerecollected, includingHIVstatusandCD4counts.UA

lactate levels and neonatal outcomes were compared between women with and

without HIV infection. The potential confounding impact of HIV treatment on

outcomeswasstudiedusingmaternalCD4counts.Univariatecomparisonsweremade

using independent t-tests for continuous outcomes and Chi-square or Fisher exact

testsforcategoricalcomparisons.

25

Image1.5CTGpriortointrapartumdeathatKalafongHospital

Itiscriticalthatinterventionscanbeimplementedbeyondaresearchsetting,andare

acceptable to clinicians and patients in daily clinical practice. Thus, Chapter ten

assesses the acceptability and barriers to implementing universal umbilical artery

lactate testing inadistricthospital in Zithulele,on theeastern capeof SouthAfrica.

During the period 16/11/2014 - 13/01/2015, we conducted a training course in

cardiotocograph(CTG)interpretation,fetalphysiology,andthesamplingandanalysing

of UA lactate, with a pre- and post-training questionnaire aimed at assessing the

barriers and facilitators to the introduction of universal UA lactate sampling. In

addition to descriptive statistics, the relationship between various attitudes and

barriersbeforeandafterthetrainingprogramwereanalysedusingFisher’sexacttest.

Theroleofgender,age,position,andlengthoftimepractisingontheidentificationof

individual barriers to the implementation of the intervention pre- and post-training

wereexploredusingordinalregressionmodels.

26

Chapterelevensummarisesthefindingsofthisthesis,thestrengthsandlimitationsof

theresearchandfuturedirectionsforimplementationandongoingresearch.

27

REFERENCES

1. Scott H, Danel I. Accountability for improvingmaternal and newborn health.BestPractResClinObstetGynaecol.2016;36:45-56.2. Lawn J, Shibuya K, Stein C. No cry at birth: global estimates of intrapartumstillbirths and intrapartum-related neonatal deaths. Bull World Health Organ.2005;83(6):409-17.3. Nordstrom L, Arulkumaran S. Intrapartum fetal hypoxia and biochemicalmarkers:areview.ObstetGynecolSurv.1998;53(10):645-57.4. Ayres-de-Campos D. Introduction: Why is intrapartum foetal monitoringnecessary - Impact on outcomes and interventions. Best Pract Res Clin ObstetGynaecol.2016;30:3-8.5. GrahamEM,RuisKA,HartmanAL,NorthingtonFJ,FoxHE.Asystematicreviewof the role of intrapartum hypoxia-ischemia in the causation of neonatalencephalopathy.AmJObstetGynecol.2008;199(6):587-95.6. Pattinson R. Saving Babies: A Perinatal Care Survey of South Africa 2000.Pretoria,SouthAfrica;2000.7. PattinsonR,Rhoda,N.Savingbabies2012-2013:NinthreportonperinatalcareinSouthAfrica.TshepesaPress,Pretoria,SouthAfrica;2014.8. LloydLG,DeWittT.NeonatalmortalityinSouthAfrica:Howarewedoingandcanwedobetter?SAMJ:SouthAfricanMedicalJournal.2013;103(8):518-9.9. Hofmeyr GJ, Haws RA, Bergström S, Lee AC, Okong P, Darmstadt GL, et al.Obstetriccareinlow‐resourcesettings:what,who,andhowtoovercomechallengesto scale up? International Journal of Gynecology & Obstetrics.2009;107(Supplement):S21-S45.10. BuchmannE, PattinsonR,NyathikaziN. Intrapartum-relatedbirth asphyxia inSouth Africa lessons From the first national perinatal care survey. S Afr Med J.2002;92(11):897-901.11. Wall SN, Lee AC, CarloW, Goldenberg R, Niermeyer S, Darmstadt GL, et al.,editors. Reducing intrapartum-related neonatal deaths in low-and middle-incomecountries—whatworks?SeminPerinatol;2010:Elsevier.12. van Bogaert LJ, Misra A. Neonatal outcome after caesarean birth for fetaldistressand/ormeconiumstaininginaSouthAfricanruralsetting.JObstetGynaecol.2008;28(1):56-9.13. Di Tommaso M, Seravalli V, Cordisco A, Consorti G, Mecacci F, Rizzello F.Comparisonoffiveclassificationsystemsforinterpretingelectronicfetalmonitoringinpredictingneonatalstatusatbirth.TheJournalofMaternal-Fetal&NeonatalMedicine.2013;26(5):487-90.14. Thorp JA, Rushing RS. Umbilical cord blood gas analysis.Obstet Gynecol ClinNorthAm.1999;26(4):695-709.15. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Benefitsof introducinguniversalumbilicalcordbloodgasandlactateanalysis intoanobstetricunit.AustNZJObstetGynaecol.2010;50(4):318-28.16. HamedHO.Intrapartumfetalasphyxia:studyofumbilicalcordbloodlactateinrelationtofetalheartratepatterns.ArchGynecolObstet.2013;287(6):1067-73.17. PattinsonR,KerberK,WaiswaP,DayLT,MussellF,AsiruddinS,etal.Perinatalmortality audit: counting, accountability, and overcoming challenges in scaling up in

28

low‐ and middle‐ income countries. International Journal of Gynecology &Obstetrics.2009;107(Supplement).

29

CHAPTERTWO:LITERATUREREVIEW

30

CHAPTEROUTLINE

Chaptertwoexplorestheexistingliteraturefortheconceptsthatformthebasisofthis

thesis. It commences with the definitions of perinatal mortality, an overview of

perinataloutcomesinmiddleincomecountries(MIC)andtheclassificationofperinatal

deathsinSouthAfrica.Thenormalphysiologyofthefetusandtheimpactofhypoxia

onthis,themeasurementofacidosisincludingwithlactate,theinterplaybetweenCTG

andacidosisisthendiscussed.Finally,qualityofcare(QoC)auditsarereviewed.

31

INTRODUCTION

Perinatal death is both a global public health concern and an individual tragedy for

millions ofwomen, their partners, and their families around theworld. The time of

childbirthposes thegreatest risk towomenandtheirbabies,andthere is significant

overlapintheaetiologyofthecausesofdeathofboth(1).AsdiscussedinChapterone,

hypoxiaduringthefirstandsecondstageoflabourplacesthefetusatriskofsignificant

morbidity andmortality(2). Intrapartum hypoxia contributes tomore than amillion

perinataldeathsgloballyeachyear(3)andthemajorityofthesedeathsoccurinLMIC,

where the lack of consistent availability of quality intrapartum care is a significant

contributing factor(4). Thus, quality of care (QoC) around the time of childbirth is

imperative to thesurvivalmothersandbabies(5).Moreover, issueswith intrapartum

carethatdonotresultinperinataldeath,havethepotentialtoresultinnearmissesor

perinatalmorbidity.

While it is clear that the causesof perinatalmortality aremultifactorial,QoC in low

resourcesettingsareasignificantcontributoryfactor.SouthAfrica,forexample,hasa

perinatalmortalityrate(PNMR)of33per1000births.Australia’sPNMRbycomparison

is6per1000.InSouthAfrica,5deathsper1000birthsareattributedtobirthasphyxia,

and more than a third of these asphyxia perinatal deaths are secondary to either

inadequate fetal heart rate monitoring or failure to recognise intrapartum fetal

compromise(6).

32

The ability to identify and manage intrapartum hypoxia is a significant clinical

challenge,particularlygiventhatmostcaseswillhavenoantepartumriskfactors,and

morethanathirdwillhavenoobviousintrapartumriskfactors(7).Intrapartumhypoxia

oftenmanifestsinlabourwithabnormalitiesinthefetalheartrate,howeverthefetal

cardiotocograph (CTG) has a limited positive predictive value of only 30% for the

outcomeofacidosis(8). There is great variability in theuseofobjectivemeasuresof

intrapartum compromise such as fetal scalp sampling and assessment of acidosis

during labour (using pH or lactate) internationally(8-10). While improving the

sensitivityoftheCTGwith intrapartumfetalscalpsamplinghasbeendescribedsince

the1960’s(withvariableresultsonitsabilitytoreduceintrapartumasphyxia)(10),its

use in LMIC is limited(11).Reasons for thisutilisation include the resources required

for fetal scalp sampling, and the contraindications to scalp sampling, including

maternalHIVinfectionandotherbloodborneviruses.

DEFINITIONSOFPERINATALMORTALITY

GLOBALPERINATALMORTALITY

Theburdenofperinatal (andmaternal)mortality isalmostexclusivelywithLMIC(12).

The most recent estimates indicate more than 2.5 million stillbirths and a similar

number of neonatal deaths annually(13-16). Additionally, there were 303,000

maternaldeathsglobally in2015(17).Bothmothersandbabiesaremost likelytodie

33

onthedayofdelivery(18,19).Therefore,anyinterventionfocusedaroundthetimeof

childbirth has the potential to impact the outcomes of both mother and baby.

Perinatalmortalityiscalculatedper1000totalbirths,usingtheformulainFigure2.1.

Thedefinition of perinatalmortality varies between countries as to the gestation at

whichafetalorneonataldeathisincludedinperinatalmortalitystatistics.Stillbirthis

definedas“deathpriortothecompleteexpulsionorextractionfromitsmother”and

earlyneonataldeathisdeathwithinthefirst7daysoflife(20).Whilethedefinitionof

the perinatal period is generally defined as any death from 22 weeks gestation

onwards,globalreportingandcomparisonsaregenerallydoneusingthedefinitionof

28weeksgestationorabirthweightof1000grams(gr)ormore(21),withbirthweight

asthepreferentialdenominator(22).ThePNMRforSouthAfricaisgenerallyquotedas

that for babies with a birth weight of 1000gr or more, although PNMR for smaller

babiesarealsorecorded(23).Conversely,perinatalmortalityinAustraliaisdefinedby

deathofatleast20weeksgestationordeathofababywithabirthweightofatleast

400grams(24).

Figure 2.1 The calculation of perinatal mortality, where early neonatal deaths are

withinthefirst7daysoflife(20).

34

Perinatal mortality can be separated in to three largely distinct time periods:

antepartum, intrapartumandneonataldeaths.The latter canbeconsideredasearly

(inthefirst7daysoflife)orlate(7-28daysoflife)neonataldeaths.Whenconsidering

theroleintrapartumcareplaysinneonataldeaths,itisthepredominantlyearlydeaths

thatareofinterest.

ANTEPARTUMMORTALITY

Antepartummortality isgenerallydefinedas“deathbeforetheonsetof labour”(25),

althoughthereisvariationbetweencountriesastowhichdeathsareincludedwithin

thisdefinition.Forexample,Australiadefinesstillbirthas“ababybornofat least20

weeksgestationor400grbirthweightwhodidnot,atanytimeafterdelivery,breathe,

orshowanyevidenceoflifesuchasaheartbeat”(26),whereasinSouthAfrica,babies

weighing500grormoreareincludedinthisdefinition(23),andICD-10definesafetal

deathasthatoccurringinababyweighingmorethan500grorbeinggreaterthan22

completed weeks gestation(22). Antepartum deaths are sometimes separated into

early(≥500grand<1000gr)orlate(≥1000gr)fetaldeaths(25).Estimatesofglobalrates

of antepartum deaths (stillbirths) are difficult given the lack of recording and

registrationofmoststillbirthsaroundtheworld(18).

35

INTRAPARTUMMORTALITY

Anintrapartumdeathisthatoccurringafterthecommencementoflabour,butbefore

completeexpulsionofthefetus(25),withmanycountriesaroundtheworldidentifying

a“freshstillbirth”asasurrogateforanintrapartumdeathintheabsenceoffetalheart

ratemonitoringthatdefinitively identifies thetimeofdeath(27,28).Theeventsthat

can lead to intrapartummortality are the focus for this research. This is particularly

pertinentasintrapartumstillbirthsinLMICrepresentthemajorityoftheglobalburden

ofthesetypesofdeathsandconsideredlargelypreventable(29,30).Upto70%ofall

stillbirths in LMIC are likely to be intrapartum deaths(19). Moreover, given that in

some LMIC attendance for antenatal care is uncommon (for example in Bangladesh

whereanyantenatalattendanceforcareoccurs in<2/3ofwomen(31)),contactwith

health care professionals in the intrapartum period may be the first medical

interaction, severely hampering the potential for obstetric care provision and risk

reduction(32).

EARLYNEONATALMORTALITY

Early neonatal deaths are those that occur before day 7 of neonatal life (25), and,

especiallyinlowresourcesettings,mayoccurasadirectconsequenceofintrapartum

careandevents(29,33,34). Indeed,mostearlyneonataldeaths inLMICoccur inthe

firstdayafterbirth,andmorethanhalfofthesefollowintrapartumasphyxia(35).Early

neonataldeathsareofparticularly importanceastheycontributeapproximately75%

36

of all post-natal deaths(36), and are responsible for 50% of the under five year

mortalityratesinLMIC(35,37-40).

PERINATALMORTALITYINMIDDLEINCOMESETTINGS

There exists a substantial body of literature on global perinatal mortality. Our trial

aimedtoconsideraninterventionthatwouldbeapplicableinmiddle-incomecountries

(MIC) where it is critical to consider the distribution of resources, such that any

interventiondoesnotcompromiseotheraspectsofcare.Perinatalmortality inthese

settings is disproportionately represented around the time of childbirth, and

interventions at this point are critical(41), however theuseof one interventionmay

resultinlessaccesstoanotherbecauseoffiniteresources.Thus,wearebeholdento

consider the scope of the problem of perinatal mortality in MIC in order to affect

changethatismeaningful,contextuallyappropriate,andpragmatic(42).

EstimatesofperinatalmortalityinMICarepresentedinTable2.1.Whileestimatescan

beinfluencedbyvaryingdefinitionsofdeaths,particularlystillbirths(43),itisclearthat

rates varywidely, from 7/1000 births inMalaysia to 111/1000 births inMauritania.

South Africa’s reported PNMR is 33/1000 births, which, it is worth noting, is 500%

higher than that in Australia(20). The impact of any intervention is related to the

frequencyofthecausativeproblem. It isevidentfromTable2.1thattherearemany

37

MICwheremortalityratescouldpotentiallybeimpactedbyaninterventionsuchasUA

lactatesampling.

Table 2.1 Perinatal mortality rates in middle income countries (Adapted from the

WorldHealthOrganization(20),exceptwhereadditionalreferencesindicated)

Country Perinatalmortalityrate (per1000births)

Country Perinatalmortalityrate (per1000births)

Country Perinatalmortality rate(per 1000births)

Albania 20 Grenada 21 Panama 15

Algeria 47 Guatemala 23 Papua NewGuinea

51

AmericanSamoa

5(44)* Guyana 40 Paraguay 20

Angola 86 Honduras 28 Peru 20

Argentina 14 India 70 Philippines 23

Armenia 29 Indonesia 30 Romania 12

Azerbaijan 58 Iran, IslamicRep.

35 RussianFederation

25

Bangladesh 50 Iraq 76 Samoa 21

Belarus 9 Jamaica 17 Sao TomeandPrincipe

61

Belize 33 Jordan 26 Serbia andMontenegro

13

Bhutan 40 Kazakhstan 57 SolomonIslands

20

Bolivia 31 Kenya 53 SouthAfrica 33

Bosnia andHerzegovina

20 Kiribati 44 SriLanka 20

Botswana 79 Kosovo 19(45) St.Lucia 20

38

Brazil 20 KyrgyzRepublic

57 St. Vincentand theGrenadines

19

Bulgaria 13 LaoPDR 57 Sudan 44

CaboVerde 30 Lebanon 34 Suriname 30

Cambodia 66 Lesotho 46 Swaziland 61

Cameroon 70 Libya 19 Syrian ArabRepublic

16

China 35 Macedonia,FYR

16(20) Tajikistan 62

Colombia 23 Malaysia 7 Thailand 20

Congo,Rep. 76 Maldives 54 Timor-Leste 65

CostaRica 13 MarshallIslands

43 Tonga 18

Coted'Ivoire 96 Mauritania 111 Tunisia 20

Cuba 14 Mauritius 17 Turkey 36

Djibouti 62 Mexico 22 Turkmenistan

39

Dominica 12 Micronesia,Fed.Sts.

20 Tuvalu 36

DominicanRepublic

28 Moldova 31 Ukraine 37

Ecuador 20 Mongolia 46 Uzbekistan 46

Egypt, ArabRepublic.

27 Morocco 31 Vanuatu 31

ElSalvador 20 Myanmar 65 Venezuela,RB

18

EquatorialGuinea

65 Namibia 46 Vietnam 37

Fiji 16 Nicaragua 23 West BankandGaza

21(46)

Gabon 59 Nigeria 86 Yemen,Rep. 44

39

Georgia 42 Pakistan 59 Zambia 56

Ghana 45 Palau 23

*Dataonlyavailableforneonataldeaths

PERINATALMORTALITYINTHEMIDDLE-INCOMECOUNTRYLITERATURE:AREVIEW

A literature review on perinatal mortality in MIC was conducted. PubMed was

searched,usingacombinationof theMeSHterms forpoverty,developingcountries,

perinatal death and perinatal mortality. (((("Poverty"{Mesh}) OR "Developing

Countries"{Mesh})AND"PerinatalDeath"{Mesh})OR("PerinatalMortality"{Mesh}OR

"Fetal Mortality"{Mesh})). Middle income countries were defined usingWorld Bank

criteria(47), and are outlined in Table 2.1. 1299 studies were identified since 2000.

Titlesandabstractsweresearchedandfulltextarticlesrelevanttoperinataloutcomes

inMICareexploredbelow.

MICmayhaveahighbaselinePNMR,andthenwithincountryhaveregionswitheven

higher PNMR and therefore face greater resource challenges in addressing these

regionaldisparities(48-54).Moreover,analysesofperinataldeathsinMICoftenshow,

in addition to a high PNMR, significant numbers of these deaths are preventable,

althoughprogressonthisisoftenlimitedbyhumanandphysicalresources(55,56).Itis

this large group of preventable perinatal deaths where efforts to improve global

perinatalmortalityshouldbetargetedandwheretheconceptofthisthesisstemmed

from.

40

PERINATALMORTALITYINROUTINEOBSTETRICCARE

One of the challenges for obstetric care in MIC is the mortality associated with

commoneventsorclinicalproblemsthatwouldbeconsideredlargelystraightforward

tomanageinhigh-incomesettings.Oneofthecontributorstothismaybetheconcept

of “fragile states” – that is, that thismortality occurs in countrieswhere theoverall

physical and human infrastructure deficits lead to worse outcomes in vulnerable

groups,includingwomenandtheirnewborns(57).

Many clinical pregnancy and intrapartum scenarios in high-income countries are

recognised as potential situations for complications, and resources formanagement

directedasappropriate.ThesameisnottrueforMIC,anditisevidentfromreviewing

theliteraturethattheprevalence,aetiology,treatmentandconsequencesofmaternal

andperinatalclinicalissuesordiseasebecomefarmorecomplexthanthesameissue

inhigh-incomesettings.

Maternal conditions in pregnancy are potentially more complex to manage inMIC,

with consequent worse maternal and perinatal outcomes. Several examples are

outlinedbelow.

41

• Hypertensive disorders, which are among the commonest maternal

complicationofpregnancy,are frequentlyassociatedwithworseoutcomes in

MIC(58).

• Non-pregnancymedical conditions complexifymanagement inMIC.Maternal

anaemia, for example, has been shown in a large systematic review to be

associatedwithadverseperinataloutcomes(59).

• Even common pregnancy occurrences, such asmultifetal gestations, perform

poorly in lowresourcesettings,AGlobalNetworkStudy insixmiddle income

countries (Kenya, Zambia, Pakistan, India, Guatemala and Argentina)

demonstrated a significant contribution of multifetal gestations to adverse

outcomes in these settings,withamaternalmortalityof 0.5%andanoverall

perinatalmortalityofaround20%(60).Similarfindingsarereportedfromother

studiesofmiddleincomecountries(61,62).

• Even common labour management issues increase in complexity in MIC. A

retrospectivereviewinDakarofvaginalbirthaftercaesareansectionrevealed

a perinatal mortality of 26.4%, largely due to abruption and uterine

rupture(63).

Suchmorbidity andmortalityof so-called routineevents is repeated throughout the

literature formortality inMIC, and not often seen in high-income countries(64-68).

This is highlighted in the rates of intrapartum asphyxia in LMIC, which is markedly

differenttohigh-incomecountries.InoneprospectivecohortstudyinPakistan,119of

the1103deliverieswereperinataldeaths,andnearlyhalfofthesewereattributedto

42

birth asphyxia(69). Audit in a tertiary unit in Nigeria showed that when macerated

stillbirths were removed from consideration, asphyxia caused the overwhelming

majorityoftheremainingperinataldeaths(70).

FACILITYBASEDDELIVERYANDPERINATALMORTALITYINMIDDLE-INCOMECOUNTRIES

Itisworthnotingthattheaimofthisresearchistoimproveintrapartumcareusinga

tool that by definition necessitates facility-based delivery, which is not universal in

MIC.Thus,theinterventionneedstobeconsideredwithinthecontextoftheproblem,

andthepotentialscopeoftheresearchisclearlyrelatedtotheissueoffacility-based

delivery,aswellasthequalityofcaredeliveredinthosefacilities.TherearemanyMIC

wherefacility-baseddeliveryisstilluncommon,andthismodelofcare,particularlyin

the absence of training of community birth attendants, increases perinatal

mortality(71-81). Nigeria, for example, has a facility-based delivery rate of only

35%(82),Ghana of up to 75%(83), Kenya of only 50%(84, 85), but countries such as

China and India nearly 100%. However, while facility-based delivery will generally

result in improved perinatal outcomes(86), the referral bias of shifting high-risk

deliveriesfromthecommunitytofacilitiesmayresultinlessimprovementbeingseen

than expected(87). In LMIC, complex cases often require transfer to higher acuity

units,increasingthelikelihoodofasphyxiaanditsconsequences.InonestudyinIndia,

the rate of intrapartum perinatal mortality secondary to asphyxia among cases

referredemergentlyinlabourwas80/1000births(79).

43

Women inMICwill often, but not always, have improved perinatal outcomeswhen

antenatalandintrapartumcareisshiftedtofacilities(88),howevertheimprovementin

perinatal mortality is often not commensurate with the increase in facility-based

delivery(89). In Bangladesh the rate of facility delivery is reported as 16%, while

perinatal mortality is up to 64.7/1000 births, with up to one in five women having

intrapartum complications(90). Hence, translocating these women to facility-based

delivery shifts the acuity of cases in those centres too. Higher acuity does not

necessarilyresultinadverseoutcomesinallMIC;Chinaforexamplehasahighfacility

birthrate(91,92)butaperinatalmortalityrateofonly7.2/1000births(93).

It is clear that the occurrence of facility-based delivery does not guarantee human

resourcestodeliverintrapartumcare(94),orifdelivered,thereisnoguaranteeofthe

qualityof intrapartumcare (18, 34, 87, 95),which is necessary to improveperinatal

outcomes(96). This concept is borne out in the literature repeatedly. There are still

unacceptablyhighratesof intrapartumstillbirthreported in institutionaldeliveries in

middle income countries; for example, Ghana reports a rate of 1% frommore than

400,000institutionalbirths(42).RegionsofIndia,whichhavenearlyuniversalcoverage

of facility-based birth, have seen the shift to facility birth impact stillbirth but not

neonatalmortalityrates(97).Similarly,astudyinTurkeycomparingratesandcausesof

perinatalmortalityfrom1998to2009showedareductioninantenatalstillbirthbutan

increaseinperinatalasphyxiarelateddeathswhenbirthswereshiftedtofacilities(98).

Onewouldpresumethatantenatalinterventionstoimprovestillbirthrates(e.g.blood

pressureandurinechecksatantenatalvisitstoidentifycasesofpre-eclampsia)areless

44

complextoimplementthanwidescaleimprovedqualityofcareinfacilities,whichmay

explain the discordance between increased facility-based births and persistent

perinatalmortality.

QUALITYOFCARE

Suboptimalqualityofcarearoundthetimeofchildbirthcontributestotheburdenof

perinatalmortalityinMIC.ThereisextensivereportingintheMICliteratureregarding

the contribution of intrapartum care (and issues with the quality of that care) to

perinatal mortality. The concept of quality of care around the time of childbirth

possiblyexplains large in-countryvariations inperinatalmortalityrates.Forexample,

Turkey has a PNMR reported as 36/1000(20), but regional studies report rates of

nearly 50/1000(99). The prevention of asphyxia and the early treatment of the

asphyxiated neonate has the potential to save millions of women’s and babies

lives(100,101),howeveraccesstocarealoneisnotnecessarilyenoughforprevention,

ifthereareissueswiththequalityofthatcare(102).Examplesofthisconceptcanbe

seenthroughouttheliteratureinanalysesofthetypeofcaregivenandtheoutcomes

ofperinatalaudit.

• TheDemocraticRepublicofCongohasaveryhighPNMR,andlowQoCinthe

maternal and newborn health services that do exist(32) and Pakistan has a

neonatalmortality rate of 55/1000 births, with the lack of access to “good”

45

newborncarebeingidentifiedtoattributingtothishighrate(103).Theconcept

ofQoC(orlackthereof)islikelysynonymouswithpreventableperinataldeath.

• An analysis of perinatal deaths in Salvador, Brazil, reported a perinatal

mortalityrateof25.8/1000butmorethan90%ofthesewerepreventable,and

moreover, nearly half of the deaths would have been prevented with QoC

aroundthetimeofchildbirth(104).

• The linkbetweenQoCandperinataldeath isseen intheratesof intrapartum

asphyxia,whichaswehavealreadysaid,usuallyoccurs intheabsenceofrisk

factors (i.e. otherwise healthy mothers and fetuses). In a large population

based study inChina, a quarterof theneonatal deathswere causedbybirth

asphyxia(105). Similarly comprehensive audit of deaths in Guatemala, the

Democratic Republic of Congo, Zambia, and Pakistan resulted in a quarter of

neonataldeathsbeingattributed tobirthasphyxia(106).Tertiary level care in

NigeriahasaPNMRofupto81/1000births,withperinatalasphyxiabeingthe

greatest contributor to this(107). InGhanabothasphyxiaandbirth injuryare

thegreatestcontributorstoearlyneonataldeaths(108),andinTanzania,athird

ofneonataldeathsareattributedtoasphyxia,despiteatleasthalfofallbirths

occurringinhealthfacilities(101,109).

The critical role of QoC can be seen if one considers the outcomes of nulliparous

comparedwithmultiparouswomeninthesameinstitutions.Nulliparouswomenoften

bear the burden of obstetric complications, particularly in MIC(66, 110-115), with

multiparity often protective against adverse obstetric outcomes(116). Hence, the

46

impactofQoCmaybemorepronouncedinnulliparouswomen.Inonecomparisonof

parity-basedmaternalandperinataloutcomesinwomenhaving institutionaldelivery

in the Democratic Republic of the Congo, nulliparous women had higher rates of

obstetricinterventionandperinataldeath(4.93%comparedwith2.8%)(117).

It is evident from the above data that persisting poor quality of intrapartum care

continues to drive high perinatal mortality rates in MIC by its contribution to

preventabledeaths(118).

CAESAREANSECTIONINMIDDLEINCOMECOUNTRIES

WhileintrapartummortalityisunacceptablyhighinMIC,itisclearthatashifttowards

caesareansection(CS)toaddressthisissueisnottheentiretyoftheanswer,andtodo

sowouldpotentiallyincreaseperinatalandmaternalmorbidityandmortalityinsome

settings(3,119-122).Thatsaid,itcanbedifficulttoanalysewhethertheindicationfor

theCSortheprocedureitselfcontributesmosttothemortalitystatistics(123).Access

toCS inMICneeds tobeapartof comprehensiveemergencyobstetric servicesand

well-functioninghealthsystems(101).Moreover, inMICsettingswheretheCSrate is

alreadyhigh, forexampleBrazil, a reduction in the ratehasbeenshown to improve

maternal and perinatal outcomes(124). In a quality audit study in Nigeria,

paradoxically, fetal mortality increased with increased CS rates(125), which may

representinfrastructureissuesintimelydeliveryorintheuseofCSformaternalrather

thanfetalindications.WhileaccesstoCSisnecessaryforsafeobstetricoutcomes(126),

47

theQoCaroundCS inMICmay result inhighperinatal andmaternalmortality(127).

This isparticularlytrue inpoorlyresourcedsettings,wherethe“decisiontodelivery”

time once an emergency caesarean section is indicated may be delayed (due to

staffingandphysicalresources)suchthatperinataloutcomesareworsened(128).

48

PERINATALMORTALITYINSOUTHAFRICA

PerinatalmortalityinSouthAfricaisapproximately33/1000births(23),withstillbirths

contributing around two thirds of this rate, and neonatal deaths comprising the

remainder. This rate is highlighted as particularly unacceptable given that neonatal

deaths (deathsprior to28daysof life)account fornearlya thirdof themortalityof

childrenunderfiveinSouthAfrica,andalargeproportionofthesefollowintrapartum

hypoxia(129-131).Moreover,individualunitorregionalstillbirthratesarereportedas

fargreater,at30-40/1000births(132,133).However,therateofperinatalmortalityis

consistentlyreportedasimprovingoverthelasttwotothreedecades(130,134,135),

although some individual facilities have not followed this trend(136). One of the

challenges in perinatalmortality reporting in someprovinces is the largenumber of

women that deliver at home, oftenwithout a skilled birth attendant(135), although

thisnumberisprogressivelydeclining(130,137).

Reporting of perinatal mortality in South Africa is predominantly derived from two

sources; the District Health Information System (DHIS) and the Perinatal Problems

IdentificationProgram(PPIP)(130).SecondarytothePerinatalProblemsIdentification

Program(PPIP)(23),SouthAfricaisuniqueamongstLMICinbothitsperinatalmortality

data collection and ability for within country comparison. PPIP is a quality of care,

facility based audit system,which captures data around perinatal deaths, aswell as

identifypotentiallymodifiablefactors inthosedeaths.Since2012 it ismandatoryfor

allpublichealthfacilitiesinSouthAfricatousethePPIPsystem(129).PPIPreportson

49

PNMRforallbabieswithabirthweightmorethan500gr,aswellasratesfordifferent

weight categories. PPIP specifically reports onbirthweight rather than gestation, as

this is commonly not definitively known at the time of delivery. This represents a

limitationofthedata(commontolowandmiddleincomecountries),wherebyoutliers

of fetal weight for gestation (for example in the case of unidentified intrauterine

growthrestriction)maybeerroneouslyincludedorexcludedinthedata.

CAUSESOFPERINATALMORTALITYINSOUTHAFRICA

In themost recent PPIP report, which is a summary of the 24months prior to the

commencementoftheauthor’sthesisresearchinSouthAfrica,datawerecollectedon

76%ofallbirthsinSouthAfrica.Inthattimeperiod,therewere1,412,355birthswith

32,662stillbirthsand14,576earlyneonataldeaths(23).

50

Image 2.1: Saving babies. The most recent report from the perinatal problems

identificationprogram

The PPIP data reports on all levels of obstetric care in South Africa, including

community health centres and district hospitals. The research for this thesis was

conductedinaProvincialTertiary(PT)Hospital.Thehospitalwheretheresearchwas

conductedis intheTshwanedistrictwithintheGautengprovinceofSouthAfricaand

additionaldataforthisregion isoutlinedbelow.Thedata inthefollowingparagraph

areextractedfromthe2012-2013SavingBabiesReport(23).

51

ThePNMRforPThospitalsinSouthAfricais55.3/1000births,presumablyareflection

ofhigheracuitycasesandreferralfromother,lowerlevelsofcare.ThePNMRinthese

unitsforbabieswithabirthweightgreaterthan1000grams,aweightatwhichthereis

a significant increase in survival (compared with lower birth weight babies)(138), is

36.16/1000 births. The primary obstetric causes of death for all perinatal mortality

cases in PT hospitals are, in order of frequency, unexplained stillbirth, hypertensive

disorders, spontaneous preterm labour, antepartum haemorrhage, and intrapartum

asphyxia. For babies greater than 1000gr at birth, spontaneous preterm labour is

removedfromthislistandtheorderoffrequencyremainsotherwiseunchanged.The

PNMR in theTshwanedistrict is23.74/1000births,and16.09/1000births forbabies

weighinggreaterthan1000gr.

Whilst this research should be widely applicable in South Africa, it is important to

consider themortality data specific to the study location prior to commencing this

thesis.KalafongHospitalisa1000bedmulti-servicehospitalinPretoriaintheprovince

ofGauteng,withmore than 6000 deliveries per year. Table 2.2 shows the perinatal

careindicators(PCI)forKalafonghospitalin2013,theyearpriortocommencementof

the research contained in this thesis. The total PNMR was 14.8/1000 births, with

substantially higher rates in certain birthweight categories, for example a PNMRof

24.9/1000forbabiesweighing2000-2499gr.

52

Table2.2:PerinatalcareindicatorsatKalafongHospital,2013

53

PERINATALMORTALITYINSOUTHAFRICA:AREVIEW

A literature review was performed to explore the reporting of causes of perinatal

mortalityinSouthAfrica,inadditiontoPPIP,relevanttothisthesisandtoconsiderthe

applicability of our intervention. PubMed was searched, using the following search

strategy, restricted to articles published since 2000; (((South Africa) AND perinatal

mortality(MeSHTerms))ORfetalmortality(MeSHTerms)),aswellasanon-Meshterm

search for South Africa AND perinatal mortality. 788 articles were identified, with

relevantstudiesexploredbelow.Additionally, theSouthAfricanJournalofObstetrics

andGynaecologywashand-searchedfrom2006onwards.Themainpointsidentifiedin

thelimitedliteratureavailableareoutlinedbelow.

• Intrapartumasphyxiaisfrequentlyreportedintheliteratureasoneofthemain

contributors to perinatal mortality in South Africa, due to both (late or no)

access to hospital-based care and access to quality care once in the hospital

system,particularlyconcerningtheuseof fetalmonitoringandrecognitionof

fetalheartrateanomalies(1,131,132,134-136,139-146).

• Inananalysisof stillbirthsavertedbycarepractises,Michalowetal reported

that nearly three quarters of those averted were done so with intrapartum

interventions,includingaccesstocomprehensiveemergencycare(130).Infants

inSouthAfricaareatsignificantriskofdeathinthefirstweekoflifefollowing

clearly preventable intrapartum hypoxia(147) and this sentiment is echoed

repeatedlyintheliterature(137,140,144).

54

• There are large in-country variations in PNMR and intrapartum care; in one

retrospective audit using the PPIP data, the contribution of intrapartum

hypoxiawasfoundtobefargreaterinoneunit intheEasternCapethanthat

reportedinthenationalstatistics(148).

• The most common causes of perinatal mortality in South Africa are usually

(with slight variations in order of frequency, often depending on the level of

care) intrapartum asphyxia, unexplained intrauterine death, spontaneous

preterm labour, antepartumhaemorrhage,hypertension, fetal anomalies and

infection(130,133-135,144,149-151).Alackofqualityorcareorappropriate

responses when these situations arise contributes to both perinatal and

maternalmortalityinSouthAfrica(152,153).

THEGLOBALCLASSIFICATIONOFPERINATALMORTALITY

While themain trial in this thesiswasbased inKalafongHospital, itwasanticipated

thattheresultswouldpotentiallybeapplicabletootherLMICsettingsinternationally.

Thus,followingonfromidentifyingthecausesofperinatalmortalitybothatKalafong

hospital and in South Africa, onemust consider classifying them, such as is done in

PPIP,inordertobeabletoidentifyareaswhereinterventiononalargerscaleislikely

toimpactthePNMR.PPIP,however,isuniquetoSouthAfricaandthuscomparisonsof

theimpactofinterventionsindifferentsettingsoutsideofSouthAfricaarepotentially

difficult. The question arises: is all globally reported perinatal mortality the same?

Estimating and comparing global perinatal mortality is particularly challenging,

55

secondaryto largevariations inregistrationdataandinterpretationsofdefinitionsof

mortality,particularlyatthelimitsofviability(154).Thisisoneofthemainreasonsthat

the WHO uses the definition of 1000gr birth weight or 28 weeks gestation for

internationalcomparisons(22).

Including PPIP, there aremore than 80 perinatal death classification systems in use

around the world(155), making comparisons between countries particularly

challenging. To that end, based on experienceswith the PPIP system,we sought to

developa global classification system forperinatal deaths.Over2014and2015, the

World Health Organization, in collaboration with partners, developed The WHO

applicationof ICD-10toperinataldeaths: ICD-perinatalmortality (ICD-PM)(156,157).

Wewere thenable to further analysis the causesofdeath in SouthAfricausing the

ICD-PM. The author of this thesis co-lead the work (with Dr Ozge Tuncalp) on

developing the ICD-PM and undertook the initially field testing of the classification

system.ThepaperspublishedaboutICD-PMbytheauthorofthisthesisareoutlinedin

Appendix2.1.AfterthedevelopmentoftheICD-PM,Ifield-testedthesystemonnearly

10,000perinataldeathsfromtheUKandSouthAfrica.

The ICD-PM classifies perinatal death according to the time period of the death

(antepartum,intrapartum,orneonatal),andlinkstheperinataldeathtothematernal

conditionatthetimeofdeath(158).AnanalysisofSouthAfricandatausingtheICD-PM

provides insights in to the timings and causes of death in one large region in South

56

Africa.Halfofalldeathsinthisanalysis(344/689)occurredintheantepartumperiod,

with74(11%)occurring intrapartum,andtheremainder (271/689,39%)occurring in

theearlyneonatalperiod(21)(Figure2.2).

The major cause of death in the antepartum period was a fetal hypoxic event,

associatedmostcommonlywiththematernalconditionsofhypertensionorplacental

abruption.Intrapartumdeathswerelargelyacutehypoxicoccurrences,inthecontext

ofotherwisehealthymothers; the implicationsof thisbeing that thesedeathsarea

consequence of intrapartum care. Early neonatal deaths are distributed among

respiratorydisorders(largelyrespiratorydistressandmeconiumaspiration),lowbirth

weightandprematurity,andhypoxicischaemicencephalopathyfollowingintrapartum

asphyxia(21).

57

Figure2.2:DistributionofperinatalmortalityacrossthreetimeperiodsinSouthAfrica.

Adapted fromAllansonetal 2016(21). “Distributionn (%)ofperinatal deaths across

the antepartum, intrapartum, and neonatal time periods in South Africa (n=689,

October2013–January2014)”

GiventhisanalysisandthePPIPdata,itappearsalargeproportionoftheintrapartum

andearlyneonataldeaths inSouthAfricaareaconsequenceof thecareprovided in

theintrapartumperiod.This isaperiodduringpregnancywherecarewithinfacilities

occurs and therefore ispotentially amenable toalterations inpractice. Thus, for the

purpose of this thesis, we focused on this period and an intervention that would

potentiallyimpacttheseoutcomes.

n=344

n=74

n=271

0

10

20

30

40

50

60

Antepartum Intrapartum Neonatal

%

Timingofperinataldeath

SouthAfrica

SouthAfrica

58

FETALPHYSIOLOGYANDTHEIMPACTOFHYPOXIA

ACIDBASEBALANCEINTHEFETUS

In order to consider the impact of hypoxaemia, hypoxia, and acidosis in the

intrapartumperiod,anunderstandingofthenormalphysiologicalstateofthefetusis

necessary.

In its normal state, the fetus metabolises glucose to energy using aerobic

processes(159).Theendproductofthisprocessistheformationofpyruvateandthen

adenosinetriphosphate(ATP)(159).Maintenanceofacid-basehomeostasisinthefetus

islargelycontrolledbytheplacenta,andtheby-productofaerobicmetabolisminthe

fetus (CO2) is transferred to, and eliminated by, the placenta(160). The placenta

thereforefunctionstoresultinawell-oxygenatedfetuswithwell-oxygenatedtissues.

Tomaintain a homeostatic state, a fetus at termor near-term gestationwill have a

cardiacoutputbetween300and480ml/min/kilogramof fetalweight,with supply to

themyocardiumandthebrain(161,162).Whenhypoxaemiaandhypoxiaoccurs,the

fetus loses the capacity to efficiently convert pyruvate toATP, and thus switches to

anaerobic metabolism. This process results in rapid consumption of glucose and

accumulationoflactatewithasubsequentfallinpH.(Figure2.3)Thisultimatelyleads

toacidosis,tissuedamage,andtheclinicalmanifestationsofintrapartumasphyxia.

59

Figure 2.3 Aerobic and anaerobic metabolism. Adapted from Yli and Kjellmer,

2016(159)

INTRAPARTUMHYPOXIA

As previously described, intrapartumhypoxia canbe considered a “deficiency in the

amount of oxygen reaching the tissues(163)” during the first and second stage of

labour. Any period of hypoxaemia (interruption in fetal oxygen supply) has the

60

potentialtoresultinhypoxia(2),dependentonavarietyoffactors,whichareoutlined

laterinthischapter.Sustainedsignificanthypoxiawilleventuallyresultinacidosis,by

themechanismsdescribedabove,whichplacethefetusatriskofsignificantmorbidity,

includingneurologicalmorbidity,anddeath(2).

THEFETALRESPONSETOHYPOXIA

While the fetus is relatively hypoxic (compared to a neonate)(101), further limiting

oxygen supply to the fetus will result in one of three sequential steps occurring;

hypoxaemia, hypoxia, asphyxia(159). Hypoxaemia is defined a reduction in arterial

oxygen levels, hypoxia a reduction in oxygen supply to tissues, and asphyxia is the

consequence of acidosis (increased tissue levels of hydrogen ions) and acidaemia

(increased blood levels of hydrogen ions) (164, 165). If both acidosis and acidaemia

prolonged,thesewillleadtoadisruptioninthenormalacidbasebalance,withafallin

pH and an accumulation of PCO2 and lactic acid ions(160). As such, acidosis and

acidaemia (frequentlyused interchangeably in the literature)arehenceforthused to

reflecttheacidoticconsequenceofhypoxia.

Disruptions in oxygen supply to the fetus are more likely to occur at the time of

labour(159),howeverthefetusisdesignedtoadapttoadegreeofhypoxiapriortothe

manifestationofanyclinicalconsequence.Duringhypoxaemia,thefetuswill increase

the flow of blood through the ductus venosus in order to maintain adequate

61

oxygenation(161).Thefetuswillalsorespondtohypoxaemiabyreducingitsheartrate,

andincreasingbloodflowtobrain,myocardium,andadrenalglands(162).Theamount

ofhypoxia a fetus canendurebefore it switches tomaladaptivemechanisms varies.

Oncethefetusreachesitsself-definedlimitofprolongedhypoxia,itwillthenmoveto

thefarlessefficientformofenergyproductionofanaerobicmetabolism(159,164).

PRODUCTIONOFLACTATEANDFETALORGANSYSTEMRESPONSETOHYPOXIAANDACIDAEMIA

The fall in p02 following a hypoxic inducing event activates an acute fetal stress

response, mediated by the hypothalamic-pituitary-adrenal (HPA) axis, whereby the

production of corticotropin-releasing hormone (CRH), arginine vasopressin (AVP),

adrenocorticotropic hormone (ACTH), and glucocorticoids all increase (166). The

responsetoacutehypoxicevents includesarise infetalcortisol,whichwillgenerally

resolvewith the resolutionof theprecipitatingevent, andwitha consequent rise in

fetal lactate(167). Animal studies show complete cord occlusion results in a

progressive andmarked rise in fetal lactate,with variable recovery tobaseline after

resolutionofthe insult(168-171).Moreover,thegrowthrestrictedfetuscanmounta

similar lactateresponsetohypoxicstress,whencomparedtoitsappropriatelygrown

counterparts(172).

62

Thecomplexinterplayofanaerobicmetabolism,lactateproductionandindividualfetal

organresponsetoperiodsofhypoxianeedstobeconsideredastheresultantpotential

forclinicalconsequencesstemsfromthisinteraction.

CARDIACSYSTEM

The autonomic nervous system (ANS) regulates the fetal heart rate(159),whichwill

lower in response to hypoxia, largely vagally mediated, with the aim of reducing

overall fetaloxygenconsumptionandconservingenergy(173,174). It is importantto

notethatthefetalcardiacoutputwillremainstableinthepresenceofhypoxiawithout

acidaemia(162),whichisclearlyanadaptivemechanismforthefetusduringlabour.

Theinitialresponsetohypoxia(showninanimalstudieswithcompletecordocclusion)

isariseinmeanarterialpressure(MAP),withacorrespondingfall inheartrate(HR).

Blood flow will be redistributed to central organs and cerebral perfusion will be

maintained.TheMAPwillfallwithcontinuedhypoxia,howeverboththeMAPandthe

HRwillhaveareboundriseafterresolutionoftheinsult(168-170,174-177).Inavery

sophisticatedadaptivemechanism, the increase in lactate levelsand theconsequent

acidaemia is a further technique for the fetus to survive a hypoxic insult, as the

acidaemia aids the transfer of oxygen from haemoglobin to tissue(173). The aim of

thesecomplexcoordinatedresponsestohypoxia is to limit the impactof“threats to

fetalhomeostasis”(174).

63

RENALSYSTEM

Thefetalkidneywillgenerallymetaboliselactate,howeverwhenexposedtohypoxia,

thekidneywill insteadmetaboliseglucoseandexcretelactate(161).Bothrenalblood

flowandurineoutputwillfallduringhypoxia,andsimilartotheotherorgansystems

thereisaperiodofrenalhypoperfusionfollowingresolutionoftheinsult(170).

CEREBRALSYSTEM

Thecerebralsystemrepresentsthemainclinicalmanifestationofintrapartumhypoxia

resultingintissuedamage,withnearlyalldefinitionsofintrapartumasphyxiaincluding

somementionofalteredneonatalneurological state(178).Thecerebralvascularbed

undergoes vasodilation during hypoxia, promoting neurological protection.(173)

However, cerebral blood flowwill decrease during cord occlusion(171), and cortical

oedemawill occur during prolonged hypoxia(175). Animal studies show evidence of

decreased neurological activity (measured via EEG) during hypoxic insults(168, 179).

Thepost-insultperiodisnotwithoutriskeither,withre-perfusionandre-oxygenation

ofcerebraltissuesanatriskperiodfordamage.

64

GASTROINTESTINALSYSTEM

Cardiac output during a hypoxic insult is preferentially shunted away from the

gastrointestinal system and redistributed to the cardiac and cerebral systems. In

animal studies, there isevidence that thehypoperfusion to thegutduringahypoxic

episode is sustained after resolution of the insult, with cardiovascular collapse

followingthe insultresulting inafailureof immediatenormalredistributionofblood

flow(168,180).This furthersupports theconcept thatevenshortperiodsofhypoxia

significant enough to cause fetal circulation redistribution may have consequences,

evenintheinstanceofapparentfetal/neonatalrecovery.

THEPRETERMFETALRESPONSETOHYPOXIAANDASPHYXIA

Attheoutset,theresearchthatformedthebasisforthisthesisaimedtoincludeboth

preterm and term gestational age births. It is therefore pertinent to consider the

response that the preterm infant has to hypoxia, particularly in the intrapartum

period.Pretermfetusesareclearlyalsoatriskofintrapartummortality(181),however

thethresholdforthepretermfetustorespondtohypoxiaappearsgreaterthanthatof

its term counterpart, which is possibly a reflection of its immature autonomic

system(174).Preterm fetusesappear tobeable to survive longerperiodsofhypoxia

thantermfetuses,howevertheimmaturityoftheirphysiologicalsystemputsthemat

greater risk of more severe clinical consequences(174), and therefore potentially

contributesignificantlytohypoxiarelatedperinataldeathsinLMIC(182).

65

LACTATEASAMEASUREMENTOFMETABOLICACIDOSIS

As outlined earlier, lactate can be expected as an end product of glucosemediated

anaerobicmetabolism(159,183-185),whichoccursinthefetusinresponsetohypoxic

stress(166, 186, 187). The metabolism of pyruvate in this cycle results in the

productionoflactate(188-190),whichbyvirtueofbeingtheendproductinthecycle,

may give the most objective measure of metabolic acidosis in the fetus(191). UA

lactateisalmostentirelyfetalinorigin(189,192,193),makingitanidealmeasurement

forassessingthe impactof intrapartumcareonthefetusandnewborn.Additionally,

UA lactate is likely to reflect metabolic acidosis(191), which is more likely than

respiratory acidosis to result in poor neonatal outcome andpermanent neurological

damage(194, 195). Practically, UA lactate is simple to measure (as outlined in

subsequentsectionsofthischapter).

The most common precipitating event for lactate production in labour is transient

umbilical cord occlusion. Fraschet al demonstrated in animal studies an increase in

fetalarteriallactateof0.56mmol/Lperminuteofcordocclusion(196).Hypoxicinsults

beyond“normal”cordocclusioncouldresult ingreaterdegreesof lactateproduction

andacidoticconsequences.

66

THECLINICALCONSEQUENCESOFHYPOXIAANDACIDAEMIA

There is a complex interplay between the pre-exposure state of the fetus, and the

extent,frequencyandintensityofexposuretohypoxiaandacidaemiathatdetermines

the clinical consequences, if any, a fetus and neonatemaymanifest(166, 197). The

fetus, even in the preterm period, has the capacity to survive prolonged periods of

acutehypoxia,withanimalstudiesexploringtheconsequencesofupto30minutesof

complete cord occlusion(168, 169, 171, 179). More chronic hypoxia is seen in the

antenatal period with conditions such as fetal growth restriction and, whilst these

fetuses have the potential to have an uncomplicated antenatal intrapartum course,

this is not always the case, and they are at significant risk for both stillbirth and

intrapartum asphyxia(198). It is the sequence of events from hypoxia to the

developmentofasphyxiathatisofparticularinteresttothisthesisproject,giventhat

intrapartumasphyxialeadstothedeathsofmillionsofbabiesinLMICannually(199).

THECONCEPTOFASPHYXIA

Intrauterine hypoxia and intrauterine asphyxia are frequently used

interchangeably(200),howevertechnically,thetwoshouldbeseparated.Hypoxiaisan

(invivo)objectiveprocess,whileasphyxia isa farmorebroad term,encompassinga

conglomerationofclinicalmeasures(201).TheICD-10groupsintrauterinehypoxiaand

birth asphyxia together, but with varied definitions. Intrauterine hypoxia includes

67

“fetal or intrauterine acidosis, anoxia, asphyxia, distress or hypoxia”. Intrapartum

asphyxia is however defined with clinical terms including “pulse less than 100 per

minuteatbirthandfallingorsteady,respirationabsentorgasping,colourpoor,tone

absent”.(22)Anobjectivemeasurementofacidosis,suchaswithcordpH,baseexcess

or lactate level, can be used where available tomake the diagnosis of intrapartum

asphyxia(200),whichisappropriatewhensustainedhypoxialeadstoacidosis,thebasis

fortheclinicalsequelae(2).

CAUSESOFACUTEHYPOXIAANDASPHYXIAINTHEINTRAPARTUMPERIOD

Inordertofurtherinterrogateintrapartumhypoxiaandasphyxia,eventsthatcanlead

to the previously described physiological changes and consequent clinical outcomes

shouldbeconsidered.Normaluterineactivitywilltemporarilydiminishumbilicalblood

supply(164), however the degree of hypoxia and clinical response to that hypoxia is

likelyminimalinanotherwisewellfetus.

Thecausesofacutehypoxiathathavethepotentialtoresultinadverseoutcomesfor

the fetus are outlined in Table 2.3, with the causes, published rates for LMIC, and

consequences shown. These are particularly important to consider given the

contributionof intrapartumevents in LMIC to acutehypoxia, asphyxia andperinatal

deaths(90, 108, 202-204), and that the babies who do not die as a result of

intrapartum events are at significant risk of adverse neonatal outcomes such as

68

hypoxic ischaemicencephalopathy(205).Furthermore, thecomplications that lead to

intrapartumperinatal(andpotentiallymaternal)mortality,areoftenmoreprevalentin

LMIC(101). Once facility-based delivery is achieved, one of the next steps is the

collection of comprehensive data so prevalence can be clarified and areas where

interventions to improve PNMR can be identified and targeted. The scarcity of

prevalence and outcome data for obstetric complications in LMIC contributes to

ongoingissueswithresourceallocation(206).

Asphyxia in high income countries is of low prevalence and as such intrapartum

intervention programs may have no or minimal impact on the rate of cases(201).

HoweverthesameisnottrueinLMIC(36),wheretheoccurrenceofperinatalhypoxia

andasphyxia isassociatedwithasignificant incidenceofobstetriccomplicationsand

resultantadversematernaloutcomes(114).Obstetriccomplications inLMIC including

haemorrhage, severe pre-eclampsia/eclampsia, sepsis, prolonged/obstructed labour,

severe abortion complications, ruptureduterus, andectopic pregnancy are reported

byBaileyetalinupto30%ofwomen(42).

As outlined previously, intrauterine hypoxic injury commences with a reduction in

cerebral blood flow following a precipitating event (whichmay be acute or chronic,

andmaternal,placentalorfetalinorigin(207)),whichinturncausesthefetustoswitch

to anaerobic metabolism(159, 208), leading to asphyxia(165, 209, 210). It is

reasonable,toadegree,toexpectthataprecipitatingeventcanbeidentifiedtowhich

69

the outcomeof acidosis can be linked.We appreciate that predicting the degree of

responseafetushastoaparticulareventispotentiallymoredifficult(160).Regardless,

givenboththeratesofasphyxiainLMICandtheobstetriccomplicationratesreported

byBaileyetal,weconsideredtheratesandconsequencesofthecommonintrapartum

eventsthatmayresultinhypoxiaandasphyxiainLMIC(Table2.3).

Table2.3.Causesand frequencyof clinicalevents that result in intrapartumhypoxia

andasphyxiainLMIC.

Cause Frequencyreported inLMIC

Consequence

Normaluterineactivity(164,165,211)

N/A TransienthypoxiawithfallsinpHinnormallabour.

Suspectedfetaldistressandperinatalmortalityinthecaseofnuchalcord(78).

Excessiveuterineactivity(tachysystole)(212,213)

Not reported,data fromhigh incomecountries

Reported at a rate of approximately 10% in aretrospectivecohortofmorethan50,000cases.2/3hadnocorrespondingfetalheartratechanges.CaseswithtachysystoleweremorelikelytohavelowApgarscoresoradmissiontothenursery

Cordprolapse(214-230)

0.3-2.3% Caesarean section is themost likely consequenceofcordprolapse,withaPNMR21.7%-47.9%.

Survivaloutsideofhospitalisrare;thelengthoftimebetweenprolapseanddeliverynegatively correlateswith adverse fetal and neonatal outcome, as theacute situation results in total asphyxia – leading tolackofbloodflowandcellularinsulttobrainstem

Placentalabruption(28,67,69,231-234)

0.08-1.4% MayresultintheneedforCS,perinatalmortality(upto65%),neonatalneurological injury,hysterectomy,maternalsepsis,andmaternalmortality(upto2%)

Uterinerupture(28,42,235-238)

0.1-1% Commonly occurs after obstructed labour.Maternalmortalityupto20%,perinatalmortalityinupto95%,peripartumhysterectomy

70

OPTIONSFORMEASURINGHYPOXIAANDITSSEQUELAE

Regardlessofidentificationoftheprecipitatingevent,hypoxia,atathresholdofboth

length and levelwhich appears to vary for individual fetuses,will eventually lead to

acidosis, and possibly ischaemia, altered neurological function, and ultimately poor

clinicaloutcome(260-262).Theacidosis in thispathwaycanbeobjectivelyquantified

usingvariousmeasurements,includingumbilicalcordlactate,pH,andbaseexcess.In

LMICsettings,theApgarscore(whenlow,althoughwithheterogeneouslydefinedcut

offs)maybeusedasaproxyforacidosis,givenitsassociatedwithclinicalasphyxiaand

poorperinataloutcomes(263,264).

PH

MeasurementofumbilicalcordpHatbirth isgenerallyconsideredthegoldstandard

formeasurementofperinatalacidosis(165,207,265-268),anda lowarterialcordpH

Pre-eclampsia/Eclampsia(28,54,58,239-259)

1.02%-11.3% Maternal (up to 18%) and perinatal death (up to81%),neonatalmorbidity (RDS,NEC, IVH, Sepsis)upto55%

Intrapartumhaemorrhage(49,90)

1.9%-4.3%

Fetalgrowthrestriction(198)

5-10% of allbirths

Responsibleforuptoathirdofallstillbirthsandthemostcommoncauseinsomesettingsofintrapartumasphyxia.

71

(defined often as less than 7.2-7.0) may be associated with adverse neonatal

outcomes(269).

BASEEXCESS(BE)

Baseexcess isa calculatedvalue, requiringmeasuredvaluesofpHandpCO2,and is

themmolofbaseneededtocorrectthemeasuredpHto7.4,assumingapCO2of40

mmHg(270).Baseexcessinthecontextofacidosisisexpressedasanegativenumber

(and isa termthat isused interchangeablywithbasedeficit),andgivesanobjective

measureofmetabolicacidosisinthefetus(271)giventhattheconsumptionofbuffers

occurswith theproductionof fetalacids followingmetabolic rather than respiratory

acidosis.Theconsumptionofbuffersandanabnormalnegativebaseexcesspotentially

occursbeforeanymeasureablechangeinpH(272).

APGARSCORE

TheApgar scorewas first proposed in 1953 by the anaesthesiologist Virginia Apgar,

andassessesfetalconditionwithascoreoutof10,derivedfrom5parameters;heart

rate, respiratory effort, reflex irritability,muscle tone, and colour(273). A lowApgar

score has also been found to correlate with suboptimal intrapartum care(274) and

with poor neonatal outcomes(275-277), and may be used as a proxy for a more

objective measurement for acidosis and birth asphyxia, particularly in low resource

72

settings(278). There are obvious limitations to using Apgar score as a measure of

acidosis,which are addressed indetail in thediscussion sectionof Chapter 5. These

include,butarenotlimitedto,thesubjectivenatureofthescore,thelackofsensitivity

of the score for adverse neurological outcome and the potential for poor intra-

observerreliabilityofthescore.However,itisamarkerforacidosisfrequentlyusedin

LMIC and thereforewarrants inclusion in the options formeasuring acidosis and its

sequelae.

LACTATE

Oneobjectivemeasurementthatthecliniciancanusetogaugetheimpactofhypoxia

andthedevelopment,ifany,ofacidaemia,istomeasurethelactatelevelofthefetus

and neonate(279). Fetal lactate is increasingly used to measure fetal acidosis and

predicting risk inboth the intrapartum setting (from the fetal scalp)(185) and in the

immediatepost-partumperiod(fromtheumbilicalcord)(280).Benefitstothisinclude

thatcomparedwithothermeasurementsofacidosis(e.g.pH),itispotentiallyeasierto

implement as a clinical tool, with a higher rate of success, due to the need for a

significantlysmallersampleofblood(11,185,281).Moreover,lactatecanbemeasured

onahandheldmeter,makingitpotentiallymoreuseableinawiderrangeofresourced

challenged settings, than other measures of acidosis, which require a blood-gas

analyser(282).IthasalsobeenshowntobecomparabletopHinmeasuringacidaemia

(192,283,284),andisthemeasurementofinterestinthisthesis.

73

MEASURINGUMBILICALLACTATE

Thebenefitofusinglactatetoassessacidosisisthatitcanbeperformedasapointof

care(POC)test;thatisonahandheldbedsidemachineasasimpletestrequiringvery

fewresources.UnlikemeasuringpH,whichgenerallyrequiresaminimumof30-90uL

ofblood(285),lactatecanbemeasuredonaPOCdeviceusingaslittleas0.5uL-15uLof

blood(286,287).

The lactate cut-off levels for fetalornewbornacidaemiameasuredonaPOCdevice

varybetweendevices(286,288).Thelactatelevelinthisthesisresearchwasmeasured

on a Roche Accutrend PlusTM© hand-held lactate meter (Rotkreuz, Switzerland),

whichmeasures lactate in a range of 0.8-22mmol/L in 60 seconds onwhole blood,

requiringasampleof15-50uL(Figures2.2aand2.2b).

Image2.2aand2.2b:AccutrendPlusTM©hand-heldlactatemeter

74

The device has been shown to be accurate in themeasurement of lactate from the

umbilical artery (as planned for our research study) at results between 1 and

20mmol/L.Ithasalsobeendemonstratedasaccurateinmeasuringfetalscalplactate,

further adding to our reasoning to use this machine in our research(289). The

Accutrend has a a coefficient of variation (CV) of 1.8-5%(287, 290) and ismeasured

using reflectancephotometry, after a lactateoxidase-mediated colorimetric reaction

occurs(291)(Figure2.4).

Figure2.4:AnexampleofthecolorimetricreactionusedbytheAccutrendmonitor,as

describedbyEmmert(292).

TIMINGOFLACTATESAMPLING

UsualpracticeatKalafonghospital is theclampingof theumbilical cordwithinone

minuteofbirth.This is important,giventhat lactatelevelshavebeendemonstrated

tochangewithdelaysinbothcordclampingandsampling.Followingvariablereports

of outcomes of delays in sampling, recommended practice included early cord

75

clamping (ideallywithinoneminute)andsamplingofthecordwithin20minutesto

benecessary[23],whereasothers findnodifference inbloodgasparameterswhen

cords are immediately clamped or delayed for greater than one to two

mintues[Tang, 2019 #632]. Dessolle and colleagues collected cord arterial lactate

samplesatdeliveryandataperiodof5-30minuteslaterandfoundacorresponding

linear rise in lactate levels in the second sample, with a rise of 0.062mmol/L per

minute(293).DePacoet alhavereportednosignificantdifferenceinumbilicalartery

pHlevelsinumbilicalcords(termand healthy)clampedimmediatelyafterbirthandin

thosewheretherewasadelayofuptotwo minutes[24].

Whiteet al describe an increase in lactate levels after 15minuteswhenbloodwas

bothstoredinasyringe,orleftinsitu(inthecord)(294).

76

METHODOFSAMPLING

Theequipmentnecessarytosample:Thenormalumbilicalcordcontainsoneveinand

two arteries, with a surrounding layer of Wharton’s Jelly, connective tissue and

amnion(295). Either artery can be sampled for the assessment of UA lactate. The

arteries are usually smaller and firmer (due to their muscular structure) than the

tortuousvein(Images2.3and2.4).

Image2.3:Asegmentofumbilicalcordwiththelargerveinclearlyvisible.

77

Image2.4:Acrosssectionofumbilicalcordshowing2arteriesand1vein.

Onceidentified,asmallcalibreneedlecanbeinsertedintoeitherartery,andasmall

amount of blood drawn with relative ease. The syringe used is often heparinised,

however as a long as the sample is measured shortly after it is drawn, a non-

heparinisedsyringecanbeused,makingtheprocesscheaper(296,297)(Image2.5).

78

Image2.5:A1mlnon-heparinisedsyringeand26-gaugeneedle.

CORDARTERIALLACTATELEVELS

Cord UA lactate is used in many settings to assess acidosis (or lack thereof) in the

immediatenewbornperiod.Cut-offvaluesusedincohortstudiesareoutlinedinTable

2.4.

79

Author Numberof included

samples

Lactateresults(mmol/L)

Chilindaetal(298) 391 Median 3.4 (IQ range 2.6-4.9), with a

cut-off of 5 for admission to the

nursery

Dessolleetal(293) 149 Mean 3.76 (SD + 1.53, range=1.1-9.8).

90thpercentile=5.72

Gjerris(192) 2554 Mean 4.63 (SD + 2.33),with amedian

of 4.30 and a cut-off for neonatal

acidosisof8

Labrecque(285) 52 Median3.7(range1.7-8.8)andacutoff

4.9 for the outcome of acidosis (as

definedbypH)

Linet(299) 200 Median 3.2, and a cut off 4.9 for the

outcomeofacidosis(asdefinedbypH)

Martin(300) 763 Mean3.71(SD+1.81),97thpercentile

oflactatewas7.54

Racinetetal(301) 867 Mean3.4(SD+0.05)

Tuuli(183) 4910 Theoptimalcutoffvalue forpredicting

neonatalmorbiditywas3.90

80

Westgren 4045 90th centile 3.21 with a sensitivity of

57% forneonatal death and specificity

of90%

Whiteetal(302) 21182 OptimalcutoffforHIE=5.68

Wibergetal(303) 13735 Optimalcutoff6.0 for5minuteApgar

<7

Table2.4:Cut-offvaluesforcohortsassessingacidosisusinglactatelevels.

FETALHEARTRATEMONITORING

Identification of the hypoxic fetus such that the consequence of acidosis can be

avoidediscentraltothisthesis.Inlabour,thesuspicionoffetalhypoxiaisoftenbased

onabnormalfetalheartratemonitoring.Inmanysettings,includingKalafonghospital

wherethemainstudyofthisthesiswasconducted,thisisdonewithacardiotocograph

(CTG),aformofelectronicfetalheartratemonitoring.

TheCTGisperformedwithtwotransducersplacedonthepregnantabdomen.Oneisa

pressure transducer that assesses contractions, such that the relationship between

fetal heart rate changes and contractions can be determined. The other transducer

measuresfetalheartrate.TheInternationalFederationofGynecologyandObstetrics

(FIGO) developed the first international consensus on fetal heart ratemonitoring in

81

1987, and thesewere still in place at the time of the clinical trial conducted in this

thesis(theywereupdatedin2015)(304).TheFIGOguidelinesforCTGcomponentsand

interpretationareoutlinedbelow(305). Itshouldbenotedthatsomecomponentsof

theCTGarenotdefinedbyFIGO,butaredefinedbytheRoyalCollegeofObstetricians

and Gynaecologists, and the research outlined in this thesis used these additional

definitions.

82

FEATURE DEFINITION

Baselinerate Baselinefetalheartrateisthemeanlevelofthefetalheartratewhenthis is stable, accelerations and decelerations being absent. It isdeterminedoveratimeperiodof5or10minandexpressedinbeatsperminute(bpm)

• Normal=110-160bpm• Tachycardia = not defined in FIGO, defined by RCOG as

>180bpm• Bradycardia<80bpm

Figure2.5CTGwiththenormalfetalbaselineHRlyingwithinthe110-160bpmareadefinedbythegreyline

Variability Oscillations of fetal heart rate around its mean level (long-termvariability). This is usually only quantitated by description of theamplitude of the oscillations around the baseline heart rate. Underphysiologicalconditionsthefetalbeat-to-beatintervalsareconstantlysubjecttosmallchanges.

• Normal=5-25bpm• Reduced<5bpmformorethan40minutes• Increased>25bpm

Figure 2.6: CTGs showing decreased, normal and increasedvariability

83

Accelerations Transient increase in heart rate of 15 bpm or more and lasting 15secondsormore

Figure2.7:CTGshowingaccelerationsabovethebaseline

Decelerations Transientepisodesof slowingof fetalheart ratebelow thebaselinelevel of more than 15 bpm and lasting 10 seconds or more.DecelerationsarenotdefinedinFIGO,butaredefinedbyRCOG

• Earlydecelerations=Uniform, repetitive, periodic slowingofFHRwithonsetearlyinthecontractionandreturntobaselineattheendofthecontraction

Figure2.8:CTGshowingearlydecelerations

• Late decelerations = Uniform, repetitive, periodic slowing ofFHRwithonsetmidtoendofthecontractionandnadirmorethan20secondsafterthepeakofthecontractionandendingafter the contraction. In the presence of a non-accelerativetracewithbaselinevariability less than5bpm, thedefinitionwouldincludedecelerationslessthan15bpm

84

Figure2.9:CTGshowinglatedecelerations

• Variable decelerations = Variable, intermittent periodicslowing of FHR with rapid onset and recovery. Timerelationshipswithcontractioncyclearevariableandtheymayoccurinisolation.

Figure2.10:CTGshowingvariabledecelerations

• Prolongeddecelerations=anabruptdecreaseinFHRtolevelsbelow the baseline that lasts at least 60–90 seconds. Thesedecelerationsbecomepathologicaliftheycross2contractions(i.e.greaterthan3minutes)

85

Figure2.11:CTGshowingprolongeddecelerations

Table 2.5: Components of the CTG. Adapted from Ayres-de-Campos and Bernades,

2010(305),andtheKEMHAdvancedFetalAssessmentprogram.

Oncethecomponentsof theCTGareassessed,onemustdetermine ifabnormalities

seenaresuspiciousorpathological,anddecideontheappropriateintervention.Thisis

wheremanydisagreementsintheliteratureandbetweencliniciansarise(306).

THECTGANDACIDOSIS:AREVIEW

Theultimateaimoffetalmonitoringinlabouristopreventacidosis,whichisthebasis

of the abnormal sequelae that follow hypoxia. The occurrence of hypoxia itself, as

outlinedpreviously,willnotresultinclinicalconsequenceandsooneneedstobeable

to intervene such that the potential complications of the intervention are balanced

againsttherisksofadverseneonataloutcomewithprogressivehypoxia. Inthestudy

relatedtothisthesis, intrapartumassessmentofacidosisthroughscalpsamplingwas

not feasible, so the focusof this reviewwasonacidosisdiagnosed fromasampleof

86

blood from the umbilical cord after birth. The concept being that taking a

measurementof acidosis on all births functions as an audit of intrapartumcare and

allows clinicians receive real-time quality of care feedback on their practice,

particularly their interpretation of and response to the fetal heart rate changes in

labour(307). Thus, in order to interrogate the interaction between fetal heart rate

monitoring and consequent acidosis, a literature review was undertaken using the

following search (((((cardiotocography) OR cardiotocogram(MeSH Terms)) OR fetal

monitoring(MeSH Terms)))) AND ((acidosis) OR (acidosis(MeSH Terms)) OR (fetal

distress(MeSH Terms)OR (neonatal outcome)OR (neonatal acidosis))). 1021 articles

were identified from2000onwards.After title and abstract review, relevant articles

areexploredbelow.

WhiletheCTGremainsstandardofcareinlabourformanywomenaroundtheworld,

there is a plethora of evidence, including randomised evidence, that, while its

specificityisat least90%(308), ithasapoorsensitivityforacidosis.Thereareseveral

issuesthatfollowonfromthis.

• SensitivityoftheCTGforpredictingacidosisandadversefetaloutcomemaybe

aslowas15%(309-313).

• Therateofintervention(oftencaesareansection)withoutpotentialbenefitto

mother or fetus is high(271, 314-319), although it is clearly sometimes both

necessaryandhelpful(320).

87

• ThePPVof theCTG foracidosis (evenwhenmeasuredusingcomputerbased

algorithms) is reportedbetween27.4%and50%(314,321-324). In fact, in the

large randomised INFANT trial of more than 47,000 cases, there was no

difference in the rate of acidosis when the interpretation of the CTG by

clinicians was aided by the addition of computer based decision support

software(325), supporting the concept that the poor sensitivity of the CTG is

notonlyaconsequenceofvariationsinclinicalinterpretationbutisalsoaresult

ofvaryingfetalcapacitytocopewithhypoxicstress.

• ThereislittletonoevidencethatinterveninginabnormalCTGwithcaesarean

delivery reduces the rate of acidosis(326, 327), although in many situations,

women more often than not are delivered by caesarean section when fetal

compromiseissuspected(328).

RespondingtotheCTGduringlabourrequiresthecliniciantoconsidereachaspectof

the heart rate (as outlined in Table 2.5) and diagnose the fetal heart rate as being

somewhereonthespectrumbetweenreassuringandpotentiallypathological.Various

interpretationtoolsforCTGsexist,howeverinterandintra-observerreliabilityofthose

tools is often poor, resulting inmarkedly differing sensitivity for predicting acidosis,

whichmakesthetaskforcliniciansdifficult(329-334).Cliniciansarefrequently,butnot

universally, able to agreewhen CTG traces are reassuring or normal, but frequently

unabletoagreewhenCTGtracesareconcerningorominous(317,335).Inareviewof

inter-observervariability,clinicianswereunabletoagreeonwhichofthenearly50%

ofsuspiciousCTGtracingswouldbeindicatedforemergencydelivery,ortopredictin

88

which2%ofcasesacidosiswasevidentondelivery(336).Kunduetaldemonstratethe

challengesofCTGinterpretationverysuccinctlywiththeirretrospectivereviewof300

cases.Thecasesweregivento7variablyexperiencedclinicianswithboththeobjective

acidosisresults(UAgasanalysis)andthedeliveryoutcomes(mode,Apgarscores,etc.)

blinded.ClinicianswereverypooratpredictingthepHbasedontheCTGappearances,

and therewasmarked inter-observer variability of CTG interpretation(337). Overall,

clinicianshavedifferentthresholdsforinterventionfortheconceptofsuspectedfetal

distress,whichmakestheinterpretationoftheCTGachallengingclinicalissue.

The individual components of the CTG (Table 2.5) are variably related to neonatal

acidosisandoutcome.Cliniciansareoftenreassuredbythepresenceofnormal fetal

heartratevariabilityandfetalheartrateaccelerations,whichlikelyexplainswhymost

can agree when CTGs are considered normal. The presence of normal variability is

associatedwithahigh likelihood (at least>85%)ofanormalpH(338,339), although

conversely in their large retrospective cohort Milson et al found a sensitivity of

reduced fetal heart rate variability for the outcome of asphyxia of only 19%(340).

Thereissomeevidencethataccelerationsinlabourareunlikelytobeassociatedwith

acidosis,althoughthereisadescribedfalsenegativerateofatleast6%(341).

TheCTGchangesthatconcernaclinicianneedtobetakenwithinthecontextof the

entireCTGandclinical scenario(342);achangesuchasprolongedbradycardia is less

likely to result inacidosisandadverseneonataloutcome if theperiodpreceding the

89

bradycardiawassuggestiveofanormalfetalphysiologicalstate(343-345).Herein lies

the challenge – even ominous CTG patterns are not likely to result in an adverse

outcome.Chuangetalshowedsensitivityforadverseneonataloutcomeforprolonged

bradycardia of only 29%(346). The frequency of occurrence of decelerations or the

lengthofaprolongedbradycardiaaremorelikelythannormalCTGstobeassociated

withneonatalacidosis,butmanycaseswillhaveneitheracidosisnoradverseclinical

outcome(347). Even when the fetus that has both decreased variability and late

decelerations,bothconsideredominouspatterns,WilliamsandGalemeaufoundarate

ofacidosisinthisscenarioof44%(339).

Even a common pregnancy occurrence such as a nuchal cord, which the clinician is

unabletomodify,canresult inCTGabnormalities,particularlyvariabledecelerations.

Again,withanotherwisenormalCTG,theseareunlikelytobeassociatedwithadverse

outcomes,howeverthecontinuoustransienthypoxiathatmaybeassociatedwiththis

clinicalscenariomayleavethefetuswithlessreservetocopewithother insults(346,

348,349).ThefrequencyordurationofabnormalCTGpatternsmaybemorelikelyto

beassociatedwithadverseneonataloutcomes(350),butthetimeofabnormalities is

clearlyacontinuum,andagaincliniciansareinthedifficultpositionofdecidingwhen

tointervene.

There is no doubt thatwomenwith adverse fetal andneonatal outcomeswill often

haveabnormalCTGs,howeverthereverseisoftennottrue(351,352).Thefetushasan

90

extraordinary capacity to cope with hypoxia and acidosis before there are clinical

manifestations(317),andasfewasaquarterofthemostpathologicaltraceswillhave

consequent neonatal encephalopathy(353). Cahill et al undertook a retrospective

studyof951womenwith“terminaldecelerations” (prolongeddecelerationswithout

recoveryofatleast120seconds)andfoundthatmorethan98%didnothaveacidosis

onUAsampling(354).Furthermore,notallcasesofneonatalencephalopathywillhave

a pre-delivery pathological CTG(355), making interpretation of CTGs to prevent

acidosisallthemoredifficult.Onreviewingcaseswithacidosisdetectedonumbilical

cord sampling, Olofsson et al found that one third of CTGs showed both normal

baselineandnormalvariability(329).

Decelerationsareofteninterpretedwithconcernbyclinicians,howeverthepresence

of decelerations may be just as common in fetuses without acidosis as in those

with(321, 356, 357). There is a clear relationship between the length of time of

decelerations and consequent acidosis, however this is a continuum that requires

clinician discretion to choose themoment for intervention(306, 358-362).Moreover

there remain issueswith the interpretationofdecelerations; the conceptof atypical

features of decelerations being associatedwithworse outcomes is not borne out in

large cohort studies(363). Even variability, which is often regarded as the most

important feature of the CTG(340, 364), has a PPV for acidaemia of, at most, 17-

67%(365-369).

91

FetaldistressisacommonindicationforcaesareansectioninbothhighandLMIC(370),

including South Africa(371, 372). The diagnosis however (often based on fetal heart

ratetracing)isnotnecessarilyborneoutintermsofadverseneonataloutcomes(319,

373, 374), with no apparent association between CS for fetal distress in one South

African study and adverse perinatal outcome(375). Indeed, a Cochrane Systematic

review comparing operative intervention with conservative management for fetal

distress includes only one RCT from 1959, with no difference in perinatal

mortality(376). Despite the absence of contemporary RCTs, there is no lack of non-

randomised evidence regarding the limited role of the CTG to improve perinatal

outcomes.

InLMIC,therearesimilardatatothosedescribedaboveonthepoorpredictabilityof

CTG. In one prospective Indian study, while the CTG abnormality correlated with

adverse outcomes, the correlation was not impacted by decision to delivery time,

further cementing the poor PPV of CTG interpretation(377). Similar findings are

replicatedinmanyLMIC(378).

• ANigerianreviewof300casesofsuspectedfetaldistressshowedthatnotonly

was the time from suspicion to caesarean section (the planned intervention)

greater than 2 hours in the majority of women, less than a quarter had

diagnosedbirthasphyxia(379).

92

• Of a 146 women having a CS for suspected fetal distress in one Indian

prospectivecohort,only2%of caseshad that suspicionborneout inacidosis

andsubsequentpooroutcome(380),and inanothercase-control study,more

thanathirdofcaseswithsuspectedfetaldistressandUAprovenacidosishada

normal clinical outcome(381). That said, the aim of intervening for hypoxia

beforeacidosisoccursisimportant(382).

• In one study in Pakistan for a trace that was defined as suspicious or

pathological(FIGO),thePPVforhypoxiawas18%andacidosis26%,inanother

similarretrospectivecohort,thesensitivityofCTGtopredictacidosisbasedon

BEwas15%(383). This canbe improvedwithpathologicalonly traces to50%

(228,384).

Asinhighincomecountrybasedstudies,thepresenceofabnormalitiesonCTGismore

likelytobeassociatedwithabnormalneonataloutcomes(367),andmoreover,onecan

findevidence inbothhigh incomeandLMIC for the introductionofCTG resulting in

improvedperinataloutcomes,includingimprovedmortality,howeverthismaybeasa

resultof trainingandawarenessaroundhypoxiaratherthanthe introductionofCTG

itself(109,385,386).

FetaldistressisfrequentlycitedintheliteratureexploringperinatalmortalityinLMIC.

Itisclearlyacomponentofintrapartumcarewhereattentioncanbefocused,however

increasingcaesareansectionratesinLMICinresponsetosuspectedfetalcompromise

doesnotnecessarilyhaveany impactonPNMR(387).Moyeretal reportan8%fetal

93

distressrateinnearly5000deliveries,withanoverallperinatalmortalityof10%(388).

While the crossover between these twodiagnoses is not clear in this study,what is

clearisthatinterventionstoaddressfetaldistressareofvalueinLMIC,acceptingthe

obviouslimitationsoffetalheartratemonitoringoutlinedabove.Therearetwomajor

issues with this concept. Firstly, the availability of CTG in LMIC is often scarce to

absent(389), however it isworthnoting thatmonitoringwithout aCTG, and instead

withahandhelddevice,maybeaseffective(orineffective)ascontinuousmonitoring,

particularlyinLMIC(101,390,391).Secondly,eveninthescenariowherefetaldistress

issuspectedbasedonheartratetracing,thecapacityto intervene(forexamplewith

instrumentaldeliveryorcaesareansection)islimitedbytheavailabilityofresourcesto

doso,and therefore thepotentialbenefitsof interventionarenotnecessarilyborne

outinneonataloutcomes(392).Thatsaid,thepoorperformancecharacteristicsofthe

CTGmeansthatprolongeddecisiontodeliveryintervalsmaynotnecessarilyresult in

adverseneonataloutcome(393).

THENEONATEWITHACIDOSIS

Acidosis,asoutlinedearlierinthischapter,isdefinedvariablyinpractice,whichresults

inheterogeneousreportingofratesofacidosis.WhileoftendeterminedusingpH,the

cut-off for this is reported ranging between7.0 and7.2.We therefore rely on large

cohortstudiestobestestimatetherateofacidosis,asdeterminedwithauthordefined

cutoffs,andacceptingthelimitationsofreporting.Whiteetalintheircohortof19646

births,reportedarateofacidosis(pH<7.0)oflessthanonepercent(394),withsimilar

94

findingsrepeatedinotherlargecohortstudies(395-397).Itisclearthatnotallacidotic

neonates have adverse clinical sequelae, and equally not all unwell neonates are

acidotic.Itshouldbenotedthatneonatalacidosisinthiscontextisconsideredacidosis

diagnosedimmediatelyafterbirthfromtheumbilicalcord,andis largelyreflectiveof

intrapartumevents.Itisimportantinintroducingatooltoobjectivelymeasureacidosis

atthetimeofbirththatoneconsiderstheadverseoutcomesofacidosisaswellashow

tomanagethe“well”newbornwithacidosis.

NEONATALACIDOSISANDADVERSEOUTCOMES

Worsening levels of acidosis are associated with increasing likelihood of neonatal

resuscitation, low Apgar scores, admission to the neonatal nursery, and hypoxic

ischaemic encephalopathy(268, 398). Acidosis can result in seizures (Williams and

Singh found a sensitivity of 73% for acidosis at birth (defined as pH<7.0) for the

outcome of neonatal seizures(399)), and long term neurological complications

includingcerebralpalsy(2).

Itisclearthattheclinicalpictureguidesthemanagementofthenewborn,ratherthan

the presence or absence of acidosis. While babies born with low Apgar scores are

often acidaemic(400), there are a host of reasons for why neonates may have low

Apgarscores.Intheircohortofnearly28,000patients,Locatellietalfoundthatonlya

thirdofneonateswith lowApgarscoresalsohadumbilicalacidaemia(401).Likewise,

95

Lavrijsenet al found in their case control study that seizuresonly occurred in up to

20% of neonateswith umbilical acidaemia at birth, with no clear difference in long

termoutcomesbetweenacidaemicandnon-acidaemiccases(402).

THEWELLNEWBORNWITHACIDOSIS

The presence of acidosis at the timeof birthwithout apparent clinical consequence

presentsachallengeforclinicians.Whileitmaybepossibletodismissthisscenario,we

couldequallyconsiderthatwhenacidosisisidentifiedonumbilicalcordsamplingand

the newborn is otherwisewell, a neonatal nearmiss has occurred. That is, that the

event that leadtoacidosishasnot resulted inadverseoutcomebutcouldhave,and

reviewof thecasesmay revealanareaofpotential improvement to reduceacidosis

ratesanditspotentialmorbidandmortalconsequences.

AsdescribedbyBonnaerensetal,theconceptofneonatalnearmissisnotwidelyused

in audits of intrapartum care, but has the potential to value-add to perinatal care,

particularlywhencasesofnearmissare frequently inwomenandneonateswithno

identifiable pre-labour risk factors(403). In undertaking universal screening for

acidosis, Bonnaerens and colleagues found amore than doubling in the number of

cases where audit was warranted and improvement of perinatal care was

possible(404).Thiswasfurtherconfirmedwhentheinitialstudygroupwasexpanded

across threeyears(405). Jonssonandcolleaguescollecteddataonmore than28,000

96

deliveries,withametabolicacidosisrateof0.68%.Whilemostneonatesdidwellinthe

postdeliveryperiod,theyestimated,basedonauditofacidosiscases,thattheacidosis

couldhavebeenavoidedinnearlyhalfofallcases(406).Itisimportanttoconsiderthat

metabolicacidosiscanbeboththecauseandconsequenceofadverseintrapartumand

neonataloutcomes.Ifallacidoticcaseshaveanidentifiablepotentialprecipitant,then

itstandstoreasonthatreviewofthesecasesmaystimulateachangeinpractice,even

where the ultimate outcome for the neonate is positive (407). The identification of

acidosisatbirth,whileuncommon,maybeasignofadditionalriskofneonataladverse

events, including respiratory distress syndrome and NICU admission, and universal

testingand reviewof casesmay stimulatealternatemanagementof suchcases(268,

408).

Inthecaseofthewellnewbornwithacidosis,isthereapotentiallong-termimpactof

the acidosis? Dani et al studied 53 infants with an unexpected cord arterial pH

between 7.0 and 7.1, and more than half were asymptomatic with no short-term

consequences (409).However, there is someevidence to suggest that the long-term

developmental outcomes, although clinically minor, of these neonates may be

differenttotheirequallyclinicallywellbuteucapniccounterpartsatvariousstagesof

childhood(159, 410). Themeasurement of acidosis (or lack thereof) on all deliveries

therefore provides the clinician with the opportunity to review the “near miss”

intrapartum events, with the aim of reducing the rates of acidosis and potentially

impactinglongtermoutcomes.

97

QUALITYOFCAREAUDITS

DEFININGQUALITYOFCAREANDQUALITYOFCAREAUDIT

QualityofCare(QoC)isdifficulttodefineasitisoftencontextual(411,412),howeverit

generallyencompassescomponentscontributingtomorethanonelevelofcare(413),

and is care far beyond the basic provision of essential and emergency medicine.

Tuncalp et al define QoC in maternal and newborn health as that reflecting the

“physical infrastructure, supplies, management, and human resources with the

knowledge,skillsandcapacitytodealwithpregnancyandchildbirth”(5).Moreover,it

is consistently repeated in recentmaternal andnewborn literature thatQoCaround

the time of childbirth has the capacity to save millions of women and newborn

lives(414-419).

Audit is defined as “a systematic review or assessment”(163), a process integral to

healthserviceprovision,andhasseveralroles:

• It allows comparisons of care within a system against an agreed

standardofcare(420).

• Itmaybeusedasanassessmentofspecificandobjectivemarkersofa

desiredoutcome (e.g.perinatalmortality as a functionof intrapartum

care)(421).

98

• It canbe amotivator for adequatedata collection,which is critical to

theusefulnessofaudit(422).

Thus, clinical audit is critical to the identification of potential problems and focused

facility auditswithin awider system can identify contextually specific health service

deficienciesandprovidetheimpetusforchange(411,422-425).

Furthermore,havingauditthatdetectsproblemsinthestepstoaparticularoutcome

(quality of care audit) is more likely to be sensitive for identifying areas that need

improvement(125, 426-428). Indeed, rather than assessing an endpoint as the sole

goal,auditcan(andonewouldargueshould)beextendedtoqualityofcareaudit in

whichmarkersoftheprocessthatleadtoanendpointcanreflecttheappropriateness

and quality of the care given(429). Merali et al undertook a systematic review of

identified avoidable factors in maternal and perinatal audit in LMIC. In addition to

identifying42avoidablefactorsindeathaudits,theyestimatednearlyathirdofdeaths

werespecificallyrelatedto“substandardhealthworkerpractice”(430).Thisconceptof

specificcriteriabasedauditallowscliniciansandresearcherstohighlightdefinedgaps

wherebyresourcescanbetargeted(431).Tothatend,thestepstoimprovingqualityof

care, includingauditasapartofthatprocess,havebecomeaglobalfocusinwomen

andnewbornhealth(417).

99

QUALITYOFCAREAUDITINMATERNALANDNEWBORNHEALTH

Comprehensive data registries in LMIC are critical in order to facilitate audit that

identifies specific clinical and systempractises,whichmustbeaddressed inorder to

affectanychangeonPNMR(432-436).ThisisofparticularimportanceinLMIC,where

datacapturearoundperinatalmortalityisfrequentlypoor(437,438),andthemajority

of avoidable perinatal mortality is attributed to the poor quality or absence of

intrapartumcare(430,439).There is someevidence thatauditalone (asan inherent

intervention)mayimproveclinicaloutcomesandmortalityrateswithameta-analysis

ofsevennonRCTtrialsshowingthattheprocessofintroducingauditwithinaspecific

systemreducesmortalitybyupto30%(440).Audithasbeenshowntoattheveryleast

improveclinicalpractice,andvariablytoimprovemortalityoutcomes,althoughthere

isaclearlackofrandomiseddataonaudit,whichisinandofitselfaninterventionthat

wouldwarrantaRCT(441,442).

At itsmostbasic level,simpledatacollectionover6years inruralNigeriaresultedin

improvedmortality statistics(242) and the registrationofwomen in ruralGuatemala

onapregnancydatabase(communitylevelaudit)resultedinareducinginallmaternal

andperinatalmortality statistics, presumably secondarily to theprocessof engaging

pregnantwomenandincreasingtheirfacility-basedbirth(51).Pragmaticaddressingof

systemproblemsidentifiedbyaudithasbeenshowntoimprovemortalityinaKenyan

study(413),althoughdesignflaws(withonlya1/3ofperinataldeathsincluded)make

theevidencedifficulttointerpret.MovingfromsimpledatacollectioninaunitinFijito

100

standardisedperinatalmortalityauditresulted identificationofhealthsystemfactors

that could be addressed to reduce preventable deaths(443). Similar findings were

showninaTanzanianstudy,whereashifttocriteriabasedauditresultedinimproved

recognitionandmanagementofsuspectedfetaldistress(109).Furthermore,thelarge

multi-countryMaternalNewbornHealthRegistry studyclearly showscomprehensive

datacollectionfeedsandauditprocessthatclearlyresults in improvedmaternaland

newbornoutcomes(434).

Theimpactofauditonresultingchangesinpracticehasthepotentialtobesignificant,

aswasthecaseintheHlabisaHealthDistrictofKwaZulu/Natal,SouthAfricain1991-

1995. Rigorous classification of avoidable perinatal deaths and audit resulting in

changestohealthservicedeliverysawareductioninperinataldeathfrom18%to0%

overafouryearperiod(444).Similarly,theuseofthePPIPsysteminanEasternCape

hospitalinSouthAfricawasassociatedwithareductioninintrapartumhypoxicrelated

deaths, attributed to both the awareness generated by audit and the addressing of

specificsystemissuesthatauditidentified(148).Theintroductionofacomprehensive

training program in audit andperinatal death classification inMoldova resulted in a

reduction inperinatalmortalityandan improvement inqualityof care(445). Further

examples includeabeforeandafterstudyexaminingthe introductionofacombined

emergency obstetric care andmonitoring and evaluation program (the Advances in

LabourandRiskManagementInternationalProgram)inKenya,whichresultedinclear

improvementsinbothmaternalandperinataloutcomes(446).

101

Itisclearthatasystemthatisbothsupportiveoftherequirementsforauditandhas

the capacity to implement changes once issues are identified is central to improved

outcomes.AnauditofvillagebasedregistriesinruralIndonesiashowedastillbirthrate

fourtimesthereportedratefromthenationalvitalstatisticsprogram(447)andsitesin

Pakistan in the Maternal Newborn Health Registry study had worsening mortality

statisticswiththeintroductionofaudit,bothapparentlysecondarytopoorqualityof

care(448).Thereismoretotheprocessofauditthansimpleregistrationandpresumed

inherent clinical response to adverse outcomes, including having the resources to

executeanyclinicalresponse.Theremustalsobeinternalinfrastructuretorunaudit,

clinicalandmanagerialstaffsupport,afeelingofresponsibilityandownershipamongst

thestaff,andtheprocessofauditmustnotbeseenaspunitive(449,450).

QoC in low resource settings is likely to contribute to poor perinatal outcomes and

QoC audit in these settings repeatedly identifies poor care around the time of

childbirthasapreventablefactorinsignificantnumbersofperinataldeaths(451).

However,itisnotentirelyclearthatQoCauditbyitselfimproveslongtermoutcomes;

is it audit alone or is there a deeper process thatmust be undertaken for it to be

effective? If thisprocess is feedback (passiveoractive),which is identifiedascritical

foraudit,doesit infacthaveasustainablelongtermeffectonmortality?(423)Other

quality improvement programs aimed at improvement care around the time of

childbirthhavebeen shown to reduceperinatal andmaternal deaths(452).However

102

the inherent flaws in theseprogramsare the lackofeffectivecontroland the issues

with reproducibility. PPIP is frequently cited in the literature as the foundation on

which improvement in perinatal care and perinatal mortality has been made in

individual units and larger districts(131, 135, 453). Belizanet al identified fourmain

drivers in the successful implementation of PPIP; drivers and teams, outreach and

supervision,perinatalreviewmeetings,andcommunicationandnetworking(454).

QUALITYOFCAREAUDITANDNEARMISSCASES

Withsubstandard intrapartumand immediatenewborncarebeing linkedtonearlya

million neonatal deaths annually(4), it is evident that there is a significant group of

survivingneonatesthatcomprisethenearmissgroup.Theconceptofusingnearmiss

cases to assess quality of care is well described in maternal and newborn health

literature(427). Defining and identifying neonatal near miss allows identification of

deficiencies inobstetricandneonatal care toguide improvement in clinicalpractice.

Moreover,thecliniciancompliancewiththismaybe improvedbythe“positiveentry

point(455)” to the process; that is the survival of the babymay encourage in depth

analysis in thequalityofcareauditprocesswhentheoutcomeofanyconcernswith

careisalreadyknowntobepositive.Thetermneonatalnearmissisbeingincreasingly

used in the literature in an attempt to identify those babies who would have

potentially died from a severe obstetric complication and thus may represent a

consequenceofdeficienciesinperinatalcare(456-458).

103

MATERNALHIVINFECTION

MaternalHIVinfectionaffectsmorethan17millionwomenovertheageof15around

theworld,withtheburdenoftheseconcentratedinlowandmiddle-incomecountries.

There is a clear crossover between women with HIV and women with adverse

pregnancy outcomes, from both HIV related and non-related causes(459). It was

important to consider the role of maternal HIV infection in any study investigating

labouroutcomes,giventhatinSouthAfrica30%ofthepopulationisinfectedwithHIV

andmorethan600,000birthsoccurtowomenwithHIVeveryyear(460).MaternalHIV

infectioninpregnancy,particularlyinwomennotreceivingantiretroviralmedication,is

associatedwithadverseperinataloutcomes,includingpretermbirthandstillbirth(461).

ThereisevidenceinSouthAfricathatwhilepregnancyisoftenprotectiveagainstnew

acquisitionofHIV, youngerwomenwithhigherparitymaybeatgreater riskofnew

infections in pregnancy(462). Thereforewe need evidence of the potential effect of

HIVonintrapartumcare.This isparticularly importantgivensomeretrospectivedata

fromthesameregionasthelocationofthiscurrentresearchsuggestingwomenwith

HIVinfectionhaveahigherrateofbirthasphyxiarelatedperinatalmortalitythantheir

HIV negative counterparts(463). The small numbers and retrospective nature of the

datameansone should interpret thiswith caution, however the concept formedan

importantconsiderationinourresearch.

104

CONCLUSION

Perinatalmortalityinlowandmiddleincomecountriesisacomplexandmultifaceted

entity. Intrapartum hypoxia contributes significantly to perinatal mortality in these

settings.Inordertounderstandthis,thephysiologyofthefetus,theunderlyingcauses

and clinical consequences of hypoxia and the capacity to measure and respond to

hypoxia must be considered. The thesis interrogates these issues and attempts to

addressthemwithaspecificfocusonSouthAfrica.

105

REFERENCES

1. Allanson ER, Muller M, Pattinson RC. Causes of perinatal mortality andassociatedmaternalcomplicationsinaSouthAfricanprovince:challengesinpredictingpooroutcomes.BMCPregnancyChildbirth.2015;15(1):37.2. Rei M, Ayres-de-Campos D, Bernardes J. Neurological damage arising fromintrapartumhypoxia/acidosis.BestPractResClinObstetGynaecol.2016;30:79-86.3. BuchmannEJ, StonesW,ThomasN.Preventingdeaths fromcomplicationsoflabouranddelivery.BestPractResClinObstetGynaecol.2016;36:103-15.4. Bhutta ZA, Das JK, Bahl R, Lawn JE, Salam RA, Paul VK, et al. Can availableinterventionsendpreventabledeathsinmothers,newbornbabies,andstillbirths,andatwhatcost?Lancet.2014;384(9940):347-70.5. Tunçalp Ӧ,WereW,MacLennan C, Oladapo O, Gülmezoglu A, Bahl R, et al.Quality of care for pregnant women and newborns—the WHO vision. BJOG.2015;122(8):1045-9.6. BuchmannE, PattinsonR,NyathikaziN. Intrapartum-relatedbirth asphyxia inSouth Africa lessons From the first national perinatal care survey. S Afr Med J.2002;92(11):897-901.7. Low JA, Pickersgill H, Killen H, Derrick EJ. The prediction and prevention ofintrapartumfetalasphyxiaintermpregnancies.AmJObstetGynecol.2001;184(4):724-30.8. PinasA,ChandraharanE.Continuouscardiotocographyduringlabour:Analysis,classificationandmanagement.BestPractResClinObstetGynaecol.2016;30:33-47.9. Prior T, Kumar S. Expert review--identification of intra-partum fetalcompromise.EurJObstetGynecolReprodBiol.2015;190:1-6.10. Jorgensen JS, Weber T. Fetal scalp blood sampling in labor--a review. ActaObstetGynecolScand.2014;93(6):548-55.11. EastCE,LeaderLR,SheehanP,HenshallNE,ColditzPB,LauR.Intrapartumfetalscalp lactate sampling for fetal assessment in thepresenceof anon-reassuring fetalheartratetrace.CochraneDatabaseSystRev.2015(5):Cd006174.12. LassiZS,MiddletonPF,BhuttaZA,CrowtherC.Strategiesforimprovinghealthcareseekingformaternalandnewbornillnessesinlow-andmiddle-incomecountries:asystematicreviewandmeta-analysis.GlobHealthAction.2016;9(1):31408.13. UNInter-agencyGroupforChildMortalityEstimation(IGME).Levelsandtrendsinchildmortality:Report2013.NewYork:UNICEF;2013.14. Akseer N, Lawn JE, KeenanW, Konstantopoulos A, Cooper P, Ismail Z, et al.Endingpreventablenewborndeaths inageneration. IntJGynaecolObstet.2015;131Suppl1:S43-8.15. Froen JF, Friberg IK, Lawn JE, Bhutta ZA, Pattinson RC, Allanson ER, et al.Stillbirths:progressandunfinishedbusiness.Lancet.2016;387(10018):574-86.16. Blencowe H, Cousens S, Jassir FB, Say L, Chou D,Mathers C, et al. National,regional,andworldwideestimatesofstillbirthratesin2015,withtrendsfrom2000:asystematicanalysis.LancetGlobHealth.2016;4(2):e98-e108.17. Alkema L, Chou D, Hogan D, Zhang S, Moller AB, Gemmill A, et al. Global,regional,andnationallevelsandtrendsinmaternalmortalitybetween1990and2015,

106

with scenario-based projections to 2030: a systematic analysis by the UNMaternalMortalityEstimationInter-AgencyGroup.Lancet.2016;387(10017):462-74.18. RoosN, vonXylander SR.Whydomaternal andnewborndeaths continue tooccur?BestPractResClinObstetGynaecol.2016;36:30-44.19. Saleem S, McClure EM, Goudar SS, Patel A, Esamai F, Garces A, et al. Aprospective studyofmaternal, fetal andneonataldeaths in low-andmiddle-incomecountries.BullWorldHealthOrgan.2014;92(8):605-12.20. WorldHealthOrganization.Neonatalandperinatalmortality:country,regionalandglobalestimates.2006.21. AllansonE,TunçalpÖ,GardosiJ,PattinsonR,FrancisA,VogelJ,etal.TheWHOapplicationof ICD‐10todeathsduringtheperinatalperiod(ICD‐PM):resultsfrompilotdatabasetestinginSouthAfricaandUnitedKingdom.BJOG.2016;123(12):2019-28.22. World Health Organization. ICD-10: international statistical classification ofdiseasesandrelatedhealthproblems:tenthrevision.2004.23. PattinsonR,Rhoda,N.Savingbabies2012-2013:NinthreportonperinatalcareinSouthAfrica.TshepesaPress,Pretoria,SouthAfrica;2014.24. Australian Institute of Health andWelfare. National PerinatalMortality DataCollection (NPMDC) 2019 [updated 16/04/2019 v9.0. Available from:https://www.aihw.gov.au/about-our-data/our-data-collections/national-perinatal-mortality-data-collection-npmdc.25. Blencowe H, Calvert Ph DC, Lawn JE, Cousens S, Campbell OM. Measuringmaternal,foetalandneonatalmortality:Challengesandsolutions.BestPractResClinObstetGynaecol.2016;36:14-29.26. GordonA, Raynes-GreenowC,McGeechanK,Morris J, JefferyH. Risk factorsfor antepartum stillbirth and the influence of maternal age in New South WalesAustralia:apopulationbasedstudy.BMCPregnancyChildbirth.2013;13(1):12.27. Allanson ER, Muller M, Pattinson RC. Causes of perinatal mortality andassociatedmaternalcomplicationsinaSouthAfricanprovince:challengesinpredictingpooroutcomes.BMCPregnancyChildbirth.2015;15:37.28. Vogel J, Souza J, Mori R, Morisaki N, Lumbiganon P, Laopaiboon M, et al.Maternal complications and perinatal mortality: findings of the World HealthOrganization Multicountry Survey on Maternal and Newborn Health. BJOG.2014;121:76-88.29. Ariff S, LeeAC, Lawn J, Bhutta ZA.Global Burden, Epidemiologic Trends, andPrevention of Intrapartum-Related Deaths in Low-Resource Settings. Clin Perinatol.2016;43(3):593-608.30. Kamath-RayneBD,Griffin JB,MoranK, JonesB,DownsA,McClureEM,etal.ResuscitationandObstetricalCaretoReduceIntrapartum-RelatedNeonatalDeaths:AMANDATEStudy.MaternChildHealthJ.2015;19(8):1853-63.31. AkterT,DawsonA,SibbrittD.Whatimpactdoesantenatalandpostnatalcarehave on neonatal deaths in low-and lower-middle-income countries? Evidence fromBangladesh.HealthCareWomenInt.2017;38(8):848-60.32. Ntambue AM,Malonga FK, Dramaix-WilmetM, Ngatu RN, Donnen P. Betterthan nothing?maternal, newborn, and child health services and perinatalmortality,Lubumbashi, democratic republic of the Congo: a cohort study. BMC PregnancyChildbirth.2016;16:89.

107

33. Chaudhury S, Arlington L, Brenan S, Kairuki AK,Meda AR, Isangula KG, et al.Cost analysisof large-scale implementationof the 'HelpingBabiesBreathe'newbornresuscitation-trainingprograminTanzania.BMCHealthServRes.2016;16(1):681.34. YegoF,D'EsteC,BylesJ,NyongesaP,WilliamsJS.Acase-controlstudyofriskfactors for fetal and early neonatal deaths in a tertiary hospital in Kenya. BMCPregnancyChildbirth.2014;14:389.35. SankarMJ,NatarajanCK,DasRR,AgarwalR,ChandrasekaranA,PaulVK.Whendonewbornsdie?Asystematicreviewoftimingofoverallandcause-specificneonataldeathsindevelopingcountries.JPerinatol.2016;36Suppl1:S1-s11.36. LehtonenL,GimenoA,Parra-LlorcaA,VentoM,editors.Earlyneonataldeath:Achallengeworldwide.SeminarsinFetalandNeonatalMedicine;2017:Elsevier.37. Engmann CM, Hodgson A, Aborigo R, Adongo PL,Moyer CA. Addressing thecontinuum of maternal and newborn care in Ghana: implications for policy andpractice.HealthPolicyPlan.2016;31(10):1355-63.38. SinhaS,AggarwalAR,OsmondC,FallCH,BhargavaSK,SachdevHS.MaternalAgeatChildbirthandPerinatalandUnderfiveMortalityinaProspectiveBirthCohortfromDelhi.IndianPediatr.2016;53(10):871-7.39. SemrauKEA,HerlihyJ,GroganC,MusokotwaneK,Yeboah-AntwiK,MbeweR,et al. Effectiveness of 4% chlorhexidine umbilical cord care on neonatalmortality inSouthern Province, Zambia (ZamCAT): a cluster-randomised controlled trial. LancetGlobHealth.2016;4(11):e827-e36.40. Sazawal S, Dhingra U, Ali SM, Dutta A, Deb S, Ame SM, et al. Efficacy ofchlorhexidineapplicationtoumbilicalcordonneonatalmortalityinPemba,Tanzania:acommunity-based randomised controlled trial. Lancet Glob Health. 2016;4(11):e837-e44.41. Engmann C, Garces A, Jehan I, Ditekemena J, Phiri M, Mazariegos M, et al.Causes of community stillbirths and early neonatal deaths in low-income countriesusingverbalautopsy:anInternational,MulticenterStudy.JPerinatol.2012;32(8):585.42. Bailey PE, AndualemW, BrunM, Freedman L, Gbangbade S, KanteM, et al.Institutional maternal and perinatal deaths: a review of 40 low andmiddle incomecountries.BMCPregnancyChildbirth.2017;17(1):295.43. Woods R. Long-term trends in fetal mortality: implications for developingcountries.BullWorldHealthOrgan.2008;86:460-6.44. TuiletaE.AmericanSamoaVitalStatisticsReport2010-2012,AmericanSamoaGovernment.2015.45. Moore JE, Uka S, Vogel JP, Timmings C, Rashid S, Gülmezoglu AM, et al.Navigating barriers: two-year follow up on recommendations to improve the use ofmaternalhealthguidelinesinKosovo.BMCPublicHealth.2016;16(1):987.46. KalterHD,KhazenRR,BarghouthiM,OdehM.Prospectivecommunity‐basedclustercensusandcase‐controlstudyofstillbirthsandneonataldeaths intheWestBankandGazaStrip.PaediatrPerinatEpidemiol.2008;22(4):321-33.47. WorldBankGroup. TheWorldBankData [Internet] 2018 [5thAugust, 2018].Availablefrom:https://data.worldbank.org/income-level/low-and-middle-income.48. AnastasiE,EkanemE,HillO,AdebayoOluwakemiA,AbayomiO,BernasconiA.Unmasking inequalities: Sub-national maternal and child mortality data from twourbanslumsinLagos,Nigeriatellsthestory.PLoSOne.2017;12(5):e0177190.

108

49. AnggondowatiT,El-MohandesAA,QomariyahSN,KielyM,RyonJJ,GipsonRF,etal.MaternalcharacteristicsandobstetricalcomplicationsimpactneonataloutcomesinIndonesia:aprospectivestudy.BMCPregnancyChildbirth.2017;17(1):100.50. Banda M, Kazembe L, Lewycka S, King C, Phiri T, Masache G, et al. Spatialmodelling of perinatal mortality inMchinji, Malawi. Spat Spatiotemporal Epidemiol.2016;16:50-8.51. Garces A, McClure EM, Hambidge K, Krebs NF, Figueroa L, Aguilar M, et al.Trends inperinataldeathsfrom2010to2013 intheGuatemalanWesternHighlands.ReprodHealth.2015;12Suppl2:S14.52. Memon ZA, Khan GN, Soofi SB, Baig IY, Bhutta ZA. Impact of a community-based perinatal and newborn preventive care package on perinatal and neonatalmortality in a remote mountainous district in Northern Pakistan. BMC PregnancyChildbirth.2015;15:106.53. Groppi L, SomiglianaE, PisaniV, IkaM,Mabor JL,AkecHN,et al.Ahospital-centeredapproachtoimproveemergencyobstetriccareinSouthSudan.IntJGynaecolObstet.2015;128(1):58-61.54. Ugwa EA, Ashimi A. An assessment of stillbirths in a tertiary hospital innorthernNigeria.JMaternFetalNeonatalMed.2015;28(13):1585-8.55. Nascimento RCS, Costa M, Braga JU, Natividade MSD. Spatial patterns ofpreventable perinatal mortality in Salvador, Bahia, Brazil. Rev Saude Publica.2017;51:73.56. Okonofua F, Randawa A, Ogu R, Agholor K, Okike O, Abdus-Salam RA, et al.Views of senior health personnel about quality of emergency obstetric care: AqualitativestudyinNigeria.PLoSOne.2017;12(3):e0173414.57. LawnJE,BlencoweH,KinneyMV,BianchiF,GrahamWJ.Evidencetoinformthefutureformaternalandnewbornhealth.Bestpractice&researchClinicalobstetrics&gynaecology.2016;36:169-83.58. BrowneJL,VissersKM,AntwiE,SrofenyohEK,VanderLindenEL,AgyepongIA,etal.Perinataloutcomesafterhypertensivedisordersinpregnancyinalowresourcesetting.TropMedIntHealth.2015;20(12):1778-86.59. RahmanMM,AbeSK,RahmanMS,KandaM,NaritaS,BilanoV,etal.Maternalanemia and risk of adverse birth and health outcomes in low- and middle-incomecountries:systematicreviewandmeta-analysis.AmJClinNutr.2016;103(2):495-504.60. Marete I, Tenge C, Pasha O, Goudar S, Chomba E, Patel A, et al. PerinatalOutcomes of Multiple Gestation Pregnancies in Kenya, Zambia, Pakistan, India,GuatemalaandArgentina:AGlobalNetworkStudy.AmJPerinatol.2014;31(2):125.61. FajoluI,EzeakaV,AdeniyiO,IrohaE,Egri-OkwajiM.PrevalenceandoutcomeofhigherordermultiplepregnanciesinLagos,Nigeria.TheJournalofMaternal-Fetal&NeonatalMedicine.2013;26(13):1342-5.62. Olamijulo J, Olaleye O. Perinatal mortality in Lagos University TeachingHospital:afiveyearreview.NigQJHospMed.2011;21(4):255-61.63. Koulimaya-Gombet CE, Diouf AA, Diallo M, Dia A, Sene C, Moreau JC, et al.[Pregnancy and delivery in patients with a personal history of cesarean section inDakar: epidemiological, clinical, therapeutic and prognostic aspects].Pan AfrMed J.2017;27:135.

109

64. Kalisa R, Rulisa S, vanRoosmalen J, van denAkker T.Maternal andperinataloutcomeafterpreviouscaesareansectioninruralRwanda.BMCPregnancyChildbirth.2017;17(1):272.65. Boatin AA, Adu-Bonsaffoh K, Wylie BJ, Obed SA. Evaluating Facility-BasedDecision-Making in Women with a Prior Cesarean Delivery and Association withMaternalandPerinatalOutcomes.MaternChildHealthJ.2017;21(9):1845-52.66. Yirgu R, Molla M, Sibley L, Gebremariam A. Perinatal Mortality Magnitude,Determinants and Causes in West Gojam: Population-Based Nested Case-ControlStudy.PLoSOne.2016;11(7):e0159390.67. Macheku GS, Philemon RN, Oneko O, Mlay PS, Masenga G, Obure J, et al.Frequency,riskfactorsandfeto-maternaloutcomesofabruptioplacentaeinNorthernTanzania: a registry-based retrospective cohort study. BMC Pregnancy Childbirth.2015;15:242.68. KaboreC,ChailletN,KouandaS,BujoldE,TraoreM,DumontA.Maternalandperinataloutcomesassociatedwithatrialoflabourafterpreviouscaesareansectioninsub-Saharancountries.BJOG.2016;123(13):2147-55.69. PerveenF,TayyabS,ZuberiBF.Risk factors forperinataldeaths inPakistan. JObstetGynaecolRes.2011;37(10):1359-64.70. Ekure E, Ezeaka V, Iroha E, Egri-Okwaji M. Prospective audit of perinatalmortalityamonginbornbabiesinatertiaryhealthcenterinLagos,Nigeria.NigerJClinPract.2011;14(1).71. Harrison MS, Goldenberg RL. Global burden of prematurity. Semin FetalNeonatalMed.2016;21(2):74-9.72. ChinkhumbaJ,DeAllegriM,MuulaAS,RobberstadB.Maternalandperinatalmortality by place of delivery in sub-Saharan Africa: a meta-analysis of population-basedcohortstudies.BMCPublicHealth.2014;14:1014.73. Ali AAA, Elgessim ME, Taha E, Adam GK. Factors associated with perinatalmortality in Kassala, Eastern Sudan: a community-based study 2010–2011. J TropPediatr.2013;60(1):79-82.74. Turab A, Ariff S, Habib MA, Ahmed I, Hussain M, Rashid A, et al. Improvedaccessibilityofemergencyobstetricsandnewborncare(EmONC)servicesformaternaland newborn health: a community based project. BMC Pregnancy Childbirth.2013;13(1):136.75. More NS, Bapat U, Das S, Alcock G, Patil S, Porel M, et al. Communitymobilization in Mumbai slums to improve perinatal care and outcomes: a clusterrandomizedcontrolledtrial.PLoSMed.2012;9(7):e1001257.76. WilsonA,GallosID,PlanaN,LissauerD,KhanKS,ZamoraJ,etal.Effectivenessof strategies incorporating training and support of traditional birth attendants onperinatalandmaternalmortality:meta-analysis.BMJ.2011;343:d7102.77. Matendo R, Engmann C, Ditekemena J, Gado J, Tshefu A, Kinoshita R, et al.Reduced perinatal mortality following enhanced training of birth attendants in theDemocraticRepublicofCongo:atime-dependenteffect.BMCMed.2011;9(1):93.78. GoudarSS,DhadedSM,McClureEM,DermanRJ,PatilVD,MahantshettiNS,etal. ENC training reduces perinatal mortality in Karnataka, India. The Journal ofMaternal-Fetal&NeonatalMedicine.2012;25(6):568-74.

110

79. Rani S, Chawla D, Huria A, Jain S. Risk factors for perinatalmortality due toasphyxia among emergency obstetric referrals in a tertiary hospital. Indian Pediatr.2012;49(3):191-4.80. YakoobMY, AliMA, AliMU, Imdad A, Lawn JE, Van Den BroekN, et al. Theeffectofprovidingskilledbirthattendanceandemergencyobstetriccareinpreventingstillbirths.BMCPublicHealth.2011;11(3):S7.81. EllisM, Azad K, Banerjee B, Shaha SK, Prost A, Rego AR, et al. Intrapartum-relatedstillbirthsandneonataldeathsinruralBangladesh:aprospective,community-basedcohortstudy.Pediatrics.2011:peds.2010-0842.82. Anastasi E, Ekanem E, Hill O, Oluwakemi AA, Abayomi O, Bernasconi A.Unmasking inequalities: Sub-national maternal and child mortality data from twourbanslumsinLagos,Nigeriatellsthestory.PLoSOne.2017;12(5):e0177190.83. EnuamehYA,OkawaS,AsanteKP,KikuchiK,MahamaE,AnsahE,etal.FactorsInfluencing Health Facility Delivery in Predominantly Rural Communities across theThree Ecological Zones in Ghana: A Cross-Sectional Study. PLoS One.2016;11(3):e0152235.84. BelladRM,BangA,CarloWA,McClureEM,MelethS,GocoN,etal.Apre-poststudyofamulti-countryscaleupofresuscitationtrainingof facilitybirthattendants:does Helping Babies Breathe training save lives? BMC Pregnancy Childbirth.2016;16(1):222.85. BucherS,KonanaO,LiechtyE,GarcesA,GisoreP,MareteI,etal.Self-reportedpracticesamongtraditionalbirthattendantssurveyedinwesternKenya:adescriptivestudy.BMCPregnancyChildbirth.2016;16(1):219.86. Berhan Y, Berhan A. Skilled health personnel attended delivery as a proxyindicatorformaternalandperinatalmortality:asystematicreview.EthiopJHealthSci.2014;24Suppl:69-80.87. NkwoPO,LawaniLO,EzugwuEC,IyokeCA,UbesieAC,OnohRC.Correlatesofpoorperinataloutcomesinnon-hospitalbirths inthecontextofweakhealthsystem:theNigerianexperience.BMCPregnancyChildbirth.2014;14:341.88. Pervin J, Moran A, Rahman M, Razzaque A, Sibley L, Streatfield PK, et al.Associationofantenatalcarewithfacilitydeliveryandperinatalsurvival–apopulation-basedstudyinBangladesh.BMCPregnancyChildbirth.2012;12(1):111.89. SinghSK,KaurR,GuptaM,KumarR.ImpactofnationalruralhealthmissiononperinatalmortalityinruralIndia.IndianPediatr.2012;49(2):136-8.90. KhanamR,BaquiAH,SyedMIM,HarrisonM,BegumN,QuaiyumA,etal.Canfacility delivery reduce the risk of intrapartum complications-related perinatalmortality?Findingsfromacohortstudy.Journalofglobalhealth.2018;8(1).91. FengXL,GuoS,HipgraveD,ZhuJ,ZhangL,SongL,etal.China'sfacility-basedbirth strategy and neonatal mortality: a population-based epidemiological study.Lancet.2011;378(9801):1493-500.92. FengXL,XuL,GuoY,RonsmansC.SocioeconomicinequalitiesinhospitalbirthsinChinabetween1988and2008.BullWorldHealthOrgan.2011;89(6):432-41.93. Liang J,MuY, Li X, TangW,WangY, Liu Z, et al. Relaxationof theone childpolicy and trends in caesarean section rates and birth outcomes in China between2012and2016:observationalstudyofnearlysevenmillionhealthfacilitybirths.BMJ.2018;360:k817.

111

94. Lori JR, Stalls S, Rominski SD. Strengthening Midwifery to End PreventableMaternal,Child,andNewbornDeaths.JMidwiferyWomensHealth.2015;60(4):343-7.95. KrukME,LeslieHH,VerguetS,MbarukuGM,AdanuRMK,LangerA.QualityofbasicmaternalcarefunctionsinhealthfacilitiesoffiveAfricancountries:ananalysisofnationalhealthsystemsurveys.LancetGlobHealth.2016;4(11):e845-e55.96. SemrauKE,HirschhornLR,MarxDelaneyM,SinghVP,SaurastriR,SharmaN,etal.OutcomesofaCoaching-BasedWHOSafeChildbirthChecklistProgramin India.NEnglJMed.2017;377(24):2313-24.97. GoudarSS,GocoN,SomannavarMS,VernekarSS,MallapurAA,MooreJL,etal.Institutional deliveries and perinatal and neonatalmortality in Southern andCentralIndia.ReprodHealth.2015;12Suppl2:S13.98. Ecevit A, Oguz SS, Tarcan A, Yazici C, Dilmen U. The changing pattern ofperinatal mortality and causes of death in central Anatolian region of Turkey. TheJournalofMaternal-Fetal&NeonatalMedicine.2012;25(9):1738-41.99. Kültürsay N, Aşkar N, Terek D, Yeniel AÖ, Köroğlu ÖA, Yalaz M, et al. TheChangeofPerinatalMortalityOverThreeDecadesinaReferenceCentreintheAegeanRegion:NeonatalMortality has decreased but FoetalMortality RemainsUnchanged.Balkanmedicaljournal.2017;34(6).100. Goldenberg RL, McClure EM, Kamath BD. Intrapartum perinatal mortality.IndianPediatr.2012;49(3):187-90.101. Wall SN, Lee AC, CarloW, Goldenberg R, Niermeyer S, Darmstadt GL, et al.,editors. Reducing intrapartum-related neonatal deaths in low-and middle-incomecountries—whatworks?SeminPerinatol;2010:Elsevier.102. RonsmansC,ChowdhuryME,KoblinskyM,AhmedA.Care seekingat timeofchildbirth, and maternal and perinatal mortality in Matlab, Bangladesh. Bull WorldHealthOrgan.2010;88:289-96.103. AhmedM,WonY.Cross-NationalSystematicReviewofNeonatalMortalityandPostnatal Newborn Care: Special Focus on Pakistan. Int J Environ Res Public Health.2017;14(12):1442.104. NascimentoRdCdS,CostaMdCN,BragaJU,NatividadeMSd.SpatialpatternsofpreventableperinatalmortalityinSalvador,Bahia,Brazil.RevSaudePublica.2017;51.105. SunL,YueH,SunB,HanL,QiM,TianZ,etal.Estimationofbirthpopulation-basedperinatal-neonatalmortalityandpretermrateinChinafromaregionalsurveyin2010.TheJournalofMaternal-Fetal&NeonatalMedicine.2013;26(16):1641-8.106. EngmannC,JehanI,DitekemenaJ,GarcesA,PhiriM,MazariegosM,etal.Analternativestrategyforperinatalverbalautopsycoding:singleversusmultiplecoders.TropMedIntHealth.2011;16(1):18-29.107. SuleimanB,MokuoluO,AdesiyunO,AdeniyiA.PatternofperinatalmortalityinbabiesdeliveredattheUniversityofIlorinTeachingHospital,Ilorin,Nigeria.WestAfrJMed.2012;31(2):102-8.108. Engmann C, Walega P, Aborigo RA, Adongo P, Moyer CA, Lavasani L, et al.StillbirthsandearlyneonatalmortalityinruralNorthernGhana.TropMedIntHealth.2012;17(3):272-82.109. Mgaya AH, Litorp H, Kidanto HL, Nyström L, Essén B. Criteria-based audit toimprove quality of care of foetal distress: standardising obstetric care at a nationalreferral hospital in a low resource setting, Tanzania. BMC Pregnancy Childbirth.2016;16(1):343.

112

110. Esike COU, Chukwuemeka UI, Anozie OB, Eze JN, Aluka OC, Twomey DE.EclampsiainruralNigeria:Theunmitigatingcatastrophe.AnnAfrMed.2017;16(4):175.111. Niang MM. [Unmet obstetrical needs for major obstetric interventions inDakar].MedSanteTrop.2015;25(3):276-9.112. BerhanY,BerhanA.Ameta-analysisofselectedmaternalandfetalfactorsforperinatalmortality.EthiopJHealthSci.2014;24Suppl:55-68.113. MouraPMSS,MaestáI,RugoloLMSS,AngulskiLFRB,CaldeiraAP,PeraçoliJC,etal.Riskfactorsforperinataldeathintwodifferentlevelsofcare:acase–controlstudy.Reproductivehealth.2014;11(1):11.114. Oliveira LC, CostaAARd. Fetal and neonatal deaths among cases ofmaternalnearmiss.RevAssocMedBras.2013;59(5):487-94.115. NankabirwaV,TumwineJK,TylleskärT,NankundaJ,SommerfeltH,ConsortiumPER. Perinatal mortality in eastern Uganda: a community based prospective cohortstudy.PLoSOne.2011;6(5):e19674.116. Bai J,WongFW,BaumanA,MohsinM.Parityandpregnancyoutcomes.AmJObstetGynecol.2002;186(2):274-8.117. Munan R, Kakudji Y, Nsambi J, Mukuku O, Maleya A, Kinenkinda X, et al.[Childbirth among primiparous women in Lubumbashi: maternal and perinatalprognosis].PanAfrMedJ.2017;28:77.118. Jacinto E, Aquino EM, Mota EL. [Perinatal mortality in the municipality ofSalvador, Northeastern Brazil: evolution from 2000 to 2009]. Rev Saude Publica.2013;47(5):846-53.119. MohammadiS,EssenB,FallahianM,TaheripanahR,SalehGargariS,KallestalC.Maternal near-miss at university hospitals with cesarean overuse: an incident case-controlstudy.ActaObstetGynecolScand.2016;95(7):777-86.120. WilsonA,LissauerD,ThangaratinamS,KhanKS,MacArthurC,CoomarasamyA.A comparison of clinical officers with medical doctors on outcomes of caesareansection in the developing world: meta-analysis of controlled studies. BMJ.2011;342:d2600.121. Chongsuvivatwong V, Bachtiar H, Chowdhury ME, Fernando S, Suwanrath C,Kor‐anantakul O, et al. Maternal and fetal mortality and complications associatedwith cesarean sectiondeliveries in teachinghospitals inAsia. JObstetGynaecolRes.2010;36(1):45-51.122. MaswimeS,BuchmannE.Causesandavoidablefactorsinmaternaldeathduetocesarean‐relatedhemorrhageinSouthAfrica.InternationalJournalofGynecology&Obstetrics.2016;134(3):320-3.123. Lumbiganon P, LaopaiboonM,Gülmezoglu AM, Souza JP, Taneepanichskul S,RuyanP,etal.Methodofdeliveryandpregnancyoutcomes inAsia: theWHOglobalsurveyonmaternalandperinatalhealth2007–08.TheLancet.2010;375(9713):490-9.124. RudgeMV,MaestáI,MouraPM,RudgeCV,MorceliG,CostaRA,etal.Thesafemotherhood referral system to reduce cesarean sections and perinatal mortality-across-sectionalstudy[1995-2006].Reproductivehealth.2011;8(1):34.125. GaladanciH,KünzelW,ShittuO,ZinserR,GruhlM,AdamsS.ObstetricqualityassurancetoreducematernalandfetalmortalityinKanoandKadunaStatehospitalsinNigeria.InternationalJournalofGynecology&Obstetrics.2011;114(1):23-8.

113

126. Chu K, Maine R, Trelles M. Cesarean section surgical site infections in sub-Saharan Africa: amulti-country study fromMedecins Sans Frontieres.World J Surg.2015;39(2):350-5.127. KinenkindaX,MukukuO,ChengeF,KakudjiP,BanzuluP,KakomaJB,etal.[Riskfactors for maternal and perinatal mortality among women undergoing cesareansectioninLubumbashi,DemocraticRepublicofCongoII].PanAfrMedJ.2017;26:208.128. OppongSA,TuuliMG,Seffah JD,AdanuRM. Is therea safe limitofdelay foremergencycaesareansectioninGhana?Resultsofanalysisofearlyperinataloutcome.GhanaMedJ.2014;48(1):24-30.129. Rhoda N, Greenfield D, Muller M, Prinsloo R, Pattinson R, Kauchali S, et al.Experiences with perinatal death reviews in South Africa—the Perinatal ProblemIdentificationProgramme: scalingup fromprogrammetoprovince tocountry.BJOG.2014;121(s4):160-6.130. MichalowJ,CholaL,McGeeS,TugendhaftA,PattinsonR,KerberK,etal.Triplereturn on investment: the cost and impact of 13 interventions that could preventstillbirths and save the lives ofmothers and babies in South Africa.BMC PregnancyChildbirth.2015;15(1):39.131. Baleta A. South Africa takes steps to reduce perinatal mortality. The Lancet.2011;377(9774):1303-4.132. NtuliST,MalanguN.An investigationof thestillbirthsata tertiaryhospital inLimpopoProvinceofSouthAfrica.Globaljournalofhealthscience.2012;4(6):141.133. PattinsonR.Whybabiesdie-aperinatalcaresurveyofSouthAfrica,2000-2002.SAfrMedJ.2003;93(6):445-50.134. TalipQ,TheronG,SteynW,HallD.TotalperinatallyrelatedlossesatTygerbergHospital–acomparisonbetween1986,1993and2006.SAfrMedJ.2010;100(4).135. Gaunt CB. Are we winning? Improving perinatal outcomes at a deeply ruraldistricthospitalinSouthAfrica.SAMJ:SouthAfricanMedicalJournal.2010;100(2):101-4.136. Allanson ER, Pattinson RC. Quality-of-care audits and perinatal mortality inSouthAfrica.BullWorldHealthOrgan.2015;93(6):424-8.137. McGeeS-A,CholaL,TugendhaftA,MubaiwaV,MoranN,McKerrowN,etal.Strategic planning for saving the lives of mothers, newborns and children andpreventingstillbirthsinKwaZulu-NatalprovinceSouthAfrica:modellingusingtheLivesSavedTool(LiST).BMCPublicHealth.2015;16(1):49.138. BallotDE,ChirwaT,RamdinT,ChirwaL,MareI,DaviesVA,etal.Comparisonofmorbidity and mortality of very low birth weight infants in a Central Hospital inJohannesburgbetween2006/2007and2013.BMCPediatr.2015;15(1):20.139. BothmaM,BuchmannEJ.Astudyoffreshstillbirthsweighing2500gormoreatthree academic hospitals in South Africa. International Journal of Gynecology &Obstetrics.2016;134(2):186-9.140. Bateman C. Stillbirths: an invisible earthquake. SAMJ: South African MedicalJournal.2011;101(6):364-6.141. Buchmann E, Pattinson R. Babies who die from labour-related intrapartumhypoxia: a confidential enquiry in South African public hospitals. Trop Doct.2006;36(1):8-10.142. Hofmeyr GJ, Mancotywa T, Silwana-Kwadjo N, Mgudlwa B, Lawrie TA,Gülmezoglu AM. Audit of a new model of birth care for women with low risk

114

pregnancies in SouthAfrica: the primary care onsitemidwife-led birth unit (OMBU).BMCPregnancyChildbirth.2014;14(1):417.143. PattinsonR.Reduction inperinatalmortality feasiblewithout incurringmajorcosts.SAfrJObstetGynaecol.2003;9(3):54-5.144. Ebrahim S, Daponte A, Guidozzi F. The impact of free antenatal care onperinatalmortality. InternationalJournalofGynecology&Obstetrics.2000;71(3):205-7.145. Woods D. Appropriate technology and education for improved intrapartumcareinunderresourcedcountries.SAfrJObstetGynaecol.2009;15(3).146. Velaphi S,PattinsonR.Avoidable factorsandcausesofneonataldeaths fromperinatal asphyxia-hypoxia in South Africa: national perinatal survey. Ann TropPaediatr.2007;27(2):99-106.147. Pattinson R, Woods D, Greenfield D, Velaphi S. Improving survival rates ofnewborninfantsinSouthAfrica.Reproductivehealth.2005;2(1):4.148. PatrickM.PerinatalmortalityatFrontierHospital,Queenstown-a6-yearauditusingthePerinatalProblemIdentificationProgramme(PPIP).SAfrJObstetGynaecol.2007;13(1):10-4.149. NgocNTN,MerialdiM,Abdel-AleemH,CarroliG,PurwarM,ZavaletaN,etal.Causes of stillbirths and early neonatal deaths: data from 7993 pregnancies in sixdevelopingcountries.BullWorldHealthOrgan.2006;84(9):699-705.150. OdendaalHJ,GebhardtGS,TheronGB.Stillbirthratesinsingletonpregnanciesin a stable population at Karl Bremer and Tygerberg hospitals over 50 years.SAfr JObstetGynaecol.2013;19(3):67-70.151. Lavin T, Preen DB, Pattinson R. Timing and cause of perinatal mortality forsmall-for-gestational-age babies in South Africa: critical periods and challenges withdetection.Maternalhealth,neonatologyandperinatology.2016;2(1):11.152. FrankK,LombaardH,PattinsonRC.Doescompletionof theEssentialSteps inManaging Obstetric Emergencies (ESMOE) training package result in improvedknowledge and skills in managing obstetric emergencies? S Afr J Obstet Gynaecol.2009;15(3).153. Pattinson R. Challenges in saving babies-Avoidable factors, missedopportunities and substandard care in perinatal deaths in SouthAfrica.SAfrMed J.2003;93(6):450-5.154. Joseph KS, Razaz N, Muraca GM, Lisonkova S. Methodological Challenges inInternationalComparisonsofPerinatalMortality.CurrEpidemiolRep.2017;4(2):73-82.155. Leisher SH, Teoh Z, Reinebrant H, Allanson E, Blencowe H, Erwich JJ, et al.Seekingorderamidstchaos:Asystematicreviewofclassificationsystemsforcausesofstillbirthandneonataldeath,2009–2014.BMCPregnancyChildbirth.2016;16(1):295.156. Allanson E, TunçalpÖ, Gardosi J, Pattinson RC, Erwich JJH, Flenady VJ, et al.Classifyingthecausesofperinataldeath.BullWorldHealthOrgan.2016;94(2):79.157. Organization WH. The WHO application of ICD-10 to perinatal deaths: ICD-perinatalmortality(ICD-PM).Geneva:WHO.2015.158. AllansonE,TunçalpӦ,Gardosi J,PattinsonR,Vogel J,ErwichJ,etal.Givingavoice tomillions: developing theWHO application of ICD‐10 to deaths during theperinatalperiod:ICD‐PM.BJOG.2016;123(12):1896-9.159. Yli BM, Kjellmer I. Pathophysiology of foetal oxygenation and cell damageduringlabour.BestPractResClinObstetGynaecol.2016;30:9-21.

115

160. AssaliNS,AssaliNS.FetalandNeonatalDisorders.SaintLouis:ElsevierScience;2013.161. Jensen A, Garnier Y, Berger R. Dynamics of fetal circulatory responses tohypoxiaandasphyxia.EurJObstetGynecolReprodBiol.1999;84(2):155-72.162. Martin CB, Jr. Normal fetal physiology and behavior, and adaptive responseswithhypoxemia.SeminPerinatol.2008;32(4):239-42.163. StevensonA.OxforddictionaryofEnglish:OxfordUniversityPress,USA;2010.164. Ayres-de-Campos D. Introduction: Why is intrapartum foetal monitoringnecessary - Impact on outcomes and interventions. Best Pract Res Clin ObstetGynaecol.2016;30:3-8.165. Fahey J, King TL. Intrauterine asphyxia: clinical implications for providers ofintrapartumcare.JMidwiferyWomensHealth.2005;50(6):498-506.166. NewbyEA,MyersDA,DucsayCA.Fetalendocrineandmetabolicadaptationstohypoxia: the role of the hypothalamic-pituitary-adrenal axis.Am J Physiol EndocrinolMetab.2015;309(5):E429-39.167. VaughanOR,DeBlasioMJ, FowdenAL.OVINEUTEROPLACENTALAND FETALMETABOLISM DURING AND AFTER FETAL CORTISOL OVEREXPOSURE IN LATEGESTATION.American JournalofPhysiology-Regulatory, IntegrativeandComparativePhysiology.2018.168. Bennet L, Rossenrode S, Gunning MI, Gluckman PD, Gunn AJ. Thecardiovascular and cerebrovascular responses of the immature fetal sheep to acuteumbilicalcordocclusion.JPhysiol.1999;517(Pt1):247-57.169. Yan EB, Baburamani AA,Walker AM,Walker DW. Changes in cerebral bloodflow, cerebral metabolites, and breathing movements in the sheep fetus followingasphyxiaproducedbyocclusionoftheumbilicalcord.AmJPhysiolRegulIntegrCompPhysiol.2009;297(1):R60-9.170. Quaedackers JS, Roelfsema V, Hunter CJ, Heineman E, Gunn AJ, Bennet L.Polyuria and impaired renal blood flow after asphyxia in preterm fetal sheep.Am JPhysiolRegulIntegrCompPhysiol.2004;286(3):R576-83.171. LotgeringFK,BishaiJM,StruijkPC,BloodAB,HunterCJ,PowerGG,etal.Ten-minute umbilical cord occlusion markedly reduces cerebral blood flow and heatproductioninfetalsheep.AmJObstetGynecol.2003;189(1):233-8.172. HolzmannM,CnattingiusS,NordströmL.Lactateproductionasaresponsetointrapartumhypoxiainthegrowth‐restrictedfetus.BJOG.2012;119(10):1265-9.173. Giussani DA. The fetal brain sparing response to hypoxia: physiologicalmechanisms.JPhysiol.2016;594(5):1215-30.174. BennetL.Sex,drugsandrockandroll:talesfrompretermfetallife.TheJournalofphysiology.2017;595(6):1865-81.175. MallardEC,WilliamsCE,JohnstonBM,GluckmanPD.Increasedvulnerabilitytoneuronal damage after umbilical cord occlusion in fetal sheep with advancinggestation.AmJObstetGynecol.1994;170(1Pt1):206-14.176. Hernandez-Andrade E, Hellstrom-Westas L, Thorngren-Jerneck K, Jansson T,LiubaK,LingmanG,etal.Perinataladaptiveresponseoftheadrenalandcarotidbloodflowinsheepfetusessubjectedtototalcordocclusion.JMaternFetalNeonatalMed.2005;17(2):101-9.

116

177. vanDijk AJ, van Loon JP, TaverneMA, Jonker FH.Umbilical cord clamping interm piglets: a useful model to study perinatal asphyxia? Theriogenology.2008;70(4):662-74.178. GrahamEM,RuisKA,HartmanAL,NorthingtonFJ,FoxHE.Asystematicreviewof the role of intrapartum hypoxia-ischemia in the causation of neonatalencephalopathy.AmJObstetGynecol.2008;199(6):587-95.179. GeorgeS,GunnAJ,WestgateJA,BrabynC,GuanJ,BennetL.Fetalheart ratevariability and brain stem injury after asphyxia in preterm fetal sheep.Am J PhysiolRegulIntegrCompPhysiol.2004;287(4):R925-33.180. Bennet L,Quaedackers JS,GunnAJ,RossenrodeS,HeinemanE.Theeffectofasphyxia on superior mesenteric artery blood flow in the premature sheep fetus. JPediatrSurg.2000;35(1):34-40.181. SarratoSZ,CastroSV,NavarroCR,SarratoGZ,LunaMS.Risksfactorsassociatedwith intra-partum foetal mortality in pre-term infants. Anales de Pediatría (EnglishEdition).2017;86(3):127-34.182. deAlmeidaMF,MoreiraLM,VazdosSantosRM,KawakamiMD,AnchietaLM,Guinsburg R. Early neonatal deathswith perinatal asphyxia in very low birthweightBrazilianinfants.JPerinatol.2015;35(11):954-7.183. TuuliMG,StoutMJ,ShanksA,OdiboAO,MaconesGA,CahillAG.Umbilicalcordarterial lactate comparedwith pH for predicting neonatalmorbidity at term.ObstetGynecol.2014;124(4):756-61.184. Morales P, Bustamante D, Espina-Marchant P, Neira-Pena T, Gutierrez-HernandezMA,Allende-CastroC,etal.Pathophysiologyofperinatalasphyxia:canwepredictandimproveindividualoutcomes?Epmaj.2011;2(2):211-30.185. EastCE,LeaderLR,SheehanP,HenshallNE,ColditzPB.Intrapartumfetalscalplactate sampling for fetalassessment in thepresenceofanon-reassuring fetalheartrate trace.CochraneDatabase of Systematic Reviews. 2010;2010(3):CD006174-1-CD-30.186. ZanardoV,MariG,deLucaF,ScambiaG,GuerriniP,StrafaceG.Lactateincordbloodand its relation to fetal gluconeogenesis in at termdeliveries.EarlyHumDev.2015;91(3):165-8.187. RossMG.Laborand fetalheart ratedecelerations: relation to fetalmetabolicacidosis.ClinObstetGynecol.2011;54(1):74-82.188. HamedHO.Intrapartumfetalasphyxia:studyofumbilicalcordbloodlactateinrelationtofetalheartratepatterns.ArchGynecolObstet.2013;287(6):1067-73.189. Nordstrom L, Arulkumaran S. Intrapartum fetal hypoxia and biochemicalmarkers:areview.ObstetGynecolSurv.1998;53(10):645-57.190. Nordstrom L, Ingemarsson I, Westgren M. Fetal monitoring with lactate.BaillieresClinObstetGynaecol.1996;10(2):225-42.191. HeinisAM,SpaandermanME,GunnewiekJM,LotgeringFK.Scalpbloodlactatefor intra-partum assessment of fetalmetabolic acidosis.ActaObstet Gynecol Scand.2011;90(10):1107-14.192. GjerrisAC,Staer-JensenJ,JorgensenJS,BergholtT,NickelsenC.Umbilicalcordbloodlactate:avaluabletoolintheassessmentoffetalmetabolicacidosis.EurJObstetGynecolReprodBiol.2008;139(1):16-20.193. Nordstrom L, Achanna S, Naka K, Arulkumaran S. Fetal andmaternal lactateincreaseduringactivesecondstageoflabour.BJOG.2001;108(3):263-8.

117

194. Low JA, Panagiotopoulos C, Derrick EJ. Newborn complications afterintrapartum asphyxia with metabolic acidosis in the preterm fetus. Am J ObstetGynecol.1995;172(3):805-10.195. Low JA, Panagiotopoulos C, Derrick EJ. Newborn complications afterintrapartumasphyxiawithmetabolicacidosis inthetermfetus.AmJObstetGynecol.1994;170(4):1081-7.196. Frasch MG, Mansano RZ, Gagnon R, Richardson BS, Ross MG. Measures ofacidosiswithrepetitiveumbilicalcordocclusionsleadingtofetalasphyxiainthenear-termovinefetus.AmJObstetGynecol.2009;200(2):200.e1-.e7.197. Low JA.Metabolic acidosis and fetal reserve.Baillieres Clin Obstet Gynaecol.1996;10(2):211-24.198. NardozzaLMM,CaetanoACR,ZamarianACP,MazzolaJB,SilvaCP,MarçalVMG,etal.Fetalgrowthrestriction:currentknowledge.ArchGynecolObstet.2017:1-17.199. Sousa S, Mielke JG. Does Resuscitation Training Reduce Neonatal Deaths inLow-ResourceCommunities?ASystematicReviewof theLiterature.AsiaPac JPublicHealth.2015;27(7):690-704.200. Endrich O, Rimle C, Zwahlen M, Triep K, Raio L, Nelle M. Asphyxia in theNewborn: Evaluating the Accuracy of ICD Coding, Clinical Diagnosis andReimbursement: Observational Study at a Swiss Tertiary Care Center on RoutinelyCollectedHealthDatafrom2012-2015.PLoSOne.2017;12(1):e0170691.201. Ensing S, Abu-Hanna A, Schaaf JM, Mol BWJ, Ravelli AC. Trends in birthasphyxia, obstetric interventions and perinatal mortality among term singletons: anationwide cohort study. The Journal of Maternal-Fetal & Neonatal Medicine.2015;28(6):632-7.202. Manandhar SR, Manandhar DS, Adhikari D, Shrestha J, Rai C, Rana H, et al.AnalysisofHealthFacilityBasedPerinatalVerbalAutopsyofElectoralConstituency2ofArghakhanchiDistrict,Nepal.JNepalHealthResCounc.2015;13(29):73-7.203. ViswanathK,PsR,ChakrabortyA,Prasad JH,MinzS,GeorgeK.AcommunitybasedcasecontrolstudyondeterminantsofperinatalmortalityinatribalpopulationofsouthernIndia.RuralRemoteHealth.2015;15(3):3388.204. NouailiEBH,ChaouachiS,Ayadi I,SaidAB,ZouariB,MarrakchiZ.Risk factorsforperinatalmortalityinaTunisianpopulation.InternationalJournalofGynecology&Obstetrics.2010;111(3):265-6.205. Kali GTJ,Martinez‐BiargeM, Van Zyl J, Smith J, RutherfordM. Therapeutichypothermia for neonatal hypoxic‐ ischaemic encephalopathy had favourableoutcomes at a referral hospital in a middle‐ income country. Acta Paediatr.2016;105(7):806-15.206. Phoa KY, Chedraui P, Perez-Lopez FR,Wendte JF, Ghiabi S, Vrijkotte T, et al.Perinatal outcome in singleton pregnancies complicated with preeclampsia andeclampsiainEcuador.JObstetGynaecol.2016;36(5):581-4.207. BobrowCS, Soothill PW. Causes and consequences of fetal acidosis.ArchDisChildFetalNeonatalEd.1999;80(3):F246-9.208. Perlman JM. Intervention strategies for neonatal hypoxic-ischemic cerebralinjury.ClinTher.2006;28(9):1353-65.209. Pin TW, Eldridge B, GaleaMP. A review of developmental outcomes of terminfants with post-asphyxia neonatal encephalopathy. Eur J Paediatr Neurol.2009;13(3):224-34.

118

210. Perlman JM. Intrapartum asphyxia and cerebral palsy: is there a link? ClinPerinatol.2006;33(2):335-53.211. Westgate JA, Wibbens B, Bennet L, Wassink G, Parer JT, Gunn AJ. Theintrapartumdecelerationincenterstage:aphysiologicapproachtotheinterpretationoffetalheartratechangesinlabor.AmJObstetGynecol.2007;197(3):236.e1-11.212. Heuser CC, Knight S, Esplin MS, Eller AG, Holmgren CM, Richards D, et al.Tachysystoleintermlabor:incidence,riskfactors,outcomes,andeffectonfetalhearttracings.AmJObstetGynecol.2013;209(1):32.e1-.e6.213. BakkerPC,KurverPH,KuikDJ,VanGeijnHP.Elevateduterineactivityincreasestheriskoffetalacidosisatbirth.AmJObstetGynecol.2007;196(4):313.e1-6.214. GibbonsC,O'HerlihyC,MurphyJ.Umbilicalcordprolapse–changingpatternsandimprovedoutcomes:aretrospectivecohortstudy.BJOG.2014;121(13):1705-8.215. Huang J-P, Chen C-P, Chen C-P, Wang K-G, Wang K-L. Term pregnancy withumbilicalcordprolapse.TaiwanJObstetGynecol.2012;51(3):375-80.216. NorwitzER,BelfortMA,SaadeGR,MillerH.CordProlapse.Oxford,UK:Wiley‐Blackwell;2010.156-7p.217. Holbrook BD, Phelan ST. Umbilical Cord Prolapse.Obstet Gynecol Clin NorthAm.2012.218. Kahana B, Sheiner E, Levy A, Lazer S,MazorM. Umbilical cord prolapse andperinataloutcomes.IntJGynaecolObstet.2004;84(2):127-32.219. Gabbay-Benziv R, MamanM,Wiznitzer A, Linder N, Yogev Y. Umbilical cordprolapse during delivery - risk factors and pregnancy outcome: a single centerexperience.Thejournalofmaternal-fetal&neonatalmedicine:theofficial journalofthe European Association of PerinatalMedicine, the Federation of Asia andOceaniaPerinatalSocieties,theInternationalSocietyofPerinatalObstetricians.2014;27(1):14.220. Wasswa EW, Nakubulwa S, Mutyaba T. Fetal demise and associated factorsfollowingumbilical cord prolapse inMulagohospital,Uganda: a retrospective study.ReproductiveHealth.2014;11:12.221. UmarU,GayaS.OutcomeofumbilicalcordprolapseatAminuKanoTeachingHospital,Kano,North-WesternNigeria.NigerianJournalofBasicandClinicalSciences.2015;12(1):20-4.222. BehbehaniS,PatenaudeV,AbenhaimHA.MaternalRiskFactorsandOutcomesof Umbilical Cord Prolapse: A Population-Based Study. Journal of Obstetrics andGynaecologyCanada.2016;38(1):23-8.223. Hasegawa J, SekizawaA, IkedaT,KoresawaM, Ishiwata I,KawabataM,etal.Clinicalriskfactorsforpoorneonataloutcomesinumbilicalcordprolapse.Thejournalofmaternal-fetal&neonatalmedicine:theofficialjournaloftheEuropeanAssociationof Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, theInternationalSocietyofPerinatalObstetricians.2016;29(10):1652.224. EsikeCOU,AlukaOC,UkanwokoC,EzeJN,TowmeyDE.O044CORDPROLAPSEIN A RURAL SOUTH EASTHERN NIGERIA HOSPITAL: A TEN‐ YEAR EXPERIENCE.InternationalJournalofGynecology&Obstetrics.2012;119:S276-S.225. ObeidatN,ZayedF,AlchalabiH,ObeidatB,El-JalladMF,ObeidatM.Umbilicalcord prolapse: A 10-year retrospective study in two civil hospitals, North Jordan. JObstetGynaecol.2010;30(3):257-60.226. EnakpeneCA,OdukogbeA-TO,Morhason-BelloIO,OmigbodunAO,ArowojoluAO.Theinfluenceofhealth-seekingbehaviorontheincidenceandperinataloutcome

119

of umbilical cord prolapse in Nigeria. International Journal of Women's Health.2010;2:177-82.227. FazilA,ZafarF,AsifaA,AkmalN,KarimA.Prevalenceandperinatalmortalityofcordprolapseandpresentation.AnnalsofKingEdwardMedicalUniversity.2016;11(3).228. KamoshitaE,AmanoK,KanaiY,MochizukiJ,IkedaY,KikuchiS,etal.Effectofthe interval between onset of sustained fetal bradycardia and cesarean delivery onlong‐term neonatal neurologic prognosis. International Journal of Gynecology &Obstetrics.2010;111(1):23-7.229. Kalu C, Umeora O. Risk factors and perinatal outcome of umbilical cordprolapse in Ebonyi StateUniversity TeachingHospital, Abakaliki,Nigeria.Niger J ClinPract.2011;14(4):413-7.230. Lamina M, Adefuye P, Akinsanya A, Odubena O. Risk factors and perinataloutcome of umbilical cord prolapse in Sagamu, Nigeria. Annals of Health Research.2016;2(2):98-.231. Downes KL, Grantz KL, Shenassa ED.Maternal, Labor, Delivery, and PerinatalOutcomesAssociatedwithPlacentalAbruption:ASystematicReview.AmJPerinatol.2017;34(10):935-57.232. Nkwabong E, Tiomela Goula G. Placenta abruption surface and perinataloutcome.JMaternFetalNeonatalMed.2017;30(12):1456-9.233. ColemanJ,SrofenyoEK,OforiEK,BrakohiapaEK,AntwiWK.Maternalandfetalprognosis in abruptio placentae at Korle-Bu Teaching Hospital, Ghana.Afr J ReprodHealth.2014;18(4):115-22.234. BerhanY.Predictorsofperinatalmortalityassociatedwithplacentapreviaandplacentalabruption:anexperiencefromalowincomecountry.Journalofpregnancy.2014;2014.235. Akintayo AA, Olagbuji BN, Aderoba AK, Akadiri O, Olofinbiyi BA, Bakare B.Emergency Peripartum Hysterectomy: A Multicenter Study of Incidence, IndicationsandOutcomesinSouthwesternNigeria.MaternChildHealthJ.2016;20(6):1230-6.236. AbasiattaiAM,UmoiyohoAJ,UtukNM,Inyang-EtohEC,AsuquoE.Emergencyperipartum hysterectomy in a tertiary hospital in southern Nigeria. Pan Afr Med J.2013;15(1).237. Rizwan N, Abbasi RM, Uddin SF. Uterine rupture, frequency of cases andfetomaternal outcome. JPMA-Journal of the Pakistan Medical Association.2011;61(4):322.238. Gupta A, Nanda S. Uterine rupture in pregnancy: a five-year study. ArchGynecolObstet.2011;283(3):437-41.239. Rezende KB, Bornia RG, Esteves AP, Cunha AJ, Amim Junior J. Preeclampsia:PrevalenceandperinatalrepercussionsinaUniversityHospitalinRiodeJaneiro,Brazil.PregnancyHypertens.2016;6(4):253-5.240. Mooij R, Lugumila J, Mwashambwa MY, Mwampagatwa IH, van Dillen J,Stekelenburg J. Characteristics and outcomes of patientswith eclampsia and severepre-eclampsia inaruralhospital inWesternTanzania:aretrospectivemedicalrecordstudy.BMCPregnancyChildbirth.2015;15:213.241. DasR,BiswasS.Eclapmsia:TheMajorCauseofMaternalMortality inEasternIndia.EthiopJHealthSci.2015;25(2):111-6.

120

242. GaladanciH,KunzelW,ZinserR,ShittuO,AdamsS,GruhlM.Experiencesof6years quality assurance in obstetrics in Nigeria - a critical review of results andobstacles.JPerinatMed.2016;44(3):301-8.243. Vigil-DeGraciaP,Rojas-SuarezJ,RamosE,ReyesO,CollantesJ,QuinteroA,etal. Incidence of eclampsia with HELLP syndrome and associated mortality in LatinAmerica.IntJGynaecolObstet.2015;129(3):219-22.244. Perveen S. Frequency and impact of hypertensive disorders of pregnancy. JAyubMedCollAbbottabad.2014;26(4):518-21.245. Kumari A, ChakrawartyA, SinghA, SinghR.Maternofoetal complications andtheir associationwithproteinuria in a tertiary carehospital of adeveloping country.Journalofpregnancy.2014;2014.246. AsıcıogluO,ŞahbazA,GüngördükK,YildirimG,AsıcıogluBB,ÜlkerV.Maternaland perinatal outcomes in women with placenta praevia and accreta in teachinghospitalsinWesternTurkey.JObstetGynaecol.2014;34(6):462-6.247. Kiondo P, Tumwesigye NM, Wandabwa J, Wamuyu-Maina G, Bimenya GS,Okong P. Adverse neonatal outcomes in women with pre-eclampsia in MulagoHospital, Kampala,Uganda: a cross-sectional study.ThePanAfricanmedical journal.2014;17(Suppl1).248. Sadaf N, Haq G, Shukar-ud-Din S. Maternal and foetal outcome in HELLPsyndromeattertiarycarehospital.JPakMedAssoc.2013;63(12):1500-3.249. ParnaF, LatifT,SultanaN,AliM,ChowdhuryS.Maternal& fetaloutcomeofeclampticpatientsadmittedinobstetrics&gynaecologydepartmentofsecondarycarehospitalinBangladesh.Mymensinghmedicaljournal:MMJ.2013;22(3):522-6.250. Vigil-DeGraciaP,Montufar-RuedaC,RuizJ.Expectantmanagementofseverepreeclampsia and preeclampsia superimposed on chronic hypertension between 24and 34 weeks’ gestation. European Journal of Obstetrics & Gynecology andReproductiveBiology.2003;107(1):24-7.251. AdamuA,EkeleB,AhmedY,MohammedB,IsezuoS,AbdullahpiA.PregnancyoutcomeinwomenwitheclampsiaatatertiarycentreinnorthernNigeria.AfrJMedMedSci.2012;41(2):211-9.252. JidoTA.Ecalmpsia:maternalandfetaloutcome.AfrHealthSci.2012;12(2):148-52.253. TukurJ,AhonsiB,IshakuSM,AraoyinboI,OkerekeE,BabatundeAO.Maternaland fetal outcomes after introduction of magnesium sulphate for treatment ofpreeclampsia and eclampsia in selected secondary facilities: a low-cost intervention.Maternalandchildhealthjournal.2013;17(7):1191-8.254. UgwuE,DimC,OkonkwoC,NwankwoT.Maternalandperinataloutcomeofseverepre-eclampsiainEnugu,NigeriaafterintroductionofMagnesiumsulfate.NigerJClinPract.2011;14(4):418-21.255. Abd El Aal DE, Shahin AY. Management of eclampsia at Assiut UniversityHospital,Egypt.IntJGynaecolObstet.2012;116(3):232-6.256. Kumar M, Meena J, Gupta U, Singh A, Jain N. Management of early onsetseverepreeclampsia inatertiaryhospital in india:doesexpectantmanagementalterperinataloutcome?IndianJMedSci.2011;65(12).257. AgidaE,AdekaB,JibrilK.PregnancyoutcomeineclampticsattheUniversityofAbuja Teaching Hospital, Gwagwalada, Abuja: a 3 year review. Niger J Clin Pract.2010;13(4).

121

258. YildirimG,GungordukK,GulA,AsiciogluO,SudolmusS,GungordukOC,etal.HELLP Syndrome: 8 years of experience from a tertiary referral center in WesternTurkey.HypertensPregnancy.2012;31(3):316-26.259. Tukur J,Muhammad Z.Management of eclampsia at AKTH: before and aftermagnesiumsulphate.NigerJMed.2010;19(1).260. Low JA. Cerebral perfusion, metabolism, and outcome. Curr Opin Pediatr.1995;7(2):132-9.261. DeppR.Perinatalasphyxia:assessing itscausalroleandtiming.SeminPediatrNeurol.1995;2(1):3-36.262. Blackstone J, Young BK. Umbilical cord blood acid-base values and otherdescriptorsoffetalcondition.ClinObstetGynecol.1993;36(1):33-46.263. OliveiraTGd,FreirePV,MoreiraFT,Moraes JdSBd,ArrelaroRC,RossiS,etal.Apgarscoreandneonatalmortality inahospital located in thesouthernareaofSãoPauloCity,Brazil.Einstein(SãoPaulo).2012;10(1):22-8.264. SabiriN,KabiriM,KarboubiL,BouzianeA,BarkatA.Risk factors forperinatalmortality at Souissi Maternity Hospital, Rabat, Morocco. International Journal ofGynecology&Obstetrics.2012;119(3):285-6.265. Thorp JA, Rushing RS. Umbilical cord blood gas analysis.Obstet Gynecol ClinNorthAm.1999;26(4):695-709.266. BlicksteinI,GreenT.Umbilicalcordbloodgases.ClinPerinatol.2007;34(3):451-9.267. YehP,EmaryK, ImpeyL. The relationshipbetweenumbilical cordarterialpHand serious adverse neonatal outcome: analysis of 51,519 consecutive validatedsamples.BJOG.2012;119(7):824-31.268. VictoryR,PenavaD,DaSilvaO,NataleR,RichardsonB.UmbilicalcordpHandbaseexcessvaluesinrelationtoadverseoutcomeeventsforinfantsdeliveringatterm.AmJObstetGynecol.2004;191(6):2021-8.269. MalinGL,MorrisRK,KhanKS.StrengthofassociationbetweenumbilicalcordpHandperinatalandlongtermoutcomes:systematicreviewandmeta-analysis.BMJ.2010;340:c1471.270. Olofsson P. Determination of base excess in umbilical cord blood at birth:accessoryorexcess?AmJObstetGynecol.2015;213(3):259-61.271. UccellaS,CromiA,ColomboG,AgostiM,BoganiG,CasarinJ,etal.Predictionof fetal base excess values at birth using an algorithm to interpret fetal heart ratetracings:aretrospectivevalidation.BJOG.2012;119(13):1657-64.272. Blechner JN. Maternal-fetal acid-base physiology. Clin Obstet Gynecol.1993;36(1):3-12.273. Apgar V. A proposal for a newmethod of evaluation of the newborn infant.CurrResAnesthAnalg.1953;32(4):260-7.274. BerglundS,PetterssonH,CnattingiusS,GrunewaldC.HowoftenisalowApgarscoretheresultofsubstandardcareduringlabour?BJOG.2010;117(8):968-78.275. Thorngren-JerneckK,HerbstA.Low5-minuteApgarscore:apopulation-basedregisterstudyof1milliontermbirths.ObstetGynecol.2001;98(1):65-70.276. Ehrenstein V, Pedersen L, Grijota M, Nielsen GL, Rothman KJ, Sorensen HT.Association of Apgar score at five minutes with long-term neurologic disability andcognitive function in a prevalence study of Danish conscripts. BMC PregnancyChildbirth.2009;9:14.

122

277. IliodromitiS,MackayDF,SmithGC,PellJP,NelsonSM.Apgarscoreandtheriskof cause-specific infant mortality: a population-based cohort study. Lancet.2014;384(9956):1749-55.278. Lawn J, Shibuya K, Stein C. No cry at birth: global estimates of intrapartumstillbirths and intrapartum-related neonatal deaths. Bull World Health Organ.2005;83(6):409-17.279. Allanson E,Waqar T,White C, Tunçalp Ö, Dickinson J. Umbilical lactate as ameasureofacidosisandpredictorofneonatalrisk:asystematicreview.BJOG.2016.280. BorrutoF,ComparettoC,WegherE,TreisserA.Screeningoffoetaldistressbyassessmentofumbilicalcordlactate.ClinExpObstetGynecol.2006;33(4):219-22.281. ParisA,Maurice-Tison S, Coatleven F,Vandenbossche F,DallayD,Horovitz J.[Interestoflactatemicro-dosageinscalpandumbilicalcordincasesofabnormalfetalheart rate during labor. Prospective study on 162 patients]. J Gynecol Obstet BiolReprod(Paris).2012;41(4):324-32.282. ReifP,LakovschekI,TappaufC,HaasJ,LangU,SchollW.Validationofapoint-of-care(POC)lactatetestingdeviceforfetalscalpbloodsamplingduringlabor:clinicalconsiderations,practicalitiesandrealities.ClinChemLabMed.2014;52(6):825-33.283. Wiberg-ItzelE,LipponerC,NormanM,HerbstA,PrebensenD,HanssonA,etal.DeterminationofpHorlactateinfetalscalpbloodinmanagementofintrapartumfetaldistress:randomisedcontrolledmulticentretrial.BMJ.2008;336(7656):1284-7.284. RamanahR,MartinA,ClementM-C,MailletR,RiethmullerD.Fetalscalplactatemicrosamplingfornon-reassuringfetalstatusduringlabor:aprospectiveobservationalstudy.FetalDiagnTher.2010;27(1):14-9.285. Labrecque L, Provencal M, Caqueret A, Wo BL, Bujold E, Lariviere F, et al.CorrelationofcordbloodpH,baseexcess,andlactateconcentrationmeasuredwithaportabledevice for identifying fetal acidosis. JObstetGynaecolCan. 2014;36(7):598-604.286. IorizzoL,KlausenT,Wiberg-ItzelE,OvinF,WibergN.UseofLactateProTM2formeasurement of fetal scalp blood lactate during labor–proposing new cutoffs fornormality,preacidemiaandacidemia:across-sectionalstudy.TheJournalofMaternal-Fetal&NeonatalMedicine.2018:1-7.287. PennellCE,TracyMB.Anewmethodforrapidmeasurementoflactateinfetalandneonatalblood.AustNZJObstetGynaecol.1999;39(2):227-33.288. WangM,ChuaSC,BouhadirL,TreadwellEL,GibbsE,McGeeTM.Point‐of‐caremeasurementoffetalbloodlactate–Timetotrustanewdevice.AustNZJObstetGynaecol.2018;58(1):72-8.289. AllenRM,BowlingFG, Jeremy J.Determining the fetal scalp lactate level thatindicates the need for intervention in labour. Aust N Z J Obstet Gynaecol.2004;44(6):549-52.290. Baldari C, Bonavolonta V, Emerenziani GP, Gallotta MC, Silva AJ, Guidetti L.Accuracy,reliability, linearityofAccutrendandLactateProversusEBIOplusanalyzer.EurJApplPhysiol.2009;107(1):105-11.291. Karagiannis MH, Reniker AN, Kerl ME, Mann F. Lactate measurement as anindicatorofperfusion.Compendium.2006;28(4):287-98.292. Emmert J, editor Analytical requirements for the measurement of lactate.WorkshopreportAccusportMannheim,Germany:BoehringerMannheimGmbH;1995.

123

293. DessolleL,LebrecJ,DaraïE.Impactofdelayedarterialcordbloodsamplingforlactateassay.Neonatology.2009;95(3):224-9.294. White CR, Mok T, Doherty DA, Henderson JJ, Newnham JP, Pennell CE. Theeffectoftime,temperatureandstoragedeviceonumbilicalcordbloodgasandlactatemeasurement:arandomizedcontrolledtrial.TheJournalofMaternal-Fetal&NeonatalMedicine.2012;25(6):587-94.295. Ferguson VL, Dodson RB. Bioengineering aspects of the umbilical cord.European Journal of Obstetrics & Gynecology and Reproductive Biology.2009;144:S108-S13.296. DuerbeckNB,ChaffinDG,SeedsJW.ApracticalapproachtoumbilicalarterypHandbloodgasdeterminations.ObstetGynecol.1992;79(6):959-62.297. Westgate J, Garibaldi JM, Greene KR. Umbilical cord blood gas analysis atdelivery:atimeforqualitydata.BJOG.1994;101(12):1054-63.298. ChilindaGK,GadamaLA,StonesW.Point-of-careumbilicalarteriallactateandnewborn outcomes in a low resource setting: cohort study. BMC Res Notes.2018;11(1):477.299. Linet T, Laporte J, GueyeH, BoogG. [Microvolume dosage of lactate in cordblood for the evaluation of the neonatal well-being]. J Gynecol Obstet Biol Reprod(Paris).2002;31(4):352-7.300. MartinA,GaillardM,MiotS,RiethmullerD,SchaalJP.[Lactatemeasurementsandacid-basebalance incordblood].JGynecolObstetBiolReprod(Paris).2003;32(8Pt1):713-9.301. Racinet C, Richalet G, Corne C, Faure P, Peresse J, Leverve X. Diagnosis ofneonatal metabolic acidosis by eucapnic pH determination. Gynecol Obstet Fertil.2013;41(9):485-92.302. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Accuratepredictionofhypoxic-ischaemicencephalopathyatdelivery:acohortstudy.TheJournalofMaternal-Fetal&NeonatalMedicine.2012;25(9):1653-9.303. Wiberg N, Källén K, Herbst A, Olofsson P. Relation between umbilical cordblood pH, base deficit, lactate, 5‐minute Apgar score and development of hypoxicischemicencephalopathy.ActaObstetGynecolScand.2010;89(10):1263-9.304. Ayres-de-CamposD,SpongCY,ChandraharanE.FIGOconsensusguidelinesonintrapartum fetal monitoring: Cardiotocography. Int J Gynaecol Obstet.2015;131(1):13-24.305. Ayres-de-CamposD, Bernardes J. Twenty-five years after the FIGO guidelinesfortheuseoffetalmonitoring:timeforasimplifiedapproach?IntJGynaecolObstet.2010;110(1):1-6.306. Makino S, Hirai C, Takeda J, Itakura A, Takeda S. Relationship between fetalheart rate patterns and a time course for evaluation of fetal well-being:“the 30minutesrule”fordecisionofmechanicaldelivery.TaiwaneseJournalofObstetricsandGynecology.2017;56(6):788-92.307. BulkmansN,LyrenasS,HallbergG,NiklassonF.Umbilicalcordbloodsampling--atoolfordeliveryqualitycontrol?ActaObstetGynecolScand.1997;76(5):419-22.308. CrovettoF,FumagalliM,DeCarliA,BafferoGM,NozzaS,DessimoneF,etal.Obstetric risk factors forpoorneonataladaptationatbirth.The JournalofMaternal-Fetal&NeonatalMedicine.2018;31(18):2429-35.

124

309. SantoS,Ayres‐de‐CamposD,Costa‐SantosC,SchnettlerW,UgwumaduA,Da Graça LM. Agreement and accuracy using the FIGO, ACOG and NICEcardiotocography interpretation guidelines. Acta Obstet Gynecol Scand.2017;96(2):166-75.310. Gamboa SM, Giménez OR, Mancho JP, Moros ML, Sada JR, Mateo SC.DiagnosticaccuracyoftheFIGOandthe5-tierfetalheartrateclassificationsystemsinthedetectionofneonatalacidemia.AmJPerinatol.2017;34(05):508-14.311. GrahamEM,AdamiRR,McKenneySL,JenningsJM,BurdI,WitterFR.Diagnosticaccuracy of fetal heart rate monitoring in the identification of neonatalencephalopathy.ObstetGynecol.2014;124(3):507.312. CostantineMM, SaadeGR, editors. The first cesarean: roleof “fetal distress”diagnosis.SeminPerinatol;2012:Elsevier.313. HadarA,SheinerE,HallakM,KatzM,MazorM,Shoham-VardiI.Abnormalfetalheartratetracingpatternsduringthefirststageoflabor:effectonperinataloutcome.AmJObstetGynecol.2001;185(4):863-8.314. Chandraharan E, Arulkumaran S. Prevention of birth asphyxia: respondingappropriately to cardiotocograph (CTG) traces. Best Pract Res Clin Obstet Gynaecol.2007;21(4):609-24.315. Grivell RM, Alfirevic Z, Gyte GM, Devane D. Antenatal cardiotocography forfetalassessment.CochraneDatabaseSystRev.2015(9):Cd007863.316. Impey L, Reynolds M, MacQuillan K, Gates S, Murphy J, Sheil O. Admissioncardiotocography:arandomisedcontrolledtrial.Lancet.2003;361(9356):465-70.317. Freeman RK. Problems with intrapartum fetal heart rate monitoringinterpretationandpatientmanagement.ObstetGynecol.2002;100(4):813-26.318. MiresG,WilliamsF,HowieP.RandomisedcontrolledtrialofcardiotocographyversusDopplerauscultationof fetalheartatadmission in labour in lowriskobstetricpopulation.BMJ.2001;322(7300):1457-60;discussion60-2.319. Dellinger EH, Boehm FH, Crane MM. Electronic fetal heart rate monitoring:earlyneonataloutcomesassociatedwithnormal rate, fetal stress,and fetaldistress.AmJObstetGynecol.2000;182(1Pt1):214-20.320. Ribak R, Harlev A, Ohel I, Sergienko R, Wiznitzer A, Sheiner E. Refusal ofemergency caesarean delivery in cases of non-reassuring fetal heart rate is anindependent risk factor for perinatal mortality. European Journal of Obstetrics &GynecologyandReproductiveBiology.2011;158(1):33-6.321. Larma JD, Silva AM, Holcroft CJ, Thompson RE, Donohue PK, Graham EM.Intrapartumelectronicfetalheartratemonitoringandtheidentificationofmetabolicacidosis and hypoxic-ischemic encephalopathy. Am J Obstet Gynecol.2007;197(3):301.e1-8.322. Vintzileos AM, Nochimson DJ, Antsaklis A, Varvarigos I, Guzman ER, KnuppelRA. Comparison of intrapartum electronic fetal heart rate monitoring versusintermittent auscultation in detecting fetal acidemia at birth. Am J Obstet Gynecol.1995;173(4):1021-4.323. Georgieva A, Redman CW, Papageorghiou AT. Computerized data‐driveninterpretationoftheintrapartumcardiotocogram:acohortstudy.ActaObstetGynecolScand.2017;96(7):883-91.

125

324. NunesI,Ayres-de-CamposD,UgwumaduA,AminP,BanfieldP,NicollA,etal.Central Fetal Monitoring With and Without Computer Analysis. Obstet Gynecol.2017;129(1):83-90.325. Brocklehurst P, Field D, Greene K, Juszczak E, Keith R, Kenyon S, et al.Computerisedinterpretationoffetalheartrateduringlabour(INFANT):arandomisedcontrolledtrial.TheLancet.2017;389(10080):1719-29.326. HendrixNW,ChauhanSP.Cesareandeliveryfornonreassuringfetalheartratetracing.ObstetGynecolClinNorthAm.2005;32(2):273-86,ix.327. KuhnertM,SchmidtS. Intrapartummanagementofnonreassuring fetalheartrate patterns: a randomized controlled trial of fetal pulse oximetry. Am J ObstetGynecol.2004;191(6):1989-95.328. Elimian A, Lawlor P, Figueroa R, Wiencek V, Garry D, Quirk JG. Intrapartumassessment of fetal well-being: any role for a fetal admission test? J Matern FetalNeonatalMed.2003;13(6):408-13.329. OlofssonP,NorénH,CarlssonA,RosénKG.IdentifyingnewbornswithumbilicalcordbloodmetabolicacidosisbyintrapartumcardiotographycombinedwithfetalECGST analysis (STAN): comparison of the new and old FIGO systems to classifycardiotocograms. The Journal of Maternal-Fetal & Neonatal Medicine. 2018(just-accepted):1-11.330. GarabedianC,ButruilleL,DrumezE,SchreiberES,BartoloS,BleuG,etal.Inter-observerreliabilityof4fetalheartrateclassifications.Journalofgynecologyobstetricsandhumanreproduction.2017;46(2):131-5.331. RhöseS,HeinisAM,VandenbusscheF,DrongelenJ,DillenJ.Inter‐andintra‐observer agreement of non‐reassuring cardiotocography analysis and subsequentclinicalmanagement.ActaObstetGynecolScand.2014;93(6):596-602.332. BlixE,OianP.Interobserveragreementsinassessing549laboradmissiontestsafterastandardizedtrainingprogram.ActaObstetGynecolScand.2005;84(11):1087-92.333. FiguerasF,AlbelaS,BoninoS,PalacioM,BarrauE,HernandezS,etal.Visualanalysisofantepartum fetalheart rate tracings: inter-and intra-observeragreementandimpactofknowledgeofneonataloutcome.JPerinatMed.2005;33(3):241-5.334. HolcroftCJ,GrahamEM,Aina-MumuneyA,RaiKK,HendersonJL,PenningDH.Cordgasanalysis,decision-to-deliveryinterval,andthe30-minuteruleforemergencycesareans.JPerinatol.2005;25(4):229-35.335. BlixE,SviggumO,KossKS,OianP.Inter-observervariationinassessmentof845labouradmissiontests:comparisonbetweenmidwivesandobstetriciansintheclinicalsettingandtwoexperts.BJOG.2003;110(1):1-5.336. ChauhanSP,KlauserCK,WoodringTC,SandersonM,MagannEF,MorrisonJC.Intrapartum nonreassuring fetal heart rate tracing and prediction of adverseoutcomes:interobservervariability.AmJObstetGynecol.2008;199(6):623.e1-5.337. Kundu S, Kuehnle E, Schippert C, von Ehr J, Hillemanns P, Staboulidou I.EstimationofneonataloutcomearterypHvalueaccordingtoCTGinterpretationofthelast 60 min before delivery: a retrospective study. Can the outcome pH value bepredicted?ArchGynecolObstet.2017;296(5):897-905.338. Parer J, King T, Flanders S, FoxM, Kilpatrick S. Fetal acidemia and electronicfetalheartratepatterns:isthereevidenceofanassociation?TheJournalofMaternal-Fetal&NeonatalMedicine.2006;19(5):289-94.

126

339. Williams KP, Galerneau F. Intrapartum fetal heart rate patterns in thepredictionofneonatalacidemia.AmJObstetGynecol.2003;188(3):820-3.340. MilsomI,LadforsL,ThiringerK,NiklassonA,OdebackA,ThornbergE.Influenceofmaternal,obstetricandfetalriskfactorsontheprevalenceofbirthasphyxiaatterminaSwedishurbanpopulation.ActaObstetGynecolScand.2002;81(10):909-17.341. Trochez RD, Sibanda T, Sharma R, Draycott T. Fetal monitoring in labor: areaccelerationsgoodenough?JMaternFetalNeonatalMed.2005;18(5):349-52.342. TharmaratnamS.Fetaldistress.BaillieresBestPractResClinObstetGynaecol.2000;14(1):155-72.343. Schifrin BS. The CTG and the timing and mechanism of fetal neurologicalinjuries.BestPractResClinObstetGynaecol.2004;18(3):437-56.344. Williams KP, Galerneau F. Fetal heart rate parameters predictive of neonataloutcomeinthepresenceofaprolongeddeceleration.ObstetGynecol.2002;100(5Pt1):951-4.345. Han JY, Nava-Ocampo AA. Fetal heart rate response to cordocentesis andpregnancy outcome: a prospective cohort. J Matern Fetal Neonatal Med.2005;17(3):207-11.346. ChuangJ,ChouCT,ChengWC,HuangLW,HwangJL,TsaiYL.Spontaneousfetalheart rate deceleration: an ominous sign for fetal outcome. Arch Gynecol Obstet.2004;269(4):254-8.347. Strachan BK, Sahota DS, van Wijngaarden WJ, James DK, Chang AM.Computerised analysis of the fetal heart rate and relation to acidaemia at delivery.BJOG.2001;108(8):848-52.348. Agrawal SK, Doucette F, Gratton R, Richardson B, Gagnon R. Intrapartumcomputerized fetal heart rate parameters and metabolic acidosis at birth. ObstetGynecol.2003;102(4):731-8.349. Kazandi M, Sendag F, Akercan F, Terek MC, Gundem G. Different types ofvariable decelerations and their effects to neonatal outcome. Singapore Med J.2003;44(5):243-7.350. WilliamsKP,GalerneauF.Comparisonof intrapartumfetalheartratetracingsinpatientswithneonatalseizuresvs.noseizures:whatarethedifferences?JPerinatMed.2004;32(5):422-5.351. BorrutoF,ComparettoC,TreisserA.Preventionofcerebralpalsyduringlabour:roleoffoetallactate.ArchGynecolObstet.2008;278(1):17-22.352. Amer-WahlinI,IngemarssonI,MarsalK,HerbstA.FetalheartratepatternsandECG ST segment changes preceding metabolic acidaemia at birth. BJOG.2005;112(2):160-5.353. SonciniE,PaganelliS,VezzaniC,GarganoG,GiovanniBattistaLS. Intrapartumfetalheartratemonitoring:evaluationofastandardizedsystemof interpretationforpredictionofmetabolicacidosisatdeliveryandneonatalneurologicalmorbidity.TheJournalofMaternal-Fetal&NeonatalMedicine.2014;27(14):1465-9.354. CahillAG,CaugheyAB,RoehlKA,OdiboAO,MaconesGA.Terminalfetalheartdecelerationsandneonataloutcomes.ObstetGynecol.2013;122(5):1070-6.355. Jonsson M, Ågren J, Nordén-Lindeberg S, Ohlin A, Hanson U. Neonatalencephalopathy and the association to asphyxia in labor. Am J Obstet Gynecol.2014;211(6):667.e1-.e8.

127

356. Galazios G, Tripsianis G, Tsikouras P, Koutlaki N, Liberis V. Fetal distressevaluation using and analyzing the variables of antepartum computerizedcardiotocography.ArchGynecolObstet.2010;281(2):229-33.357. Ghi T, Morganelli G, Bellussi F, Rucci P, Giorgetta F, Rizzo N, et al.Cardiotocographicfindingsinthesecondstageoflaboramongfetusesdeliveredwithacidemia:acomparisonoftwoclassificationsystems.EuropeanJournalofObstetrics&GynecologyandReproductiveBiology.2016;203:297-302.358. Martí Gamboa S, Lapresta Moros M, Pascual Mancho J, Lapresta Moros C,CastánMateoS.Decelerationareaandfetalacidemia.TheJournalofMaternal-Fetal&NeonatalMedicine.2017;30(21):2578-84.359. TriebwasserJE,ColvinR,MaconesGA,CahillAG.Nonreassuringfetalstatusinthe second stage of labor: fetal monitoring features and association with neonataloutcomes.AmJPerinatol.2016;33(07):665-70.360. TranquilliAL,BiaginiA,GrecoP,DiTommasoM,GiannubiloSR.Thecorrelationbetweenfetalbradycardiaareainthesecondstageoflaborandacidemiaatbirth.TheJournalofMaternal-Fetal&NeonatalMedicine.2013;26(14):1425-9.361. Cahill AG, Roehl KA, Odibo AO, Macones GA. Association and prediction ofneonatalacidemia.AmJObstetGynecol.2012;207(3):206.e1-.e8.362. GiannubiloSR,BuscicchioG,Gentilucci L,PallaGP,TranquilliAL.Decelerationarea of fetal heart rate trace and fetal acidemia at delivery: a case-control study. JMaternFetalNeonatalMed.2007;20(2):141-4.363. Cahill AG, Roehl KA, Odibo AO, Macones GA. Association of atypicaldecelerationswithacidemia.ObstetGynecol.2012;120(6):1387-93.364. Ugwumadu A. Understanding cardiotocographic patterns associated withintrapartum fetal hypoxia and neurologic injury. Best practice & research Clinicalobstetrics&gynaecology.2013;27(4):509-36.365. GarciaGS,MarianiNetoC,AraujoJuniorE,GarciaRL,NardozzaLM,MoronAF.Fetal acidemia prediction through short-term variation assessed by antepartumcomputerized cardiotocography in pregnant women with hypertension syndrome.ArchGynecolObstet.2008;278(2):125-8.366. LowJA,VictoryR,DerrickEJ.Predictivevalueofelectronicfetalmonitoringforintrapartumfetalasphyxiawithmetabolicacidosis.ObstetGynecol.1999;93(2):285-91.367. AmsumangS,WuttikonsammakitP,ChamnanP,ChuenwattanaP,PhaophanA,Mongkolchat N, et al. Association Between Intrapartum Cardiotocogram and EarlyNeonatal Outcomes in a Tertiary Hospital in Thailand. Journal of the MedicalAssociationofThailand=Chotmaihetthangphaet.2017;100(4):365-73.368. KapayaH,JacquesR,RahaimN,AnumbaD.“Doesshort-termvariationinfetalheart rate predict fetal acidaemia?” A systematic review and meta-analysis. TheJournalofMaternal-Fetal&NeonatalMedicine.2016;29(24):4070-7.369. HolzmannM,WretlerS,CnattingiusS,NordströmL.Cardiotocographypatternsandriskofintrapartumfetalacidemia.JPerinatMed.2015;43(4):473-9.370. De Los Santos-GarateA,Villa-GuillenM,Villanueva-GarciaD,Vallejos-RuizM,Murguia-Peniche M. Perinatal morbidity and mortality in late-term and post-termpregnancy. NEOSANO perinatal network's experience in Mexico. J Perinatol.2011;31(12):789.

128

371. AdamY,MwinyogleeJ,MasukuB,NicolaouE.Anevaluationoftheindicationsfor caesarean sections at Chris Hani Baragwanath Academic Hospital. S Afr J ObstetGynaecol.2018;24(1):8-11.372. Makhanya V, Govender L, Moodley J. Utility of the Robson Ten GroupClassification System to determine appropriateness of caesarean section at a ruralregionalhospitalinKwaZulu-Natal,SouthAfrica.SAfrMedJ.2015;105(4):292-5.373. BenzouinaS,BoubkraouiMelM,MrabetM,ChahidN,KharbachA,El-HassaniA, et al. Fetal outcome in emergency versus elective cesarean sections at SouissiMaternityHospital,Rabat,Morocco.PanAfrMedJ.2016;23:197.374. OettleCA,OdendaalHJ.Umbilicalarterybloodgases innewbornsofhigh-riskmothers delivered within 6 hours of antenatal fetal heart monitoring. S Afr Med J.2000;90(7):705-6.375. van Bogaert LJ, Misra A. Neonatal outcome after caesarean birth for fetaldistressand/ormeconiumstaininginaSouthAfricanruralsetting.JObstetGynaecol.2008;28(1):56-9.376. Hofmeyr GJ, Kulier R. Operative versus conservative management for'fetaldistress' in labour.TheCochranedatabaseof systematic reviews.2012(6):CD001065-CD.377. Roy KK, Baruah J, Kumar S, Deorari AK, Sharma JB, Karmakar D. Cesareansectionforsuspectedfetaldistress,continuousfetalheartmonitoringanddecisiontodeliverytime.IndianJPediatr.2008;75(12):1249-52.378. ShiekhSM,KamruddinA,SetnaF,RiazT.Roleofpathologicalcardiotocographyinevaluatingfetalwell-being.JCollPhysiciansSurgPak.2006;16(6):404-7.379. Adanikin AI, Awoleke JO. Clinical suspicion, management and outcome ofintrapartum foetal distress in a public hospital with limited advanced foetalsurveillance.TheJournalofMaternal-Fetal&NeonatalMedicine.2017;30(4):424-9.380. GangwarR,ChaudharyS.CaesareanSectionforFoetalDistressandCorrelationwith Perinatal Outcome. The Journal of Obstetrics and Gynecology of India.2016;66(1):177-80.381. Perveen F, Khan A, Ali T, Rabia S. Umbilical Cord Blood pH in IntrapartumHypoxia.JCollPhysiciansSurgPak.2015;25(9):667-70.382. Ayres-de-Campos D. Introduction: why is intrapartum foetal monitoringnecessary–impact on outcomes and interventions. Best Practice & Research ClinicalObstetrics&Gynaecology.2016;30:3-8.383. ParveenS.Umbilicalcordarterialbloodbaseexcessasgoldstandardforfoetalwell being screening test validity at termdelivery. JPMAThe Journal of thePakistanMedicalAssociation.2010;60(5):347.384. Tasnim N, Mahmud G, Akram S. Predictive accuracy of intrapartumcardiotocographyintermsoffetalacidbasestatusatbirth.JCollPhysiciansSurgPak.2009;19(10):632-5.385. LawrenceER,QuarshieEL,LewisKF,PeltzmanT,AndersonFJ. IntroductionofcardiotocographmonitoringimprovesbirthoutcomesinwomenwithpreeclampsiainGhana.InternationalJournalofGynecology&Obstetrics.2016;132(1):103-4.386. Katsuragi S, Parer JT, Noda S, Onishi J, Kikuchi H, Ikeda T. Mechanism ofreduction of newborn metabolic acidemia following application of a rule-based 5-category color-coded fetal heart rate management framework. The Journal ofMaternal-Fetal&NeonatalMedicine.2015;28(13):1608-13.

129

387. GeidamAD,AuduBM, KawuwaBM,Obed JY. Rising trend and indications ofcaesarean section at the university ofMaiduguri teaching hospital, Nigeria.Ann AfrMed.2009;8(2).388. Moyer CA, Kolars CK, Oppong SA, Bakari A, Bell A, Busingye P. Predictors ofstillbirths and neonatal deaths in rural western Uganda. Int J Gynaecol Obstet.2016;134(2):190-3.389. LawrenceER,QuarshieEL,LewisKF,PeltzmanT,AndersonFJ. IntroductionofcardiotocographmonitoringimprovesbirthoutcomesinwomenwithpreeclampsiainGhana. International journal of gynaecology and obstetrics: the official organ of theInternationalFederationofGynaecologyandObstetrics.2016;132(1):103-4.390. MartisR,EmiliaO,NurdiatiDS,BrownJ.Intermittentauscultation(IA)offetalheartrateinlabourforfetalwell‐being.TheCochraneLibrary.2017.391. ErsdalHL,MdumaE,SvensenE,SundbyJ,PerlmanJ.Intermittentdetectionoffetalheartrateabnormalitiesidentifyinfantsatgreatestriskforfreshstillbirths,birthasphyxia, neonatal resuscitation, and early neonatal deaths in a limited-resourcesetting: a prospective descriptive observational study at Haydom Lutheran Hospital.Neonatology.2012;102(3):235-42.392. Adanikin AI, Awoleke JO. Clinical suspicion, management and outcome ofintrapartum foetal distress in a public hospital with limited advanced foetalsurveillance.Thejournalofmaternal-fetal&neonatalmedicine:theofficialjournalofthe European Association of PerinatalMedicine, the Federation of Asia andOceaniaPerinatalSocieties,theInternationalSocietyofPerinatalObstet.2017;30(4):424-9.393. Tomlinson J, LucasD.Decision-to-delivery interval: Is30min themagic time?Whatistheevidence?Doesitwork?BestPractice&ResearchClinicalAnaesthesiology.2017;31(1):49-56.394. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Benefitsof introducinguniversalumbilicalcordbloodgasandlactateanalysis intoanobstetricunit.AustNZJObstetGynaecol.2010;50(4):318-28.395. Norén H, Carlsson A. Reduced prevalence of metabolic acidosis at birth: ananalysisof establishedSTANusage in the total populationofdeliveries in a Swedishdistricthospital.AmJObstetGynecol.2010;202(6):546.e1-.e7.396. Maisonneuve E, Audibert F, Guilbaud L, Lathelize J, JousseM, Pierre F, et al.Riskfactorsforsevereneonatalacidosis.ObstetGynecol.2011;118(4):818-23.397. RacinetC,PeresseJF,RichaletG,CorneC,OuelletP.[NeonataleucapnicpHatbirth:Applicationinacohortof5392neonates].GynecolObstetFertil.2016;44(9):468-74.398. JonssonM,Agren J,Norden-LindebergS,OhlinA,HansonU.Suboptimalcareand metabolic acidemia is associated with neonatal encephalopathy but not withneonatalseizuresalone:apopulation-basedclinicalaudit.ActaObstetGynecolScand.2014;93(5):477-82.399. Williams KP, Singh A. The correlation of seizures in newborn infants withsignificant acidosis at birth with umbilical artery cord gas values. Obstet Gynecol.2002;100(3):557-60.400. HoganL, IngemarssonI,Thorngren-JerneckK,HerbstA.Howoftenisa low5-minApgarscoreintermnewbornsduetoasphyxia?EuropeanJournalofObstetrics&GynecologyandReproductiveBiology.2007;130(2):169-75.

130

401. LocatelliA,IncertiM,GhidiniA,GrecoM,VillaE,PaterliniG.Factorsassociatedwithumbilical artery acidemia in term infantswith lowApgar scores at 5min.Eur JObstetGynecolReprodBiol.2008;139(2):146-50.402. LavrijsenSW,UiterwaalCS,StigterRH,deVriesLS,VisserGH,GroenendaalF.Severe umbilical cord acidemia and neurological outcome in preterm and full-termneonates.BiolNeonate.2005;88(1):27-34.403. Ronsmans C, Cresswell JA, Goufodji S, Agbla S, Ganaba R, Assarag B, et al.CharacteristicsofneonatalnearmissinhospitalsinBenin,BurkinaFasoandMoroccoin2012-2013.TropMedIntHealth.2016;21(4):535-45.404. BonnaerensA,ThaensA,MesensT,VanHolsbekeC,DeJongeE,GyselaersW.Identificationofneonatalnearmissbysystematicscreeningformetabolicacidosisatbirth.Facts,views&visioninObGyn.2011;3(4):281.405. KerkhofsC,DeBruynC,MesensT,TheyskensC,VanhoestenbergheM,BruneelE, et al. Identification of peripartum near-miss for perinatal audit. Facts Views VisObgyn.2014;6(4):177-83.406. Jonsson M, Norden-Lindeberg S, Ostlund I, Hanson U. Metabolic acidosis atbirth and suboptimal care--illustration of the gap between knowledge and clinicalpractice.BJOG.2009;116(11):1453-60.407. MorganJL,CaseyBM,BloomSL,McIntireDD,LevenoKJ.MetabolicAcidemiainLiveBirthsat35WeeksofGestationorGreater.ObstetGynecol.2015;126(2):279-83.408. SabolBA,CaugheyAB.Acidemiainneonateswitha5-minuteApgarscoreof7orgreater-Whataretheoutcomes?AmJObstetGynecol.2016;215(4):486.e1-6.409. DaniC,BresciC,BertiE,LoriS,DiTommasoMR,PratesiS.Shorttermoutcomeof term newborns with unexpected umbilical cord arterial pH between 7.000 and7.100.EarlyHumDev.2013;89(12):1037-40.410. WibergN,KlausenTW,TyrbergT,NordströmL,Wiberg-ItzelE.Infantoutcomeat four years of age after intrapartum sampling of scalp blood lactate for fetalassessment.Acohortstudy.PLoSOne.2018;13(3):e0193887.411. NyamtemaAS,deJongAB,UrassaDP,vanRoosmalenJ.Usingaudittoenhancequality of maternity care in resource limited countries: lessons learnt from ruralTanzania.BMCPregnancyChildbirth.2011;11(1):94.412. BohrenMA,TitiloyeMA,KyaddondoD,HunterEC,OladapoOT,TunçalpÖ,etal. Defining quality of care during childbirth from the perspectives of Nigerian andUgandanwomen:Aqualitativestudy.InternationalJournalofGynecology&Obstetrics.2017;139(S1):4-16.413. VandenBroekN,GrahamW.Qualityofcareformaternalandnewbornhealth:theneglectedagenda.BJOG.2009;116(s1):18-21.414. OladapoOT,SouzaJP,BohrenMA,TunçalpÖ,VogelJP,FawoleB,etal.WHOBetterOutcomesinLabourDifficulty(BOLD)project: innovatingtoimprovequalityofcarearoundthetimeofchildbirth.Reproductivehealth.2015;12(1):48.415. LawnJE,BlencoweH,OzaS,YouD,LeeAC,WaiswaP,etal.EveryNewborn:progress, priorities, and potential beyond survival. The Lancet. 2014;384(9938):189-205.416. BohrenMA,OladapoOT,TunçalpÖ,WendlandM,VogelJP,TikkanenM,etal.Formative research and development of innovative tools for “Better Outcomes inLabourDifficulty”(BOLD):studyprotocol.Reproductivehealth.2015;12(1):50.

131

417. Adeniran A, Likaka A, Knutsson A, Costello A, Daelmans B, Maliqi B, et al.Leadership, action, learning and accountability to deliver quality care for women,newbornsandchildren.BullWorldHealthOrgan.2018;96(3):222.418. WilundaC,PutotoG,DallaRivaD,ManentiF,AtzoriA,CaliaF,etal.Assessingcoverage,equityandqualitygapsinmaternalandneonatalcareinsub-saharanAfrica:anintegratedapproach.PLoSOne.2015;10(5):e0127827.419. TraversCP,CarloWA.HowtoSave1MillionLivesinaYearinLow-andMiddle-IncomeCountries.Neonatology.2017;111(4):431-6.420. Group SAEDCW. Every death counts: use ofmortality audit data for decisionmakingtosavethelivesofmothers,babies,andchildreninSouthAfrica.TheLancet.2008;371(9620):1294-304.421. Pirkle CM, Dumont A, Traore M, Zunzunegui M-V. Validity and reliability ofcriterionbasedclinicalaudit toassessobstetricalqualityofcare inWestAfrica.BMCPregnancyChildbirth.2012;12(1):118.422. LindmarkG,Langhoff-RoosJ,editors.Regionalqualityassessment inperinatalcare.SeminNeonatol;2004:Elsevier.423. Mancey-Jones M, Brugha RF. Using perinatal audit to promote change: areview.HealthPolicyPlan.1997;12(3):183-92.424. Nakibuuka V, Okong P,Waiswa P, Byaruhanga R. Perinatal death audits in aperi-urbanhospitalinKampala,Uganda.AfrHealthSci.2012;12(4):435-42.425. YoungP,HamiltonR,HodgettS,MossM,RigbyC,JonesP,etal.Reducingriskbyimprovingstandardsofintrapartumfetalcare.JRSocMed.2001;94(5):226-31.426. Mant J. Process versus outcome indicators in the assessment of quality ofhealthcare.IntJQualHealthCare.2001;13(6):475-80.427. TunçalpÖ,HindinMJ,Adu‐BonsaffohK,AdanuRM.Assessmentofmaternalnear‐miss and quality of care in a hospital‐ based study in Accra, Ghana.InternationalJournalofGynecology&Obstetrics.2013;123(1):58-63.428. BuchmannE.Towardsgreatereffectivenessofperinataldeathaudit in low‐andmiddle‐incomecountries.BJOG.2014;121:134-6.429. DonabedianA.Evaluating thequalityofmedical care.TheMilbankQuarterly.2005;83(4):691-729.430. Merali HS, Lipsitz S, HeveloneN, Gawande AA, Lashoher A, Agrawal P, et al.Audit-identified avoidable factors in maternal and perinatal deaths in low resourcesettings:asystematicreview.BMCPregnancyChildbirth.2014;14(1):280.431. Browne JL, vanNievelt SW, Srofenyoh EK,GrobbeeDE, Klipstein-GrobuschK.Criteria-based audit of quality of care to women with severe pre-eclampsia andeclampsiainareferralhospitalinAccra,Ghana.PLoSOne.2015;10(4):e0125749.432. ElAminS,Langhoff-Roos J,BödkerB,BakrAA,AshmeigAL, IbrahimSA,etal.Introducing qualitative perinatal audit in a tertiary hospital in Sudan. Health PolicyPlan.2002;17(3):296-303.433. KodkanyBS,DermanRJ,HonnungarNV,TyagiNK,GoudarSS,MastiholiSC,etal. Establishment of aMaternal Newborn Health Registry in the BelgaumDistrict ofKarnataka,India.ReprodHealth.2015;12Suppl2:S3.434. GoudarSS,StolkaKB,Koso-ThomasM,HonnungarNV,MastiholiSC,RamadurgUY, et al. Data quality monitoring and performance metrics of a prospective,population-basedobservationalstudyofmaternalandnewbornhealthinlowresourcesettings.ReprodHealth.2015;12Suppl2:S2.

132

435. ShresthaS,SharmaA,UpadhyayS,RijalP.Perinatalmortalityaudit.NepalMedCollJ.2010;12(4):257-9.436. Barbuscia DM, Rodrigues-Júnior AL. Completeness of data on live birthcertificatesanddeath certificates forearlyneonatal and fetaldeaths in theRibeirãoPretoRegion,SãoPauloState,Brazil,2000-2007.CadSaudePublica.2011;27(6):1192-200.437. Almeida MFd, Alencar GP, Schoeps D, Minuci EG, Silva ZPd, Ortiz LP, et al.Quality of information registered on fetal deaths certificates in São Paulo,SoutheasternBrazil.RevSaudePublica.2011;45(5):845-53.438. PhillipsJ,MillumJ.Valuingstillbirths.Bioethics.2015;29(6):413-23.439. KidantoHL,MogrenI,vanRoosmalenJ,ThomasAN,MassaweSN,NystromL,etal. IntroductionofaqualitativeperinatalauditatMuhimbiliNationalHospital,DaresSalaam,Tanzania.BMCPregnancyChildbirth.2009;9(1):45.440. PattinsonRC,SayL,MakinJ,BastosMH.Criticalincidentauditandfeedbacktoimproveperinatalandmaternalmortalityandmorbidity.TheCochraneLibrary.2005.441. Kerber KJ,MathaiM, Lewis G, Flenady V, Erwich JJ, Segun T, et al. Countingeverystillbirthandneonataldeaththroughmortalityaudittoimprovequalityofcarefor every pregnantwoman and her baby.BMC Pregnancy Childbirth. 2015;15 Suppl2:S9.442. PattinsonRC,SayL,MakinJ,BastosMH.Criticalincidentauditandfeedbacktoimprove perinatal andmaternalmortality andmorbidity. The Cochrane database ofsystematicreviews.2005(4):CD002961.443. Raman S, Iljadica A, Gyaneshwar R, Taito R, Fong J. Improvingmaternal andchildhealthsystems inFiji throughaperinatalmortalityaudit. Int JGynaecolObstet.2015;129(2):165-8.444. WilkinsonD.Reducingperinatalmortalityindevelopingcountries.HealthPolicyPlan.1997;12(2):161-5.445. Stratulat P, CurteanuA, Caraus T, PetrovV,Gardosi J. The experience of theimplementationofperinatalauditinMoldova.BJOG.2014;121Suppl4:167-71.446. SpitzerRF,SteeleSJ,CaloiaD,ThorneJ,BockingAD,Christoffersen‐DebA,etal.One‐yearevaluationof the impactofanemergencyobstetricandneonatalcaretrainingprograminWesternKenya. InternationalJournalofGynecology&Obstetrics.2014;127(2):189-93.447. BurkeL,SuswardanyDL,MichenerK,MazurkiS,AdairT,ElmiyatiC,etal.Utilityof local health registers in measuring perinatal mortality: a case study in ruralIndonesia.BMCPregnancyChildbirth.2011;11(1):20.448. PashaO, SaleemS,Ali S,Goudar SS,GarcesA, Esamai F, et al.Maternal andnewbornoutcomesinPakistancomparedtootherlowandmiddleincomecountriesintheGlobalNetwork'sMaternalNewbornHealthRegistry:anactive,community-based,pregnancysurveillancemechanism.ReprodHealth.2015;12Suppl2:S15.449. Drife JO,editorPerinatalaudit in low-andhigh-incomecountries.Seminars inFetalandNeonatalMedicine;2006:Elsevier.450. NyamtemaAS,UrassaDP,PembeAB,KisangaF,vanRoosmalenJ.Factors forchange inmaternalandperinatalauditsystems inDaresSalaamhospitals,Tanzania.BMCPregnancyChildbirth.2010;10(1):29.

133

451. MusafiliA,PerssonLA,BaribwiraC,PafsJ,MulindwaPA,EssenB.CasereviewofperinataldeathsathospitalsinKigali,Rwanda:perinatalauditwithapplicationofathree-delaysanalysis.BMCPregnancyChildbirth.2017;17(1):85.452. Goodman DM, Ramaswamy R, Jeuland M, Srofenyoh EK, Engmann CM,OlufolabiAJ,etal.ThecosteffectivenessofaqualityimprovementprogramtoreducematernalandfetalmortalityinaregionalreferralhospitalinAccra,Ghana.PLoSOne.2017;12(7):e0180929.453. FlenadyV,WojcieszekAM,EllwoodD,LeisherSH,ErwichJJH,DraperES,etal.,editors.Classificationof causesandassociatedconditions for stillbirthsandneonataldeaths.SeminarsinFetalandNeonatalMedicine;2017:Elsevier.454. Belizán M, Bergh A-M, Cilliers C, Pattinson RC, Voce A. Stages of change: Aqualitative study on the implementation of a perinatal audit programme in SouthAfrica.BMCHealthServRes.2011;11(1):243.455. Tunçalp Ö, Souza J. Maternal near‐miss audits to improve quality of care.BJOG.2014;121(s4):102-4.456. Say L. Neonatal nearmiss: a potentially useful approach to assess quality ofnewborncare.JPediatr(RioJ).2010;86(1):1-2.457. Pileggi-Castro C, Camelo JS, Jr., Perdona GC, Mussi-Pinhata MM, Cecatti JG,MoriR,etal.Developmentofcriteriaforidentifyingneonatalnear-misscases:analysisoftwoWHOmulticountrycross-sectionalstudies.BJOG.2014;121Suppl1:110-8.458. NardelloDM,GuimaraesAM,BarretoID,GurgelRQ,RibeiroER,GoisCF.Fetalandneonataldeathsof childrenofpatientsclassifiedasnearmiss.RevBrasEnferm.2017;70(1):104-11.459. Chevalier MS, King CC, Ellington S, Wiener J, Kayira D, Chasela CS, et al.MaternalandneonataloutcomesamongwomenwithHIVinfectionandtheirinfantsinMalawi.IntJGynaecolObstet.2017;137(3):282-9.460. ChweneyagaeD,Delis-JarrosayN,FarinaZ,FawcusS,GodiN,KhaoleN,etal.The impact of HIV infection on maternal deaths in South Africa. S Afr J ObstetGynaecol.2012;18(3):70-6.461. WediCO,KirtleyS,HopewellS,CorriganR,KennedySH,HemelaarJ.Perinataloutcomes associated with maternal HIV infection: a systematic review and meta-analysis.LancetHIV.2016;3(1):e33-48.462. ChettyT,VandormaelA,ThorneC,CoutsoudisA.IncidentHIVduringpregnancyand early postpartum period: a population-based cohort study in a rural area inKwaZulu-Natal,SouthAfrica.BMCPregnancyChildbirth.2017;17(1):248.463. PattinsonR,HulsbergenM,VanHoorickL.TheeffectofmaternalHIVinfectiononmaternal conditions and perinatal deaths in southwest Tshwane. Facts, views &visioninObGyn.2010;2(4):227.

134

CHAPTERTHREE:THESISOUTLINEANDMETHODOLOGYOFTHE

RESEARCH

135

THESISOUTLINE

This thesis is centredon twodata set analyses, a systematic reviewand two clinical

trials.

Bothclinicaltrialswereconductedin2014.Thefirstofwhichwasabeforeandafter

trial, conducted at KalafongHospital, Pretoria, South Africa (Image 3.1). The second

was an implementation research trial, conducted at Zithulele Hospital, South Africa.

Thedesignofeachofthesestudiesisoutlinedbelow.Themethodsforthebeforeand

afterstudyarerelevantforthemanuscriptsinChapters7-9,andthemethodsforthe

implementationtrialarerelevantforChapter10.

Image3.1:KalafongHospitalbynight,2014

136

Thedataforthesetrialsweresupportedbyalargescaleretrospectivecohortstudyof

thecausesofperinatalmortalityinSouthAfrica(Chapter4),asystematicreviewand

meta-analysisofthediagnostictestaccuracyoflactate(Chapter5),andtheimpactof

largescaleperinatalaudit(Chapter6).

137

OUTLINEOFTHERESEARCHINCHAPTERFOUR

The research relevant to this thesis was predominantly based in South Africa. As

outlined below, the clinical trials conducted were specific to two South African

hospitals.Thepotentialapplicabilityoftheresearchinthelongtermwasanimportant

consideration from thebeginningof this thesis. Therefore, itwas critical to consider

thecausesofperinatalmortalityinSouthAfricaasawhole.Themostaccuratemethod

ofexecutingthiswaswithananalysisofdatafromonespecificprovince,Mpumalanga.

Mpumalanga is the second smallest province in South Africa, covering an area of

76,495squarekilometres,withapopulationof4,039,939asatthelastcensus(2011).

Mpumalangaisuniqueinthatallpublicobstetricfacilitiesintheprovincecollectdata

on every birth and death using the Perinatal Problems Identification Program (PPIP)

(version3)asoutlinedinChapter2.

AIM

TodeterminethecausesofperinatalmortalityinSouthAfrica

138

METHODOLOGY

A retrospective analysis of the PPIP database from 1st October 2013 to 31st January

2014,inclusive,wasundertaken

INCLUSIONCRITERIA

- Late fetaldeaths (babiesweighing1000gror≥28weeksgestation)andearly

neonataldeaths(deathsupto7daysofneonatallife).

DATAEXTRACTED

- Foreachindividualdeath,thefollowingdatawereextracted:

o Maternalage

o Parity

o SyphilisandHIVstatus

o Gestationalage

o BirthweightforgestationusingcountryspecificTheroncharts

o Certaintyoruncertaintyofgestation

o Conditionofthenew-born(bornaliveandearlyneonataldeath(ENND),

freshstillborn,maceratedstillborn)

o Primaryobstetriccauseofdeath(allcases)

o Primaryneonatalcauseofdeath(ENNDs).

o Maternalconditionspresentatthetimeofdeath

139

o Avoidablefactorsidentifiedineachdeath

Weexaminedthefrequenciesofmaternalandfetalorneonatalcharacteristicsinlate

fetal deaths and analysed the relationship betweenmaternal condition and fetal or

neonatal outcomes. IBM SPSS Statistics 22.0 was used for data analysis. Frequency

distributionswereperformedforoutcomesummaries.Fisher’sexacttestonbivariate

correlationwasusedtointerrogatetherelationshipbetweenmaternalconditionsand

obstetriccausesofdeathandthetimingoflateperinataldeath.Analysisoutcomesare

summarisedusing risk ratios and95%confidence intervals.All testswere two tailed

andp-values<0.05wereconsideredsignificant.

ETHICS

ThePPIPprogramhasethical approval from theUniversityofPretoria. Thedataare

collected with permission from the South African Department of Health. All public

obstetricfacilitiesinSouthAfricahavebeenmandatedtousethePPIPprogramsince

2012.All data is deidentifiedprior to entry in thePPIPdatabase. ThePPIP technical

task teamapproved this secondaryanalysis.TheMpumalangaDepartmentofHealth

grantedpermissionforanalysisandpublicationoftheProvince’sPPIPdata.

140

OUTLINEOFTHERESEARCHINCHAPTERFIVE

Chapterfiveisasystematicreviewofthediagnostictestaccuracyofumbilicallactate

in the setting of acidosis and neonatal outcomes. The protocol for this review is

outlinedbelow.

OBJECTIVESOFTHEREVIEW

1. Tocompareumbilicallactatewithotherassessmentsofacidosis(pH,base

excess,pO2,pCO2)throughcorrelationforthemeasuringtheoutcomeof

acidosis

2. Tocomparethesensitivityandspecificityofumbilicalarterylactatewithother

assessmentsofacidosisinpredictingneonataloutcomes

CRITERIAFORCONSIDERINGSTUDIESINTHEREVIEW

GENERALINCLUSION/EXCLUSIONCRITERIA

All identified studies, irrespective of language, published orwritten from1990-2013

willbeassessed.

Exclusioncriteria:

141

- Studiesoutsidetheinclusiondates

- Studieswithnodataavailable

TYPEOFSTUDYDESIGNSFORDIAGNOSTICTESTACCURACY

Theidealstudydesignfordiagnostictestperformanceiscross-sectionalwherethetest

isperformedonconsecutivelyassessedpatientsiscross-classifiedwithdiseasestatus

andareferencetest.Althoughlessideal,randomisedcontroltrialsthatusepreviously

independentlyassessedtestswillalsobeconsidered.

TYPEOFPARTICIPANTS

Women delivering in a facility with the ability to undertake an assessment of fetal

acidosisandneonatalriskandneonatesintheimmediatepost-deliveryperiod

TYPEOFOUTCOMES

Acidosis (defined by the study being assessed); typically pH <7.2 or <7.15, BE <15,

p02<10,pCO2>70(1)

Neonatal risk / outcome (defined by the study); potentially APGAR scores <7 at 5

minutes, admission to the neonatal intensive care nursery, need for neonatal

resuscitation,diagnosisofhypoxicischaemicencephalopathy

142

SEARCHSTRATEGYFORIDENTIFICATIONOFSTUDIES

Thereviewwillhaveno languagerestrictions.ReviewerswillbeabletoreadEnglish,

IndonesianandFrench.AnassociatedresearcherwillreadreviewsinTurkish.Reports

inotherlanguageswillbetranslatedintoEnglish.Thereviewwillcoverpublishedand

unpublishedstudiesfrom1990-2013

Electronic databases that will be search are: PubMed/Medline, CINAHL, EMBASE.

Search strategieswill be customised to the subject heading and search structure of

individualdatabases.

Otherelectronicsearching:Aninternetsearchusing“GoogleScholar”searchingengine

withtheterms“lactate”,“fetalacidosis”,“predict*neonatalrisk”

Handsearching:Referencelistsfromreviewarticleswillbesearched.

Initially a comprehensive search to identify existing systematic reviews was

undertakenusingtheHuntandMcKibbonmethod

1. Meta-analysis(pt)

2. Meta-anal:(tw)

3. Metaanal:(tw)

4. Quantitative:review:ORquantitative:overview:(tw)

5. Systematic:review:ORsystematic:overview:(tw)

6. Methodological:review:ORmethodologic:overview:(tw)

7. Review(pt)ANDmedline(tw)

8. 1OR2OR3OR4OR5OR6OR7

143

Contenttermswerethenaddedwithmedlineexplosion

TW: fetal; neonatal; risk; lactate; intrapartum; umbilical; acidosis; asphyxia; labour;

fetal;cord;lacticacid;outcome

Contentterms:fetal;neonatal;risk;lactate;intrapartum;umbilical;acidosis;asphyxia;

outcomes

Usingthismethod,therewerenorelevantexistingsystematicreviewsidentified

CUSTOMISEDSEARCHSTRATEGY:

PubMed/Medlinesearchstrategy:

The lists the MeSH Terms (Medline Subject Headings) and keywords for the three

concepts:UmbilicalArtery;Lactate;NeonatalRisk/Outcome

Notes:

Removing“FetalBlood”anditsvariantsremovesmanyresultsrelatingto“fetalscalp

blood”.Although“FetalBlood”isthepreferredMESHtermforCordBloodorUmbilical

CordBlood

AddingthefollowingsearchstringwithinPubMedremovesreferencesthatarereferto

animalsandnothumans.

NOT(“animals”[MESHTerms]NOT“humans”[MESHTerms])

ConceptswithMESHTermsandKeywords

144

UmbilicalArteryConcept

"umbilicalarteries"[MeSHTerms]

“umbilicalartery”

“umbilicalarteries”

"umbilicalcord"[MeSHTerms]

"umbilicalcordblood"

“umbilicalcordartery”

“cordblood”

"fetalblood"[MeSHTerms]

"fetalblood"

“foetalblood”

("umbilical arteries"[MeSH Terms] OR “umbilical artery” OR “umbilical arteries” OR

"umbilicalcord"[MeSHTerms]OR"umbilicalcordblood"OR“umbilicalcordartery”OR

“cordblood”OR"fetalblood"[MeSHTerms]OR"fetalblood"OR“foetalblood”)

WithoutFetalBlood

("umbilical arteries"[MeSH Terms] OR “umbilical artery” OR “umbilical arteries” OR

"umbilicalcord"[MeSHTerms]OR"umbilicalcordblood"OR“umbilicalcordartery”OR

“cordblood”)

145

LactateConcept

"lacticacid"[MeSHTerms]

"lacticacid"

"lactates"[MeSHTerms]

lactate

lactates

"acidosis,lactic"[MESHTerms]

“lacticacidosis”

("lacticacid"[MeSHTerms]OR"lacticacid"OR"lactates"[MeSHTerms]ORlactateOR

lactatesOR"acidosis,lactic"[MESHTerms]OR“lacticacidosis”)

NeonatalRisk/OutcomeConcept

"neonataloutcome"

"neonataloutcomes”

(("infant,newborn"[MESHTerms]ORneonat*)AND"riskfactors"[MESHTerms])

“FetalHypoxia”[MESH]

“fetalhypoxia”

146

“foetalhypoxia”

AsphyxiaNeonatorum[MESH]

“birthasphyxia”

"fetalasphyxia"

"foetalasphyxia"

“intrapartumhypoxia”

"intrapartumfetalasphyxia"

“intrapartumstillbirths”

“intrapartum-linkedneonataldeaths”

"PerinatalMortality"[MeSHTerms](2008+)

“perinatalmortality”

“perinatalmorbidity”

“Neonatalnearmiss”

“InfantMortality[MESH]”(UsedforPerinatalMortality1990-2007)

“Neonatalmortality”

"APGARscore"[MeSHTerms]

“APGARscore”

"fetaldistress"[MeSHTerms]

147

“fetaldistress”

“foetaldistress”

"Infant,newborn,diseases/mortality"[MeshTerms]

"RespiratoryDistressSyndrome,Newborn"[MeSHTerms]

("neonatal outcome"OR "neonatal outcomes"OR (("infant, newborn"[MESH Terms]

OR neonat*) AND "risk factors"[MESH Terms]) OR “Fetal Hypoxia”[MESH Terms]OR

“fetal hypoxia” OR “foetal hypoxia” OR “Asphyxia Neonatorum”[MESH Terms] OR

“birthasphyxia”OR"fetalasphyxia"OR"foetalasphyxia"OR“intrapartumhypoxia”OR

"intrapartum fetal asphyxia" OR “intrapartum stillbirths” OR “intrapartum-linked

neonataldeaths”OR"PerinatalMortality"[MeSHTerms]OR“perinatalmortality”OR

“perinatalmorbidity”OR“Neonatalnearmiss”OR“InfantMortality”[MESHTerms]OR

“Neonatal mortality” OR "APGAR score"[MeSH Terms] OR “APGAR score” OR "fetal

distress"[MeSH Terms]OR “fetal distress”OR “foetal distress”OR "Infant, newborn,

diseases/mortality"[Mesh Terms] OR "Respiratory Distress Syndrome,

Newborn"[MeSHTerms])

DescriptionsofMeSHTerms

MeSHHEADING:UMBILICALARTERIES

SCOPE: Specialized arterial vessels in the umbilical cord. They carry waste and

deoxygenated blood from the FETUS to the mother via the PLACENTA. In humans,

thereareusuallytwoumbilicalarteriesbutsometimesone.

148

YEARofENTRY:68(64)

PREVIOUSINDEXING:UmbilicalCord/BS(1966-1967)

UsedFor:

umbilicalartery

arteryumbilical

umbilicalarteries

arteriesumbilical

MeSHHEADING:UMBILICALCORD

SCOPE: The flexible rope-like structure that connects a developing FETUS to the

PLACENTA in mammals. The cord contains blood vessels, which carry oxygen and

nutrientsfromthemothertothefetusandwasteproductsawayfromthefetus.

NOTE:CORDBLOODseeFETALBLOODisavailable

UsedFor:

cordumbilical

umbilicalcords

umbilicalcord

cordsumbilical

MeSHHEADING:FETALBLOOD

SCOPE: Blood of the fetus. Exchange of nutrients andwaste between the fetal and

maternal blood occurs via the PLACENTA. The cord blood is blood contained in the

umbilicalvessels(UMBILICALCORD)atthetimeofdelivery.

149

YEARofENTRY:75

PREVIOUSINDEXING:UmbilicalCord(1966-1974)

UsedFor:

bloodsfetal

cordblood

bloodfetal

cordbloodumbilical

bloodscord

fetalbloods

bloodcord

umbilicalcordbloods

umbilicalcordblood

cordbloodsumbilical

cordbloods

bloodsumbilicalcord

bloodumbilicalcord

fetalblood

MeSHHEADING:LACTATES

SCOPE:SaltsorestersofLACTICACIDcontainingthegeneralformulaCH3CHOHCOOR.

NOTE:donotconfusewithLACTATEseeLACTICACID

UsedFor:

lactates

150

MeSHHEADING:LACTICACID

SCOPE:Anormal intermediate in the fermentation (oxidation,metabolism)of sugar.

The concentrated form is used internally to prevent gastrointestinal fermentation.

(FromStedman,26thed)

NOTE:donotconfuseXrefLACTATEwithLACTATES

YEARofENTRY:97;wasLACTICACID(NM)1981-96

SEARCHNOTE:useLACTICACID(NM)tosearchLACTICACID1981-96

PREVIOUSINDEXING:Lactates(1981-1996)

MeSHHEADING:ACIDOSIS,LACTIC

SCOPE: Acidosis caused by accumulation of lactic acid more rapidly than it can be

metabolized. It may occur spontaneously or in association with diseases such as

diabetesmellitus,leukemia,orliverfailure.

NOTE:accumulationoflacticacid

YEARofENTRY:87

PREVIOUSINDEXING:Acidosis(1966-1986);Lactates(1966-1986)

UsedFor:

acidosislactic

lacticacidosis

MeSHHEADING:ASPHYXIANEONATORUM

SCOPE:Respiratoryfailureinthenewborn.(Dorland,27thed)

151

NOTE: do not use /congen & do not coord with INFANT, NEWBORN, DISEASES; tag

INFANT,NEWBORN

REFERENCES:

UsedFor:

asphyxianeonatorum

MeSHHEADING:FETALHYPOXIA

SCOPE:DeficientoxygenationofFETALBLOOD.

NOTE:donotconfusewithASPHYXIANEONATORUM

YEARofENTRY:2006(1975)

PREVIOUSINDEXING:FetalDiseases(1966-1974)

UsedFor:

hypoxiafetal

fetalanoxia

anoxiafetal

fetalhypoxia

MeSHHEADING:INFANTMORTALITY

SCOPE: Postnatal deaths from BIRTH to 365 days after birth in a given population.

Postneonatalmortalityrepresentsdeathsbetween28daysand365daysafterbirth(as

definedbyNationalCenterforHealthStatistics).Neonatalmortalityrepresentsdeaths

frombirthto27daysafterbirth.

UsedFor:

152

infantmortality

mortalitiesinfant

neonatalmortalities

infantmortalities

neonatalmortality

mortalityneonatal

postneonatalmortality

mortalityinfant

mortalitiesneonatal

MeSH HEADING: PERINATAL MORTALITY (Note: Used from 2008+; Infant Mortality

usedpriortothis)

SCOPE:Deathsoccurringfromthe28thweekofGESTATIONtothe7thdayafterbirth.

NOTE: a statistical concept and not a substitute for /mortal with FETAL DISEASES;

INFANT,NEWBORN,DISEASESorspecificdisease;specifygeographyifpertinent

YEARofENTRY:2008;useINFANTMORTALITY1990-2007

PREVIOUSINDEXING:InfantMortality(1965-2007)

REFERENCES:

SeeRelated:

INFANTMORTALITY

UsedFor:

mortalitiesperinatal

perinatalmortality

153

perinatalmortalities

mortalityperinatal

MeSHHEADING:FETALDISTRESS

SCOPE:Anonreassuringfetalstatus(NRFS) indicatingthattheFETUSiscompromised

(American College of Obstetricians and Gynecologists 1988). It can be identified by

sub-optimal values in FETAL HEART RATE; oxygenation of FETAL BLOOD; and other

parameters.

NOTE:donotconfusewithRESPIRATORYDISTRESSSYNDROME,NEWBORN(postnatal)

MeSHHEADING:RESPIRATORYDISTRESSSYNDROME,NEWBORN

SCOPE:AconditionofthenewbornmarkedbyDYSPNEAwithCYANOSIS,heraldedby

suchprodromalsignsasdilatationofthealaenasi,expiratorygrunt,andretractionof

the suprasternal notch or costal margins, mostly frequently occurring in premature

infants, childrenof diabeticmothers, and infants deliveredby cesarean section, and

sometimeswithnoapparentpredisposingcause.

Customisedsearchstrategy:

Embasesearchstrategy:

SubjectHeadings

ThefollowingareSubjectTermsusedwithinEmbase:

MAINTERM:umbilicalartery

154

UsedFor:

allantoicartery

arteriaumbilicalis

springleumbilicalartery

umbilicalarteries

umbilicalartery,persisting

umbilicalcordartery

umbilicusartery

MAINTERM:umbilicalcordblood

UsedFor:

blood,umbilical

cordblood

cordblood,umbilical

umbilicalblood

umbilicalcordfluid

umbilicalcordplasma

MAINTERM:LacticAcid

MAINTERM:lactatebloodlevel

UsedFor:

bloodlactate

bloodlacticacid

155

lactateplasmalevel

lactateserumlevel

lacticacidbloodlevel

lacticacidserumlevel

plasmalactate

plasmalacticacid

serumlactate

serumlacticacid

MAINTERM:lacticacidosis

UsedFor:

acidosis,lactic

acidosis,lacticacid

congenitallacticacidosis

lactacidosis

lactateacidosis

lacticacidacidosis

lacticacidosis,congenital

MAINTERM:fetusoutcome

UsedFor:

fetaloutcome

outcome,fetus

MAINTERM:perinatalmortality

156

UsedFor:

mortality,perinatal

perinataldeathrate

MAINTERM:newbornmortality

UsedFor:

mortality,neonatal

neonatalmortality

neonatalsurvival

neonatusmortality

newbornsurvival

postneonatalmortality

survival,newborn

MAINTERM:perinatalmorbidity

UsedFor:

morbidity,perinatal

perinataldisease

MAINTERM:fetusdistress

UsedFor:

distress,fetal

fetaldistress

foetaldistress

157

foetusdistress

intrauterinedistress

MAINTERM:perinatalasphyxia

UsedFor:

asphyxia,perinatal

birthasphyxia

MAINTERM:fetushypoxia

UsedFor:

fetalanoxia

fetalasphyxia

fetalhypoxemia

fetalhypoxia

fetusanoxia

fetusasphyxia

fetusasphyxiation

foetalhypoxia

foetushypoxia

hypoxia,fetus

intrauterineasphyxia

intrauterinehypoxia

MAINTERM:neonatalrespiratorydistresssyndrome

UsedFor:

158

neonatalrds

neonatalrespiratorydistress

newbornrespiratorydistress

newbornrespiratorydistresssyndrome

respiratorydistresssyndrome,newborn

respiratorydistress,newborn

METHODSOFTHEREVIEW

SCREENINGANDDATAEXTRACTIONFORMS

SCREENINGQUESTIONSFORABSTRACTANALYSIS:

SCREENINGQUESTIONS

No Yes

1 Does the report include data on umbilical artery lactate and another

methodofassessingfetalacidosis?⃝ □

2 Does the report compare the tests in either their ability to predict

acidosisortheirabilitytopredictpotentialneonatalrisk?⃝ □

3 Isthestudyacross-sectionaldesignorrandomisedcontroltrial? ⃝ □

159

4 Areacidosisand/orneonataloutcomedefined? ⃝ □

Note:Ifanyanswerfallsintoacirclethenthestudywillbeexcluded

DATAEXTRACTIONFORM:

DATAEXTRACTIONFORM

1 Studyidentificationnumber #

2 Nameofdataextractor

3 Dateofextraction __/__/__

4 Lastnameoffirstauthor

5 Countryoforigin

6 Published

7 Yearofpublication ____

8 Languageofpaper

9 Studyperiod __/__/__-__/__/__

10 Studydesign

11 Methodofsampling

160

12 Whattestsarebeingevaluated?

12 Numberofeligiblesubjects

13 Samplesizestudied

14 Whatisthestudysetting?

15 Description of the characteristics of the study

population

16 Description of the health characteristics of the

women

17 Description of the health characteristics of the

fetus

18 Is there information regarding the number of

non-validsamples

Yes/No

19 Is the incremental value of the test being

comparedtootherroutinetests?

Yes/No

20 Isthedecisiontoperformthereferencestandard

independentofthetestresult?

Yes/No

21 Ifnot,whatpercentwerenotverified __%

22 Wasthereavalidreferencestandard? Yes/No

23 Arethetestsandreferencesstandardsmeasured Yes/No

161

independently?

24 Are tests measured independently of other

clinicalandtestinformation?

Yes/No

25 If tests are being compared, have they been

assessed independently in the same patients or

doneinrandomlyallocatedpatients?

Yes/No

26 Whatistheprimaryoutcome?

27 Whattest(s)islactatecomparedto?

28 Whatistherbetweenlactateandthetest(s)?

29 What is the sensitivity and specificity of lactate

fortheoutcome?

30 What is the sensitivity and specificity of the

comparisonfortheprimaryoutcome?

MANAGEMENTOFSTUDIESANDDATAEXTRACTION

Endnote bibliographic software will be used to store the citations identified in the

search, keep track of them and delete duplicates. Electronic searches can be

downloadeddirectly in toEndnote.Reports fromothersourcescanbeentered in to

Endnote manually. Duplicates will be deleted and each study will be assigned an

identificationnumberforreview.

162

All citations identified fromtheelectronicsearchstrategieswillbeevaluated initially

withthescreeningformonthebasisoftitlesandabstracts,whereavailable.Records

notmeetingthescreeningcriteriawillbediscarded.Fulltextarticlesoftheremaining

citationswillbeobtained.Fulltextarticleswillalsobeobtainedforcitationswherethe

initial screeningcouldnotbecompletedbecauseof lackof information in the title /

abstract.Alistoftheirrelevantrecordswillbeavailableuponrequest.

The full text articleswill be evaluatedusing the screening form. Studies excludedat

thispoint in the reviewwillbe listed separately,with the reason for theirexclusion.

Data from the remaining studies will be extracted by reviewer EA using the data

extraction form.A secondand third reviewerwill be availablewhendata cannotbe

extracted from the first review and where there are queries about a manuscript.

Attempts will be made to contact authors where data is missing or requiring

clarification. Datawill bemanually entered in a database and processedwith SPSS

software.

CRITICALAPPRAISALOFSTUDIES

Includedstudieswillbe interrogated todeterminedifference in studycharacteristics

usingthedataextractionform.Therewill thenbeaqualityassessmentofstudiesby

threereviewersusingthefollowingcriteria:

- Highqualityifall4criteriaaremet

- Mediumqualityif2or3criteriaaremet

163

- Lowqualityif1ornocriteriaaremet

ADEQUATE INADEQUATE

Sampling Consecutive,random Selectgroups

Population Type and characteristic of

populationspecified

Notspecified

Followup Reportedand<20%loss Notreportedor>20%loss

Definitionofoutcome Definition of acidosis and

neonatal outcome

reported

Definition of acidosis and

neonatal outcome not

reported

Assessmentofbias:

- Hasselectionbiasbeenminimised?

o Haveconsecutivepatientsbeenenrolled?

o Studiesthatonlyincludedcaseswithconfirmedneonataloutcome(e.g.

HIE)arelikelytooverestimatesensitivityandspecificity

- Wasfollowupforprimaryoutcomesadequate?

o Wereallthoseenrolledshouldbeverifiedbythereferencestandard

(e.g.pH)

o Wereallthoseenrolledincludedintheanalysis?

o Wasthereferencestandardapplieddifferentlytothetestpositivesand

thetestnegatives?

164

- Applicability

o Howtestswereperformede.g.kitsfromdifferentmanufacturers?What

thresholdwasusedtodifferentiatepositivefromnegativetests?

- Hasmeasurementormisclassificationbiasbeenminimised?

o Wasavalidatedreferencestandardused?

o Iftestsarebeingcompared,havetheybeenassessedindependently?

o Werebothtestsdoneonthesameindividual?

ANALYSISPLAN

Twomethods of analysis are anticipated as there are likely twomodes of reported

data; those that compare lactate and other measures of acidosis (e.g. pH) using a

correlationcoefficient(r)andthosethatreportthesensitivityandspecificityoflactate

topredictneonataloutcomecomparedtoothermeasuresofacidosis.

Analysisofstudieswithcorrelationcoefficients

- Meta-analysisusingfixedandrandomeffectmodels(MonteCarlosimulation)

Analysisofstudieswithsensitivityandspecificity

- InitialsimpleplotofsensitivityandspecificitywithCI

- Thenasynthesisofstudyresultsforsensitivityandspecificitywillbedoneby

fittingasummaryROC(SROC)

o Truepositiverate(sensitivity)andthefalsepositiverate(1-specificity)

arefirsttransformedthroughthelogarithmoftheirodds

165

o Theregressionanalysisisdone

§ D=a+bS

§ D=(logitTPR)-(logitFPR)=log(oddsratio)

§ S=(logitTPR)+(logitFPR)=log(oddsproduct),whichisaproxyfor

thethreshold

§ a=estimatedlinearintercept

§ b=estimatedregressioncoefficient(slope)

§ RelationshipcanbeplottedwithDontheverticalaccessandS

onthehorizontal

o Theresultsareback-transformedandplottedinthestandardROC

format

Themanuscript of the reviewwill bewritten by EA. JanDickinson, CW and TWwill

reviewthedraftsandapprovethefinalmanuscript.

DESCRIPTIONOFSTUDIES

Atablewiththecharacteristicsoftheincludedstudieswillbepreparedcontainingthe

methods, participants, and outcomes studied. This tablewill show the studies listed

chronologicallybyyearofpublication,listedalphabeticallybythesurnameofthefirst

author.

166

OUTLINEOFTHERESEARCHINCHAPTERSIX

The South African PPIP database is a nationally supported program, which was

voluntaryfromthe1990s–2012andcompulsorythereafter,thatallowsprovidersof

perinatalcaretoenterandanalyseperinatalbirthsanddeathsandmaternaldeathsin

detail.Asof2014,94%ofhospitalcontributedatatoPPIPandmorethan73%ofbirths

inallinstitutionsarecaptured.

AIM

ToassesstheimpactoncontinuousauditusingPPIPonperinatalmortalityinallSouth

Africanobstetricunitswithatleastfiveyearsofcontinuousauditdata.

METHODOLOGY

- IdentifysitesfromthePPIPdatabasewithatleastfiveyearsofcontinuousaudit

(morethan80%compliancewithmonthlysubmissionofdata) fromthepoint

offirstusingthePPIPprogram.

- Undertake an analysis of trend in PNMR (defined as fetal deaths and early

neonatal deathsper1000births) forbirthswithweights greater than1000gr

overthefirstfiveyearsofauditinthesesites.

167

- AlldatawasanalysedusingEpiinfoversion7andSPSSversion22.Themonthly

perinatalmortalityrateforeachsiteforthefirstfiveyearswasobtained.The

datawassmoothedbyusingtwelvemonthmovingaverages.Achi-squaretest

forlineartrend(extendedMantel-Haenszel)wasundertakenoneachsite.AM2

statistic for each sitewas calculatedwith one degree of freedomwith linear

trend(eitherincreasingordecreasing)beingrelatedtothefirstPNMRratefor

eachsite (thebeginningofaudit)withtheconstraintbeingthat theratesare

monotonic.Ap-valueof<0.05wasconsideredsignificant.

- TofurtherinterrogatethereasonsforthedifferencesinPNMR,asubgroupof

siteswhoonlybeganauditingafter2006wasformedfromtheinitial163sites.

ETHICS

The PPIP program has ethical approval from the University of Pretoria. The data is

collected with permission from the South African Department of Health. The PPIP

technicaltaskteamapprovedthissecondaryanalysis.

ANALYSIS

All datawas analysedusing Epiinfo version 7 and SPSS version 22. The trendof the

PNMRusing themonthlyPNMR (defined forPPIPas fetaldeathsandearlyneonatal

deathsper1000births)overfiveyearswasanalysedbyperformingaChisquaretest

fortrendonthe12monthmovingaverages.

168

OUTLINEOFTHERESEARCHINCHAPTERSSEVEN,EIGHTANDNINE

BACKGROUND

ResearchintoUAlactatesamplinghasshownthatintroductionoftheproceduretoall

deliveries in a unit produces progressive improvement in neonatal biochemical

markersand reducedadmission tohigh care facilitiesover time, independentof the

rate of obstetric intervention(1). However, previous studies into implementation of

universal UA lactate sampling and its effect on neonatal outcomes have been

performed in resource-rich environments, usually in developed countries and

particularlyintertiaryandteachingfacilities(2).Thismainaimofthisprojectseeksto

identify the impact of implementation of cord lactate sampling into a SouthAfrican

hospital,withcomparativelylimitedresourcesandaPNMRof14.8/1000births.

STUDYOBJECTIVE

This study aimed to investigate whether UA cord lactate sampling correlates with

neonatal outcome at Kalafong Hospital, Pretoria, South Africa, and to determine

whether the use of lactate sampling and cardiotocograph (CTG) interpretation

training combinedwith education in lactate physiology as a functionof intrapartum

careresultsinachangeinclinicalpracticeand/orneonataloutcomes.

169

PRIMARY AIMS

1. To determine whether UA cord lactate sampling correlates with neonatal

outcomeatKalafongHospital.

2. To determine whether the introduction of UA cord lactate sampling at

KalafongHospital , with concomitant staff education inCTGinterpretation

and lactate physiologyresults inachange inclinicalpractice, including

a. Achange in instrumentaldeliveryrates

b. Achangeincaesareansectionrates

3. To determine whether the introduction of UA cord lactate sampling with

concomitant staff education in CTG interpretation and lactate physiology

resultsinanimprovementinlactatelevelsovertimeatKalafonghospital.

4. To determine whether the introduction of UA cord lactate levels with


results in a change in the rate of neonatal admission tohighdependencyor

intensivecareat Kalafonghospital.

5. To determine whether the introduction of UA cord lactate levels with


resultsinachangeintherateofneonatalmortalityatKalafong Hospital.

170

SECONDARY AIMS

1. To compare UA lactate levels in neonates of HIV positive and negative

womenat KalafongHospital.

2. To compare UA lactate levels in preterm and term neonates at Kalafong

hospital.

METHODOLOGY

This studywas conducted in two phases. Initially a cohort studywas undertaken to

determine whether UA cord lactate sampling correlates with neonatal outcome at

KalafongHospital. The results from this cohortwere servedas the control group for

thesecondphaseofthestudy,whichtestedtheinterventionofUAlactatesamplingin

combination with staff education in lactate physiology and CTG interpretation on

neonataloutcomes.

171

PHASE1

STUDYDESIGNANDSTUDYSETTING

During the period 3rd March 2014 to the 12th November 2014, we conducted a

prospectivecohortstudyofUAlactatelevelsinthelabourwardatKalafongHospital.

Kalafong Hospital is a tertiary hospital in Pretoria, South Africa, affiliated with the

UniversityofPretoria.In2014,therewere6,499deliveriesintheunit.

The study group consisted of prospectively collected samples of umbilical artery

lactateobtainedbytheprimaryresearcher(EA),oradoctorintheunittrainedbythe

primaryresearcher.TheUAwas identifiedbysite,andknowledgeoftheanatomyof

theumbilicalcord.AsUAsamplinghadnotbeenperformedintheunitbefore,theaim

of this cohort studywas to determine the optimal cut-off of lactate level for a pre-

definedsetofoutcomesintheunit.Theoutcomeschosenweretheneedforneonatal

resuscitation,neonataladmissiontothenursery,andApgarscoresat1and5minutes.

STUDYPOPULATION

Allwomenaimingtohaveavaginaldeliverywereeligibletobeconsentedforthetrial,

regardlessoftheirultimatemodeofdelivery.Womenofanygestationwhereaviable

172

outcomewasexpectedwereeligibletobeincludedinthestudy.AtKalafonghospital

thisincludesexpectedbirthweightsof600grormore.

STUDYPROCEDURESANDDATACOLLECTED

Aftermaternalconsentwasobtained(Appendix2.2),theumbilicalcordwasclamped

andcutwithinoneminuteofbirth.Asmallarterialbloodsample(upto1ml)wasthen

taken from a double clamped segment of the remaining umbilical cord, prior to

deliveryoftheplacenta.Thecordbloodsamplewastakenwithin15minutesofbirth,

regardlessofmodeofdelivery.The lactate levelwasmeasured(mmol/L)onaRoche

Accutrend PlusTM© hand-held lactate meter (Rotkreuz, Switzerland). The handheld

lactatemachinewasavailablewithinthebirthunitandthepatient’sdetailsandcord

lactateresultwasrecorded.Casereviewthenoccurredtocorrelatethepatientdetails

andcordlactateresultwiththeprimaryendpoints.

Introductionofanewprocessintoalowresourcesettingmandatesbotheaseofuse

andfinancialefficiency.Heparinisedsyringesareusedinmanysettingstoallowfora

lapseoftimebetweenobtainingandtestingthesample.Non-heparinisedsyringesare

howevermorefinanciallysoundandcanbeusedforUAsampleswithoutanychange

inthebiochemicalmeasurements,providedthesampleistestedshortlyaftertaking(3,

4).

173

Thecordlactateresultwaskeptblindtoclinicians(midwives,obstetricregistrarsand

obstetricconsultants)managing intrapartumcare,allowingtheentirestudygroupto

serveasthecontrolgroupforthesecondstageoftheproject.

Adatasheetwascompletedforeachcaseusingthebirthregistry,thepatient’scase

file,andtheneonatalnurseryadmissiondata (Appendix2.3).Thefollowingvariables

wererecorded:

DATACOLLECTED

The variables collected were defined by the clinicians in the unit. For example

gestation was taken as recorded (regardless of the method) or if a case was

documentedtohavepre-eclampsiathanthiswastakenasread.

1. Maternaldata

a. Maternalage

b. Gravidity, parity

c. Gestation

d. HIVstatus

e. HIVtreatment

174

f. Lengthoffirststage

g. Lengthofsecondstage

h. Complications of pregnancy

• Including:Abruption,adolescentpregnancy,advancedmaternalage,

anaemia, antepartum haemorrhage, breech presentation, cervical

incompetence, decreased fetal movements, eclampsia, gestational

diabetes mellitus, HELLP syndrome, hypertension (not pre-

eclampsia), intrauterine growth restriction, fetal macrosomia,

maternalmedicalconditionsnotrelatedtopregnancy,noantenatal

care, oligohydramnios, pre-eclampsia, placenta praevia, prelabour

and / or preterm rupture of membranes, syphilis, thyrotoxicosis,

tuberculosis, urinary tract infection, uterine anomaly, venous

thromboembolicevent

i. Complications of labour

j. Modeofbirth

k. Dateandtimeofbirth

l. Decisiontodeliverytime

2. Neonataldata

a. Birthweight

175

b. Cordlactateresult

c. Dateandtimeofcordlactatesample

d. Needforneonatalresuscitation

e. Apgarscore

f. Neonataloutcome

• Admissiontothenursery

• Diagnosisforadmission

• Hypoxic IschaemicEncephalopathy

• Deathbeforedischarge

SAMPLESIZECALCULATION

A sample size calculation was performed for both phase one and two. The

estimationof sample sizewasperformedusing standardisednormal lactatemean

and standard deviation. A sample size of 265 in each of the pre and post

intervention groups would achieve an 80% power to detect a difference of 0.25

standard deviations (approximately 8% difference in means) in lactate

measurementsbetweengroupsusinga two-sided sample t-test at5%significance

level.Thissamplesizewouldalsoachieveatleast80%powertodetectanR-squared

of 0.05 attributed to one or more independent variable/s while simultaneously

176

adjusting for multiple other relevant covariates with an R-squared of 0.2 in a

multiple linear regression analysis. Assuming a 15% attrition rate, the sample size

wasincreasedto305ineachgroupforatotalsamplesizeof610(PowerandSample

Size(PASS2008)).

PHASE TWO

The aim of this arm was to determine whether the introduction of the

interventions of sampling of UA lactate in addition to education in CTG

interpretation,fetalphysiology,andacidbasebalanceresultedinachangein

clinicalpracticeand/orneonataloutcomes.

STUDYDESIGNANDSTUDYSETTING

Thisisthesameasforphaseone

STUDYPOPULATION

AllmidwivesanddoctorsinvolvedinintrapartumcareatKalafongHospital.

177

Allwomenaimingtohaveavaginaldeliverywereeligibletobeconsentedforthetrial,

regardlessoftheirultimatemodeofdelivery.Womenofanygestationwhereaviable

outcomewasexpectedwereeligibletobeincludedinthestudy.AtKalafonghospital

thisincludesexpectedbirthweightsof600grormore.

STUDYPROCEDURESANDDATACOLLECTED

Beforethecompletionofthecohortoutlinedinphaseone,allmidwivesandregistrars

in the unit underwent a training course in fetal physiology, lactate physiology and

cardiotocograph(CTG)interpretation,conductedbytheprimaryresearcher(EA).

TEACHINGINTERVENTION

All staff members involved in CTG interpretation and intrapartum care

participated in the study and the teaching intervention. The teaching coursewas

adapted from the King Edward Memorial Hospital (KEMH) Advanced Fetal

Assessment(AFA) course resources (with permission).

Therewere18midwivesintheunit,managingnearly6500deliveriesperyear.Staffing

ratios and resources were therefore stretched and so the teaching intervention

needed to be delivered in one session to minimise the time away from work and

178

ensure allmidwiveswere trained.As such, the threepart teaching interventionwas

modified for themidwives to fit in toone90-minutesession.Conversely,47doctors

across the study period participated in the course, with the staff resource capacity

allowingfortheteachinginterventiontobedeliveredacrossthree90-minutesessions.

Theteachinginterventionwasdividedintofivecomponents

1.Baselinestaffdemographicsandpre-test

• Demographicsincluded

o Clinicalrole(Midwiferyornursingstudent,midwife,professionalnurse,

communityservicenurse)or(Intern,registrar)

o Previoustraining inCTGinterpretation(none,medicalschoolteaching,

ESMOEcourse)

Thedoctorscompleteda19questionandthemidwivescompleteda10questionpre-

teachingtesttoassessbaselineknowledge.

• Thepre-teachingtestassessed:

o Baseline confidence in CTG interpretation

o Baselineconfidenceinunderstandingfetalphysiology

o Baselineconfidenceininterpretationofcordlactatevalues

179

2.Fetalphysiology

The fetal physiology lecture outlined the history of fetal heart ratemonitoring, the

physiological control of the fetal heart, and the physiological mechanisms behind

specificfetalheartratepatterns.

3.Fetalacidbasestatus

The fetal acid base lecture outlined the fetal response to labour, fetal acid base

physiologyandnormal ranges forcomponentsofacidbase, thecomponentsofcord

acidbaseassessment,andtheroleoflactateinintrapartummonitoringandcare.

4.CTGpatternrecognition

TheCTGpattern recognition lecturegaveanoverviewofelectronic fetalmonitoring,

explained the differences between antenatal and intrapartum monitoring, outlined

CTG pattern recognition, and explored different clinical responses for the CTG

patterns.

5.Post-test

Thedoctorscompleteda33questionandthemidwivescompleteda14questionpost-

teaching test to assess acquisition of knowledge. Pre and post test scores were

comparedandanalysisofchangeinmedianswithstudentst-testperformed.

180

Atcompletionof theteachingintervention,eligiblewomenwhogaveconsenthadUA

lactatesamplestaken,withthesameprocessasdescribedforphaseone.From21stJuly

2014 to 12th November 2014, lactate sampling occurred, with the results made

immediatelyavailabletothecliniciansmanagingtheintrapartumcareofwomeninthe

labourward.Daily reviewof lactate results occurred everyweekdaymorning at the

department team meeting, involving all medical staff (interns, registrars, and

consultants).Alllactatesamplesobtainedintheprevious24hourswerereviewedand

cases with abnormal lactate levels were discussed, the CTG reviewed and the case

outcomes (including need for NICU admission and adverse neonatal outcome)

discussed. Clinical practice components of the cases where changes could be

considered were explored briefly. In addition, an in-depth review of interesting or

challenging cases occurred twice weekly, facilitated by the primary researcher (EA).

Casereflection,education,andoptionsforchanges infuturepracticewerediscussed

indetailatthesesessions.

DATACOLLECTED

Thisisthesameasforphaseone.

181

Image3.2AKalafonghospitalobstetricsregistrartestingaUAsample.

ETHICS

ThisstudywasapprovedbytheHumanEthicsResearchCommitteeofTheUniversity

ofWesternAustralia,Perth,WesternAustraliaonthe09/10/2014(Referencenumber

RA/4/1/6581),andtheUniversityofPretoria,Pretoria,SouthAfricaonthe01/10/2014

(Referencenumber7/2014).

182

OUTLINEOFTHERESEARCHINCHAPTERTEN

BACKGROUND

The hypothesis of the first trialwas thatUA lactate sampling as a clinical audit tool

wouldresultinanimprovementinseveralaspectsofintrapartumcare.Ifthisproved

to be the case, the tool needed to be applicable outside the research setting in an

LMICunit.Thusweaimedtoaddress the feasibilityof introducingthetool inclinical

practiceatZithuleleHospitalontheeasterncapeinSouthAfrica.

Implementingumbilicalcordsamplingintonon-teachingfacilitiesinSouthAfricasuch

asZithuleleHospitalispotentiallybeneficialbutlikelytoposeanumberofchallenges

whenconsideringthelimitedresourceswithinthesesettings.Whiteetalinterviewed

midwives and other professionals involved inmaternity care in a resource rich high

income country setting, and found that more than 50% considered universal cord

blood gas or lactate analysis to be difficult to introduce due to insufficient time

following delivery to perform the sample, increasedworkload and encroachment of

technology onto birth(5). Overall the majority of participants (67.3%) agreed that

umbilical cord blood gas/lactate analysis was beneficial to maternity care with

proposedbenefits beingobjectivemeasurementof neonatal status to assist in care,

objectiveprotectionformedico-legalissuesandimprovedopportunitiesforauditand

teaching.

183

STUDYOBJECTIVE

Todeterminefeasibilityofintroducinguniversalumbilicalarterycordlactatesampling

intoadistrictlevel,ruralhospitalinSouthAfrica.

PRIMARYAIMS

1. To Identify attitudes and barriers towards implementation of universal cord

lactatesampling.

2. ToimplementatrainingmodulecombiningCTGinterpretationandcordlactate

sampling to assess pre and post training confidence and competence of

midwiferyandmedicalstaff.

METHODOLOGY

Thisstudywasabeforeandafterimplementationstudy,consistingofabaselinestaff

survey, a training program in CTG interpretation, fetal physiology, and the practical

stepsandinterpretationofUAlactatesampling,andfinallyaposttrainingsurvey.The

studywasconductedfromNovember2014toJanuary2015.

184

STUDYSETTING

ZithuleleHospitalisaruraldistricthospitalintheEasternCapeofSouthAfrica.Despite

itsresourcepoorenvironmentitperformsbetterthanmostfacilitiesinitscategoryin

theareaofneonataloutcomes. Theperinatalmortality rate in2013was25.2/1000

births,ofwhich25.4%wasdueto intrapartumasphyxia.Monthlyperinatalmortality

meetings with case reflection are attended by midwives, doctors and hospital

managers.

INFORMEDCONSENT

Midwivesanddoctorsweregivenwritteninformationbeforetakingpartinthestudy.

Participationwas voluntary.Written informationwas also provided towomen from

whocordbloodwasextractedforthepurposeoftrainingandwrittenconsentwasbe

obtained(Appendix9.2).

PRE-INTERVENTION

Allmidwives and doctors were invited to consent to participate in a survey, as the

baselineforthestudy.

185

• Demographics

• PreviousexperiencewithCTG

• Previousexperiencewithcasereflection/qualityofcareaudit

• Attitudes towards perinatal audit, CTG monitoring and cord lactate blood

sampling

• Confidenceintheattainmentofcordlactateblood

• ConfidenceinthepotentialbenefitsofknowingUAlactatelevels

• Perceived barriers to implementation of CTG/lactate training and universal

measurementofcordlactatevalues

Questionnairesincluded20statementswithresponseoptionsofstronglyagree,agree,

neutral, disagree and strongly disagree. Baseline demographics were taken. See

proposedquestionnaireincluded,adaptedfromWhiteetal.,‘Attitudesandbarriersto

the introduction of umbilical cord blood gas and lactate analysis at birth’ (with

permission)(5).

INTERVENTION

We then performed a modified CTG training program, which explained lactate

physiologyandthebenefitsofmeasuringit,adaptedfromtheKingEdwardMemorial

HospitalAdvancedFetalAssessmentcourseresources(withpermission).Thisincluded

lecturesonfetalphysiology,fetalacidbasestatus,andCTGpatternrecognition(asper

themethodologyfor“Abeforeandafterstudyoftheimpactonobstetricandperinatal

186

outcomes following the introduction of an educational package of fetal heart rate

monitoringeducationcoupledwithumbilicalarterylactatesamplinginalowresource

settinglaborwardinSouthAfrica”).

A teaching session on the steps to sample the umbilical artery and test the lactate

formed the second componentof the intervention. Thiswasdoneonumbilical cord

post-delivery with patient consent for its use as a training tool. Teaching was

undertaken by an obstetric clinician who was trained in cord sampling in Western

Australiainaunitwherealldeliveriesundergopairedcordsamplingatbirth.Segments

of cord (usedwithmaternal permission)was used to demonstrate cord anatomy as

well as samplingof theartery.All participantswere taught and supervised sampling

theumbilicalartery.

POSTINTERVENTION

After the training program, an identical survey as in the pre-interventionphasewas

presented to midwifery and medical staff and pre and post intervention responses

compared.

187

CHAPTERFOUR:CAUSESOFPERINATALMORTALITYANDASSOCIATED

MATERNALCOMPLICATIONSINASOUTHAFRICANPROVINCE:

CHALLENGESINPREDICTINGPOOROUTCOMES

188

CHAPTEROVERVIEW

Asoutlined inChapter two, theburdenofglobalperinatalmortality liesheavilywith

the LMICs. Itwas important thatweascertained the causesofperinatalmortality in

SouthAfricainordertobeginconsideringthepotentialroleourinterventionmayplay.

Oneofthemajorchallenges inLMICisthepaucityofcomprehensivedatacollection,

whichcanbeusedto identifyareaswhereresourcescanbetargeted.SouthAfrica is

unique in its comprehensiveuseof thePPIPauditprogram,whichallows systematic

dataanalysisrelatedtoperinatalmortality.TheProvinceofMpumalangaisparticularly

notable in that facility-based delivery is almost universal, and every facility

comprehensively participates in QoC audit using PPIP. Furthermore, the units also

record thematernal condition associatedwith every perinatal death, allowing an in

depthanalysisofallaspectsofthecontributorstoperinatalmortalityintheProvince.

189

190

191

192

193

194

195

196

CHAPTERFIVE:UMBILICALLACTATEASAMEASUREOFACIDOSISAND

PREDICTOROFNEONATALRISK:ASYSTEMATICREVIEW

197

CHAPTEROVERVIEW

HavingascertainedinChapter3thatintrapartumasphyxiacontributesto25%oflate

stillbirthsandearlyneonataldeathsinSouthAfrica,itwasthenimportanttoconsider

the accuracy of the planned intervention. The gold standard assessment of acidosis

fromtheumbilicalarteryatbirthispH.However,pHassessmenttypicallynecessitates

abloodgasanalyser,whichusuallyrequiresupto90uLofblood.Thisisnotfeasiblein

low resources settings for the following reasons. Firstly the cost of obtaining and

maintainingabloodgasanalyser isprohibitive inmanyLMIC.Secondly,given lactate

can bemeasured using a hand-heldmeter with as little as 5uL of blood, there is a

markeddifference inconsumablecosts intermsofsyringes(6,7).Finally, theeaseof

collection in terms of human resources (training and time) renders lactate

measurementmorefeasibleinresourcesstretchedareas,suchasSouthAfrica.

That said, the decision to assess acidosis at delivery with lactate rather than pH

necessitatesconfidenceinthediagnosticaccuracyoflactateintheassessmentofthe

targetoutcome. Lactatehas repeatedlybeen shown to correlatewithumbilical cord

pH (the gold standard assessment of acidosis) and to have reasonable accuracy in

predictingpoorneonataloutcome(7-18).Therefore,thenextstepinthisthesiswasto

synthesise the evidence on the diagnostic test accuracy of umbilical cord lactate by

comparing it with other assessments of acidosis in the immediate newborn and by

assessingtheabilityofumbilicalcordlactatetopredictneonataloutcomes.

198

199

200

201

202

203

204

205

206

207

208

209

REFERENCESFORCHAPTERINTRODUCTION

1. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Benefitsof introducinguniversalumbilicalcordbloodgasandlactateanalysis intoanobstetricunit.AustNZJObstetGynaecol.2010;50(4):318-28.2. Allanson E,Waqar T,White C, Tunçalp Ö, Dickinson J. Umbilical lactate as ameasureofacidosisandpredictorofneonatalrisk:asystematicreview.BJOG.2016.3. DuerbeckNB,ChaffinDG,SeedsJW.ApracticalapproachtoumbilicalarterypHandbloodgasdeterminations.ObstetGynecol.1992;79(6):959-62.4. Westgate J, Garibaldi JM, Greene KR. Umbilical cord blood gas analysis atdelivery:atimeforqualitydata.BJOG.1994;101(12):1054-63.5. White CR, Kohan R, Doherty DA, Newnham JP, Pennell CE. Attitudes andbarriers to the introductionofumbilical cordbloodgasand lactateanalysis atbirth.AustNZJObstetGynaecol.2013;53(3):271-6.6. WestgrenM,KrugerK,EkS,GrunevaldC,KublickasM,NakaK,etal. Lactatecomparedwith pH analysis at fetal scalp blood sampling: a prospective randomisedstudy.BrJObstetGynaecol.1998;105(1):29-33.7. Labrecque L, Provencal M, Caqueret A, Wo BL, Bujold E, Lariviere F, et al.CorrelationofcordbloodpH,baseexcess,andlactateconcentrationmeasuredwithaportabledevice for identifying fetal acidosis. JObstetGynaecolCan. 2014;36(7):598-604.8. SuTY,ReeceM,ChuaSC.Lactatestudyusingumbilicalcordblood:agreementbetween Lactate Pro hand-held devices with blood gas analyser and evaluation oflactatestabilityovertime.AustNZJObstetGynaecol.2013;53(4):375-80.9. GjerrisAC,Staer-JensenJ,JorgensenJS,BergholtT,NickelsenC.Umbilicalcordbloodlactate:avaluabletoolintheassessmentoffetalmetabolicacidosis.EurJObstetGynecolReprodBiol.2008;139(1):16-20.10. Linet T, Laporte J, GueyeH, BoogG. [Microvolume dosage of lactate in cordblood for the evaluation of the neonatal well-being]. J Gynecol Obstet Biol Reprod(Paris).2002;31(4):352-7.11. MartinA,GaillardM,MiotS,RiethmullerD,SchaalJP.[Lactatemeasurementsandacid-basebalance incordblood].JGynecolObstetBiolReprod(Paris).2003;32(8Pt1):713-9.12. Racinet C, Richalet G, Corne C, Faure P, Peresse JF, Leverve X. [Diagnosis ofneonatal metabolic acidosis by eucapnic pH determination]. Gynecol Obstet Fertil.2013;41(9):485-92.13. TuuliMG,StoutMJ,ShanksA,OdiboAO,MaconesGA,CahillAG.Umbilicalcordarterial lactate comparedwith pH for predicting neonatalmorbidity at term.ObstetGynecol.2014;124(4):756-61.14. WestgrenM,DivonM,HoralM,IngemarssonI,KublickasM,ShimojoN,etal.Routinemeasurementsofumbilicalarterylactatelevels inthepredictionofperinataloutcome.AmJObstetGynecol.1995;173(5):1416-22.15. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Accuratepredictionofhypoxic-ischaemicencephalopathyatdelivery:acohortstudy.JMaternFetalNeonatalMed.2012;25(9):1653-9.

210

16. Wiberg N, Kallen K, Herbst A, Olofsson P. Relation between umbilical cordblood pH, base deficit, lactate, 5-minute Apgar score and development of hypoxicischemicencephalopathy.ActaObstetGynecolScand.2010;89(10):1263-9.17. WattWF,TanKH,YeoGSH.Umbilicalcord lactate:apreliminarystudyof130termbabies.SingaporeJournalofObstetricsandGynaecology.2002;33(2):42-8.18. HamedHO.Intrapartumfetalasphyxia:studyofumbilicalcordbloodlactateinrelationtofetalheartratepatterns.ArchGynecolObstet.2013;287(6):1067-73.

211

CHAPTERSIX:QUALITYOFCAREAUDITSANDPERINATALMORTALITY

INSOUTHAFRICA

212

CHAPTEROVERVIEW

Theaimofthisthesiswastoconsidertheroleof introducingumbilicalartery lactate

samplinginanLMIC.AsoutlinedinChapters1-3,theintentionwasforthesamplingto

serveasfeedbacktocliniciansprovidingcare,suchthatcasereflectioncanoccurand

ideallypracticealtered,ultimatelywiththegoalofimprovingclinicaloutcomes.Thus,

the intervention is essentially a quality of care audit intervention. Once we had

determined that it was possible to use UA lactate as a measure of acidosis, it was

importanttoconsidertherolethatQoCauditcanplay, ifany,onperinatalmortality.

Therefore, the following study was set up to retrospectively analyse the impact of

qualityofcareperinataldeathauditontheratesofperinatalmortalityinSouthAfrica,

andthisisreportedinthemanuscriptthatfollows.

213

214

215

216

217

218

CHAPTERSEVEN:THEINTRODUCTIONOFUMBILICALCORDLACTATE

MEASUREMENTANDASSOCIATEDNEONATALOUTCOMESINASOUTH

AFRICANTERTIARYHOSPITALLABOURWARD

219

CHAPTEROVERVIEW

Having established the contribution of intrapartumhypoxia to perinatalmortality in

South Africa, the diagnostic accuracy of umbilical artery lactate, and the potential

benefitofQoCauditinSouthAfrica,itwasthenimportanttoassesstheutilityofUA

lactatesamplinginoursetting.AstherearelittledataontheuseoflactateinLMIC,we

neededtoestablish itsutility inmeasuring intrapartumhypoxia inoursetting,rather

thanrelyingonthepublishedcut-offsforlactatelevelsfromhighincomesettings.

220

221

222

223

224

225

226

227

CHAPTEREIGHT:ABEFOREANDAFTERSTUDYOFTHEIMPACTON

OBSTETRICANDPERINATALOUTCOMESFOLLOWINGTHE

INTRODUCTIONOFANEDUCATIONALPACKAGEOFFETALHEARTRATE

MONITORINGEDUCATIONCOUPLEDWITHUMBILICALARTERYLACTATE

SAMPLINGINALOWRESOURCESETTINGLABOURWARDINSOUTH

AFRICA.

228

CHAPTEROVERVIEW

ItisevidentfromChapter7thattheoptimalcut-offforumbilicalarterylactateinour

setting is similar to that inhigh-income settings. We thenaimed to assesswhether

training of doctors and midwives in fetal physiology, acid-base balance and CTG

interpretation in addition to umbilical artery lactate sampling improved intrapartum

outcomes.

AsoutlinedinChapter3,thesecondphaseoftheumbilicalarterylactatesamplingtrial

involvedaneducationalinterventionwherebytheauthorofthisthesistraineddoctors

andmidwivesprovidingintrapartumcareinfetalphysiology,fetalandneonatallactate

andacidbasebalance, and interpretationofCTGsand their relationship to acidosis.

Participants completed a pre-and post-training test to demonstrate acquisition of

knowledge. Fifty-three doctors participated in training,with 47 (89%) completing all

three sessions and the pre-and post-tests. Mean test scores significantly improved

from43%to70%,p<0.001.(Table8.1)All18midwivesintheunitcompletedtraining

aswellaspreandpost-tests,withanimprovementinmeanscoresfrom38%to59%,

p<0.001.(Table8.2)

229

Table8.1:Preandposttestscoresfordoctors

230

Table8.2:Preandposttestscoresformidwives

There is a plethora of evidence that training such as we conducted increases

knowledge(1), however, while the improvement in test scores is reassuring, the

ultimate aim of the research is a change in clinical practice and improvement in

intrapartumoutcomes.Theremainderofthischapterexploresthisphenomenon,and

hasbeensubmittedandisunderreviewbyBMCPregnancyChildbirthandisformatted

forthisjournal’sspecifications.

231

REFERENCEFORCHAPTERINTRODUCTION

1. CalvertKL,McGurganPM,DebenhamEM,GratwickFJ,MaourisP.Emergencyobstetricsimulationtraining:howdoweknowwherewearegoing,ifwedon'tknowwherewehavebeen?AustNZJObstetGynaecol.2013;53(6):509-16.

232

ABEFOREANDAFTERSTUDYOFTHEIMPACTONOBSTETRICANDPERINATALOUTCOMES

FOLLOWINGTHEINTRODUCTIONOFANEDUCATIONALPACKAGEOFFETALHEARTRATE

MONITORINGEDUCATIONCOUPLEDWITHUMBILICALARTERYLACTATESAMPLINGINALOW

RESOURCESETTINGLABOURWARDINSOUTHAFRICA.

AllansonER1,2,PattinsonRC2,NathanEA1,DickinsonJE1

1. Division of Obstetrics and Gynaecology, Faculty of Health and Medical Sciences,

UniversityofWesternAustralia,Crawley,WA,Australia.

2. SAMRC/UP Maternal and Infant Health Care Strategies Unit, Department of

ObstetricsandGynaecology,UniversityofPretoria,Pretoria,SouthAfrica.

EmmaRAllanson

Division of Obstetrics and Gynaecology, Faculty of Health and Medical Sciences,

UniversityofWesternAustralia,Crawley,WA,Australia.

(M550)TheUniversityofWesternAustralia

35StirlingHwyCRAWLEYWA6009

+61432199297

[email protected]

233

RobertCPattinson

UniversityofPretoria,KalafongCampus

SAMRC/UPMaternalandInfantHealthCareStrategies,UnitPrivateBagX323Arcadia,

SouthAfrica,0007

[email protected]

ElizabethANathan

DivisionofObstetricsandGynaecology,FacultyofHealthandMedicalSciences,



[email protected]

JanEDickinson




[email protected]

234

Correspondingauthor:ERAllanson




+61432199297

[email protected]

Wordcount:3242

CONFLICTSOFINTEREST

Theauthorsdeclarenoconflictsofinterest

235

ABSTRACT

INTRODUCTION

Rates of caesarean section (CS) are increasing and abnormal fetal heart rate tracing

andconcernaboutconsequentacidosis remainoneof themostcommon indications

forprimaryCS.Umbilicalartery(UA)lactatesamplingprovidesclinicianswithpointof

care feedbackonCTG interpretationand intrapartumcareandmay result inaltered

futurepractice.

MATERIALSANDMETHODS

From 3rd March - 12th November 2014 we undertook a before and after study in

Pretoria, South Africa, to determine the impact of introducing a clinical package of

fetal heart rate monitoring education and prompt feedback with UA cord lactate

sampling,usingahand-heldmeter,onmaternalandperinataloutcomes.

RESULTS

936 consecutive sampleswere analyzed (pre n=374 and post n=562). Therewas no

difference in mean lactate (4.6 mmol/L [95%CI 4.4-4.8] compared with 4.9 mmol/L

[95%CI 4.7-5.1], p=0.089). Suspected fetal compromise was reduced in the post-

interventionperiod:30·2%vs22·1%,aOR0·71,95%CI0·52-0·96,p=0·027.Cesarean

236

sectionratesweresignificantlyreducedintheunivariateanalysis:pre-40·3%vspost-

intervention31·6% (p=0·007). This reduction remained significantwhenadjusted for

previouscaesareansection,primiparity,maternalHIVinfectionandpretermbirth(aOR

0·72,95%CI0·54-0·98,p=0·035).Neonataloutcomesdidnotdifferbetween the two

groups.

CONCLUSION

Theintroductionofaclinicalpracticepackageoffetalheartratemonitoringeducation

combined with routine UA cord lactate sampling has the potential to reduce the

caesarean section rate without increasing adverse neonatal outcomes in a low-

resourcesetting.

KEYWORDS

Lactate;caesarean;SouthAfrica;maternal;neonatal

KEYMESSAGE

Both rising caesarean section rates and high perinatalmortality are issues of global

concern.We have demonstrated, using a simplemethod of testing umbilical artery

237

lactateatbirth,thatitispossibletoreducecaesareansectionrateswithoutworsening

neonatalmorbidityinalowresourcesetting.

238

INTRODUCTION

Theratesofcaesareansection(CS)areincreasingglobally(1),althoughboththetotal

rates and the rate of rise vary considerably between low,middle, and high income

countries(2). While there is little agreement about the optimal rate of CS, there is

some evidence that increasing rates above a certain threshold may not result in a

corresponding improvement in maternal and perinatal mortality(3). Moreover, CS,

while a life-saving procedure that allwomenmust have access towhen required, is

alsoassociatedwithpotentialadversematernaloutcomesinboththeindexandfuture

pregnancies(4). Balancing the indication for CS with its potential risks is a critical

clinicalandpublichealthchallenge.

Abnormal fetal heart rate tracing or suspected fetal distress with concern for

consequenthypoxiaandacidaemiaremainsoneof themostcommon indications for

thefirstCS(withratescitedbetween10-32%)(5-8).However,theratesofpathological

acidaemia in neonates generally is far lower than this level, and although there are

documentedbenefits, theuseofelectronic fetalheart ratemonitoring in labourhas

not been shown to improve neonatal mortality(9). It is important to consider fetal

physiologyintheintrapartumperiodandcardiotocograph(CTG)interpretationingreat

detail given the need to balance the role that intrapartum hypoxia plays in global

perinatal deaths, with the role that potentially unnecessary CS play in maternal

morbidityandmortality.

239

Objectivemeasurementsof intrapartumfetalhypoxia(suchaspHor lactatesampled

from the fetal scalp) may improve upon the known limited specificity of CTG and

reducetherateofunnecessarycaesareansections(10,11).Thecostoftheequipment

to do this, alongside high rates of maternal HIV infection makes this prohibitive in

someunitsaroundtheworld.Universalumbilicalcord-gassamplingafterdeliveryhas

been suggested as amethod to improve the specificity of the CTG(12). Paired cord

arterial and venous gas sampling is both time and resource intensive however,

particularly outside of high-income country settings. To that end, it is possible to

introduceinexpensivebedsidepointofcareumbilicalartery(UA)lactatetesting,which

provides clinicians feedback on intrapartum care and potentially alters practice for

futurecases,whichmaypartlymitigatethepoorspecificityoftheCTG.Therefore,we

conductedabeforeandafterstudyinatertiarylabourwardinPretoria,SouthAfrica,

to determine the impact of introducing UA lactate sampling, in conjunction with

education on CTG interpretation and fetal and lactate physiology, on maternal and

perinataloutcomes.

MATERIALSANDMETHODS

Kalafong Hospital is a tertiary hospital in Pretoria, South Africa, affiliated with the

UniversityofPretoria. In2014,therewere6,499birthsintheunit.Duringtheperiod

3rd March to the 12th November 2014 we conducted a before and after study to

determine the impact of umbilical artery lactate sampling, combined with an

240

educational program on fetal heart rate monitoring, on maternal and neonatal

outcomes.

Thestudygroups includedprospectivelycollectedsamplesofumbilicalartery lactate

obtainedbytheprimaryresearcher(EA),oradoctorintheunittrainedbytheprimary

researcher.WehavepresentedthespecificsofUAsamplingintheunit,andthedata

points collected previously(13). Briefly, after maternal consent was obtained, the

umbilicalcordwasclampedandcutwithinoneminuteofbirth.Asmallarterialblood

sample (<0.5uL) was then taken from a double clamped segment of the remaining

umbilical cord, prior to delivery of the placenta. The lactate level was measured

(mmol/L) on a Roche Accutrend PlusTM© hand-held lactate meter (Rotkreuz,

Switzerland).Thecoefficientofvariation(CV)forthislactatemeteris1.8-3%(14).All

women aiming to have a vaginal birth were eligible to be consented for the trial,

regardlessoftheirultimatebirthmode.

Duringtheperiod3rdMarchto18thofJuly,2014,aprospectivecohortofwomenwere

consented to be enrolled in the study. As this was a blinded cohort, this was a

convenience sample, obtained within the capacity of a single researcher (EA). The

resultsofthelactatesamplinginthiscohortwereblindedtotheclinicians(midwives,

registrars, and consultants) managing the intrapartum care of women. This cohort

served as the baseline for the unit, the results of which have been published

previously(13).Beforethecompletionofthecohort,allmidwivesandregistrarsinthe

241

unit underwent a training course in fetal physiology, lactate physiology and

cardiotocograph(CTG)interpretation,conductedbytheprimaryresearcher(EA).The

courseconsistedofapre-test,a seriesofdidactic lectures, interactiveapplicationof

knowledgewith CTG examples, and a post-test. Subsequently, from21st July to 12th

November 2014, lactate sampling continued, with the results made immediately

availabletothecliniciansmanagingtheintrapartumcareofwomeninthelaborward.

Furthermore,all lactatesamplesobtainedintheprevious24hourswerereviewedat

the dailymorning labourwardmeetings of allmedical staff (interns, registrars, and

consultants),andtwiceweekly,anin-depthreviewofinterestingorchallengingcases

wasfacilitatedbytheprimaryresearcher(EA).DailyreviewsinvolvedcorrelatingCTG

interpretationpre-birthwithneonataloutcomes,includingtheUAlactate.Duringthe

two study periods, basic data on all births (regardless of recruitment to the study)

were collected, including number of births, mode of birth, and admissions to the

neonatalnursery(alllevels).

STATISTICSANDANALYSIS

The estimation of sample size was performed using standardised normal lactate

mean and standard deviation. A sample size of 265 in each of the pre and post

intervention groups would achieve an 80% power to detect a difference of 0.25

standard deviations (approximately 8% difference in means) in lactate

measurementsbetweengroupsusinga two-sided sample t-test at5%significance

level.Thissamplesizewouldalsoachieveatleast80%powertodetectanR-squared

242

of 0.05 attributed to one or more independent variable/s while simultaneously

adjusting for multiple other relevant covariates with an R-squared of 0.2 in a

multiple linear regression analysis. Assuming a 15% attrition rate, the sample size

was increased to 305 in each group for a total sample size of 610. ([Power and

SampleSize(PASS2008)].

Toassessthelactateresults,dataweretransformedtothenaturallogarithmtocorrect

normality and summarised as geometric means and 95% confidence intervals (CI).

Linear regression analysis was conducted on the log transformed lactate

measurementsandpresentedasestimatedmeaneffectsand95%CI’s.Amultivariate

analysis on lactate levels, adjusted for primiparity, HIV, fetal problems in pregnancy

and preterm birth, was done with lactate measurements log transformed for the

analysis. Summary values of estimated mean effects and 95% confidence intervals

were back transformed. Outcomes were summarised using means and standard

deviations or medians, interquartile ranges and ranges for continuous data and

frequencydistributionsforcategoricaldata.Basedonourpreviouslypublishedcohort

study, a cut-off for an abnormal lactate was assumed to be 5·45 mmol/L(13).

Univariate comparisons were made using independent t-tests for continuous

outcomes and Chi-square or Fisher exact tests for categorical comparisons.

Multivariable logistic regression analysis was performed on maternal and neonatal

outcomes including fetal distress, cesarean section, resuscitation, Apgar scores and

admissiontospecialcarenursery.Resultsweresummarisedwithunadjusted(OR)and

adjustedoddsratios(aOR)and95%CI’s.

243

Analyses of outcomes were adjusted for preterm birth, primiparity, HIV, previous

uterine surgery or maternal problems in pregnancy (maternal outcome), and fetal

problems during pregnancy (neonatal outcomes). Maternal problems in pregnancy

includedbeinganadolescent,havingadvancedmaternalage(>35),anaemia,cervical

incompetence, eclampsia, gestational diabetes mellitus, HELLP syndrome,

hypertensionwithoutpre-eclampsia,maternalmedicalconditions,noantenatalcare,

pre-eclampsia, preterm labour, syphilis, thyrotoxicosis, tuberculosis, urinary tract

infection,uterineanomaly,orvenousthromboembolicevent.Apre-plannedsubgroup

analysiswas conducted on infantswith birthweight <2500g comparedwith ≥2500g

and neonatal outcomes were assessed. A Bonferroni adjustment was applied such

that the significance level for each comparison of pre- and post-intervention groups

within birth weight strata was set to 0·025. All tests were two-sided and a p-value

<0·05was considered statistically significant for theoverall analysis. SPSS statistical

softwarewasusedindataanalysis(version22·0,Armonk,NY:IBMCorp).

Aprotocolwasdevelopedforthisstudy;itisnotpublishedbutavailableuponrequest.

ETHICS

This study was approved by the ethics committees of The University of Western

Australia Human Research Ethics Committee on the 26/02/2014 (Reference number

244

RA/4/1/6581), the University of Pretoria, Pretoria, South Africa on the 10/02/2014


RESULTS

397sampleswerecollectedinthepre-interventionperiodfrom2196plannedvaginal

births and 2436 total births. 597 samples were collected in the post-intervention

period from1978plannedvaginalbirthsand2232 totalbirths.Afterexclusionof54

twinpregnancies,atotalof936sampleswereanalysed(pren=374andpostn=562).

MATERNALCHARACTERISTICS

Maternal characteristics did not differ between groups. (Table 1) Thematernal HIV

prevalenceinthepre-interventiongroupwas21·8%(n=79),withantiretroviraltherapy

(both prophylactic antiretroviral therapy and highly active antiretroviral treatment)

used in 96·2%·of women antenatally. The maternal HIV prevalence in the post-

interventiongroupwas23%(n=123),withantiretroviraltherapyin92·4%ofwomen.A

comparison of the maternal and neonatal outcomes between the HIV and non-HIV

infectedmothershasbeenpublishedelsewhere(15).

245

Datasummariesaremean(±SD),median(IQR,R)orN(%),asappropriate

PREGNANCYANDLABOUR

The pre-intervention group had a higher rate of maternal-related pregnancy

complications (41·8% vs 22·4%, p<0·001) and were also more likely to have

intrapartumcomplications (55·6%vs45·9%,p=0·004), fetaldistress (30·2%vs22·1%,

p=0·005), other labor complications (3·5% vs 1·4%, p=0·038) and be transferred

intrapartum from a referral unit (19·5% vs 7·3%, p<0·001). The intrapartum

complications are outlined in Table 2. Other labor complications included were

infrequent in both groups and included twin pregnancy diagnosed intrapartum, the

development of isolated intrapartum hypertension, acute twin to twin transfusion

syndromeand singleunusual occurrences such as “patient foundon antenatalward

Table1:Maternalcharacteristicsinpre-andpost-interventiongroups

Pre-intervention(N=374) Post–intervention

(N=562)

p-value

Maternalage(y) 27·4(±6·1) 27·4(±6·6) 0·897

Gestation(w) 38·8(±3·1)(n=345) 38·8(±3·2)(n=504) 0·963

Gravidity 2(1-3;1-7) 2)1-3;1-8) 0·672

Parity 1(0-2;0-5) 1(0-2;0-7) 0·820

Nulliparous 136(36·7%) 195(36·2%) 0·899

Previousuterinesurgery 53(14·7%) 56(10·6%) 0·072

HIVstatus 79(21·8%) 123(23·0%) 0·655

HIVtreatment 75(96·2%) 109(92·4%) 0·369

246

withfetalheaddelivered”,“patientpushingfromfourcentimetres”,maternalcollapse

atfulldilatation,andfailedinstrumentaldeliveryatlocalclinicpriortotransfer.

Table2:Intrapartumcomplicationsinpre-andpost-interventiongroups

Pre-intervention

N=374

N(%)

Post–intervention

N=562

N(%)

p-value

Any 208(55·6%) 258(45·9%) 0·004

Abruption 7(1·9%) 7(1·2%) 0·440

Chorioamnionitis 0 4(0·7%) 0·155

Delayedprogress1ststage 40(10·7%) 85(15·1%) 0·051

Delayedprogress2ndstage 15(4·0%) 23(4·1%) 0·950

Intrapartumhaemorrhage 7(1·9%) 12(2·1%) 0·779

Malpresentation 8(2·1%) 14(2·5%) 0·728

Meconiumstainedamnioticfluid 27(7·2%) 40(7·1%) 0·953

Augmentationoflabour 13(3·5%) 25(4·4%) 0·460

Shoulderdystocia 9(2·4%) 7(1·2%) 0·180

Fetaldistress 113(30·2%) 124(22·1%) 0·005

Other* 13(3·5%) 8(1·4%) 0·038

Intrapartumtransfer 73(19·5%) 41(7·3%) <0·001

*Other=seenoteinmainbodyoftext.

247

LACTATERESULTS

Alactateresultwasrecordedfor351(94%)ofwomenrecruitedinthepre-intervention

groupand542(96%)ofwomenrecruitedpost-intervention.Therewasnodifferencein

themeanlactateresult(4.6mmol/L[95%CI4.4-4.8]comparedwith4.9mmol/L[95%CI

4.7-5.1], p=0.089) or in babies born with a lactate greater than 5.45 mmol/L (122

(34.8%) compared with 210 (38.7%), p=0.228). There was no difference in mean

lactate levels between the two groupswhen adjusted for primiparity, HIV infection,

fetalproblemsinpregnancyandpretermbirth(unadjustedOR1.05[95%CI0.99-1.13],

comparedwithanadjustedOR1.06[95%CI0.99-1.13],p=0.108).

BIRTHCHARACTERISTICS

Pre-intervention women were more likely to have a caesarean section due to

suspected fetal distress (21·4% vs14·9%, p=0.011) and vaginal birth after caesarean

(VBAC) in labour (8·0% vs 2·5%, p<0·001). Other indications for caesarean delivery

were similar between the two groups. Suspected fetal distress during labor was

reduced in thepost-interventionperiod:pre-30·2%vspost-intervention22·1%,aOR

0·71,95%CI0·52-0·96,p=0·027.Caesareansectionratesweresignificantlyreducedin

the univariate analysis: pre- 40·3% vs post-intervention 31·6% (p=0·007). This

reduction remained significant when adjusted for previous caesarean section,

primiparity, maternal HIV infection and preterm birth (aOR 0·72, 95%CI 0·54-0·98,

p=0·035).

248

NEONATALCHARACTERISTICS

Theneonatalcharacteristics in thetwogroupsareoutlined inTable3.Therewasno

difference in neonatal outcomes when adjusted for primiparity, HIV infection, fetal

problems in pregnancy and pretermbirth. When stratified by birthweight (<2500g

and≥2500g),therewasnodifferenceinanyoutcomes, includinglactateresults.(See

Table4)

249

Table3:Neonatalcharacteristicspre-andpost-interventiongroups

N*

Pre-intervention

N*

Post–intervention

p-value

Preterm(<37w) 345 64(18·6%) 504 99(19·6%) 0·691

Birthweight(g) 363 3033·9(±623·5)

520 3006·4(±641·2)

0·527

Lactate>5·45 122(34·8%) 210(38·7%) 0·228

Resuscitation 344 71(20·6%) 532 129(24·2%) 0·214

Apgar<7at1min 365 45(12·3%) 526 83(15·8%) 0·149

Apgar<7at5min 366 20(5·5%) 526 35(6·7%) 0·468

Special care nurseryadmission

374 56(15·0%) 562 80(14·2%) 0·754

Admissiontype 367 558

None 318(86·6%) 480(86·0%)

Highdependencyunit 8(2·2%) 22(3·9%) 0·387

Intensivecareunit 14(3·8%) 15(2·7%)

Neonatalward 27(7·4%) 41(7·3%)

Datasummariesaremean(±sd)orN(%),asappropriate*Nreflectscasesforwhichdatawereavailable

forthevariable

250

Table4:Neonataloutcomesstratifiedbybirthweight<2500gand≥2500g.

Birthweight<2500g Birthweight≥2500g

Pre-intervention

Post–intervention

p-value

Pre-intervention

Post-intervention

p-value

n=52 n=78 n=311 n=442

Birth weight(g)

1937.48(±556.1)

1836.35(±534.2)

0.300 3217.2(±409.6) 3212.9(±386.3)

0.884

n=46 n=68 n=291 n=417

Gestation(w) 35.4(±4.7) 34.6(±4.4) 0.372 39.3(±2.4) 39.5(±2.3) 0.454

Preterm(<37w)

25(54.3%) 47(69.1%) 0.109 38(13.1%) 47(11.3%) 0.472

N=52 N=75

Lactate1 4.4(3.8-5.2) 4.9(4.4-5.4) 0.263 4.7(4.4-4.9) 4.8(4.6-5.1) 0.333

Lactate>5.45 19(36.5%) 27(36.0%) 0.950 99(34.3%) 165(38.8%) 0.215

Resuscitation 22(43.1%) 34(44.2%) 0.909 47(16.5%) 90(20.5%) 0.181

Apgar<7 at 1min

14(27.5%) 24(32.0%) 0.585 30(9.8%) 56(12.8%) 0.209

Apgar <7 at 5min

7(13.7%) 16(21.3%) 0.278 12(3.9%) 19(4.3%) 0.771

Special carenurseryadmission

25(48.1%) 41(52.6%) 0.616 30(9.6%) 38(8.6%) 0.621

Admissiontype

None 27(55.1%) 37(48.7%) 0.687 281(91.5%) 402(91.2%) 0.664

Ward 15(30.6%) 24(31.6%) 6(2.0%) 13(2.9%)

Highdependencyunit /Intensive careunit

7(14.3%) 15(19.7%) 20(6.5%) 26(5.9%)

Data summaries are mean (±SD) or N (%), as appropriate. P-values less than 0.025 consideredstatisticallysignificantwithaBonferronicorrection.1Dataaresummarisedasgeometricmeanand95%confidenceintervals.

251

MODEOFDELIVERYANDNEONATALOUTCOMESFORALLDELIVERIES

Therewere 2,436births during the entire pre-interventionperiod, and 2,232during

thepost-interventionperiod.Duringthepre-interventionperiodtherewere932/2436

cesarean deliveries (38·3%), which decreased in the post-intervention period to

653/2232 (29·3%, P<0·001). There was no change in rate of elective caesarean

sections; 240/2436 (9.9%) to 254/2232 (11.4%, P=0.09). There was no difference in

instrumental births 73/2436 (3%) to 81/2232 (3.6%, P=0.227). Emergency caesarean

deliveries decreased from 692/2436 (28·4%) to 399/2232 (17·0%, P<0·001), and

neonataladmissionsdecreasedfrom450/2436(18·5%)to349/2232(15·6%,P=0·010).

Adiagnosiswasrecordedfor360(80%)admissionsinthepre-interventionperiodand

278 (80%) of admissions in the post-operative period. The commonest reasons for

admission in the pre-intervention period were respiratory distress syndrome (RDS)

(19.7%), neonatal jaundice (NNJ) (13.1%), congenital infections (8.6%), hypoglycemia

(7.5%),hypoxicischaemicencephalopathy(HIE)(7.2%),andprematurity(7.2%).Inthe

post intervention period the commonest reasons were RDS (19.1%), NNJ (12.2%),

hyalinemembranedisease(HMD)(10.1%),transienttachypneaofthenewborn(TTN)

(10.1%),lowbirthweight(LBW)(8.6%)andHIE(7.9%).

DISCUSSION

This study explored the impact of an educational program coupled with the

introduction of UA lactate sampling in a low-resource setting on intrapartum

252

outcomes. Previous studies have been based in high-income settings(16) and itwas

notclearwhetherthepreviouslyobservedbenefitsinearlierstudieswouldtranslateto

alowerresourcecentre.Whiletherewasnosignificantdifferenceinthecordlactate

results between the two groups, we found a reduction in the CS rate without

worseninganyofneonataloutcomes.Thelargestcontributortothisreductionwasin

CS for suspected fetal compromise. A significant reduction post-intervention was

observed inCS,emergencyCS, andneonatal admissions forallwomen regardlessof

whethertheywererecruitedtothestudyforlactatemeasurementornot.Whilethese

overallresultsarenotadjustedforconfoundingfactors,theyarereassuring;thefalling

CSratehasnotworsenedfetalandneonataloutcomes,and,basedonthereductionin

overall neonatal unit admission rate, may have improved them. The impact of an

educational program on fetal physiology and intrapartum heart rate monitoring

demonstratedbenefitsthatweretranslatedtoallwomeninthelabourwardandwe

speculate this should be sustainedwhen coupledwith the introduction of universal

cordlactatesamplingforongoingclinicianoutcomefeedback.

There is evidence that the lack of appropriate interpretation of fetal monitoring in

South Africa contributes to the unacceptably high intrapartum and early neonatal

mortality rates(17, 18). Equally, there is evidence that theprocessofqualityof care

audit and the feedback to cliniciansof theUA lactate resultsmay improvematernal

and neonatal outcomes(12, 19). Given the reduction in suspected fetal compromise

without an increase in adverse neonatal outcomes that occurred following our

intervention,wehypothesise that theobjective feedbackavailable to clinicians from

253

the UA lactate result highlighted recognition on CTG interpretation of the “true”

acidemic cases. This theorymaybe supportedby the trend to lessneonatalnursery

admissionsinthestudygroupandthesignificantlylessadmissionsoverallinthepost-

intervention study period, regardless of trial recruitment status. In addition, we

observed an increase in delayed labour progress in the post-intervention group

without worse neonatal outcomes; it may be that less early interventions for

suspectedfetaldistressallowedotherpotentialcomplicationsoflongerlabourslabor

tobecomeapparent.

Interestingly,therewerelessmaternalproblemsinthepostinterventionperiod,which

plausibly could have resulted in fewer fetuses at risk of compromise. There were,

however, significantly less intrapartum transfers from referral facilities in the post-

interventionperiod. Thenatureof an intrapartum transfer indicates thewomenare

morelikelytohavematernalrelatedpregnancyproblems,necessitatingup-referralto

a tertiary unit(20, 21), and this may have accounted for some of the reduction in

maternalcomplicationsinthepost-interventionperiod.

TherewasnosignificantchangeintheelectiveCSrate,withthereductioninourstudy

almost exclusively intrapartum non-elective CS, which remained significant when

adjustedforprimiparity.Thisiswherewebelievethestrengthofthisinterventionlies

–inpreventingthefirstCS,particularconsideringevidencesuchasthatofVillaretal,

demonstrating that intrapartum caesarean section in a sample of nearly 100000

254

deliveries in 8 Latin American countries was associated with an increased risk in

severalmaternalandneonatalmorbidityandmortality(22).TheRobsonclassification

for caesarean section is widely used tomonitor CS rates internationally(23), with a

particularfocusonreducingcaesareansectioninnulliparouswomen(group1and2in

theRobsonclassification)andthosewithapreviouscaesareansection,(group5),with

thelatteraclearconsequenceoftheformer(1).Althoughheterogeneousinthequality

ofincludedstudies,arecentmeta-analysishasdemonstratedthatqualityofcareaudit

usingtheRobsonclassificationsystemtomonitorcesareansectionratescanresultina

reduction in CS(24). In our studywe introduced a quality of care audit process and

providedinstantfeedbacktothecliniciansintheunit,withareductioninCSinwomen

ingroups1and2oftheRobsonclassification.

Reduction inCS isofparticular importance inSouthAfrica,aswomeninthiscountry

are more likely to die following a CS compared with a vaginal birth(25). This is

particularly concerning as 22·6% of deliveries in South Africa are by CS, and in

provincialtertiaryhospitals,asistheunitinthisstudy,therateis35·2%(26).Moreover,

maternaldeathsrelatedtoCShavebeenincreasing,anddisturbingly,largenumbersof

theseareduetoperioperativehaemorrhage(27),withsurgical traumabeingoneof, if

notthemostcommon,reasonfordeath(28,29).ThereductioninCSinourstudy(bothin

theincludedcasesandoverallduringthestudyperiod)hasthepotentialforimmediate

and downstream effects for future pregnancies, as well as long-term outcomes for

womenandtheirfamilies.

255

Finally, testing forpotential fetal acidemiawith lactateonahandheld,pointof care

devicehasseveraladvantages.Firstly;itmaybemorefeasible(botheconomicallyand

in ease of use) than other measures (such as paired arterial and venous cord gas

sampling), especially in the limited resource setting. Secondly, the hand-held device

correlateswellwithlaboratoryanalysers,addingvaliditytotheprocessofpointofcare

testing(30,31).

STRENGTHS

To our knowledge this is the largest assessment of an intervention using umbilical

lactatemeasurementanditsimpactinalowormiddleincomecountryonintrapartum

obstetric management(16), and is likely representative of the unit as a whole. The

presenceofasingleresearcheronthelaborward(EA)tooverseethetrainingofhealth

care providers in sample collection and ensure complete data collection throughout

thestudyperiodaddstotheinternalvalidityoftheresults.

LIMITATIONS

It is difficult to separate the impact of the fetal physiology and CTG interpretation

trainingwiththelactatesamplingonthechangeinpracticefollowingtheintervention,

and indeed it is likely the combination of both that has created a management

package. All staffwho cared forwomen inour studyhadpreviously completed the

essential steps in themanagement of obstetric emergencies (esmoe) course, which

256

includes a module on ctg interpretation. this suggests that the addition of the ua

lactateandthequalityofcareauditprocessaroundthiswasthemainfactorinpractice

change. however, ongoing revision of ctg interpretation is an essential element in

labourwardpracticeandwerecommendregularcasereviewandeducationprograms.

Aspartof this study,we introducedpointof care testing,andalthough trainingwas

undertaken for all involved in the collection, we do not have paired samples (both

arterial and venous) to provide confirmation of arterial sampling.We are, however,

reassured by the fact that venous lactate is shown to be predictive of arterial lactic

acidaemia,andthat thepreviousoptimalcut-off for lactateweobtainedusing these

sampleswasverysimilartoalargecohortincludingpairedsamples(13).

CONCLUSION

The introduction of UA lactate has the ability to significantly reduce the CS rate

without increasing adverse neonatal outcomes in a low-resource setting. Further

researchtoexpandthebodyofevidenceontheuseofUAlactatesamplinginlowand

middle-incomecountriesisneeded.

257

ABBREVIATIONS

aOR Adjustedoddsratio

AVD Assistedvaginaldelivery

CI Confidenceintervals

CS Caesareansections

CTG Cardiotocograph

CV Coefficientofvariation

HIE Hypoxicischaemicencephalopathy

HIV Humanimmunodeficiencyvirus

HMD Hyalinemembranedisease

LBW Lowbirthweight

NNJ Neonataljaundice

OR Oddsratio

RDS Respiratorydistresssyndrome

TTN Transienttachypnoeaofthenewborn

UA Umbilicalartery

VBAC Vaginalbirthaftercaesarean

258

DECLARATIONS

ETHICSAPPROVALANDCONSENTTOPARTICIPATE

This study was approved by the ethics committees of The University of Western

Australia Human Research Ethics Committee on the 26/02/2014 (Reference number

RA/4/1/6581), the University of Pretoria, Pretoria, South Africa on the 10/02/2014


Allparticipantsconsentedtothestudy.

Availabilityofdataandmaterials:

The datasets used and/or analysed during the current study are available from the

correspondingauthoronreasonablerequest.

COMPETINGINTERESTS

Theauthorsreportnoconflictsofinterest

259

FUNDING

EAisaPhDcandidatefundedbytheUniversityofWesternAustraliawithanAustralian

PostgraduateAward,andanAthelstanandAmySawMedicaltop-upscholarship,and

by the Women and Infants Research Foundation with a Gordon King Doctor of

Philosophyscholarship.

AUTHORS'CONTRIBUTIONS

ERAdesignedthestudy,supervisedtheconductofthestudy,collectedthedata,wrote

the initial manuscript, and approved the final manuscript. EAN contributed to the

designofthestudy,analysedthedata,andapprovedthefinalmanuscript.RCPandJED

contributed to the design of the study, supported the conduct of the study, and

approvedthefinalmanuscript.

ACKNOWLEDGEMENTS

The authors would like to thank the midwives, doctors, and women of Kalafong

Hospitalfortheirsupportinrunningthisstudy.

260

REFERENCES

1. VogelJP,BetránAP,VindevoghelN,SouzaJP,TorloniMR,ZhangJ,etal.Useofthe Robson classification to assess caesarean section trends in 21 countries: asecondary analysis of two WHO multicountry surveys. The Lancet Global health.2015;3(5):e260-e70.2. Betrán AP, Ye J, Moller A-B, Zhang J, Gülmezoglu AM, Torloni MR. Theincreasing trend in caesarean section rates: global, regional and national estimates:1990-2014.PLoSOne.2016;11(2):e0148343.3. BetranAP,TorloniMR,Zhang J, Ye J,MikolajczykR,Deneux-TharauxC,et al.Whatistheoptimalrateofcaesareansectionatpopulationlevel?Asystematicreviewofecologicstudies.Reproductivehealth.2015;12(1):57.4. Lumbiganon P, LaopaiboonM,Gülmezoglu AM, Souza JP, Taneepanichskul S,RuyanP,etal.Methodofdeliveryandpregnancyoutcomes inAsia: theWHOglobalsurveyonmaternalandperinatalhealth2007–08.TheLancet.2010;375(9713):490-9.5. CaugheyAB,CahillAG,GuiseJ-M,RouseDJ,ObstetriciansACo,Gynecologists.Safe prevention of the primary cesarean delivery. Am J Obstet Gynecol.2014;210(3):179-93.6. Spong CY, Berghella V,Wenstrom KD,Mercer BM, SaadeGR. Preventing thefirstcesareandelivery:summaryofajointEuniceKennedyShrivernationalinstituteofchild health and human development, society for maternal-fetal medicine, andAmerican college of obstetricians and gynecologists workshop. Obstet Gynecol.2012;120(5):1181.7. BarberEL,LundsbergL,BelangerK,PettkerCM,FunaiEF,IlluzziJL.Contributingindicationstotherisingcesareandeliveryrate.ObstetGynecol.2011;118(1):29.8. Chauhan S, Beydoun H, Hammad I, Babbar S, Hill J, Mlynarczyk M, et al.Indications for caesarean sections at≥ 34 weeks among nulliparous women anddifferential composite maternal and neonatal morbidity. BJOG. 2014;121(11):1395-402.9. AlfirevicZ,DevaneD,GyteGM.Continuouscardiotocography(CTG)asaformofelectronicfetalmonitoring(EFM)forfetalassessmentduringlabour.TheCochraneLibrary.2013.10. Wiberg-ItzelE,LipponerC,NormanM,HerbstA,PrebensenD,HanssonA,etal.DeterminationofpHorlactateinfetalscalpbloodinmanagementofintrapartumfetaldistress:randomisedcontrolledmulticentretrial.BMJ.2008;336(7656):1284-7.11. MorinC,ChartierM,BounanS,HatemG,GoffinetF,LeRayC.FetalscalppHduring labor:Which threshold for intervention? Journalofgynecologyobstetricsandhumanreproduction.2017;46(2):183.12. White CR, Doherty DA, Newnham JP, Pennell CE. The impact of introducinguniversalumbilicalcordbloodgasanalysisandlactatemeasurementatdelivery.AustNZJObstetGynaecol.2014;54(1):71-8.13. Allanson E, Pattinson R, Nathan E, Dickinson J. The introduction of umbilicalcord lactate measurement and associated neonatal outcomes in a South Africantertiary hospital labor ward. The Journal of Maternal-Fetal & Neonatal Medicine.2017:1-7.

261

14. Baldari C, Bonavolonta V, Emerenziani GP, Gallotta MC, Silva AJ, Guidetti L.Accuracy,reliability, linearityofAccutrendandLactateProversusEBIOplusanalyzer.EurJApplPhysiol.2009;107(1):105-11.15. AllansonE, R, Pattinson,RC,Nathan, EA,Dickinson, JE. The impactofHIVonperinataloutcomesinaSouthAfricanlabourward.2017.16. Allanson E,Waqar T,White C, Tunçalp Ö, Dickinson J. Umbilical lactate as ameasureofacidosisandpredictorofneonatalrisk:asystematicreview.BJOG.2016.17. Buchmann EJ, Pattinson RC. Babieswho die from labour-related intrapartumhypoxia: a confidential enquiry in South African public hospitals. Trop Doct.2006;36(1):8-10.18. BothmaM,BuchmannEJ.Astudyoffreshstillbirthsweighing2500gormoreatthree academic hospitals in South Africa. International journal of gynaecology andobstetrics: the official organ of the International Federation of Gynaecology andObstetrics.2016;134(2):186-9.19. Allanson ER, Pattinson RC. Quality-of-care audits and perinatal mortality inSouthAfrica.BullWorldHealthOrgan.2015;93(6):424-8.20. GandhiM,WelzT,RonsmansC.SevereacutematernalmorbidityinruralSouthAfrica.InternationalJournalofGynecology&Obstetrics.2004;87(2):180-7.21. SchoonMG.Impactofinter-facilitytransportonmaternalmortalityintheFreeStateProvince.SAMJ:SouthAfricanMedicalJournal.2013;103(8):534-6.22. VillarJ,CarroliG,ZavaletaN,DonnerA,WojdylaD,FaundesA,etal.Maternaland neonatal individual risks and benefits associated with caesarean delivery:multicentreprospectivestudy.BMJ.2007;335(7628):1025.23. BetranAP,VindevoghelN,SouzaJP,GulmezogluAM,TorloniMR.Asystematicreviewof the Robson classification for caesarean section:whatworks, doesn'tworkandhowtoimproveit.PLoSOne.2014;9(6):e97769.24. Boatin AA, Cullinane F, TorloniMR, BetranAP. Audit and feedback using theRobson classification to reduce caesarean section rates: a systematic review. BJOG.2017.25. Gebhardt S, editor Preparing for caesarean delivery: review. Obstetrics andGynaecologyForum;2015:InHousePublications.26. PattinsonR,Rhoda,N.Savingbabies2012-2013:NinthreportonperinatalcareinSouthAfrica.TshepesaPress,Pretoria,SouthAfrica;2014.27. FawcusS,PattinsonRC,MoodleyJ,MoranNF,SchoonMG,MhlangaRE,etal.Maternal deaths from bleeding associated with caesarean delivery: A nationalemergency.SAfrMedJ.2016;106(5):53-7.28. MaswimeS,BuchmannE.Causesandavoidablefactorsinmaternaldeathdueto cesarean-related hemorrhage in South Africa. Int J Gynaecol Obstet.2016;134(3):320-3.29. Gebhardt GS, Fawcus S,Moodley J, Farina Z.Maternal death and caesareansection in South Africa: results from the 2011-2013 saving mothers report of thenational committee for confidential enquiries into maternal deaths. SAMJ: SouthAfricanMedicalJournal.2015;105(4):287-91.30. Labrecque L, Provencal M, Caqueret A, Wo BL, Bujold E, Lariviere F, et al.CorrelationofcordbloodpH,baseexcess,andlactateconcentrationmeasuredwithaportabledevice for identifying fetal acidosis. JObstetGynaecolCan. 2014;36(7):598-604.

262

31. Karon BS, Scott R, Burritt MF, Santrach PJ. Comparison of lactate valuesbetween point-of-care and central laboratory analyzers. Am J Clin Pathol.2007;128(1):168-71.

263

CHAPTERNINE:IMPACTOFMATERNALHIVONUMBILICALCORD

LACTATEMEASUREMENTATDELIVERYINASOUTHAFRICANLABOUR

WARD

264

CHAPTEROVERVIEW

Asoutlined inChapter8,21%of thewomen included inour trialwereHIVpositive.

ThereareconflictingdataontheroleofHIVinfectioninintrapartumoutcomes.Itwas

thereforeimportanttoconsideriftheperinataloutcomesinourstudywereimpacted

bymaternalHIVstatus.

265

266

267

268

269

270

CHAPTERTEN:ATTITUDESTOWARDSTHEIMPLEMENTATIONOF

UNIVERSALUMBILICALARTERYLACTATEANALYSISINASOUTH

AFRICANDISTRICTHOSPITAL

271

CHAPTEROVERVIEW

The capacity to translate research into practice is potentially difficult in all areas of

medicine,nonemoresothanmaternalandnewbornhealthinLMIC,wherehighrates

of mortality persist despite clear evidence on the steps to prevention(1). Having

establishedtheutilityandeffectivenessofintroducingumbilicalarterylactate(withina

packageof interventionsincludingCTGeducation) inaresearchsettinginanLMIC, it

wasnextpertinent to consider the implementationof the tool inpractice. It is clear

thatoneofthepotentialchallengestoanynewtool inLMICsettings isthepotential

forburdenonalready limitedresources.Asoutlined inchapter3, thiscomponentof

theresearchwasconductedatZithuleleHospital intheeasterncapeofSouthAfrica,

whereweassessedstaffattitudestowardstheintroductionofUAlactatesampling.

272

273

274

275

276

277

278

CHAPTERELEVEN:DISCUSSION

279

OVERVIEW

Thisthesisconsistsofcomprehensiveseriesofclinicaltrialsanddatasetanalysesthat

demonstratethatthe introductionofumbilicalcordartery lactatesampling isuseful,

feasibleandacceptableinaresourcepoorMICsetting.Ihavesystematicallyexplored

the causes of perinatalmortality in South Africa, the diagnostic test accuracy ofUA

lactate,theroleofQoCauditinperinatalmortalityinSouthAfrica,andtheeffectsof

the implementation on UA lactate sampling. The key findings are summarised as

follows:

• In a comprehensive data set representative of a total obstetric population in

South Africa, 49.5% and 74.2% of women having a late stillbirth or early

neonataldeathwerehealthy(nocomplicationsofpregnancyor labour)anda

quarteroftheperinataldeathsfollowedintrapartumasphyxia.

• Inthefirstsystematicreviewofthediagnostictestaccuracyofumbilicallactate

for measuring acidosis and predicting neonatal outcome, including 38,284

patients, lactatewas shown to correlatewith pH (pooled ES -0.650 95% CI -

0.663to-0.637,p<0.01),BE(pooledES-0.71095%CI -0.721—0.699,p<0.01),

andApgarscoreat5minutes(pooledES0.395%CI0.193-0.407p<0.01).Using

thecommonthresholdoutcomeofHIE,lactatehasasensitivityof69.7%anda

specificityof93%.

• In the largest analysis of the impact of QoC audit on perinatalmortality,we

interrogatedthechangesinPNMRin163SouthAfricanfacilitieswithatleast5

280

yearsofcontinuousQoCauditdata.3,406,347birthsand85,728deathswere

included and 29% of facilities had a decrease in their PNMRwith 5 years of

continuousQoCaudit.Facilitieswithincreasingmortalityappearedlesslikelyto

haveacomprehensiveQoCauditprocess.

• Ina largecohortof946babieswithaUAlactatesample,wedeterminedthat

theoptimal cut-off for lactate for theoutcomeofneonatal resuscitationwas

5.46mmol/L (sensitivity 68%, specificity 72%), for neonatal admission was

4.95mmol/L (sensitivity61%,specificity59%),andforanApgarscoreof<7at

fiveminuteswas5.65mmol/L(sensitivity64%,specificity69%).

• In a before and after trial assessing the impact of introducing UA lactate

sampling,wefoundnodifferenceinmeanlactate(4.6mmol/L{95%CI4.4-4.8}

comparedwith4.9mmol/L{95%CI4.7-5.1},p=0.089).Howeversuspectedfetal

compromisewasreducedinthepost-interventionperiod:30·2%vs.22·1%,aOR

0·71, 95% CI 0·52-0·96, p=0·027 without worsening neonatal outcomes.

Caesarean section rateswere significantly reduced in the univariate analysis:

pre- 40·3% vs. post-intervention 31·6% (p=0·007). This reduction remained

significantwhenadjustedforpreviouscaesareansection,primiparity,maternal

HIV infection and preterm birth (aOR 0·72, 95%CI 0·54-0·98, p=0·035).

Moreover,emergency caesareandeliveries in theunit asawhole (not just in

patients recruitedtothetrial)decreasedfrom692/2436(28·4%)to399/2232

(17·0%,P<0·001), andneonatal admissionsdecreased from450/2436 (18·5%)

to349/2232(15·6%,P=0·010).

• Inthe21.6%ofwomeninourbeforeandaftertrialthatareHIVpositive,there

was no impact of either their disease or the use of HAART on the preterm

281

delivery rate, mode of delivery, neonatal resuscitation rate, 1 or 5 minute

Apgar score below 7, or the rate of having a UA lactate level greater than

5.45mmol/L.

• InthefinalstudyassessingtheattitudesandbarrierstointroducingUAlactate

sampling,themajorityofmidwivesanddoctorsprovidingintrapartumcareina

SouthAfricandistricthospitalwerepositiveaboutboththeroleofUAlactate

analysis as well as the potential benefits it provides. Training aided in

overcoming some of the perceived barriers to implementing UA lactate

sampling.

APPLICABILITYOFTHERESEARCH

THEMANAGEMENTOFINTRAPARTUMASPHYXIA

MostperinataldeathsinLMICarepreventable(2)anditisreiteratedrepeatedlyinthe

literature that the use of QoC measures is associated with improved maternal and

neonataloutcomes.Thisisparticularlytrueofintrapartumstillbirthsandintrapartum

relatedneonataldeaths(3,4),as is the focusof the research in this thesis. It is clear

that intrapartum hypoxia and asphyxia need to remain a focus of global efforts to

reduce perinatal mortality, as we have demonstrated in the analysis of causes of

perinataldeathinSouthAfrica(Chapter3)aswellasintheapplicationoftheICD-PM

to perinatal deaths (Appendix 2.1). It isworth noting that in high-income countries,

282

intrapartumasphyxiaisalsooftenassociatedwithstaff-relatedqualityofcareissues(5-

9),andsothereispotentialapplicabilityofthisthesisoutsideofanLMIC.

THEUSEANDBENEFITSOFUALACTATESAMPLING

Theuseof umbilical artery lactate in LMIC is only applicable to facility-basedbirths,

which remain uncommon in many LMIC(10). However, there are many places

internationallywithsimilarratesoffacility-basedbirthandintrapartumissuesasSouth

Africa. InBrazil, forexample,more thana thirdofperinataldeathsaresecondary to

asphyxia,andfacility-basedbirthisusual(11).Wewouldthereforeanticipatethatthis

trial could be replicated in other resource-limited settings. Equally, umbilical cord

assessmentofacidosis isusedfrequently inhigh-incomecountriesandwewereable

to conduct a large meta-analysis showing that this can conducted using lactate

measurementonasimplehandheldmachine,expandingtheapplicabilityofthisthesis.

While there is need for complex multifactorial approaches to improve perinatal

mortality inLMIC(12,13),theprocessofUAlactatesamplingcanformapartofthis.

However, itwasimportantthatweconsiderthebalanceof impactonresourceswith

thepotential for improvedoutcomeswhen implementingour trial. Simplemeasures

suchasuseofthepartograminlabour,orbasicneonatalresuscitationarecomponents

ofQoCthatarepositivelyassociatedwithmaternalandperinatalsurvival,andthese

basicmeasuresarecriticalbeforeconsiderationofcomplex technologicalanswers to

perinatalmortality(14-19).Thefeasibilityof introducingUA lactatesamplingoutlined

283

inChaptersixinadditiontothepositiveattitudesofstaffseeninthestudyinChapter

ninemakesitpossiblethatthisinterventioncouldbeintroducedalongsideotherQoC

measures. However, oversaturation of existing resources in combination with

deficiencies in QoC contributes to adverse maternal and perinatal outcomes in

LMIC(20)(Figure10.1).

Image11.1CrowdedantenatalwardatKalafongHospital

WhileIsawsomebenefitwiththeintroductionofUAlactatesampling(lower

caesareandeliveriesandreducedadmissions to thenursery), therewasno

apparent change in neonatal acidosis or mortality. It was important to

consider that the benefits of introduction were greater than any risks. In

284

Chapter2IshowedthePerinatalCareIndicators(PCI)forKalafongfor2013,

prior to the intervention and outlined below are the same PCI for 2014,

duringwhichourtrialwasconducted(Table11.1).Itwasreassuringtonote

that PNMR was lower in 2014 than in 2013, so while there was not a

significant result in our trial for this outcome, the introduction of the

samplingdidnotworsenKalafong’soutcomes.

Table11.1PerinatalcareindicatorsforKalafongHospital,2014

ItisalsoreassuringtoobservethattheUAlactatelevelsinwomenwithHIVinfection

in this thesiswere no differentwhen comparedwithwomenwithout HIV, and that

their neonatal outcomes were similarly comparable. The burden of HIV distribution

285

overlapswiththeglobalburdenofadverseperinataloutcomes,andUAsamplingcan

beconfidentlyimplementedinthesesettings.

TRANSLATINGRESEARCHINTOCLINICALPRACTICE

Itwas clear during the implementationof this quality of care intervention that staff

ownershipoftheprocesscontributedtothesuccessof thetrial.Bothmidwiferyand

medicalstaffbecameinvestedintheprocess,andcouldseethefruitsoftheir labour

on a daily basis (Image 11.2). A quality improvement program in India similarly

described health care training and ownership of a maternal and newborn quality

improvementprogramthat resulted insignificant results(21).The introductionofUA

lactatesampling inothersettingswillclearlyrequirestaffbuy-inandcommitmentto

eachpartoftheQoCauditprocess,ratherthanjustthesamplingitself.

286

Image 11.2 The primary author of this thesis and SisterMaria, Matron of Kalafong

Hospitallabourward

THEIMPLEMENTATIONOFQOCAUDIT

Audit can objectively identify areas where QoC interventions can be targeted.

However, our analysis demonstrated that audit alone does not necessarily improve

mortalitygiven thevariation indirectionof trends inmortalityof163 siteswith five

years of continuous audit. Thus, the assumption is false that once a problem is

identified health care providers will make an automatic correction and solve the

problem.TheintroductionoftheUAlactatesamplingandthedailyandweeklyreview

of cases in this thesis was a QoC audit process, which appeared to have some

beneficialoutcomes, includingareduction intherateofCS.Moreover, thereal-time

qualityofcarefeedbackmayhaveotherbenefits;immediatepostnatalcareimproves

probability of neonatal survival in sub-Saharan Africa(22) and so the process of

287

obtaining a lactate sample may identify an additional safety net for clinicians to

recognisetheat-riskneonate.

Thereremainseveralchallengestoimplementingaudits.Inorderforaudittohaveany

effectonperinatalcare,itmustbeimplementedinasystemthatissupportiveofthe

requirements for audit. An assessment in Dar es Salaam hospitals (with a hospital

basedMMRof1602/100000livebirthsandaPNMRof123/1000livebirths)showeda

lack of internal infrastructure for audit as well as clinical staff support and

responsibility.This resulted inanessentially ineffectiveprocess–withvery fewstaff

surveyedbeingabletonameasingleoutcome/changeasaresultofaudit(23).Audit

must leadintodatareportingtoappropriatehealthcareprofessionalsfollowedbya

processofimplementingchangetomakeitarelevanttool.QualitativeanalysisofPPIP

showsthattheabilitytosustainauditandimplementchangesrequiresmotivatedstaff

(amulti-levelandmulti-disciplinaryapproach) toconductandbe responsible for the

program(24). This is one of the potential challenges of audit as a tool to improve

perinatalmortality–ifitisonlyintroducedasapointintimecapturetoolratherthan

asanagentofprogresstochange,thenanyeffectshavethepotentialtobeveryshort

lived.

288

STRENGTHSOFTHERESEARCH

Thestrengthsofthisresearchlieinthelargeclinicaltrialssupportedbycomprehensive

dataanalysesoftherelatedevidence.Theassessmentofcausesofperinatalmortality

was from largepopulationbaseddata, truly reflectiveof theusual stateofobstetric

outcomesinSouthAfrica.Themeta-analysisofthediagnostictestaccuracyoflactate

samplingincludeabroadsearchtermsystematicreview,withmorethan40,000titles

andabstractsscreenedtoensureanycohortassessinglactateasameasureofacidosis

was captured. More than 38,000 women were included in the analysis in the first

comprehensivesystematic reviewofusing lactateasameasureofacidosis.TheQoC

auditdata focusedon facilitieswithat least fiveyearsof continuousaudit,andwith

morethanthreemillionbirthsandnearly100,000deathsincluded,thelargenumbers

likelyreducedtheimpactofanyindividualfacilityconfoundersontheimpactofQoC

audit.AssessingtheuseandimpactofUAlactateinoursettingwas,toourknowledge,

the largest cohort ofUA lactate samples in an LMIC,with cut-off findings similar to

high-incomecountries.TheintroductionofUAsamplingwaspartofacomprehensive

QoC audit program, including focused educations and daily and weekly review

meetings, which was known to be important in the successful implementation of

qualityofcareauditprogramsinSouthAfrica(24).Equallyourlargecohortmakesthe

assessmentof the impactofHIV infectiononUA lactateresultsmorerobust.Finally,

wehaveconsideredthechallengesoftranslatingresearchintopracticeinLMIC

289

LIMITATIONSOFTHERESEARCH

Much of our supporting data (causes of perinatal mortality and the impact of QoC

audit) was retrospective and therefore subject to confounders that we could not

control for. Thiswas particularly true for the analysis of the impact of audit,where

there were no overt differences in the data between the facilities with increasing

comparedtodecreasingmortality.Clearlythereissomethingdifferentoccurringhere,

butwehadnocapacitytointerrogatethisissue.

TheuseofUA lactatemeasuringrequiresbothphysicalandhumanresources.While

onthesurfacetheseseeminsignificant,inlowresourcesettingsthisislesslikelytobe

thecase.Asurveyof53deliveryunitsintheDemocraticRepublicofCongofoundthat

notonehadthealloftheequipmentnecessarytoexecutebasicemergencyobstetric

andneonatalcare(25).Moreover, theavailabilityofCTG in thesesettings is far from

universal(26). These deficiencies are replicated innumerable times in the literature

acrossAfrica(27,28).Wethereforerecognisethat inLMICthebasicstepsneededto

reduce the high perinatalmortality burdenwould and should take priority over the

introduction of umbilical artery sampling. However there is great variability in units

withinLMICandwehaveshownbothfeasibilityandeffectivenessof introducingthis

tool.

290

Therewerepragmaticcomponentstomain lactatestudy,whichcouldbeconsidered

limitations.Onlyumbilicalartery lactatesweremeasured,and theabsenceofpaired

arterial and venous sampling may have resulted in inadvertent venous sampling in

somecases.Whileoursystematicreviewshowedthatvenouslactatecorrelatedwith

neonatal acidosis, this limitation potentially altered the lactate results in the study.

Furthermore, the sampling of lactateswas essentially a convenience sample, within

thelimitationsofboththemainresearcheraswellasthelimitedhumanresourcesof

the Kalafong obstetric unit. The resultsmust be interpretedwithin the potential for

sampling error given only a proportion of deliveries during the study period had an

umbilicalartery lactatesampling.Thatsaid,therewasevidencethattheintervention

potentially impacted the care in theunit as awhole (reduction in caesarean section

ratesinalldeliveries)andsothesamplingandoutcomesmaybereflectiveoftheunit

overallduringthestudyperiod.

It isalsoclear thatwecannotseparate theCTGtrainingcomponentof thetrialwith

thelactatesamplingintermsoftheimpactonmaternalandneonataloutcomes.We

know that training in fetal monitoring in large cohort studies is associated with

improvedneonataloutcomes(29).However the lactatesamplingcanonlyadd to the

understanding of the fetalmonitoring,which is critical if themonitoring is to be of

benefitwithcorrectclinicalmanagementwhereappropriate(30,31).

291

FUTUREDIRECTIONSFORRESEARCH

Sincetheconductofourclinicaltrial,weareawareofasimilarcohortstudyinMalawi

being commenced to review umbilical arterial lactate and neonatal outcomes(32).

There is large variance in the levels of intrapartum resources, care andoutcomes in

LMIC and ongoing large cohort studies looking at the role of UA lactate to improve

intrapartumcarewouldbeappropriate.Weinitiallyhypothesisedthattheintroduction

of UA lactate sampling would result in an improvement in lactate levels over time,

however this was not borne out. It may be because our sample size was relatively

small compared to other large cohorts that have proved that hypothesis(33), and

further testing in larger LMICcohortswouldbeappropriate. Equallyweaccept that

thecomplexfacetsofbothantenatalandintrapartumcareinoursettingmaylimitthe

impactofthissingleintervention.

CONCLUSION

The initial concept for this thesis stemmed from the overwhelming burden of

intrapartumhypoxiaandasphyxiatoperinatalmortalityinLMIC.Aquarterofperinatal

deathsinSouthAfricafollowintrapartumasphyxiaandit ispossibleinthesesettings

to introduce quality of care audit that decrease perinatal mortality. The use of

umbilical artery lactate is both feasible and effective in South Africa, and can

contributetoimprovedmaternalandperinataloutcomes.

292

REFERENCES

1. Allanson ER, Tunçalp Ö, Vogel JP, Khan DN, Oladapo OT, Long Q, et al.Implementation of s in health facilities: a systematic review of cluster randomisedtrials.BMJglobalhealth.2017;2(2):e000266.2. TheL.Yemen'ssilentkillers.Lancet.2017;389(10070):672.3. Ariff S, LeeAC, Lawn J, Bhutta ZA.Global Burden, Epidemiologic Trends, andPrevention of Intrapartum-Related Deaths in Low-Resource Settings. Clin Perinatol.2016;43(3):593-608.4. CarloWA,McClureEM,ChombaE,ChakrabortyH,HartwellT,HarrisH,etal.Newborn care training of midwives and neonatal and perinatal mortality rates in adevelopingcountry.Pediatrics.2010:peds.2009-3464.5. Sadler LC, Farquhar CM, Masson VL, Battin MR. Contributory factors andpotentiallyavoidableneonatalencephalopathyassociatedwithperinatalasphyxia.AmJObstetGynecol.2016;214(6):747.e1-.e8.6. Eggermont M. Intrapartum care and substandard care: juridicalrecommendationstoreducetheriskofliability.ArchGynecolObstet.2015;292(1):87-95.7. Andreasen S, Backe B, Øian P. Claims for compensation after alleged birthasphyxia: a nationwide study covering 15 years. Acta Obstet Gynecol Scand.2014;93(2):152-8.8. Evers AC, Brouwers HA, Nikkels PG, Boon J, VAN EGMOND‐ LINDEN A,Groenendaal F, et al. Substandard care in delivery‐related asphyxia among terminfants:prospectivecohortstudy.ActaObstetGynecolScand.2013;92(1):85-93.9. BerglundS,GrunewaldC,PetterssonH,CnattingiusS.Riskfactorsforasphyxiaassociated with substandard care during labor. Acta Obstet Gynecol Scand.2010;89(1):39-48.10. Khanam R, Ahmed S, Creanga AA, Begum N, Koffi AK, Mahmud A, et al.AntepartumcomplicationsandperinatalmortalityinruralBangladesh.BMCPregnancyChildbirth.2017;17(1):81.11. AlmeidaMFB,KawakamiMD,MoreiraLMO,SantosR,AnchietaLM,GuinsburgR. Early neonatal deaths associated with perinatal asphyxia in infants >/=2500g inBrazil.JPediatr(RioJ).2017;93(6):576-84.12. Pasha O, McClure EM, Wright LL, Saleem S, Goudar SS, Chomba E, et al. Acombinedcommunity-andfacility-basedapproachtoimprovepregnancyoutcomesinlow-resource settings: a Global Network cluster randomized trial. BMC Med.2013;11(1):215.13. OsrinD,ProstA.Perinatalinterventionsandsurvivalinresource-poorsettings:whichwork,whichdon't,whichhavethejuryout?:BMJPublishingGroupLtd;2010.14. GetiyeY,FantahunM.Factorsassociatedwithperinatalmortalityamongpublichealth deliveries in Addis Ababa, Ethiopia, an unmatched case control study. BMCPregnancyChildbirth.2017;17(1):245.15. Kamath-RayneBD,Griffin JB,MoranK, JonesB,DownsA,McClureEM,etal.ResuscitationandObstetricalCaretoReduceIntrapartum-RelatedNeonatalDeaths:AMANDATEStudy.MaternChildHealthJ.2015;19(8):1853-63.

293

16. YismaE,DessalegnB,AstatkieA,FessehaN.CompletionofthemodifiedWorldHealthOrganization (WHO) partograph during labour in public health institutions ofAddisAbaba,Ethiopia.Reproductivehealth.2013;10(1):23.17. OdendaalHJ,GebhardtGS,TheronGB.Stillbirthratesinsingletonpregnanciesin a stable population at Karl Bremer and Tygerberg hospitals over 50 years.SAfr JObstetGynaecol.2013;19(3):67-70.18. Gans-Lartey F, O'Brien BA, Gyekye FO, Schopflocher D. The relationshipbetweentheuseofthepartographandbirthoutcomesatKorle-Buteachinghospital.Midwifery.2013;29(5):461-7.19. Alfirevic Z, Edwards G, Platt MJ. The impact of delivery suite guidelines onintrapartum care in 'standard primigravida'. Eur J Obstet Gynecol Reprod Biol.2004;115(1):28-31.20. MgayaA,Hinju J, KidantoH. Is timeof birth a predictor of adverseperinataloutcome?Ahospital-basedcross-sectional study ina low-resource setting, Tanzania.BMCPregnancyChildbirth.2017;17(1):184.21. SarinE,Kole SK,PatelR, SoodenA,Kharwal S, SinghR, et al. Evaluationof aquality improvement intervention for obstetric and neonatal care in selected publichealthfacilitiesacrosssixstatesofIndia.BMCPregnancyChildbirth.2017;17(1):134.22. SinghK,BrodishP,HaneyE.Postnatalcarebyprovidertypeandneonataldeathinsub-SaharanAfrica:amultilevelanalysis.BMCPublicHealth.2014;14:941.23. NyamtemaAS,UrassaDP,PembeAB,KisangaF,vanRoosmalenJ.Factors forchange inmaternalandperinatalauditsystems inDaresSalaamhospitals,Tanzania.BMCPregnancyChildbirth.2010;10(1):29.24. Belizán M, Bergh A-M, Cilliers C, Pattinson RC, Voce A. Stages of change: Aqualitative study on the implementation of a perinatal audit programme in SouthAfrica.BMCHealthServRes.2011;11(1):243.25. Ntambue AM, Malonga FK, Cowgill KD, Dramaix-Wilmet M, Donnen P.Emergencyobstetricandneonatalcareavailability,use,andquality:across-sectionalstudy in the city of Lubumbashi, Democratic Republic of the Congo, 2011. BMCPregnancyChildbirth.2017;17(1):40.26. LawrenceER,QuarshieEL,LewisKF,PeltzmanT,AndersonFJ. IntroductionofcardiotocographmonitoringimprovesbirthoutcomesinwomenwithpreeclampsiainGhana. International journal of gynaecology and obstetrics: the official organ of theInternationalFederationofGynaecologyandObstetrics.2016;132(1):103-4.27. KrukME,LeslieHH,VerguetS,MbarukuGM,AdanuRMK,LangerA.QualityofbasicmaternalcarefunctionsinhealthfacilitiesoffiveAfricancountries:ananalysisofnationalhealthsystemsurveys.LancetGlobHealth.2016;4(11):e845-e55.28. Lawani LO, Iyoke CA, Onoh RC, Nkwo PO, Ibrahim IA, Ekwedigwe KC, et al.Obstetricbenefitsofhealthinsurance:Acomparativeanalysisofobstetricindicesandoutcome of enrollees and non-enrollees in southeast Nigeria. J Obstet Gynaecol.2016;36(7):946-9.29. Brown L, Permezel M, Holberton J, Whitehead C. Neonatal outcomes afterintroduction of a national intrapartum fetal surveillance education program: aretrospective cohort study. The Journal of Maternal-Fetal & Neonatal Medicine.2017;30(15):1777-81.30. Woods D. Appropriate technology and education for improved intrapartumcareinunderresourcedcountries.SAfrJObstetGynaecol.2009;15(3).

294

31. Williams KP, Galerneau F. Intrapartum fetal heart rate patterns in thepredictionofneonatalacidemia.AmJObstetGynecol.2003;188(3):820-3.32. ChilindaGK,GadamaLA,StonesW.Point-of-careumbilicalarteriallactateandnewborn outcomes in a low resource setting: cohort study. BMC Res Notes.2018;11(1):477.33. White CR, Doherty DA, Henderson JJ, Kohan R, Newnham JP, Pennell CE.Benefitsof introducinguniversalumbilicalcordbloodgasandlactateanalysis intoanobstetricunit.AustNZJObstetGynaecol.2010;50(4):318-28.

295

APPENDICES

296

APPENDICESFORCHAPTER2

APPENDIX2.1:PAPERSBYTHETHESISAUTHORONTHEDEVELOPMENTOFTHEICD-PM

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329


330

APPENDIX3.1:PATIENTINFORMATIONANDCONSENTFORMFORTHE“INTRODUCTIONOF

UNIVERSALCORDLACTATESAMPLINGANDCORRELATIONWITHNEONATALOUTCOMES”

331

332

333

334

APPENDIX3.2:DATACOLLECTIONSHEET

335

336

APPENDIX3.3:PRETRAININGCTGTESTFORDOCTORS(QUESTIONS)

PretrainingCTGtest

KalafongHospital

Name:______________________________________

Demographics(pleasecircle)

Male/female

Traininglevel:

Internyear1

Internyear2

Registraryear1/2/3/4/5

Other__________________

PreviousCTGtraining:

337

Noformaltraining

ESMOElecture

Other__________________

1) Earlydecelerationsare:

a) Duetoinhibitionofthesympatheticnervoussystem

b) Usuallybelow100beats/min

c) Alwaysphysiological

d) Reproducedbypressureovertheposteriorfontanelle

2) Latedecelerationsduetoreflexvagalstimulationusuallyareassociatedwith:

a) Fetaltachycardia

b) Fetalbradycardia

c) Increasedbaselinevariability

d) Normalbaselinevariability

e) Decreasedbaselinevariability

3) Themostimportantfeatureoflatedecelerationsis:

338

a) Thefrequencyoftheiroccurrence

b) Thedepthofthedecelerations

c) Durationofthedecelerations

d) Thetimefromonsetofcontractiontoonsetofdeceleration

e) Thetimefromcessationofthecontractiontocessationofthedeceleration

4) Theaccelerationswhich frequentlyprecedeandfollowvariabledecelerationsare

saidtoresultfrom:

a) Adrenalinerelease

b) Atropinerelease

c) Umbilicalarteryocclusion

d) Umbilicalveinocclusion

e) Totalcordocclusion

5) WhichofthefollowingisNOTimportantintheinterpretationofthe

significanceofvariabledecelerations:

a) Baselinevariability

b) Thedepthofthedecelerations

339

c) Thedurationofthedecelerations

d) Anincreasingbaselinerate

e) Thedurationfromonsetofthecontractionstoonsetofthedecelerations

6) Which of the following tests for the assessment of fetal wellbeing has been

validated through rigorous scientific investigation, and found to reduce perinatal

mortality?

a) Antenatalelectronicfetalmonitoring

b) Intrapartumelectronicfetalmonitoring

c) Amnioticfluidvolumeassessment

d) FetalumbilicalarteryDopplerassessmentinlowriskpregnancies

e) FetalumbilicalarteryDopplerassessmentinhighriskpregnancies

340

7) Thefollowingcordbloodgasresultsuggests:

UmbilicalArtery UmbilicalVein

pH 6.983 6.999

pO2 10.5 24.5

pCo2 100.6 92.8

BaseExcess -11.2 -10.8

a) Acuterespiratoryacidosis

b) Acutemetabolicacidosis

c) Chronicrespiratoryacidosis

d) Chronicmetabolicacidosis

e) Chronicmixedrespiratoryandmetabolicacidosis

8) Intermittent fetal heart rate monitoring has the same outcomes as continuous

monitoringif:

a) 1:1dedicatedmidwife

b) Allwomenhaverupturedmembranes(SROM,ARM)

c) Allbabieswithmeconiumhavecontinuousmonitoring

341

d) All suspected abnormalities on intermittent monitoring are confirmed by

continuousfetalmonitoring

e) Alloftheabove

9) WhichofthefollowingassociationswithasinusoidalFHRtraceisNOTtrue:

a) Vasapraevia

b) Feto-maternalhaemorrhage

c) Normalbaselinevariability

d) Rhisoimmunisation

e) Normalbaselinerate

10) Inanaerobicmetabolism

a) ThetissuepHrises

b) TheFHRpatternisalwaysunchanged

c) 29timesmoreenergyisproducedduringthedegradationofeachglucose

molecule

d) Lacticacidaccumulates

e) ThetissuepHdoesnotchange

342

11) Thefrequencyofepisodesof fetalbreathingmovements (chestwallmovements)

onultrasoundexaminationisdecreasedinallthefollowingconditionsEXCEPT:

a) Labour

b) Normalfetaloxygenation

c) Hypoglycaemia

d) CNSdepressantdrugs

e) Fetalacidosis

12) Followingdeliveryanumbilicalarterialbloodgasanalysisisreportedas:

Umbilicalartery Umbilicalvein

pH 7.08 7.34

pO2 14.6 20.4

pCO2 76.7 38.6

BaseExcess -9.1 -4.8

a) Acuterespiratoryacidosis

b) Acutemetabolicacidosis

343

c) Chronicrespiratoryacidosis

d) Acutemixedrespiratoryandmetabolicacidosis

e) Chronicmixedrespiratoryandmetabolicacidosis

13) TheInternationalCerebralPalsyTaskforceconsensusstatementoutlinesessential

criteriathatmustbepresenttoconsiderthatacute Intrapartumhypoxiawasthe

causeofcerebralpalsy.Whicharethecorrectstatements?

1) Evidenceofametabolicacidosisinintrapartumfetal,umbilicalarterialcordor

earlyneonatalblood(pH<7.00andbasedeficit>=12mmol/L)

2) Earlyonsetofsevereormoderateneonatalencephalopathyininfantsof>=34

weeksgestation

3) Cerebralpalsyofthespasticquadriplegicordyskinetictype

4) Evidenceofametabolicacidosisinintrapartumfetal,umbilicalarterialcordor


(a) 1,2,3

(b) 2,3,4

(c) 2and4

(d) 1and3

(e) Alloftheabove

344

14) Followingdeliveryanumbilicalarterybloodgasanalysisisreportedas:

pH7.04

pO210

pCO250

BaseExcess–18

Thisdataisconsistentwith

a) Respiratoryacidosis

b) Mixedrespiratoryandmetabolicacidosis

c) Metabolicacidosis

d) Normalacidbasebalance

345

346

347

348

349

350

APPENDIX3.4:PRETRAININGCTGTESTFORDOCTORS(ANSWERS)

1. C

2. D

3. A

4. D

5. E

6. E

7. E

8. E

9. C

10. D

11. B

12. A

13. A

14. C

15. C

16. E

17. B

18. C

19. E

351

APPENDIX3.5:POSTTRAININGCTGTESTFORDOCTORS

PosttrainingCTGtest

KalafongHospital

1. FetalsleepstatesmaybeassociatedwithanyorallofthefollowingEXCEPT

a. AbsentVariability

b. NormalFHRvariability

c. Absentfetalbreathingmovements

d. Latedecelerations

e. Reducedfetalmovements

2. Earlydecelerationsare

a. Duetoinhibitionofthesympatheticnervoussystem

b. Usuallybelow100beats/min

c. Alwaysphysiological

d. Reproducedbypressureovertheposteriorfontanelle

e. Common

352

3. Whichofthefollowingisnotarecognisedcauseoflatedecelerations

a. Fetalheadcompression

b. Uterinehypertonia

c. Maternalhypotension

d. Uterinerupture

e. Placentalfailure

4. Latedecelerationsduetoreflexvagalstimulationusuallyareassociatedwith

a. Fetaltachycardia

b. Fetalbradycardia

c. Increasedbaselinevariability

d. Normalbaselinevariability

e. Decreasedbaselinevariability

5. Themostimportantfeatureoflatedecelerationsare

a. Thefrequencyoftheiroccurrence

b. Thedepthofthedecelerations

353

c. Durationofthedecelerations

d. Thetimefromonsetofcontractiontoonsetofdeceleration

e. Thetimefromcessationofthecontractiontocessationofthedeceleration

6. Theaccelerationswhichfrequentlyprecedeandfollowvariabledecelerationsare

saidtoresultfrom

a. Adrenalinerelease

b. Atropinerelease

c. Umbilicalarteryocclusion

d. Umbilicalveinocclusion

e. Totalcordocclusion

7. Regardingphysiologicalfetalheartratemechanismswhichoneofthefollowing

associationsisFALSE:

a. Earlydecelerationsareduetoheadcompression

b. Variabledecelerationsareduetocordcompression

c. Variabledecelerationsareduetofetalhypoxia

d. Latedecelerationsmaybeduetoareflexreceptorresponse

354

e.Latedecelerationsarealwaysduetomyocardialischaemia

8. WhenperforminganantenatalCTGafter10-20minutesnon-reactivitythemost

appropriateactionisto

a. Recommenddelivery

b. Informthemedicalofficer

c. Manuallypalpatethefetus(ifnotcontraindicated)andadministeradrink

tothemother

d. Repeatthetestthefollowingday

e. Performavaginalexamination

9. Which of the following tests for the assessment of fetal wellbeing has been

validated through rigorous scientific investigation, and found to reduce perinatal

mortality?

a. Antenatalelectronicfetalmonitoring

b. Intrapartumelectronicfetalmonitoring

c. Amnioticfluidvolumeassessment

d. FetalumbilicalarteryDopplerassessmentinlowriskpregnancies

e. FetalumbilicalarteryDopplerassessmentinhighriskpregnancies

355

10. Thefollowingcordbloodgasresultsuggests:

A V

pH 6.983 6.999

pO2 10.5 20.4

pCO2 100.6 92.8

BE -10.8 -11.2

a. Acuterespiratoryacidosis

b. Acutemetabolicacidosis

c. Chronicrespiratoryacidosis

d. Chronicmetabolicacidosis

e. Chronicmixedrespiratoryandmetabolicacidosis

11.Indicationsforcontinuousfetalmonitoringincludeallthefollowingexcept

a) Maternalautoimmunecondition(SLE)

b) Pregnancyinducedhypertension

c) Oligohydramnios

356

d) Allwomeninlabourat40weeksgestation

e) labourinduction

12. Intermittent fetal heart ratemonitoring has the same outcomes as continuous

monitoringif

a) 1:1dedicatedmidwife

b) Allwomenhaverupturedmembranes(SROM,ARM)

c) Allbabieswithmeconiumhavecontinuousmonitoring

d) All suspected abnormalities on intermittentmonitoring are confirmed by

continuousfetalmonitoring.

e) Alloftheabove

13.WhichofthefollowingassociationswithasinusoidalFHRtraceisNOTtrue

a) Vasapraevia

b) Feto-maternalhaemorrhage

c) Normalbaselinevariability

d) Rhisoimmunisation

e) Normalbaselinerate

14.Inanaerobicmetabolism

a. ThetissuepHrises

357

b. TheFHRpatternisalwaysunchanged

c. 29 timesmoreenergy isproducedduring thedegradationofeachglucose

molecule

d. Lacticacidaccumulates

e. ThetissuepHdoesnotchange

15. Followingdeliveryanumbilicalarterialbloodgasanalysisisreportedas

A V

pH 6.76 6.87

pO2 8.2 15.2

pCO2 50.2 58.4

BE -21.7 -20.1

Thesedataareconsistentwith

a. Respiratoryalkalosis

b. Acuterespiratoryacidosis

c. MixedRespiratory&Metabolicacidosis

d. Metabolicalkalosis

e. ChronicMetabolicacidosis

358

16. The frequency of episodes of fetal breathing movements (chest wall

movements)onultrasoundexaminationisdecreasedinallthefollowingconditions

EXCEPT

a. Labour

b. Normalfetaloxygenation

c. Hypoglycaemia

d. CNSdepressantdrugs

e. Fetalacidosis

17.WhichoneofthefollowingisNOTincludedintheBiophysicalprofile?

a. Fetalbreathingmovements

b. ReactiveCTG

c. FetalRapidEyeMovements

d. Fetalmovements

e. Fetaltone

359

18.Followingdeliveryanumbilicalarterialbloodgasanalysisisreportedas

A V

pH 7.08 7.34

pO2 14.6 20.4

pCO2 76.7 38.6

BE -9.1 -4.8

a. Acuterespiratoryacidosis

b. Acutemetabolicacidosis

c. Chronicrespiratoryacidosis

d. Acutemixedrespiratoryandmetabolicacidosis

e. Chronicmixedrespiratoryandmetabolicacidosis

19.Currentlythemostsensitivereflectionofbirthasphyxiais

a. umbilicalarterypH

b. Apgarscore

c. Neonataladmission

d. FirstneonatalpH

360

e. Fetalheartratepatterninlast60minutesoflabour

20.TheInternationalCerebralPalsyTaskforceconsensusstatementoutlinesessential

criteriathatmustbepresenttoconsiderthatacuteIntrapartumhypoxiawasthecause

ofcerebralpalsy.

Whicharethecorrectstatements?

1. Evidenceofametabolicacidosisinintrapartumfetal,umbilicalarterialcordor


2. Earlyonsetofsevereormoderateneonatalencephalopathyininfantsof>=34

weeksgestation

3. Cerebralpalsyofthespasticquadriplegicordyskinetictype

4. Evidenceofametabolicacidosisinintrapartumfetal,umbilicalarterialcordor


a. 1,2,3

b. 2,3,4

c. 2and4

d. 1and3

e. Alloftheabove

361

21.Regardingumbilicalarterylactate

a. Samplingshouldonlybeperformedoncordfromfulltermfetuses

b. Ismostlymaternalinorigin

c. IscomparabletopHinpredictingtheneonateatrisk

d. Indicateswhichneonatesneedresuscitation

22.Thisdataisconsistentwith

A V

pH 7.08 7.34

pO2 14.6 20.4

pCO2 76.7 38.6

BE -12.1 -11.8

a. AcuteRespiratoryacidosis

b. AcuteMetabolicacidosis

c. Normalacidbasebalance

d. ChronicMetabolicacidosis

e. Mixedrespiratoryandmetabolicacidosis

362

23.Inlabour,thepresenceofafetalheartaccelerationduringavaginalexamination

indicates:

a. Anacidoticfetus

b. Labourisprogressingnormally

c. Acaesareansectionshouldbeperformed

d. Normalfetaloxygenationin99%ofcases

e. Uterinehyper-stimulation

363

364

365

366

367

368

369

370

371

APPENDIX3.6:POSTTRAININGCTGTESTFORDOCTORS(ANSWERS)

1. D

2. C

3. A

4. D

5. A

6. D

7. E

8. C

9. E

10. E

11. D

12. E

13. C

14. D

15. E

16. B

17. C

18. A

19. A

20. A

21. C

372

22. E

23. D

24. C

25. B

26. D

27. E

28. E

29. C

30. B

31. E

32. C

33. E

373

APPENDIX3.7:PRETRAININGCTGTESTFORMIDWIVES(QUESTIONS)

KalafongHospital

PretrainingCTGtest


Male/female

Traininglevel:

Student

Midwife

Professionalnurse

Communityservice

Other___________________

PreviousCTGtraining

Noformaltraining

ESMOElecture

374

Other__________________

1. Earlydecelerationsare:

a. Duetoinhibitionofthesympatheticnervoussystem

b. Usuallybelow100beats/min

c. Alwaysphysiological

d. Reproducedbypressureovertheposteriorfontanelle

2. Themostimportantfeatureoflatedecelerationsis:





e. The time from cessation of the contraction to cessation of the

deceleration

3. Theaccelerationswhich frequentlyprecedeand followvariabledecelerations

aresaidtoresultfrom:

a. Adrenalinerelease

b. Atropinerelease

c. Umbilicalarteryocclusion

d. Umbilicalveinocclusion

e. Totalcordocclusion

375

4. Which of the following is NOT important in the interpretation of the

significanceofvariabledecelerations:

a. Baselinevariability


c. Thedurationofthedecelerations

d. Anincreasingbaselinerate

e. The duration from onset of the contractions to onset of the

decelerations

5. Intermittentfetalheartratemonitoringhasthesameoutcomesascontinuous

monitoringif:

a. 1:1dedicatedmidwife

b. Allwomenhaverupturedmembranes(SROM,ARM)

c. Allbabieswithmeconiumhavecontinuousmonitoring

d. All suspected abnormalities on intermittentmonitoring are confirmed

bycontinuousfetalmonitoring

e. Alloftheabove

6. Regardingumbilicalarterylactate





376

377

378

379

380

APPENDIX3.8:PRETRAININGCTGTESTFORMIDWIVES(ANSWERS)

1. C

2. A

3. D

4. E

5. E

6. C

7. C

8. E

9. C

10. E

381

APPENDIX3.9:POSTTRAININGCTGTESTFORMIDWIVES

PosttrainingCTGtest

KalafongHospital

1. FetalsleepstatesmaybeassociatedwithanyorallofthefollowingEXCEPT

a. AbsentVariability

b. NormalFHRvariability

c. Absentfetalbreathingmovements

d. Latedecelerations

e. Reducedfetalmovements

2. Whichofthefollowingisnotarecognisedcauseoflatedecelerations

a. Fetalheadcompression

b. Uterinehypertonia

c. Maternalhypotension

d. Uterinerupture

f. Placentalfailure

382

3. Themostimportantfeatureoflatedecelerationsare





f. Thetimefromcessationofthecontractiontocessationofthedeceleration

4. Intermittent fetal heart rate monitoring has the same outcomes as continuous

monitoringif

a. 1:1dedicatedmidwife

b. Allwomenhaverupturedmembranes(SROM,ARM)

c. Allbabieswithmeconiumhavecontinuesmonitoring

d. All suspected abnormalities on intermittent monitoring are confirmed by

continuousfetalmonitoring.

e. Alloftheabove

5.Currentlythemostsensitivereflectionofbirthasphyxiais

a. umbilicalarterypH

b. Apgarscore

c. Neonataladmission

383

d. FirstneonatalpH

e. Fetalheartratepatterninlast60minutesoflabour

6.Regardingumbilicalarterylactate





7. In labour, thepresenceofa fetalheartaccelerationduringavaginalexamination

indicates:

a. Anacidoticfetus

b. Labourisprogressingnormally

c. Acaesareansectionshouldbeperformed

d. Normalfetaloxygenationin99%ofcases

e. Uterinehyper-stimulation

384

385

386

387

388

389

390

391

APPENDIX3.10:POSTTRAININGCTGTESTFORMIDWIVES(ANSWERS)

1. D

2. A

3. A

4. E

5. A

6. C

7. D

8. C

9. B

10. E

11. E

12. E

13. C

14. E

392


393

APPENDIX5.1:SEARCHSTRATEGY

PUBMED 1 ("umbilical arteries"{MeSH Terms} OR “umbilical artery” OR “umbilical

arteries” OR "umbilical cord"{MeSH Terms} OR "umbilical cord blood" OR“umbilicalcordartery”OR“cordblood”)NOT(“animals”{MESHTerms}NOT“humans”{MESHTerms})

36795

2 ("lactic acid"{MeSHTerms}OR "lactic acid"OR "lactates"{MeSHTerms}ORlactateOR lactates OR "acidosis, lactic"{MESH Terms} OR “lactic acidosis”)NOT(“animals”{MESHTerms}NOT“humans”{MESHTerms})

70437

3 ("neonataloutcome"OR"neonataloutcomes"OR(("infant,newborn"{MESHTerms} OR neonat*) AND "risk factors"{MESH Terms}) OR “FetalHypoxia”{MESHTerms}OR“fetalhypoxia”OR“foetalhypoxia”OR“AsphyxiaNeonatorum”{MESH Terms} OR “birth asphyxia” OR "fetal asphyxia" OR"foetalasphyxia"OR“intrapartumhypoxia”OR"intrapartumfetalasphyxia"OR “intrapartum stillbirths” OR “intrapartum-linked neonatal deaths” OR"Perinatal Mortality"{MeSH Terms} OR “perinatal mortality” OR “perinatalmorbidity”OR“Neonatalnearmiss”OR“InfantMortality”{MESHTerms}OR“Neonatalmortality”OR"APGARscore"{MeSHTerms}OR“APGARscore”OR"fetal distress"{MeSH Terms} OR “fetal distress” OR “foetal distress” OR"Infant,newborn,diseases/mortality"{MeshTerms}OR"RespiratoryDistressSyndrome, Newborn"{MeSH Terms}) NOT (“animals”{MESH Terms} NOT“humans”{MESHTerms})

81649

4 1OR2OR3 1847545 1AND3 33576 2AND3 4947 1AND2 4098 1OR6 48039 EMBASE 1 Umbilicalartery,umbilicalcordblood,lacticacid,lacticacidosis 12 Lacticacidandfetusoutcome 153 Lacticacid,perinatalasphyxia,fetushypoxia 64 Lacticacid,perinatalasphyxia 875 Lacticacid,fetushypoxia 916 Umbilicalartery,umbilicalcordblood,lacticacid 247 Umbilicalcordbloodandfetusoutcome 1618 Umbilicalarteryandfetusoutcome 2609 1OR2OR3OR4OR5OR6OR7OR8 651 COCHRANE 1 Lactate 53262 Neonat* 88663 Umbilical 90614 2OR3 99765 1AND4 1872 CLINICALTRIALS.GOV 1 "umbilical arteries” OR “umbilical artery” OR “umbilical arteries” OR

"umbilical cord" OR "umbilical cord blood" OR “umbilical cord artery” OR“cordblood”OR“lactate”

1247

394

APPENDIX5.2:DATAFROMEACHINCLUDEDSTUDY,INCLUDINGTHETHRESHOLDSFOREACH

OUTCOME

395

Abessolo Chou Gjerris Labrecque Linet Martin Racinet Westgren White Wiberg Tuuli

Lactate foracidosispH

Sample size52tp9fp4tn 37 fn 2lactate>4.9pH<7.2

Samplesize200tp8 fp 15 tn169 fn 8lactate >6pH>7.15

Lactate foracidosisusingBE

Sample size52tp5fp0fn3tn44

Samplesize 44 tp4 fp 2 tn35 fn 4lactate >6,BE>-8

Lactate forHIE (studydefinedreference)

Sample size58. tp 15, fp10,tn33,fn0.high lactatedefined as4500-5100 - 2SDabove this.sens100,spec77

See studyfordef. forlactate>3.21 tp 1fp 404 tn3635 fn 5sample4045

Lactate>5.68samplesize21182tp 30 fp2497 tn18644fn12

Lactate>10tp3fp 314tn12609fn3

Lactate forpoorAPGAR

<7 at5/60 tp24 fp304 tn13293

396

fn95

Lactate forany poorneonataloutcome(studieddefinedreference)

Compositeneonataloutcome tp45 fp 4258tn 3598 fn9 forlactate>3.90

CorrelationRandR2forlactate andpH

Samplesize2554,r=-0.73;p<0.001

Sample size52 r=-0.678;p<0.001;r2=0.52

Samplesize 200r=-0.693;p<0.001;r2=0.48

Samplesize 410r=0.538p<0.001

Samplesize 867r=-0.75,sig

Samplesize 4045r2=0.20;p<0.001

CorrelationRandR2forlactate andAPGAR at5/60

Samplesize81.r=-.258;p=0.0010

Samplesize 200r2=0.1

CorrelationRandR2forlactate andSBE

Samplesize2554,r=-0.76;p<0.001

Sample size52 r=-0.719;p<0.001;r2=0.46

Samplesize 44 atrisk r=-0.837;p<0.0001

Samplesize 410r=-0.586p<0.001

Samplesize 867r=-0.79,sig

Samplesize 4045r2=0.29;r=-0.53p=0.001

CorrelationRandR2for

Samplesize

397

lactate andABE

2554,r=-0.83;p<0.001

398

APPENDIX5.3CHARACTERISTICSOFINCLUDEDSTUDIES(N=12)

Studycharacteristics

Countryandsetting

Patientsampling &#ofpatients

Patientcharacteristics Indextests Target condition andreferencesstandards

Flowandtiming

Abessolo2009

Gabon.Tertiaryhospital

Consecutive.81

Low risk, non-anomalouslivebirths

Umbilical venouslactate

Neonatal Apgar at 5minutes. Umbilicalvenous superoxidasedismutase andglutamateperoxidase

Consecutivedeliveries,samplestaken immediatelypost-delivery fromclamped cord. Alltestsperformedatthe same time.Allsamples takenfrom umbilicalvein.

Chou1998

Taiwan.Tertiaryhospital.

Consecutive.128

Consecutive patientsincluded, but classified byrisk. Low risk: no maternalmedical or obstetricproblems with a non-anomalousneonate,withnoevidence of prenatal orperinatal insult, at an

Umbilical arteriallactate, pyruvate,andlactate/pyruvate(LP)ratio

Neonatalencephalopathy(classified according toSarnat and Sarnat(1))UmbilicalarterialpH

Consecutivepatients with anarterial blood gas,samples takenfrom a clampedcord, using aheparinisedsyringe stored on

399

appropriate weight forgestation. High risk:maternal complication (atleast one of placentapraevia,placentalabruption,breech, APH, pre-eclampsia/eclampsia, IUGR,diabetes, cephalopelvicdisproportion, or ROM > 48hours)with a neonatalwithat least one of history ofabnormal fetal heart rate,meconiumaspiration,orlowApgar.

ice. Arterial bloodgasdonewithin5-10 minutes ordelivery, serumstored -20 andlactate andpyruvate donewithin 6-12 hoursofdelivery

Gjerris2008

Denmark.Universityhospital

Consecutive.2554

Allwomenregardlessofriskwithasingletonpregnancy

Umbilical arteriallactate

Umbilical arterial pH,ABE,SBE

Consecutivepatients had asampletakenafterdelivery from adouble clampedsegment ofarterialcord,usingheparinisedsyringes. Samplesweremeasuredonaverage6minutesafterobtaining.

400

Labrecque2014

Canada.Universityhospital

Consecutive.52

Successive singletondeliveries


Umbilical arterial pH,BE

Consecutivedeliveries with asample of arteriallactatetakenfroma double clampedsegment of cord,immediately post-delivery,testedona portable and alab analyser(lactate) and labanalyseralone(pHandBE)

Linet2002 France.Universityhospital

Consecutive.200

No inclusion characteristicsgiven


Asphyxia UmbilicalarterialpHandBE

Consecutivedeliveries withumbilical arterialsamples takenfrom a doubleclamped segmentof cord. Sampletaken before thefirst respiration,sample analysedwithin5/60

Martin2003

France.Universityhospital

Consecutive.763

No inclusion characteristicsgiven


Umbilical arterial pHandBE

Consecutivedeliveries,umbilical arterysampletakenfroma double clamped

401

segment of cordbeforethefirstcry

Racinet2013

France.Privateclinic

Consecutive.867

Low risk deliveries, >34weeksgestation


Umbilical arterial pH,BE,CO2

Consecutivedeliveries, withsampletakenfromclamped segmentof umbilical arteryafterdelivery

Tuuli2014 USA.Universityhospital

Consecutive.4910

Women who laboured atdelivery with a term, non-anomaloussingletonfetus


Composite neonataloutcome: Neonataldeath, intubation,mechanical ventilation,MAS, HIE, need forhypothermictreatmentUmbilicalarterialpH

Consecutivedeliveries, with asample collectedimmediately afterdelivery, beforeknowledge ofneonataloutcome. Sampleanalysed in thelaboratory

Watt

2002

Singapore.Tertiaryhospital

Unclear.130

Deliveriesattermgestation Umbilical arterialand venouslactate

Fetal distress on CTGUmbilical arterial andvenouspHandBE

Samples collectedpost-delivery inheparinisedsyringes,measured withinonehour

402

Westgren1995

Sweden.Universityhospitals

Consecutive.4045

Alldeliverieseligible Umbilical arteriallactate

NICU admission, MAS,IRDS, assistedventilation,neurologicalabnormality, neonataldeaths UmbilicalarterialpH,pCO2,pO2,HCO3,BE

Consecutive,samples takenimmediately afterdelivery from adouble clampedsegment of cord,all analysisperformed within5minutes

White2012

Australia.Tertiaryhospital

Consecutive.21182

All modes of delivery, allmaternal ages, andconditions,withdeliveryofafetus(s)>36weeksgestation

Umbilical arterialand venouslactate

Hypoxic ischemicencephalopathy(classified according toSarnat andSarnat(56))Umbilicalarterial and venous pHandBE

Consecutivedeliveries, with asampletakenfroma double clampedsegment of cord,straight afterbirth, using aheparinisedsyringe, stored onice

Wiberg2010

Sweden.Universityhospitals

Consecutive.13753

All planned vaginaldeliveries, with accuratedatingof gestationage, andasingletonfetus

Umbilical arterialand venouslactate

5/60 Apgar <7 or <4;HIEstage2-3Umbilicalarterial and venous pHandBE

Consecutivedeliveries, with asample from adouble clampedsegments of cord,taken immediatelyafter delivery andbefore newbornsfirstbreath

403


404

APPENDIX10.1QUESTIONNAIRE:ATTITUDESANDBARRIERSTOWARDSIMPLEMENTATIONOF

UNIVERSALCORDLACTATESAMPLING

Thisquestionnaireisanonymousandallyourresponseswillbetreatedconfidentially.

For thedemographic questionplease circle the corresponding answerwhilst for the

remainderofquestionsplease tick theanswer thatbestagreeswithyouropinionor

situation.


1. Currentpositioninwhichyouprovidematernitycare:

Midwife Doctor

Nurse

2. Age

20to29years 30to39years

40to49years ≥50years

405

3. Gender

Male Female

4. Lengthoftimepracticinginmaternitycare:

<1year 1-5years

6-10years >10years

PreviousexperienceofCTGuse(pleasecircle)

1. WhatisyourpreviousCTGtrainingexperience:

None

Atworkonly

Aspartofmedicalormidwiferydegree

Specifictrainingcourse

406

2. Ticktheresponsewhichbestagreeswithyouropinion

Strongly

Agree

Agree

Neu

tral

Disagree

Strongly

Disagree

I feel confident to identify the

normalfeaturesofaCTG

I feel confident to identify the

abnormalfeaturesofaCTG

PreviousExperienceofCaseReflection:

1.Ticktheresponsewhichbestagreeswithyouropinion

Strongly

Agree

Agree

Neu

tral

Disagree

Strongly

Disagree

I regularly attend monthly

perinatalmortalitymeetings

Perinatal mortality meetings are

effective in reducing adverse

neonataloutcomes

In attending perinatal mortality

meetings,Ihavelearntnewthings

In attending perinatal mortality

meetings I have changed my

407

practice

I feel thatmore couldbedone to

reducebadoutcomesduetobirth

asphyxia

PreviousExperienceofUmbilicalCordBloodGas/LactateAnalysis

1. Have you previously had experience in undertaking umbilical cord lactate

analysis?

Yes No

2. Ifyouhavehadpreviousexperienceinutilisingumbilicalcordlactateanalysisin

whichmannerwasitutilised:

Universal(e.g.alldeliverieshaveumbilicalcordbloodsamplesanalysed)

Selective(e.g.cordbloodsamplesareonlyanalysedonsomedeliveries)

Notapplicable

3. Ticktheresponsewhichbestagreeswithyouropinion

StronglyAgree

Agree

Neu

tral

Disagree

StronglyDisa

gree

I feel confident to take blood from the

408

umbilicalarteryofaplacenta

I feel confident to interpret the results of a

cordlactateresult

SpecificBenefitsofUmbilicalCordBloodGas/LactateAnalysis

Ticktheresponsethatbestagreeswithyouropinion

Anumberofpotentialbenefitsofumbilical

cordbloodgas/lactateanalysishavebeen

suggested:

StronglyAgree

Agree

Neu

tral

Disagree

StronglyDisa

gree

Objective measurement of status at birth

assistsinneonatalcare

An objective record of status at birth is

protective for staff in the event of medico-

legalissues

Umbilicalcordbloodgasorlactateanalysisis

cost-effective

Improved opportunities for audit and

teachingofmaternitycarepractice

BarrierstoImplementationofUniversalUmbilicalCordBloodGas/LactateAnalysis

Ticktheresponsethatbestagreeswithyouropinion

409

Anumberofpotentialimpediments

totheintroductionofuniversal

umbilicalcordbloodgas/lactate

analysishavebeensuggested: StronglyAgree

Agree

Neu

tral

Disagree

StronglyDisa

gree

Expense

Insufficienttimefollowingdelivery

Increased workload on

midwifery/medicalstaffmembers

Lackofnecessaryequipment

Lackofnecessaryexpertise

Encroachmentof technology into the

birth process (or ‘medicalisation’ of

birth)

Decreasedpatientcontacttime

FurtherComments

Doyouhaveanyfurthercommentsonumbilicalcordbloodgas/lactateanalysis?

410

APPENDIX10.2PATIENT/PARTICIPANT’SINFORMATIONLEAFLET&INFORMEDCONSENTFORM

PATIENT/PARTICIPANT’SINFORMATIONLEAFLET&INFORMED

CONSENTFORMFORCLINICALTRIAL/INTERVENTIONRESEARCH

TRIALTITLE:Feasibilityofumbilicallactatetestingasaclinicaltoolinalowandmiddle-

incomesetting

INTRODUCTION

Youareinvitedtovolunteerforaresearchstudy.Thisinformationleafletistohelpyou

todecide ifyouwould liketoparticipate.Beforeyouagreetotakepart inthisstudy

youshouldfullyunderstandwhatisinvolved.Ifyouhaveanyquestions,whicharenot

fullyexplained in this leaflet,donothesitate toask the investigator.Youshouldnot

agreetotakepartunlessyouarecompletelyhappyaboutalltheproceduresinvolved

WHATISTHEPURPOSEOFTHISSTUDY?

This studyaims to teachmidwiferyandmedical staffhow tomeasure the levelof a

substancecalled lactate in thebaby’sumbilicalcord,which isusedtodeterminethe

healthandwellbeingofnewbornbabies.

EXPLANATIONOFPROCEDURESTOBEFOLLOWED

411

The labouranddeliverywillproceedas itnormallywould.Afteryourbabyhasbeen

delivered theumbilical cord is clampedandcutand thebabywillbemanaged,as it

normallywouldbe.Theplacentawillthenbedelivered,as itnormallywouldbe.The

doctorwillthentakeasmallsamplefromaclampedsectionoftheumbilicalcordafter

itisdelivered.

HASTHESTUDYRECEIVEDETHICALAPPROVAL?

This clinical trial Protocolwas submitted to the Faculty of Health Sciences Research

Ethics Committee, University of Pretoria and written approval has been granted by

thatcommittee.ThestudyhasbeenstructuredinaccordancewiththeDeclarationof

Helsinki (lastupdate:October2008),whichdealswith the recommendationsguiding

doctors in biomedical research involving human/subjects. A copy of the Declaration

maybeobtainedfromtheinvestigatorshouldyouwishtoreviewit.

WHATAREYOURRIGHTSASAPARTICIPANTINTHISSTUDY?

Yourparticipation in this trial is entirely voluntary and you can refuse toparticipate

without stating any reason. Your withdrawal will not affect your access to other

medicalcare.

ISALTERNATIVETREATMENTAVAILABLE?

As this isnota study lookingat treatment,yourcarewillbeasnormalwhetheryou

choosetoparticipateornot.

412

MAY ANY OF THESE TRIAL PROCEDURES RESULT IN DISCOMFORT OR

INCONVENIENCE?

Theumbilicalcordwillhavealreadybeenremovedfromthebaby(asnormal)before

thesampleistakenandsonodiscomfortwillbefeltbyeitheryouoryourbaby.

WHATARETHERISKSINVOLVEDINTHISSTUDY?

Therearenomedicalriskstoyouoryourbaby.

ARETHEREANYWARNINGSORRESTRICTIONSCONCERNINGMYPARTICIPATIONIN

THISSTUDY?

No

INSURANCEANDFINANCIALARRANGEMENTS

N/A

SOURCEOFADDITIONALINFORMATION

Ifyouhaveanyquestionsregardingthetrial,youmaycontactDrKateGrobickiatany

time.Thetelephonenumberis0818062213

CONFIDENTIALITY

Allinformationobtainedduringthecourseofthisstudyisstrictlyconfidential.Asthe

sample of the umbilical cord will be used for training purposes only, no personal

informationwill begathered fromyouother than thatof your signedconsent form.

Theseformswillbeusedfornootherpurposesotherthanforconsent.Theumbilical

413

cordwillbedisposedwiththerestofyourplacenta immediatelyafterthetraining is

completed.

INFORMEDCONSENT

IherebyconfirmthatIhavebeeninformedbytheinvestigator,DrKateGrobickiabout

the nature, conduct, benefits and risks of the study “Feasibility of umbilical lactate

testingasaclinicaltoolinalowandmiddleincomesetting”.Ihavealsoreceived,read

and understood the above written information (Patient Information Leaflet and

InformedConsent)regardingthestudy.

Imay,atanystage,withoutprejudice,withdrawmyconsentandparticipation inthe

study. I have had sufficient opportunity to ask questions and (ofmy own free will)

declaremyselfpreparedtoparticipateinthestudy.

Patient'sname(Pleaseprint)

Patient'ssignature Date

I, Dr …………………………………. herewith confirm that the above patient has been

informedfullyaboutthenature,conductandrisksoftheabovestudy.

414

Investigator'sname(Pleaseprint)

Investigator'ssignature Date

Witness'sname*

Witness'ssignature Date

*Consentprocedureshouldbewitnessedwheneverpossible.

VERBAL PATIENT INFORMED CONSENT (applicable when patients cannot read or

write)

I,theundersigned,Dr……………………………………………,havereadandhaveexplainedfully

tothepatient,named……………………………………………………………..and/orherrelative,the

patientinformationleaflet,whichhasindicatedthenatureandpurposeofthestudyin

which I have asked the patient to participate. The patient indicated that she

understands that she will be free to withdraw from the study at any time for any

reason.

415

Iherebycertifythatthepatienthasagreedtoparticipateinthisstudy.

Patient'sName(Pleaseprint)

Investigator'sName(Pleaseprint)

Investigator'sSignature Date

Witness'sName(Pleaseprint)

Witness'sSignature Date

416

THE INTRODUCTION OF UMBILICAL ARTERY LACTATE ...

Documents