seminars in logy june2007 recurring complications of pregnancy

SEMINARS INPERINATOLOGY

T O P I C S F O R 2 0 0 6

O P T I M I Z I N G C A R E A N D O U T C O M E S F O R L A T E P R E T E R M( N E A R - T E R M ) I N F A N T S : P A R T 1

Tonse N. K. Raju, MD

O P T I M I Z I N G C A R E A N D O U T C O M E S F O R L A T E P R E T E R M( N E A R - T E R M ) I N F A N T S : P A R T 2

Tonse N. K. Raju, MD

A D V A N C E S I N N E O N A T O L O G Y : S E L E C T E D P R O C E E D I N G S O F T H EI N T E R N A T I O N A L P E R I N A T A L C O L L E G I U M

William Oh, MD, and Harry Bard, MD

B P D : S T A T E O F T H E A R T

Vineet Bhandari, MD, DM

C E S A R E A N D E L I V E R Y O N M A T E R N A L R E Q U E S T

Catherine Y. Spong, MD, and Uma M. Reddy, MD, MPH

I N H E R I T E D R E S P I R A T O R Y D I S O R D E R S O F T H E N E O N A T E

Lawrence M. Nogee, MD, and Aaron Hamvas, MD

T O P I C S F O R 2 0 0 7

N E O N A T A L I N F E C T I O N S

Robert S. Baltimore, MD, and Hal B. Jensen, MD

T H E E V I D E N C E B A S E S U P P O R T I N G N U T R I T I O N A LP R A C T I C E S F O R V E R Y L O W B I R T H W E I G H T I N F A N T S

Richard A. Ehrenkranz, MD, and Brenda B. Poindexter, MD, MS

R E C U R R I N G C O M P L I C A T I O N S O F P R E G N A N C Y

John C. Smulian, MD, MPH

C O A G U L A T I O N D I S O R D E R S I N T H E P E R I N A T A LP E R I O D

Michael J. Paidas, MD

P A I N

K. J. S. Anand, MD, and Richard W. Hall, MD

O R G A N T R A N S P L A N T A T I O N A N D R E P R O D U C T I O N

Lloyd Ratner, MD, and Mary D’Alton, MD

.h

Pogewrtwcpptcrcmnp“co

te

Volume 31, Number 3 June 2007

0d

Introduction

vtmwhqmtp

iarpdtfc

posits

. .Whoever wishes to foresee the future must consult the past; foruman results ever resemble those of preceding times.

Niccoló Machiavelli (The Discourses, 1517)

regnancy is associated with many unique health prob-lems that occur at no other time in life and in no specialty

ther than Obstetrics. Complications can be medical or sur-ical and may affect the mother, the baby, or both. What isven more unique is that we have the opportunity to followomen through successive pregnancies, each of which is at

isk for either occurrence or recurrence of these complica-ions. Whereas prediction of complications for nulliparousomen with no pregnancy track record is notoriously diffi-

ult, it is becoming clear that a history of a pregnancy com-lication is the greatest predictor of a future pregnancy com-lication. It is the role of research in complication recurrenceo grow our understanding of the heterogeneity of pregnan-y-associated diseases and give insight into etiologies andisk. This will ultimately lead to better clinical prediction,ounseling, and management. As clinicians, we should notanage these at-risk women the same as women who haveever had a pregnancy or who have had previous normalregnancy outcomes. We also should avoid the practice ofanecdotal medicine,” where wide variations in clinical carean be driven by either insufficient experience or the traumaf a previous adverse event.This issue of Seminars in Perinatology is organized around

he theme of recurrent pregnancy complications. There are
ight articles by recognized experts, of which seven are de-
146-0005/07/$-see front matter © 2007 Elsevier Inc. All rights reserved.oi:10.1053/j.semperi.2007.03.007

oted to some of the most significant pregnancy complica-ions encountered by clinicians. (There clearly are manyore pregnancy complications with a tendency to recur,hich deserve the same careful scrutiny as those reviewedere.) Because recurrence research is complicated and re-uires different thinking about epidemiologic and statisticalethodologies, there is included an important eighth article

hat addresses challenges specific to the study of recurrentregnancy complications.The selected topics covered in this issue illustrate how

mportant the concept of recurrence is to Obstetrics. Theserticles review what is known about the epidemiology ofecurrence for each specific complication, proposed etiologicathways, prevention options, and management recommen-ations for successive pregnancies. Each article also is meanto highlight knowledge gaps and areas with a pressing needor research. Unfortunately, our knowledge gaps about re-urrence are wide for many of these conditions.

It is my hope that this issue of Seminars in Perinatology willrovide information to directly help with the clinical care ofur highest risk women. I also hope that these articles willtimulate further research on the recurrence of complicationsn successive pregnancies and promote collaborations be-ween epidemiologists and clinicians to advance our under-tanding of this very important area of medicine.

John C. Smulian, MD, MPH
Guest Editor
125

FSW

NPedp

bggoptsrsIwltg1se

D

A

1

etal Growth Restriction andubsequent Pregnancy Risksendy L. Kinzler, MD, and Lillian Kaminsky, MD

Fetal growth restriction can result from a variety of intrinsic or extrinsic insults, resultingfrom maternal, fetal, and placental factors. Determining the underlying cause of poor fetalgrowth can be difficult but is essential for assessing potential risks for future pregnancies.Importantly, recurrence risks greatly depend on these underlying conditions. Understand-ing these risks can allow more appropriate patient counseling and may influence manage-ment strategies to optimize future pregnancies.Semin Perinatol 31:126-134 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS fetal growth restriction, recurrence risks

ETgttDtdrscve

FCtupmaFcgadtpwt

ormal fetal growth is dependent on complex interac-tions between the fetal, placental, and maternal units.

oor fetal growth can result from a variety of intrinsic orxtrinsic insults. Determining the underlying cause can beifficult but is essential for assessing potential risks for futureregnancies.Small for gestational age (SGA) infants are defined as those

orn with a weight �10th percentile for gestational age. Fetalrowth restriction (FGR) can be defined as an abnormalrowth trend, which is less than the genetic growth potentialf the individual fetus and which is always pathological. Thisathological FGR may affect only 3% to 5% of births. Impor-antly, not all SGA babies have FGR since some may be con-titutionally small. Conversely, a fetus still may be growthestricted if it is above the 10th percentile in weight if it isubstantially smaller than its growth potential would predict.t is important to accurately determine as best as possiblehether a previous pregnancy was complicated by patho-

ogic FGR or merely a constitutionally SGA infant. Informa-ion that might be helpful for establishing a diagnosis of fetalrowth restriction in a previous pregnancy is listed in Table. Once the diagnosis of FGR is established, an assessment ofubsequent pregnancy risk can be performed based on thetiology of the initial case.

ivision of Maternal Fetal Medicine, Department of Obstetrics, Gynecologyand Reproductive Sciences, UMDNJ-Robert Wood Johnson MedicalSchool, New Brunswick, NJ.

ddress reprint requests to Wendy L. Kinzler, MD, Department of Obstet-rics, Gynecology and Reproductive Sciences, Clinical Academic Build-ing, 2nd Floor, 125 Paterson Street, New Brunswick, NJ 08901. E-mail:

[email protected]

26 0146-0005/07/$-see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1053/j.semperi.2007.03.004

tiology-Specific Risks for FGRhere are multiple factors which can adversely influence fetalrowth. In broad terms, contributing factors can be intrinsico the fetus, they can be specific to the uteroplacental unit, orhey may be the result of underlying maternal conditions.etermining which of these was responsible for FGR can be

he most difficult but the most useful part of the evaluationuring a woman’s subsequent pregnancy. Prior medicalecords, including prenatal labs and notes, previous ultra-ound reports, maternal and infant hospital records, and pla-ental pathology, should be requested and carefully re-iewed, with particular attention being made to the followingtiologic pathways.

etal Causeshromosomal aberrations are a well-established cause of fe-

al growth restriction and are estimated to be responsible forp to 20% of cases.1 Early onset of growth restriction, theresence of polyhydramnios, and the presence of structuralalformations all increase the likelihood of a chromosomal

bnormality. Triploidy is most common in severe cases ofGR when �26 weeks gestation, and Trisomy 18 is the mostommon aneuploidy noted with severe FGR after 26 weeksestation.1 However, other trisomies, deletions/additions,nd ring chromosomes have all been associated with variousegrees of fetal growth abnormalities. Fetal aneuploidy is nothe only chromosomal abnormality to be associated withoor fetal growth. Confined placental mosaicism is presenthen the cytogenetics of the placental mass are different than

he cytogenetics of the fetus. It occurs either as the result of a
eiotic rescue of a trisomic embryo or due to a mitotic

poopfi

ainiwthc(gaira

tspt

hdiafbcniichi

tncqm

PUispmtmtotb1ithvhtc

itgottcdamAesrt

fihflcfmd

tcbal

TF

Fetal growth restriction 127

ostzygotic error.2 It has been found in approximately 15%f intrauterine growth restriction cases, compared with �2%f appropriately grown fetuses. The recurrence risk of aneu-loidy is approximately 1%, but a recurrence risk for con-ned placental mosaicism has not been established.Nonaneuploid genetic syndromes have also been associ-

ted with fetal growth restriction but may be more difficult todentify without a known family history and/or detailed ge-etic evaluation. Lower birth weights are also often identified

n infants born with structural malformations, especiallyhen cardiac, even in the presence of a normal fetal karyo-

ype.3 Uniparental disomy (UPD) is the inheritance of twoomologous chromosomes from only one parent. There arelinical syndromes which are known to be the result of UPDSilver Russell Syndrome, for example) and have significantrowth restriction as part of their phenotype.4 Although anssociation with poor fetal growth has been established, UPDs an uncommon finding in most cases.5 The risk for FGRecurrence from these causes is dependent on the specificssociated condition.

Congenital infections, namely rubella, cytomegalovirus,oxoplasmosis, herpes simplex, and varicella, have been as-ociated with FGR.6-9 Despite this clear association, the pro-ortion of growth restriction attributed to congenital infec-ion is low (5%),10 and they are not expected to recur.

For a variety of reasons, fetuses within multiple gestationsave an increased incidence of growth restriction. The inci-ence of growth restriction in twins is 15% to 25%,11-13 mak-

ng multiple gestations responsible for approximately 5% ofll cases of FGR. There is a significant increase in the rate ofetal growth abnormalities in direct relationship to the num-er of fetuses present.14 This may be related to a reduction inell size or a relative reduction in nutrient supply as theutritional demands of the fetuses increase. This form of FGR

s usually mild. Nevertheless, other contributing factors maynclude an increased incidence of placental and umbilicalord abnormalities (velamentous insertions), a greater likeli-ood of structural malformations and vascular anastomoses

able 1 Information Useful for Establishing a Diagnosis ofGR in a Previous Pregnancy

● Previous birth weight and gestational age to determinebirth weight centile

● Antepartum complications (bleeding, multiple gestation,congenital abnormalities, preeclampsia)

● Maternal pre-pregnancy BMI and weight gain duringpregnancy

● Social history (tobacco, alcohol, or illicit drug use)● Medication exposure● Prenatal fetal growth trends and ultrasound reports● Medical complications (chronic hypertension, IDDM,

SLE, anemia, asthma, etc.)● Birth weights of other pregnancies● Maternal birth weight● Neonatal complications (respiratory, metabolic,

congenital abnormalities, NICU admission, and lengthof stay)

n monozygotic multiples, and an increased incidence of ma- s

ernal complications linked with poor fetal growth. There areo data to assess recurrence risks for FGR when the indexase is a multiple gestation. However, the risks in a subse-uent singleton pregnancy are likely to be increased mini-ally, if at all.

lacental Factorsmbilical cord and placental abnormalities are frequently

dentified in pregnancies complicated by poor fetal growth. Aingle umbilical artery is present in approximately 0.4% ofregnancies15 and has been associated with fetal structuralalformations, particularly cardiac. In cases of an isolated

wo-vessel cord, FGR is estimated to be up to twice as com-on as in pregnancies with three-vessel cords.16 Velamen-

ous cord insertions, which enter the fetal membranes insteadf the placental parenchyma, have a reported incidence of 1%o 2%.15 They have been associated with higher rates of lowirth weight (OR 2.3), small-for-gestational age infants (OR.5), and preterm delivery (OR 2.1)17 compared with umbil-

cal cords normally inserted into the placenta. This suggestshat at least some of these low-birth-weight and SGA birthsave FGR. Circumvallate placentation is defined as an ele-ated edge of the placenta �50% of the circumference, andas been associated with poor fetal growth and prematuri-y.18 These factors are generally considered to have low re-urrence rates.

Placental bleeding at any trimester places a pregnancy atncreased risk for adverse outcomes. The incidence of firstrimester intrauterine hematoma is reported to be 3% in theeneral population with increased risks of hypertensive dis-rders of pregnancy (2- to 4-fold), abruption (5-fold), pre-erm delivery (2-fold), and FGR (2.4-fold).19 Abruption aterm is also significantly associated with FGR (2-fold in-reased risk), stillbirth, preterm delivery, and pregnancy-in-uced hypertension.20,21 Placental bleeding at any gestationalge in a pregnancy complicated by FGR may be a clinicalanifestation of a chronic placental disorder that can recur.lthough the majority of low-birth-weight neonates deliv-red from pregnancies complicated by placenta previa aremall due to preterm gestational age at birth, at least oneeport estimated that as much as 3.7% of FGR is attributableo placenta previa.22

Microscopically, many placental lesions have been identi-ed in cases of abnormal fetal growth. Placental infarctionsave been associated with FGR and abnormal fetal bloodow.23 Placental examination from term pregnancies compli-ated by idiopathic FGR also found a high incidence of in-arction (24%).24 The placenta can also be the target of im-

une-mediated injury, such as occurs with massive fibrineposition and chronic villitis.Massive perivillous fibrin deposition, previously referred

o as maternal floor infarct, is a significant placental lesionharacterized by a heavy deposition of fibrin in the deciduaasalis and extending into the intervillous space, preventingppropriate maternal–fetal exchange of nutrients. It is be-ieved to be the result of an immune-mediated maternal re-
ponse and has been associated with high rates of poor fetal

o6

lfFtfdtwf

MMgcovasf

rccuwWhd

rtaafhr

cgtiawbaapwfdvwip

arc

hvsasaCipgmnctsfac2cnaT

tctIsbaccFairriltlc

EAOetsgs

128 W.L. Kinzler and L. Kaminsky

utcomes, including a 15% to 40% fetal death rate, a 30% to0% preterm birth rate, and a 50% to 100% rate of FGR.25,26

Another potential immune-mediated lesion is chronic vil-itis of unknown etiology (VUE). Chronic villitis has beenound in approximately 30% of pregnancies complicated byGR.24,27 Other, rare placental abnormalities include placen-al mesenchymal dysplasia, which is a placenta vascular mal-ormation that is often confused clinically with a partial hy-atiform mole. This abnormality can predispose to placentalhrombosis. In one series, 50% of the cases were associatedith FGR and 43% of the pregnancies were complicated by a

etal death.28

aternal Factorsultiple maternal factors have been implicated in poor fetal

rowth. Careful review of the maternal history for potentialontributing factors can sometimes identify nutritional dis-rders, anemia and maternal hypoxia-related conditions, en-ironmental exposures (particularly tobacco or cocaine use),nd conditions with maternal vascular disease. Any reviewhould include a family history, as it has been shown thatamilial factors influence the risk of SGA births.29

Malnutrition is an uncommon, but nonetheless important,isk factor for poor fetal growth, especially in developingountries. The effect of starvation was best studied in a Dutchohort during the famine of 1944 to 1945. Pregnant womennder such severe nutritional deprivation experiencedeight loss and a drop in birth weight by 250 to 300 g.30

omen with eating disorders have also been shown to haveigher rates of SGA births.31 Correction of such nutritionaleficiency is likely to reduce recurrence risk.Maternal conditions leading to hypoxemia can lead to a

eduction in fetal growth potential. Studies have inconsis-ently linked maternal anemia, including sickle cell anemia,nd low birth weight.32-34 Chronic respiratory disease, suchs asthma, can also lead to decreased fetal growth throughetal hypoxia.35 Likewise, women with congenital cyanoticeart disease and those living at high altitude are at higherisk for lower birth weight babies.36,37

Maternal substance abuse, including tobacco, alcohol, andocaine, is an important preventable cause of poor fetalrowth. The mechanism may involve direct toxic damages ofhese substances as well as associated comorbidities, such asnadequate nutrition, maternal infections, etc. Smoking isssociated with an increased risk of delivering a low-birth-eight or SGA infant in a dose-dependent fashion.38,39 It haseen estimated that approximately 20% of low-birth-weightnd SGA births are attributable to maternal smoking.40 Badand coworkers estimated that, if smoking could be com-letely prevented during pregnancy, 13.8% of FGR casesould be eliminated.41 FGR is one of the major features of

etal alcohol syndrome, which is also associated with facialysmorphia and central nervous system disorders.42 The ad-erse effects of prenatal cocaine exposure on fetal growth43 asell as on placental abruption, preterm birth, and intrauter-

ne fetal demise have been well established. Various thera-
eutic drugs, such as warfarin, folic acid antagonists, and c
nticonvulsants, can also be associated with FGR. The risk forecurrence of FGR for these exposures is likely linked to theirontinued presence or their discontinuation.

Clinical maternal vascular disease secondary to chronicypertension, renal disease, diabetes mellitus, and collagenascular disease, especially when complicated by preeclamp-ia, is the most common cause of impaired fetal growth,ccounting for nearly a third of FGR cases.2 Both preeclamp-ia and FGR may share a similar pathophysiology involvingbnormal placentation leading to placental insufficiency.44

hronic hypertension is associated with a two- to threefoldncrease in the rate of FGR, an association mostly due to theresence of superimposed preeclampsia.45 In fact, whenrowth restriction cases due to preeclampsia, smoking, andalnutrition are excluded, diabetes and hypertension mayo longer be independent risk factors.46 The association ofhronic hypertension and poor fetal growth may also be par-ially mediated by the effect of antihypertensive medications,uch as beta-blockers.47 The risk of FGR in pregnancies af-ected by pregestational diabetes is dependent on the severitynd duration of the disease.45 The incidence of FGR inhronic renal disease has been reported to be as high as3%.48 Patients with systemic lupus have an eightfold in-reased risk for developing FGR,45,49-51 with the highest risksoted in women diagnosed before pregnancy and those withctive renal and central nervous system involvement.49-51

hese risk factors will persist in subsequent pregnancies.Inherited thrombophilias are a group of genetic conditions

hat increase the risk of thromboembolic disease. They in-lude Factor V Leiden (FVL), prothrombin G20210A muta-ion, protein C deficiency, protein S deficiency, antithrombinII deficiency, and MTHFR mutation, as well as other, lesstudied, familial conditions. Theoretically, placental throm-osis in patients with inherited thrombophilias may lead ton increased risk for FGR, although the published data areonflicting. A number of studies have demonstrated an asso-iation between maternal inherited thrombophilias andGR.52-55 However, other studies were unable to show anssociation.56-59 Based on the conflicting current data, inher-ted thrombophilia in isolation does not seem to be a majorisk factor for most cases of FGR. However, the contributingole of thrombophilias in combination with other risk factors,ncluding previous FGR, may be important. Antiphospho-ipid antibody syndrome (with or without underlying sys-emic lupus) is an acquired thrombophilia, which has beeninked to several recurrent adverse pregnancy outcomes, in-luding FGR, fetal death, and recurrent miscarriages.60,61

valuation Options tossess FGR Etiologies and Risksnce the diagnosis of FGR has been confirmed and the above

tiologies have been evaluated by history and examination,here may continue to be gaps in knowledge that may have aignificant impact on further management. Some of theseaps, however, can be filled with the addition of a few simpleteps. If not already done, consider having the previous pla-
ental slides reviewed for histologic abnormalities by a peri-

npptiupoclbwatGfagidwia

tbamomtppadcid

totldl

ATtttRFsevi

wipdi1eetnceiof

ahasicfISstt

tehnFTvcl

PTocpair

sstmbctt


atal pathologist. If there is evidence of a prior abruption, orlacental thromboses and/or infarctions are noted, it is im-ortant to assess for maternal risk factors that may contributeo coagulation abnormalities. Testing for acquired and inher-table thrombophilias is recommended in these cases, partic-larly when the growth restriction is severe, associated withreterm delivery, or noted in the presence of a family historyf thromboembolic disease. Specific evaluations would in-lude maternal testing for lupus anticoagulant, anticardio-ipin antibodies (IgG and IgM), and �2-glycoprotein-1 anti-odies. It is important to remember that 10% to 15% ofomen with systemic lupus erythematosus will have second-

ry antiphospholipid antibody syndrome and should beested if their antiphospholipid status is not already known.62

enetic thrombophilia testing should include an evaluationor deficiencies in protein C, protein S, and antithrombin IIInd the presence of the Factor V Leiden and prothrombinene mutations. If immune-mediated placental dysfunctions suspected, testing for the antiphospholipid antibody syn-rome and screening for other autoimmune disorders arearranted. There is insufficient information to support test-

ng the affected offspring for the presence of thrombophiliast this time.

If the prior child was found to have structural malforma-ions (prenatally or postnatally diagnosed), and/or there haveeen concerns about development delay, genetic counselingnd possible evaluation by a pediatric geneticist is recom-ended. Review of prior records, including autopsy reports

r infant discharge summaries, can be very helpful, as parentsay not be fully aware of known or suspected diagnoses. If

here has been a documented or suspected case of aneu-loidy, parental karyotypes should be considered. This isarticularly important for cases of aneuploidy other than theutosomal trisomies thought to be the result of meiotic non-isjunction, such as unbalanced translocations and ringhromosomes. A formal genetics evaluation is also essentialn detecting and counseling about nonaneuploid genetic syn-romes.In women with an underlying medical condition, assessing

he severity of the disorder can also provide insight into thengoing obstetrical risks. This would include determininghe presence of renal dysfunction in women with systemicupus or other autoimmune processes and evaluating theegree of vascular and/or end-organ disease in women with

ong-standing diabetes or chronic hypertension.

ssess Risks for a Subsequent Pregnancyhe diagnosis and/or risk factors that have been identified in

he above steps will allow appropriate counseling regardinghe risk(s) to subsequent pregnancies (Table 2). It is clear thathe data on etiology-specific FGR recurrence risks are sparse.ecently, it has been recognized that women with a history ofGR are at risk for multiple adverse pregnancy outcomes inubsequent pregnancies that include recurrent FGR, pre-clampsia, and abruption.63 These risks are particularly rele-ant in the setting of an acquired or inherited thrombophil-
a.64-68 The recurrence rate of adverse pregnancy outcomes in r
omen with untreated inherited or acquired thrombophilias approximately 66% to 83%.63,69 A cohort study of 491atients with a history of adverse pregnancy outcomes hasemonstrated that the presence of a maternal thrombophilia

s associated with fetal loss after 14 weeks (OR 3.4, 95% CI.9-6.1), abruption (OR 3.6, 95% CI 1.4-9.1), and pre-clampsia (OR 3.2, 95% CI 1.2-8.6).70 Women who deliv-red infants �3rd percentile may be three times more likelyo have initial or recurrent preeclampsia in the next preg-ancy compared with those who deliver infants �10th per-entile.71 They may also be at increased risk of fetal demise,specially when the previous fetal growth abnormalities weredentified at early gestational ages.72 Interestingly, of the 23%f women experiencing recurrent SGA in one study, the pooretal growth was not more severe the second time.73

Previous abruption is a risk factor for FGR, as well as otherdverse obstetrical outcomes. Recurrent placental abruptionas been observed in 22% of subsequent pregnancies.74 Aftern abruption, the risk of a SGA infant is 18.5%, the risk of apontaneous preterm birth is 36%, and the risk of pregnancy-nduced hypertension is 6%.75 A previous SGA birth in-reases the risk of abruption by 1.6-fold and by 2.5- to 6.0-old if there is preexisting chronic hypertension or diabetes.76

f a first delivery has been complicated by a preterm birth,GA infant, or perinatal death, the risk of an abruption in theecond pregnancy is 7/1000 if no abruption was present inhe initial pregnancy and 33/1000 if there was a prior abrup-ion.77

If there has been a history of FGR and placental infarction,he risk of recurrent growth restriction is high. It has beenstimated to be 61% when two or more prior pregnanciesave been similarly affected.78 Recurrent villitis has beenoted in 17% of gestations, and has been associated withGR, pregnancy loss, preterm delivery, and postnatal death.here is an even higher rate of recurrence of massive inter-illous fibrin deposition and poor fetal growth (67%).79 Pla-ental mesenchymal dysplasia is a rare finding with an un-ikely risk of recurrence.

revention of Recurrent FGRhe importance of a thorough evaluation for potential etiol-gies in evaluating a woman’s risk of adverse obstetrical out-omes as the result of a prior FGR birth cannot be overem-hasized. A comprehensive assessment will lead toppropriate counseling and, in many cases, will allow themplementation of targeted risk-specific strategies to reduceecurrence (Fig. 1).

If there is an opportunity for preconceptional care, ithould focus on the elimination of known maternal expo-ures (cocaine, smoking, alcohol), folic acid supplementationo reduce the risk of congenital malformations, and the opti-ization of maternal medical conditions. Low prepregnancy

ody mass index (BMI) is a risk factor for FGR. If a womanontinues to have a suboptimal BMI in the next pregnancy,his risk persists. However, an increase in maternal BMI be-ween pregnancies has been associated with a decreased SGA
isk.80 Although prospective trials of weight gain are lacking,

paotwpm

ttcm

sagpaisihba

saotataldobdaiat

nii9wsv

TP

FASTUUASCMMMM

SM

IP

PSVCPVM


atients should be counseled about this potentially modifi-ble risk factor. Other nutritional supplements, such as fishil, are unproven.81 In cases of underlying medical condi-ions (diabetes, systemic lupus, hypertension, asthma), theoman’s health should be optimized before subsequentregnancies and actively managed during the pregnancy in aultidisciplinary fashion.If assisted reproduction is required, there should be cau-

ious use of ovulation induction agents and attention paid tohe number of embryos transferred as it relates not only to thehance of successful implantation, but also to the risks ofultifetal gestations.Any pregnancy at risk for a fetal growth abnormality

hould be screened in the first trimester to establish early andccurate gestational dating. Subsequent assessments of fetalrowth trends will depend heavily on accurate dating. Whenossible, a crown–rump length in the first trimester is best. Ifsecond trimester ultrasound alone is available due to a delay

n presenting for prenatal care, the transcerebellar diameterhould be measured, as it provides the most accurate datingn the 2nd and even 3rd trimesters.82 For women at theighest risk for FGR, serial fetal growth ultrasounds shoulde considered at approximately 4- to 6-week intervals tossess fetal growth trends.

able 2 Recurrence Risks of Etiologies and FGR Based on Etregnancy1,3,10-13,16-18,20,21,24,25,27,28,45,48,52-55

Previous Pregnancy ConditionRisk of R

Factor in

etalutosomal trisomy 1% or maternex chromosome abnormality <1%riploidy <1%nbalanced translocation, de novo <1%nbalanced translocation, inherited Variableutosomal recessive conditions 25%tructural malformations, isolated Variable, 3-5%ongenital infection <1%ultifetal gestation Variable, 3% oaternalalnutrition Situation depeaternal conditions leading to hypoxemia Variable depe

ubstance abuse Situation depeaternal vascular disease (chronichypertension, renal disease, diabetesmellitus, collagen vascular disease)

Generally pers

nherited or acquired thrombophilia Generally perslacental abruption 22%

lacentalingle umbilical artery 0.4%elamentous cord insertion 1-2%ircumvallate placentation 6%lacental infarction 24%illitis 17%assive perivillous fibrin deposition 67%

In those couples at risk for having an aneuploid conceptus, O

everal options for screening and prenatal diagnosis are avail-ble. In couples at high risk of recurrent autosomal trisomiesr in those in which a balanced translocation has been iden-ified, in vitro fertilization with preimplantation genetic di-gnosis can be offered. Others may opt for assisted reproduc-ion with the use of a gamete donor. Once a pregnancy ischieved, early prenatal diagnosis with either chorionic vil-ous sampling or amniocentesis is available. For women un-ecided about invasive testing, prenatal screening should beffered. First trimester combined nuchal translucency andiochemical screening is an excellent start, with a subsequentetailed sonographic fetal anatomy survey. For some non-neuploid genetic syndromes, genetic testing is available. It ismportant to involve a provider specialized in genetics tossist in coordinating this testing and also to aid in targetinghe fetal ultrasound.

Detailed sonography should also assess umbilical cord ab-ormalities, such as single umbilical artery. Placental cord

nsertion should be visualized to rule out a velamentous cordnsertion. The placental cord insertion can be identified in9% of cases with gray scale and color Doppler ultrasound,83

ith identification of a velamentous insertion having a sen-itivity of 100%, specificity of 99.8%, a positive predictivealue of 83%, and a negative predictive value of 100%.84

of the Poor Fetal Growth in the Previous

rence of Etiologicequent Pregnancy

Risk of SGA withRecurrent Risk Factor

-related risk Abnormality-specific riskAbnormality-specific risk100%Abnormality-specific riskAbnormality-specific riskAbnormality-specific risk

ultifactorial inheritance Abnormality-specific riskInfection-specific risk

births Up to 25%

Minimal risk if treatedon condition Largely unknown, but likely

higher than background riskSubstance-specific riskLargely unknown, but consider

very high risk (at least 50%)

Up to 30-83% if untreatedAt least 18%, even without

recurrent abruption

Up to 7%15-20%Largely unknown61% untreated53%50-100%

iology

ecurSubs

al age

for m

f live

ndentnding

ndentistent

istent

ther placental anomalies, such as placenta previa and cir-

cdaob

sfl�pmvpcp

pmbawlab

ooibFfc

tarhcpsAwano

aoa

gnanc


umvallate placentation, should be noted. The sonographiciagnosis of placental abruption relies on the identification ofthickened placenta, a hematoma (retroplacental, subchori-nic, or preplacental),85 or the presence of intraamnioticlood.Uterine artery Doppler velocimetry has been utilized as a

creening tool for pregnancies at high risk of complicationsrom ischemic placental disease. At 20 weeks’ gestation, bi-ateral diastolic notching and mean resistance index of

90th%ile had a positive predictive value of 57% for severereeclampsia � FGR and a 93% positive predictive value forild or severe disease.86 Even in women with previous ad-

erse pregnancy outcomes on low-dose aspirin therapy, theresence of a diastolic notch at 23 weeks gestation was asso-iated with a higher rate of vascular complications such asreeclampsia and FGR (31% versus 5%).87

There may be a modest benefit of low-dose aspirin use inregnancies at high risk for poor fetal growth from specificaternal or placental conditions. Although a benefit has not

een demonstrated by all studies,88 many have demonstratedsignificant reduction in the risk of FGR among high-riskomen treated with low-dose aspirin.89,90 A meta-analysis of

ow-dose aspirin use suggested an 18% reduction in FGR, butmuch greater effect (OR 0.35) when therapy was instituted

Identify petiol

Maternal

Maternal Hypoxia-related conditions

Maternal Vascular Disease

Substance Abuse

Thrombophilias

Malnutrition Increase BMI Nutrition consult

Optimize underlying disease

Counseling/DetoxSmoking cessation

Optimize underlying disease

Low-d

Heparin

Fetal GRestri

GPaPT

PlacentalAbruption

Preeclampsia

Uterine artery

Figure 1 Management of subsequent pre

efore 17 weeks gestation. Given the excellent safety profile v

f low-dose aspirin use in pregnancy, it seems reasonable toffer this to women at significant risk of recurrent FGR, start-ng as early in the pregnancy as possible. It has been found toe especially beneficial in those with a history of recurrentGR and placental infarction, reducing the incidence of FGRrom 61% in the untreated to 13% in the treated pregnan-ies.78

In the setting of antiphospholipid antibody syndrome (ei-her primary or secondary), women should receive low-dosespirin and prophylactic heparin during pregnancy. Severalandomized control trials have investigated this issue andave demonstrated improved outcomes compared with pla-ebo or with aspirin alone.91-94 The optimal management ofregnancies complicated by recurrent chronic villitis or mas-ive fibrin deposition has not been definitively determined.lthough low-dose aspirin and/or heparin have been utilizedith some benefit,79 these lesions are not the result of a co-

gulation abnormality and may be best treated with intrave-ous immunoglobulin95 due to the suspected immune etiol-gy.In high-risk women with inherited thrombophilias and

dverse pregnancy events, heparin has been used to improvebstetrical outcomes. Heparin prophylaxis (compared withspirin) has been shown to improve live birth rates (86%

e

Fetal Placental

osomal malities

hromosomal malities

Umbilical Cord Abnormalities

Immune-mediated Injury (fibrin deposition, chronic villitis)

A

1st trimester CRL Serial growth ultrasounds

Ultrasoundevaluation

IVIG

counseling diagnosis (CVS, ntesis) screeningd ultrasound

ers

Placental infarction

Low-doseASA ± heparin

y based on presumed etiology of FGR.

robablogy

ChromAbnor

Non-cAbnor

ose AS

rowth ction

eneticrenatalmniocerenatalargete

Doppl

ersus 29%) and reduce the incidence of FGR (10% versus

3Gapwnnosa

TDFlvctigsste

CIffwmrmrfd

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3


0%) in women heterozygous for either FVL or prothrombin20210A mutation or with protein S deficiency.96 Other tri-ls examining the use of heparin and/or low-dose aspirin forrevention of recurrent adverse pregnancy outcomes inomen with genetic thrombophilias had relatively smallumbers of subjects, were observational, included heteroge-eous groups of patients, had various dosages of heparin, andften used historical controls.63,69,97-99 Nevertheless, thesetudies suggest that there may be some benefit to using hep-rin in selected circumstances.

argeted Areas for Future Researchespite the vast amount of research that has been done onGR, there are many knowledge gaps that persist, particu-

arly in the areas of defining etiology-specific risks and inter-entions for preventing recurrence. The heterogeneity of theondition has hindered our ability to determine optimalreatments for individual cases. In addition, our understand-ng of the complex biologic–environmental interactions andenetic susceptibilities that exist is limited. Future researchhould attempt to use risk-specific inclusion criteria andhould take into consideration the array of adverse outcomeshat can result from similar underlying conditions (ie, isch-mic placental disease).

onclusionsn summary, poor fetal growth can result from a myriad ofetal, placental, and maternal conditions. Since many of theseactors can persist throughout subsequent pregnancies,omen should be counseled and managed appropriately toinimize future adverse outcomes. It is also important to

ecognize that risk factors for FGR overlap the risk factors forany other obstetrical concerns, such as recurrent miscar-

iage, preeclampsia, abruption, and fetal death. Therefore,uture pregnancies should be given close attention to theevelopment of these possible events.

eferences1. Snijders RJ, Sherrod C, Gosden CM, et al: Fetal growth retardation:

associated malformations and chromosomal abnormalities. Am J Ob-stet Gynecol 168:547-555, 1993

2. Creasy RLR: Maternal–Fetal Medicine. Philadelphia, PA, Saunders,2005

3. Khoury MJ, Erickson JD, Cordero JF, et al: Congenital malformationsand intrauterine growth retardation: a population study. Pediatrics 82:83-90, 1988

4. Hannula K, Lipsanen-Nyman M, Kristo P, et al: Genetic screening formaternal uniparental disomy of chromosome 7 in prenatal and postna-tal growth retardation of unknown cause. Pediatrics 109:441-448,2002

5. Kotzot D, Lurie IW, Mehes K, et al: No evidence of uniparental disomy2, 6, 14, 16, 20, and 22 as a major cause of intrauterine growth retar-dation. Clin Genet 58:177-180, 2000

6. Brown ZA, Vontver LA, Benedetti J, et al: Effects on infants of a firstepisode of genital herpes during pregnancy. N Engl J Med 317:1246-1251, 1987

7. Meyberg-Solomayer GC, Fehm T, Muller-Hansen I, et al: Prenatal ul-trasound diagnosis, follow-up, and outcome of congenital varicella syn-drome. Fetal Diagn Ther 21:296-301, 2006

8. Donner C, Liesnard C, Content J, et al: Prenatal diagnosis of 52 preg- 3

nancies at risk for congenital cytomegalovirus infection. ObstetGynecol 82:481-486, 1993

9. Daffos F, Forestier F, Capella-Pavlovsky M, et al: Prenatal managementof 746 pregnancies at risk for congenital toxoplasmosis. N Engl J Med318:271-275, 1988

0. Khan NA, Kazzi SN: Yield and costs of screening growth-retarded in-fants for torch infections. Am J Perinatol 17:131-135, 2000

1. Arbuckle TE, Wilkins R, Sherman GJ: Birth weight percentiles by ges-tational age in Canada. Obstet Gynecol 81:39-48, 1993

2. Houlton MC, Marivate M, Philpott RH: The prediction of fetal growthretardation in twin pregnancy. Br J Obstet Gynaecol 88:264-273, 1981

3. Secher NJ, Kaern J, Hansen PK: Intrauterine growth in twin pregnan-cies: prediction of fetal growth retardation. Obstet Gynecol 66:63-68,1985

4. Sherer DM, Divon MY: Fetal growth in multifetal gestation. Clin ObstetGynecol 40:764-770, 1997

5. Bjoro K Jr: Vascular anomalies of the umbilical cord. I. Obstetric im-plications. Early Hum Dev 8:119-127, 1983

6. Predanic M, Perni SC, Friedman A, et al: Fetal growth assessment andneonatal birth weight in fetuses with an isolated single umbilical artery.Obstet Gynecol 105:1093-1097, 2005

7. Heinonen S, Ryynanen M, Kirkinen P, et al: Perinatal diagnostic eval-uation of velamentous umbilical cord insertion: clinical, Doppler, andultrasonic findings. Obstet Gynecol 87:112-117, 1996

8. Rolschau J: Circumvallate placenta and intrauterine growth retarda-tion. Acta Obstet Gynecol Scand Suppl 72:11-14, 1978

9. Nagy S, Bush M, Stone J, et al: Clinical significance of subchorionic andretroplacental hematomas detected in the first trimester of pregnancy.Obstet Gynecol 102:94-100, 2003

0. Ananth CV, Berkowitz GS, Savitz DA, et al: Placental abruption andadverse perinatal outcomes. J Am Med Assoc 282:1646-1651, 1999

1. Sheiner E, Shoham-Vardi I, Hallak M, et al: Placental abruption in termpregnancies: clinical significance and obstetric risk factors. J MaternFetal Neonatal Med 13:45-49, 2003

2. Ananth CV, Demissie K, Smulian JC, et al: Relationship among placentaprevia, fetal growth restriction, and preterm delivery: a population-based study. Obstet Gynecol 98:299-306, 2001

3. Laurini R, Laurin J, Marsal K: Placental histology and fetal blood flow inintrauterine growth retardation. Acta Obstet Gynecol Scand 73:529-534, 1994

4. Salafia CM, Vintzileos AM, Silberman L, et al: Placental pathology ofidiopathic intrauterine growth retardation at term. Am J Perinatol9:179-184, 1992

5. Bane AL, Gillan JE: Massive perivillous fibrinoid causing recurrentplacental failure. Br J Obstet Gynaecol 110:292-295, 2003

6. Mandsager NT, Bendon R, Mostello D, et al: Maternal floor infarction ofthe placenta: prenatal diagnosis and clinical significance. Obstet Gy-necol 83:750-754, 1994

7. Bjoro K Jr, Myhre E: The role of chronic non-specific inflammatorylesions of the placenta in intra-uterine growth retardation. Acta PatholMicrobiol Immunol Scand [A] 92:133-137, 1984

8. Pham T, Steele J, Stayboldt C, et al: Placental mesenchymal dysplasia isassociated with high rates of intrauterine growth restriction and fetaldemise: a report of 11 new cases and a review of the literature. Am J ClinPathol 126:67-78, 2006

9. Svensson AC, Pawitan Y, Cnattingius S, et al: Familial aggregation ofsmall-for-gestational-age births: the importance of fetal genetic effects.Am J Obstet Gynecol 194:475-479, 2006

0. Stein AD, Ravelli AC, Lumey LH: Famine, third-trimester pregnancyweight gain, and intrauterine growth: the Dutch Famine Birth CohortStudy. Hum Biol 67:135-150, 1995

1. Kouba S, Hallstrom T, Lindholm C, et al: Pregnancy and neonataloutcomes in women with eating disorders. Obstet Gynecol 105:255-260, 2005

2. Murphy JF, O’Riordan J, Newcombe RG, et al: Relation of haemoglobinlevels in first and second trimesters to outcome of pregnancy. Lancet1:992-995, 1986

3. Steer P, Alam MA, Wadsworth J, et al: Relation between maternal

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6

7

7

7

7

7

7

7

7

7

7

8


haemoglobin concentration and birth weight in different ethnic groups.Br Med J 310:489-491, 1995

4. Lu ZM, Goldenberg RL, Cliver SP, et al: The relationship betweenmaternal hematocrit and pregnancy outcome. Obstet Gynecol 77:190-194, 1991

5. Sheiner E, Mazor M, Levy A, et al: Pregnancy outcome of asthmaticpatients: a population-based study. J Matern Fetal Neonatal Med 18:237-240, 2005

6. Jensen GM, Moore LG: The effect of high altitude and other risk factorson birthweight: independent or interactive effects? Am J Public Health87:1003-1007, 1997

7. Galan HL, Rigano S, Radaelli T, et al: Reduction of subcutaneous mass,but not lean mass, in normal fetuses in Denver, Colorado. Am J ObstetGynecol 185:839-844, 2001

8. Hammoud AO, Bujold E, Sorokin Y, et al: Smoking in pregnancy re-visited: findings from a large population-based study. Am J ObstetGynecol 192:1856-1862, discussion 1862-1863, 2005

9. Spinillo A, Capuzzo E, Nicola SE, et al: Factors potentiating the smok-ing-related risk of fetal growth retardation. Br J Obstet Gynaecol 101:954-958, 1994

0. U.S. Department of Health and Human Services. Women and Smoking:A Report of the Surgeon General. Public Health Service, Office of theSurgeon General, 2001

1. Bada HS, Das A, Bauer CR, et al: Low birth weight and preterm births:etiologic fraction attributable to prenatal drug exposure. J Perinatol25:631-637, 2005

2. Floyd RL, O’Connor MJ, Sokol RJ, et al: Recognition and prevention offetal alcohol syndrome. Obstet Gynecol 106:1059-1064, 2005

3. Bada HS, Das A, Bauer CR, et al: Gestational cocaine exposure andintrauterine growth: maternal lifestyle study. Obstet Gynecol 100:916-924, 2002

4. Ness RB, Sibai BM: Shared and disparate components of the pathophys-iologies of fetal growth restriction and preeclampsia. Am J Obstet Gy-necol 195:40-49, 2006

5. Bernstein PS, Divon MY: Etiologies of fetal growth restriction. ClinObstet Gynecol 40:723-729, 1997

6. Villar J, Carroli G, Wojdyla D, et al: Preeclampsia, gestational hyper-tension and intrauterine growth restriction, related or independentconditions? Am J Obstet Gynecol 194:921-931, 2006

7. Abalos E, Duley L, Steyn DW, et al: Antihypertensive drug therapy formild to moderate hypertension during pregnancy. Cochrane Databaseof Systematic Reviews 2001:CD002252

8. Stettler RW, Cunningham FG: Natural history of chronic proteinuriacomplicating pregnancy. Am J Obstet Gynecol 167:1219-1224, 1992

9. Yasmeen S, Wilkins EE, Field NT, et al: Pregnancy outcomes in womenwith systemic lupus erythematosus. J Matern Fetal Med 10:91-96,2001

0. Rahman P, Gladman DD, Urowitz MB: Clinical predictors of fetal out-come in systemic lupus erythematosus. J Rheumatol 25:1526-1530,1998

1. Julkunen H, Jouhikainen T, Kaaja R, et al: Fetal outcome in lupuspregnancy: a retrospective case-control study of 242 pregnancies in112 patients. Lupus 2:125-131, 1993

2. Kupferminc MJ, Rimon E, Ascher-Landsberg J, et al: Perinatal outcomein women with severe pregnancy complications and multiple throm-bophilias. J Perinat Med 32:225-227, 2004

3. Kupferminc MJ, Many A, Bar-Am A, et al: Mid-trimester severe intra-uterine growth restriction is associated with a high prevalence ofthrombophilia. Br J Obstet Gynaecol 109:1373-1376, 2002

4. Martinelli P, Grandone E, Colaizzo D, et al: Familial thrombophilia andthe occurrence of fetal growth restriction. Haematologica 86:428-431,2001

5. Howley HE, Walker M, Rodger MA: A systematic review of the associ-ation between factor V Leiden or prothrombin gene variant and intra-uterine growth restriction. Am J Obstet Gynecol 192:694-708, 2005

6. Infante-Rivard C, Rivard GE, Yotov WV, et al: Absence of association ofthrombophilia polymorphisms with intrauterine growth restriction.N Engl J Med 347:19-25, 2002

7. Salomon O, Seligsohn U, Steinberg DM, et al: The common prothrom- 8

botic factors in nulliparous women do not compromise blood flow inthe feto-maternal circulation and are not associated with preeclampsiaor intrauterine growth restriction. Am J Obstet Gynecol 191:2002-2009, 2004

8. Franchi F, Cetin I, Todros T, et al: Intrauterine growth restriction andgenetic predisposition to thrombophilia. Haematologica 89:444-449,2004

9. Infante-Rivard C, Rivard GE, Guiguet M, et al: Thrombophilic poly-morphisms and intrauterine growth restriction. Epidemiology 16:281-287, 2005

0. Yasuda M, Takakuwa K, Tokunaga A, et al: Prospective studies of theassociation between anticardiolipin antibody and outcome of preg-nancy. Obstet Gynecol 86:555-559, 1995

1. Levine JS, Branch DW, Rauch J: The antiphospholipid syndrome.N Engl J Med 346:752-763, 2002

2. Alarcon-Segovia D: Clinical manifestations of the antiphospholipidsyndrome. J Rheumatol 19:1778-1781, 1992

3. Paidas MJ, Ku DH, Arkel YS: Screening and management of inheritedthrombophilias in the setting of adverse pregnancy outcome. Clin Peri-natol 31:783-805, 2004

4. Alfirevic Z, Roberts D, Martlew V: How strong is the association be-tween maternal thrombophilia and adverse pregnancy outcome? A sys-tematic review. Eur J Obstet Gynecol Reprod Biol 101:6-14, 2002

5. Robertson L, Wu O, Langhorne P, et al: Thrombophilia in pregnancy: asystematic review. Br J Haematol 132:171-196, 2006

6. Kupferminc MJ, Eldor A, Steinman N, et al: Increased frequency ofgenetic thrombophilia in women with complications of pregnancy.N Engl J Med 340:9-13, 1999

7. Lim W, Crowther MA, Eikelboom JW: Management of antiphospho-lipid antibody syndrome: a systematic review. J Am Med Assoc 295:1050-1057, 2006

8. ACOG Practice Bulletin #68: Antiphospholipid syndrome. Obstet Gy-necol 106:1113-1121, 2005

9. Stella CL, Sibai BM: Thrombophilia and adverse maternal-perinataloutcome. Clin Obstet Gynecol 49:850-860, 2006

0. Roque H, Paidas MJ, Funai EF, et al: Maternal thrombophilias are notassociated with early pregnancy loss. Thromb Haemost 91:290-295,2004

1. Rasmussen S, Irgens LM, Albrechtsen S, et al: Predicting preeclampsiain the second pregnancy from low birth weight in the first pregnancy.Obstet Gynecol 96:696-700, 2000

2. Salihu HM, Sharma PP, Aliyu MH, et al: Is small for gestational age amarker of future fetal survival in utero? Obstet Gynecol 107:851-856,2006

3. Kuno N, Itakura A, Kurauchi O, et al: Decrease in severity of intrauter-ine growth retardation in subsequent pregnancies. Int J Gynaecol Ob-stet 51:219-224, 1995

4. Furuhashi M, Kurauchi O, Suganuma N: Pregnancy following placentalabruption. Arch Gynecol Obstet 267:11-13, 2002

5. Rasmussen S, Albrechtsen S, Dalaker K: Obstetric history and the riskof placenta previa. Acta Obstet Gynecol Scand 79:502-507, 2000

6. Rasmussen S, Irgens LM, Dalaker K: A history of placental dysfunctionand risk of placental abruption. Paediatr Perinat Epidemiol 13:9-21,1999

7. Rasmussen S, Irgens LM, Albrechtsen S, et al: Women with a history ofplacental abruption: when in a subsequent pregnancy should specialsurveillance for a recurrent placental abruption be initiated? Acta Ob-stet Gynecol Scand 80:708-712, 2001

8. Wallenburg HC, Rotmans N: Prevention of recurrent idiopathic fetalgrowth retardation by low-dose aspirin and dipyridamole. Am J ObstetGynecol 157:1230-1235, 1987

9. Fuke Y, Aono T, Imai S, et al: Clinical significance and treatment ofmassive intervillous fibrin deposition associated with recurrent fetalgrowth retardation. Gynecol Obstet Invest 38:5-9, 1994

0. Cheng CJ, Bommarito K, Noguchi A, et al: Body mass index changebetween pregnancies and small for gestational age births. Obstet Gy-necol 104:286-292, 2004

1. Olsen SF, Secher NJ, Tabor A, et al: Randomised clinical trials of fish oil

8

8

8

8

8

8

8

8

9

9

9

9

9

9

9

9

9

9


supplementation in high risk pregnancies. Fish Oil Trials In Pregnancy(FOTIP) Team. Br J Obstet Gynaecol 107:382-395, 2000

2. Chavez MR, Ananth CV, Smulian JC, et al: Fetal transcerebellar diam-eter measurement with particular emphasis in the third trimester: areliable predictor of gestational age. Am J Obstet Gynecol 191:979-984,2004

3. Sepulveda W, Rojas I, Robert JA, et al: Prenatal detection of velamen-tous insertion of the umbilical cord: a prospective color Doppler ultra-sound study. Ultrasound Obstet Gynecol 21:564-569, 2003

4. Nomiyama M, Toyota Y, Kawano H: Antenatal diagnosis of velamen-tous umbilical cord insertion and vasa previa with color Doppler im-aging. Ultrasound Obstet Gynecol 12:426-429, 1998

5. Nyberg DA, Cyr DR, Mack LA, et al: Sonographic spectrum of placentalabruption. AJR Am J Roentgenol 148:161-164, 1987

6. Chan FY, Pun TC, Lam C, et al: Pregnancy screening by uterine arteryDoppler velocimetry: which criterion performs best? Obstet Gynecol85:596-602, 1995

7. Haddad B, Uzan M, Breart G, et al: Uterine Doppler wave form and theprediction of the recurrence of pre-eclampsia and intra-uterine growthretardation in patients treated with low-dose aspirin. Eur J Obstet Gy-necol Reprod Biol 62:179-183, 1995

8. Low-dose aspirin in prevention and treatment of intrauterine growthretardation and pregnancy-induced hypertension. Italian study of as-pirin in pregnancy. Lancet 341:396-400, 1993

9. Ebrashy A, Ibrahim M, Marzook A, et al: Usefulness of aspirin therapyin high-risk pregnant women with abnormal uterine artery Dopplerultrasound at 14-16 weeks pregnancy: randomized controlled clinicaltrial. Croat Med J 46:826-831, 2005

0. Uzan S, Beaufils M, Breart G, et al: Prevention of fetal growth retarda-

tion with low-dose aspirin: findings of the EPREDA trial. Lancet337:1427-1431, 1991

1. Kutteh WH: Antiphospholipid antibody-associated recurrent preg-nancy loss: treatment with heparin and low-dose aspirin is superior tolow-dose aspirin alone. Am J Obstet Gynecol 174:1584-1589, 1996

2. Rai R, Cohen H, Dave M, et al: Randomised controlled trial of aspirinand aspirin plus heparin in pregnant women with recurrent miscarriageassociated with phospholipid antibodies (or antiphospholipid antibod-ies). Br Med J 314:253-257, 1997

3. Farquharson RG, Quenby S, Greaves M: Antiphospholipid syndromein pregnancy: a randomized, controlled trial of treatment. Obstet Gy-necol 100:408-413, 2002

4. Empson M, Lassere M, Craig JC, et al: Recurrent pregnancy loss withantiphospholipid antibody: a systematic review of therapeutic trials.Obstet Gynecol 99:135-144, 2002

5. Chang P, Millar D, Tsang P, et al: Intravenous immunoglobulin inantiphospholipid syndrome and maternal floor infarction when stan-dard treatment fails: a case report. Am J Perinatol 23:125-129, 2006

6. Gris JC, Mercier E, Quere I, et al: Low-molecular-weight heparin versuslow-dose aspirin in women with one fetal loss and a constitutionalthrombophilic disorder. Blood 103:3695-3699, 2004

7. Kupferminc MJ, Fait G, Many A, et al: Low-molecular-weight heparinfor the prevention of obstetric complications in women with thrombo-philias. Hypertens Pregnancy 20:35-44, 2001

8. Brenner B, Hoffman R, Blumenfeld Z, et al: Gestational outcome inthrombophilic women with recurrent pregnancy loss treated by enox-aparin. Thromb Haemost 83:693-697, 2000

9. Grandone E, Brancaccio V, Colaizzo D, et al: Preventing adverse ob-stetric outcomes in women with genetic thrombophilia. Fertil Steril
78:371-375, 2002

PG

Ptdtapbnfeasaarwd

*

†

‡

A

0d

reeclampsia Recurrence and Preventionary A. Dildy III, MD,* Michael A. Belfort, MD, PhD,† and John C. Smulian, MD, MPH‡

Women with a previous pregnancy complicated by preeclampsia have an increased risk forrecurrence in subsequent pregnancies. For severe preeclamptic women in an initial preg-nancy, recurrence rates for any type of preeclampsia are very high, approaching 50% insome studies. Significant maternal and fetal complications are more common in recurrentpreeclampsia compared with an initial episode. For women who have experienced apregnancy complicated by preeclampsia, a systematic evaluation for underlying risk factorsmay identify a specific pathway suitable for a specific intervention. Although some progresshas been made in developing potential therapeutic options to prevent preeclampsia recur-rence, there is a great need for better data to determine who will benefit most from anyspecific therapy.Semin Perinatol 31:135-141 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS preeclampsia, eclampsia, risk factors, recurrence, prevention, HELLP syndrome

PMoqdrvp(tenp

TPArarw(sao

wc

reeclampsia complicates approximately 5% to 10% ofnulliparous pregnancies1 and is consistently among the

op three causes of maternal death in both developed andeveloping countries.2-4 Two-thirds of cases will be mild andhe other third severe in degree.1 Preeclampsia is considereddisease of nulliparous women, as it is twice as common inrimigravidas as it is in women who have previously givenirth.5 It is well known that women with a previous preg-ancy complicated by preeclampsia have an increased riskor recurrence in subsequent pregnancies. For severe pre-clamptic women in an initial pregnancy, recurrence rates forny type of preeclampsia are very high, approaching 50% inome studies. Significant maternal and fetal complicationsre more common in recurrent preeclampsia compared withn initial episode. Thus, accurate and thorough counselingegarding recurrence risks and potential preventive measuresill assist women and their caregivers to make importantecisions pertaining to future childbearing.

Department of Obstetrics and Gynecology, LSU Health Sciences Center,New Orleans, LA.

Department of Obstetrics and Gynecology, University of Utah Health Sci-ences Center, Salt Lake City, UT.

Division of Maternal Fetal Medicine, Department of Obstetrics, Gynecologyand Reproductive Sciences, UMDNJ-Robert Wood Johnson MedicalSchool, New Brunswick, NJ.

ddress reprint requests to Gary A. Dildy III, MD, Maternal Fetal MedicineCenter, St. Mark’s Hospital, 1140 East 3900 South, Suite 390, Salt Lake

lCity, Utah 84124. E-mail: [email protected]


reeclampsia Risk Factorsany factors for preeclampsia have been described in the

bstetrical literature, and the majority will persist in subse-uent pregnancies (Table 1).6 Preeclampsia tends to be aisease of first pregnancy in women with no other obviousisk factors; however, underlying medical conditions withascular or renal implications (diabetes mellitus, chronic hy-ertension) and conditions with increased trophoblast massmultifetal gestation or hydrops fetalis) substantially increasehe risk. As preeclampsia is likely a syndrome of multipletiologies and many underlying factors persist across preg-ancies, a significant risk factor for future preeclampsia is arior history of preeclampsia.

he Epidemiology ofreeclampsia Recurrencenumber of studies have examined the risk for preeclampsia

ecurrence in subsequent pregnancies, and all have indicatedsignificantly increased risk (Table 2). The highest risks for

ecurrence are found most consistently when the initial caseas preterm, severe, or complicated by eclampsia, HELLP

hemolysis, elevated liver enzymes, and low platelet count)yndrome, or fetal growth restriction. However, good datare still relatively sparse because definitions for preeclampsiaften vary from study to study.Campbell and coworkers studied a population of pregnant

omen (n � 29,851) whose first recorded pregnancy oc-urred between the years 1967 and 1978 in Aberdeen, Scot-
and and had �2 subsequent pregnancies during that same
135

tmutn

dnttl

T

P

F

M

CGPVN

A

G

OMFFA

A

A*

T

CS

vS

S

C

AS

T

136 G.A. Dildy III, M.A. Belfort, and J.C. Smulian

ime period (n � 6637).7 Women were categorized as nor-otensive (68.0%), mildly preeclamptic (26.3%), protein-ric preeclamptic (5.6%), and eclamptic (0.2%). They foundhat the overall incidence of preeclampsia in a second preg-ancy was less than that in a first pregnancy, but was depen-

able 1 The Strength of the Association of Selected Risk Fac

Risk Factor Associated with Preeclampisa

reeclampsia in a previous pregnancy

irst pregnancy

ultifetal gestation

hronic hypertensionestational diabetesregestational diabetesascular and connective tissue diseaseephropathyUrinary tract infection

ntiphospholipid antibody syndrome

enetic factors (eg, thrombophilias)Factor V Leiden heterozygosityProthrombin heterozygosityMTHFR homozygosityHyperhomocysteinemiabesity (BMI >35 kg/m2) aternal age >35 years

amily history of preeclampsiaetal malformationbnormal maternal serum markers

(AFP, hCG, uE3, Inhibin A)Inhibin A >2.0 MOM2 abnormal markers

frican-American race

bbreviations: AFP, alpha fetoprotein; HCG, human chorionic gonaPresented as odds ratio (OR) and 95% confidence intervals (CI).

able 2 Summary of Studies that Present the Risk for Recurr

Author Study Popu

ampbell7 Preeclampsia (n � 279)ibai9 Second trimester severe pre

an Rijn8 Preeclampsia with delivery <ullivan 12 HELLP (n � 161)

ibai11 HELLP (n � 192)

hames13 HELLP with delivery <28 we

delusi14 Eclampsia (n � 64)ibai16 Eclampsia (n � 366)

rogstad17 Preeclampsia singleton (n �
Preeclampsia twins (n � 325)
ent on the outcome of the first pregnancy. If the first preg-ancy was complicated simply by proteinuric preeclampsia,he incidence in the second pregnancy was 7.5%, whereashose who were normotensive in the first pregnancy had aow rate of proteinuric preeclampsia in the second pregnancy

r Preeclampsia*

Reference OR (95% CI)

Hnat18 3.88 (2.98-5.05)Duckitt48 7.19 (5.85-8.83)Conde-Agudelo49 2.38 (2.28-2.49)Duckitt48 2.91 (1.28-6.61)Sibai50 2.62 (2.03-3.38)Conde-Agudelo49 2.10 (1.90-2.32)Duckitt48 2.93 (2.04-4.21)Conde-Agudelo49 1.99 (1.78-2.22)Conde-Agudelo49 1.93 (1.66-2.25)Duckitt48 3.56 (2.54-4.99)Stamilio51 6.9 (1.1-42.3)

Abi-Said52 4.23 (1.27-14.06)Robertson53 2.73 (1.65-4.51)Duckitt48 9.72 (4.34-21.75)Robertson53

2.19 (1.46-3.27)2.54 (1.52-4.23)1.37 (1.07-1.76)3.49 (1.21-10.11)

Sibai1 3.38 (1.91-6.00)Conde-Agudelo49 1.67 (1.58-1.77)Duckitt48 2.90 (1.70-4.93)Conde-Agudelo49 1.26 (1.16-1.37)Dugoff54

2.39 (1.75-3.26)3.65 (2.79-4.78)

Tucker55 1.2 (0.8-1.7)

n; uE3, unconjugated estriol.

f Preeclampsia

Rate of Recurrence

Preeclampsia 7.5%psia (n � 169) Any preeclampsia 65%

<28 weeks 21%28-36 weeks 21%37-40 weeks 24%

eeks Preeclampsia 25%Preeclampsia 43%HELLP 27%Preeclampsia 19%HELLP 3%

� 62) Preeclampsia 55%HELLP 6%Eclampsia 16%Preeclampsia 22%Eclampsia 2%

0) Preeclampsia 14.1%

tors fo

dotropi

ence o

lation

eclam

34 w

eks (n

19,96
Preeclampsia 6.8%

osiu6b

dCogiwg

cOFweo2

PRS1tpqd1hc(itcb(i(adptoptc

dsndfa

qdtapwtmTnwr1m

PEqmeOfes2fv

ca3egtdsetTwlo

PAAecddNwccw

Preeclampsia recurrence and prevention 137

f 0.7%. However, women who had proteinuric preeclamp-ia in conjunction with a low-birth-weight (�2500 g) infantn their first pregnancy had double the incidence of protein-ric preeclampsia in their second pregnancy (11.9% versus.6%), compared with similar women with a normal infantirth weight during first pregnancy.Van Rijn and coworkers studied primiparous women who

elivered between 1993 and 2002 at the University Medicalenter Utrecht in The Netherlands who had a history of earlynset preeclampsia resulting in delivery before 34 weeks ofestation.8 Preeclampsia recurred in 25% (30/120) of womenn their second pregnancy. Five percent delivered before 34eeks of gestation and 17% between 34 and 37 weeks ofestation.

Sibai and colleagues9 reported subsequent pregnancy out-omes in women with severe second trimester preeclampsia.f these 125 women, 108 had 169 subsequent pregnancies.or the subsequent pregnancies, approximately one-thirdere normotensive and two-thirds were complicated by pre-

clampsia. Of the women with preeclampsia, approximatelyne-third developed a recurrence at �28 weeks, one-third at8 to 36 weeks, and one-third at 37 to 40 weeks.

reeclampsiaecurrence After HELLP Syndrome

ibai and coworkers10 described a retrospective analysis of12 women with HELLP syndrome from 1977 to 1985. Inhis series, 38 women had 49 subsequent pregnancies.10 Oneatient (2.6%) had recurrent HELLP syndrome in 2 subse-uent pregnancies, both complicated by abruption and fetaleath. These investigators extended their series to the years977 to 1992, with follow up on 341 patients, of which 152ad subsequent pregnancies.11 Maternal complications in-luded preeclampsia (19%), recurrent HELLP syndrome3%), and placental abruption (2%). Perinatal complicationsncluded preterm birth (21%), intrauterine growth restric-ion (12%), and perinatal death (4%). Those with preexistinghronic hypertension had higher rates of preeclampsia (75%)ut no significant increase in recurrent HELLP syndrome5%). Perinatal complications such as preterm birth (80%),ntrauterine growth restriction (45%), placental abruption20%), and perinatal death (40%) were significantly highermong chronic hypertensives with previous HELLP syn-rome. Interestingly, all of the above conditions, includingreeclampsia, preterm birth, fetal growth restriction, placen-al abruption, and perinatal death, are considered to haveverlapping etiologic mechanisms that relate to abnormallacentation. Thus, a pregnancy with preeclampsia in an ini-ial pregnancy appears to be at risk for these other pregnancyomplications and should be managed accordingly.

A retrospective study of 481 women with HELLP syn-rome between the years 1980 and 1991 analyzed 161 of 195ubsequent pregnancies in 122 patients.12 Of these 161 preg-ancies, 43% had preeclampsia and 27% had HELLP syn-rome. A previous delivery �32 weeks of gestation was a riskactor for recurrence of prematurity at a similar gestational
ge secondary to preeclampsia in 61% of cases. d
Chames and coworkers reported the outcomes of subse-uent pregnancies in women with a history of HELLP syn-rome for which delivery occurred at �28 weeks of gesta-ion.13 Women were delivered in Memphis, TN (1984-1998)nd Lexington, KY (1994-1998). Data were available in 69atients; there were 76 subsequent pregnancies among 48omen, of which 62 progressed beyond 20 weeks of gesta-

ion. Preeclampsia developed in 55% (34), of which 7 wereild and 27 were severe. Recurrent HELLP occurred in 6%.here were no cases of eclampsia; however, significant peri-atal complications were frequent. Preterm birth (�37eeks of gestation) occurred in 53%. Newborns were growth

estricted in 27%, and the overall perinatal mortality rate was1%. Women with chronic hypertension had greater overallorbidity.

reeclampsia Recurrence After Eclampsiaclampsia recurrence is to some extent dependent on ade-uacy of prenatal care and peripartum practices, includingethods to control hypertensive crisis and prevention of

clamptic seizures vis-a-vis magnesium sulfate. Adelusi andjengbede14 reported a prospective study of 64 eclamptics

rom Ibadan, Nigeria of whom 16% experienced recurrentclampsia despite the benefit of antenatal care. Chesley’seminal account of eclampsia recurrence during the early0th century reported a recurrence risk range of 0% to 21%rom published series, with an approximately 5% risk foriable gestations.15

Sibai and coworkers studied 223 women whose pregnan-ies were complicated by eclampsia between the years 1977nd 1989, with an average follow up of 7.2 years.16 Of the66 subsequent pregnancies, 22% were complicated by pre-clampsia and 1.9% by eclampsia. Within the nulliparousroup, women who had eclampsia before 37 weeks of gesta-ion in the index pregnancy had significantly higher inci-ences of preeclampsia and poor perinatal outcome in sub-equent pregnancies, compared with those who hadclampsia at or beyond 37 weeks of gestation. Of the normo-ensive women, 10% had chronic hypertension on follow up.he highest incidence of chronic hypertension was in thoseith eclampsia at �30 weeks of gestation (18%) and the

owest incidence (5%) in those with eclampsia at �37 weeksf gestation.

reeclampisa Recurrencefter Multiple Gestationslthough multiple gestations are considered at risk for pre-clampsia, it is not clear whether women developing thisomplication are clearly at risk for recurrence in the sameegree as for singleton pregnancies. Only 1 study has ad-ressed this issue. Trogstad and coworkers examined theorway birth registry from 1967 to 1998, including 550,218omen.17 For women with a previous singleton pregnancy

omplicated by preeclampsia, the recurrence rate was 14.1%ompared with the recurrence rate for twins of 6.8%, whichas much closer to the population risk. This suggests that
ifferent potential etiologies for preeclampsia present differ-

edt

CRHwrtcapsaetasaa

PRTrspvld

hiwmfwsf(mip

peudicbtnt

tmt

pttsatdric

itstctp

topniiiwnolimtlcg

actTlamfpmhthegw


nt risks for recurrence. This is particularly important whenesigning a recurrence prevention strategy, since no singleherapy would be expected to target all potential etiologies.

linical Features ofecurrent Preeclampsia

nat and coworkers18 reported on 2934 nulliparous womenith an initial episode of preeclampsia and 598 women with

ecurrent preeclampsia. Two of the more severe presenta-ions of this disorder, eclampsia and HELLP syndrome, oc-urred in the nulliparous cases with an incidence of 0.34%nd 0.21%, respectively. The same severe preeclampsia com-lications examined in the women with recurrent preeclamp-ia were approximately 5-fold higher: 1.67% for eclampsiand 1.00% for HELLP syndrome. In addition, recurrent pre-clampsia was more commonly associated with other perina-al complications, such as preterm birth (67% versus 33%),bruptio placentae (8% versus 2%), and fetal death (7% ver-us 1%). Therefore, recurrent preeclampsia is clearly associ-ted with more severe disease and with more severe associ-ted morbidities.

reeclampsia Pathogenesiselevant to Recurrence

o understand the rationale and interpret some of the largeandomized controlled trials aimed at preventing preeclamp-ia, it is important to understand proposed theories of theathophysiology of this disorder. Most of the etiologies in-olve some form of physiologic change that promotes vascu-ar endothelial damage, relative placental hypoxia, and oxi-ative stress injury.1. Abnormal angiogenesis and resultant placental ischemia

ypothesis: In this hypothesis poor second wave trophoblastnvasion of the spiral arteries leads to placental ischemiahich causes increased deportation of trophoblast into theaternal circulation which results in endothelial cell dys-

unction.19 Failed second wave invasion has been associatedith many factors, including abnormal circulating levels of

oluble fms-like tyrosine kinase 1 (sFlt-1), placental growthactor (PlGF), and vascular endothelial growth factorVEGF),20 maternal infection with CMV,21 prolactin abnor-alities,22 and deficiencies of trace metals.23 Gene abnormal-

ties linked with trophoblast invasiveness have also been pro-osed to influence preeclampsia risk.2. Very low-density lipoprotein (VLDL) versus toxicity-

reventing activity (TxPA) hypothesis: Arbogast and cowork-rs,24 proposed that preeclampsia may result from an un-sual accumulation of VLDL which leads to endothelialamage. This hypothesis was based on the observation that,

n many women who subsequently develop preeclampsia,irculating free fatty acids (FFA) are increased 15 to 20 weeksefore the onset of disease. These FFA adversely affect endo-helial physiology leading to vasoconstriction. In some preg-ant women (who are known to have low albumin concen-
rations), the burden of transporting extra FFA from adipose f
issue to the liver as a response to the increased energy de-ands of pregnancy may reduce the concentration of TxPA

o a point where the VLDLs cause endothelial injury.24,25

3. Hyperdynamic disease hypothesis: According to the hy-erdynamic disease model,26 early in pregnancy, preeclamp-ic patients have an elevated cardiac output with compensa-ory vasodilatation. As the disease progresses, there is aubsequent hemodynamic crossover to low cardiac outputnd high resistance circulation coinciding with the onset ofhe clinical syndrome. During the hyperdynamic phase, theilated systemic terminal arterioles and renal afferent arte-ioles may expose capillary beds to systemic pressures andncreased flow, eventually leading to endothelial cell injuryharacteristic of preeclampsia injury.27

4. Immune/immunogenetic maladaptation hypothesis: Thenteraction between decidual leukocytes and invading cy-otrophoblast cells is essential for normal trophoblast inva-ion and development. Immune maladaptation may causehe shallow invasion of the spiral arteries by the endovascularytotrophoblast cells, resulting in endothelial cell dysfunc-ion mediated by an increased decidual release of cytokines,roteolytic enzymes, and free radical species.28

5. The genetic hypothesis: It has been hypothesized thathe development of preeclampsia–eclampsia may be basedn a single recessive gene or a dominant gene with incom-lete penetrance, dependent on the fetal genotype. However,o specific genes have been identified and no defined inher-

tance pathway has been determined sufficient to clearlydentify a specific gene pathway. There are data to supportncreased risk of preeclampsia in women who themselvesere born of a preeclamptic pregnancy.29 Women born of aormal pregnancy, but whose mother had preeclampsia withne her other pregnancies, also have an increased risk.30 Esp-in and coworkers29 showed that not only do maternal genet-cs play a role, but men born of a preeclamptic pregnancy are

ore likely to father children from a preeclamptic pregnancyhan those born of normal pregnancy. Skjaerven and col-eagues suggest that the genes that determine maternal sus-eptibility to preeclampsia are different from the paternalenes that trigger preeclampsia through the fetus.30

6. Genetic-conflict hypothesis: In both flowering plantsnd mammalian pregnancies, there is genetic conflict due toompeting interests of maternal and paternal genes regardinghe volume of nutrients transferred from mother to fetus.31

he more resources a fetus is able to take from its mother, thearger it will be at birth and the better its chances for survivalnd reproduction. However, the greater the nutritional de-ands of the pregnancy, the greater the cost to the mother’s

uture reproductive potential. Paternal genes in the fetus andlacenta may seek to maximize the transfer of nutrients fromother to baby (because the mother’s future offspring mayave a different father); maternal genes on the other handend to moderate the flow of resources in an effort to preserveer reproductive potential for future pregnancies. This inher-nt competition may contribute to genetic imprinting; that is,enes that behave differently in an organism depending onhether they were inherited from the mother or from the

ather. In mice, the paternal genes control the growth of the

pspmcictrc(ipgpcp

REoBpapwp

PRWcseisgnlst

irmaetwlgctgd

TtO

H

TA

EB

U

SA

T

A

C

M

FV

H

A


lacenta, whereas maternal genes are predominantly respon-ible for embryo formation. The placenta—as an agent ofaternal genes—invades the maternal tissues to parasitizeaternal blood supply and support fetal growth. The genetic

onflict theory posits when for whatever reason placentations abnormal and fetal growth is adversely impacted, the pla-enta somehow activates genes that enhance blood flow tohe fetus at a cost to the mother. Preeclampsia may wellepresent such a situation, and recent data show that in thisondition the placenta produces excess amounts of sFlt1soluble FMS like tyrosine kinase 1), leading to endothelialnjury and vasoconstriction. Given the low resistance of thelacental bed when maternal vasoconstriction occurs, areater proportion of the maternal blood is shunted to thelacental circulation. Data show that elevation in sFLT1 oc-urs well before the development of the clinical condition ofreeclampsia.20

able 3 Proposed Evaluation Options for Women with a His-ory Suggesting a High Rate of Preeclampsia Recurrence inrder to Target Interventions

istory & physical examinationChronic diseasesFamily historyOther historical risk factors for recurrenceReview of placental pathology from previous affected

pregnancy (eg, thrombi or infarcts, especially with fetalgrowth restriction, may benefit from low dose heparin)

hrombophilia evaluation (inherited and acquired)ntinuclear antibody if clinically suspicious for

autoimmune diseasearly glucola if clinically appropriateaseline labs for later comparisonComplete blood count, platelets, serum creatinine, BUN,

uric acid, liver enzymes, urinalysis, 24-hour urine fortotal protein content

terine artery Doppler screening to determine the intensityof ultrasound surveillance needed later in thepregnancy

erial ultrasounds for evaluation of fetal growthntepartum fetal surveillance if indications present

able 4 Reviews and Randomized Clinical Trials for Preeclam

Agent Study Pop

spirin Coomarasamy33 High riskDuley32 High risk

alcium Hofmeyr34 Meta-analysisMeta-analysis

agnesium Spatling35 General low-rSibai36 Normotensive

ish oil Makrides37 All riskitamins C�E Poston41 High risk

Rumbold42 Nulliparous weparin Mello46 Angiotensin c

polymorphisnonthrombowith history

bbreviations: CI, confidence intervals; NS, not significant.

ecommendedvaluation and Managementf Women at Risk for Recurrence

ecause there are a significant number of risk factors andotential etiologic mechanisms leading to preeclampsia, anttempt should be made to identify those risk factors andathways that offer opportunities for intervention. In thisay, a rational pathway-specific strategy for prevention ofreeclampsia recurrence can be developed. (Table 3).

revention ofecurrent Preeclampsiaomen with any of the following characteristics should be

onsidered at particularly high risk for recurrent preeclamp-ia and stand to benefit the most from prevention strategies,ven if there is only a modest effect on recurrence rates. Thesenclude preeclampsia that was complicated by high degree ofeverity, preterm birth, HELLP syndrome, eclampsia, fetalrowth restriction, abruptio placenta, oligohydramnios, peri-atal death, a strong family history, or a history of vascular

esions on placental histology. Women with any of thesehould be targeted for the most intensive etiologic evalua-ions and condition-specific therapy when warranted.

Only a few interventions have been sufficiently well-stud-ed to demonstrate any consistent efficacy for preeclampsiaecurrence prevention (Table 4). Antiplatelet therapy (pri-arily low-dose aspirin) has been studied most vigorously,

nd appears to provide a modest effect in preventing pre-clampsia. Two published literature reviews have indicatedhat there is a 14-19% reduction in preeclampsia in high-riskomen (especially those with previous preeclampsia) using

ow-dose aspirin, generally 81 mg/d.32,33 It has been sug-ested that aspirin works through a variety of mechanisms,hiefly by increasing the prostacyclin to thromboxane ratio inhe vascular endothelium and by reducing sensitivity to an-iotensin II. However, other actions are also likely. Aspirinoes not appear to be effective in preventing preeclampsia

ecurrence Prevention

on N Odds Ratio (95% CI)

12,416 0.86 (0.79-0.94)33,439 0.81 (0.75-0.88)

isk 15,206 0.48 (0.33-0.69)risk 587 0.22 (0.12-0.42)

568 NSgravidas 374 NS

1,683 0.86 (0.59-1.27)2,410 0.97 (0.80-1.17)1,877 1.20 (0.82-1.75)

ting enzyme

womeneclampsia

80 0.26 (0.08-0.86)

psia R

ulati

low rhigh

iskprimi

omenonverm inphilicof pre

innc

dwgp9t0iTw

asiaespeisrai

dwslwvtn

szewdlw(vTDtgo

Ssmo

pracdrtpta

FCvirdrnpwFttp1limbao

R

1


n low-risk women. Importantly, low-dose aspirin doesot appear to have an appreciably increased risk of mater-al, fetal, or neonatal complications. For at-risk pregnan-ies, it probably can be stopped at 36 weeks.

Calcium supplementation up to 2 g per day appears toecrease the incidence of preeclampsia significantly (12 trialsith 15,206 women, RR � 0.48, 95% CI 0.33-0.69), with areater effect in high-risk women that include those withrevious preeclampsia (5 trials with 587 women, RR � 0.22,5% CI 0.12-0.42) and those with low baseline calcium in-ake (7 trials with 10,154 women, RR � 0.36, 95% CI 0.18-.70).34 It may work via parathyroid hormone, by reducing

ntravascular calcium and lowering vascular contractility.he use of calcium to prevent preeclampsia in low-riskomen is not supported by current literature.Other potential therapies have been less well-studied, but

vailable data are disappointing. Two small trials of magne-ium sulfate showed no significant reduction in preeclampsiancidence.35,36 Likewise, fish oil supplements, progesterone,nd garlic have not been proven beneficial.37-39 Because pre-clampisa has been linked with an increase in oxidativetress, several therapies with antioxidants have been pro-osed. Most recently, vitamin C and vitamin E have beenvaluated in large prospective randomized trials for prevent-ng preeclampsia. An early review that examined initial trialsuggested a potential benefit.40 However, two recent, largerandomized trials by Poston and coworkers39 and Rumboldnd coworkers41,42 have shown no such benefit and, in fact,ndicated that there may actually be some potential harm.

Unfortunately, to date there have been no published ran-omized trials of therapy to prevent recurrence for womenith previous preeclampsia who have a thrombophilia. Case

eries and anecdotal reports suggest a potential benefit ofow-dose heparin (either unfractionated or low moleculareight).43,44 This approach seems reasonable since placentalascular lesions are more common when there is a significanthrombophilia,45 but prospective randomized trials areeeded to support this as a standard of care.Interestingly, heparin has been used to prevent preeclamp-

ia recurrence in women with the angiotensin-converting en-yme insertion/deletion polymorphism and a history of pre-clampsia. Mello and coworkers randomized 80 subjectsith a history of preeclampsia who were without chronicisease or a demonstrable thrombophilia to either low mo-

ecular weight heparin prophylaxis or no treatment.46 Thereere significant reductions in recurrence of preeclampsia

28% versus 7%), onset of preeclampsia �34 weeks (21%ersus 2%), and fetal growth restriction (44% versus 10%).he authors also monitored uterine artery blood flow byoppler and noted a significant improvement in uterine ar-

ery blood flow in the heparin treatment subjects. This sug-ests that heparin may have a therapeutic role for preventionf preeclampsia recurrence, but that role needs clarification.At present there is no accepted consensus in the United

tates regarding the best approach in preventing preeclamp-ia in the general population or in high-risk women. Theost recent American College of Obstetricians and Gynecol-

gists (ACOG) Practice Bulletin addressing preeclampsia,

ublished in 2002, recommends no specific prophylacticegimen to prevent preeclampsia.6 However, based on avail-ble data, it would seem reasonable, at the minimum, to offeralcium (2 g per day) and/or low-dose aspirin (81 mg peray) to high-risk women, given the perceived risk–benefitatio. Heparin, especially low-molecular-weight prepara-ions, may also be considered in the presence of previousreeclampsia with a significant thrombophilia, with placen-al vascular lesions, or after failure of calcium and low-dosespirin therapy.

uture Directionlearly, more research is needed to identify and refine inter-entions to prevent the recurrence of preeclampsia. By focus-ng on this particularly high-risk population who have al-eady had preeclampsia, it may be easier to successfullyevelop prevention strategies that can be applied to the lowerisk women who lack obvious risk factors. Clinical areaseeding significant clarification currently include thrombo-hilias in recurrent preeclampsia and identification of theomen most likely to benefit from heparin preparations.urther insight into the pathophysiology of preeclampsia athe cellular and genomic level will likely create new oppor-unities for prevention. Recent reports looking at angiogenicroteins, such as circulating soluble fms-like tyrosine kinase(sFlt-1), placental growth factor (PlGF), vascular endothe-

ial growth factor (VEGF), and endoglin, have seeminglydentified a pathogenic role for these factors in the develop-

ent of preeclampsia.20,41 It is not clear yet whether these cane useful for refining risk more precisely or if they mightllow potential intervention points for therapy, but this areaf research appears to have significant promise.47

eferences1. Sibai BM, Ewell M, Levine RJ, et al: Risk factors associated with pre-

eclampsia in healthy nulliparous women. The Calcium for Preeclamp-sia Prevention (CPEP) Study Group. Am J Obstet Gynecol 177:1003-1010, 1997

2. Why Mothers Die: Report on Confidential Enquiries into MaternalDeaths in the United Kingdom 1994-1996. London, HMSO, 1998

3. Chang J, Elam-Evans LD, Berg CJ, et al: Pregnancy-related mortalitysurveillance–United States, 1991-1999. MMWR Surveill Summ 52:1-8, 2003

4. Ujah IA, Aisien OA, Mutihir JT, et al: Factors contributing to maternalmortality in north-central Nigeria: a seventeen-year review. Afr J Re-prod Health 9:27-40, 2005

5. Trupin LS, Simon LP, Eskenazi B: Change in paternity: a risk factor forpreeclampsia in multiparas. Epidemiology 7:240-244, 1996

6. ACOG practice bulletin. Diagnosis and management of preeclampsiaand eclampsia. Number 33, January 2002, American College of Obste-tricians and Gynecologists

7. Campbell DM, MacGillivray I, Carr-Hill R: Pre-eclampsia in secondpregnancy. Br J Obstet Gynaecol 92:131-140, 1985

8. van Rijn BB, Hoeks LB, Bots ML, et al: Outcomes of subsequent preg-nancy after first pregnancy with early-onset preeclampsia. Am J ObstetGynecol 195:723-728, 2006

9. Sibai BM, Mercer B, Sarinoglu C: Severe preeclampsia in the secondtrimester: recurrence risk and long-term prognosis. Am J Obstet Gy-necol 165:1408-1412, 1991

0. Sibai BM, Taslimi MM, el-Nazer A, et al: Maternal-perinatal outcome
associated with the syndrome of hemolysis, elevated liver enzymes, and

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5


low platelets in severe preeclampsia-eclampsia. Am J Obstet Gynecol155:501-509, 1986

1. Sibai BM, Ramadan MK, Chari RS, et al: Pregnancies complicated byHELLP syndrome (hemolysis, elevated liver enzymes, and low plate-lets): subsequent pregnancy outcome and long-term prognosis. Am JObstet Gynecol 172:125-129, 1995

2. Sullivan CA, Magann EF, Perry KG Jr, et al: The recurrence risk of thesyndrome of hemolysis, elevated liver enzymes, and low platelets (HELLP)in subsequent gestations. Am J Obstet Gynecol 171:940-943, 1994

3. Chames MC, Haddad B, Barton JR, et al: Subsequent pregnancy out-come in women with a history of HELLP syndrome at �28 weeks ofgestation. Am J Obstet Gynecol 188:1504-1507, discussion 1507-1508, 2003

4. Adelusi B, Ojengbede OA: Reproductive performance after eclampsia.Int J Gynaecol Obstet 24:183-189, 1986

5. Chesley LC: Eclampsia: the remote prognosis. Semin Perinatol 2:99-111, 1978

6. Sibai BM, Sarinoglu C, Mercer BM: Eclampsia. VII. Pregnancy outcomeafter eclampsia and long-term prognosis. Am J Obstet Gynecol 166:1757-1761, discussion 1761-1763, 1992

7. Trogstad L, Skrondal A, Stoltenberg C, et al: Recurrence risk of pre-eclampsia in twin and singleton pregnancies. Am J Med Genet A 126:41-45, 2004

8. Hnat MD, Sibai BM, Caritis S, et al: Perinatal outcome in women withrecurrent preeclampsia compared with women who develop pre-eclampsia as nulliparas. Am J Obstet Gynecol 186:422-426, 2002

9. Brosens I: A study of the spiral arteries of the decidua basalis in nor-motensive and hypertensive pregnancies. J Obstet Gynaecol Br Com-monw 71:222-230, 1964

0. Levine RJ, Maynard SE, Qian C, et al: Circulating angiogenic factors andthe risk of preeclampsia. N Engl J Med 350:672-683, 2004

1. Grahame-Clarke C: Human cytomegalovirus, endothelial function andatherosclerosis. Herpes 12:42-45, 2005

2. Parra A, Ramirez-Peredo J: The possible role of prolactin in preeclamp-sia: 2001, a hypothesis revisited a quarter of century later. Med Hy-potheses 59:378-384, 2002

3. Acikgoz S, Harma M, Harma M, et al: Comparison of angiotensin-converting enzyme, malonaldehyde, zinc, and copper levels in pre-eclampsia. Biol Trace Elem Res 113:1-8, 2006

4. Arbogast BW, Leeper SC, Merrick RD, et al: Which plasma factors bringabout disturbance of endothelial function in pre-eclampsia? Lancet343:340-341, 1994

5. Lorentzen BEM, Clausen T, Henriksen T: Fasting serum free fatty acidsand triglycerides are increased before 20 weeks of gestation in womenwho later develop preeclampsia. Hypertens Pregnancy 13:103-109,1994

6. Easterling TR, Benedetti TJ: Preeclampsia: a hyperdynamic diseasemodel. Am J Obstet Gynecol 160:1447-1453, 1989

7. Bosio PM, McKenna PJ, Conroy R, et al: Maternal central hemodynam-ics in hypertensive disorders of pregnancy. Obstet Gynecol 94:978-984, 1999

8. Dekker GA, Sibai BM: Etiology and pathogenesis of preeclampsia: cur-rent concepts. Am J Obstet Gynecol 179:1359-1375, 1998

9. Esplin MS, Fausett MB, Fraser A, et al: Paternal and maternal compo-nents of the predisposition to preeclampsia. N Engl J Med 344:867-872, 2001

0. Skjaerven R, Vatten LJ, Wilcox AJ, et al: Recurrence of pre-eclampsiaacross generations: exploring fetal and maternal genetic components ina population based cohort. Br Med J 331:877, 2005

1. Haig D: Altercation of generations: genetic conflicts of pregnancy. Am JReprod Immunol 35:226-232, 1996

2. Duley L, Henderson-Smart DJ, Knight M, et al: Antiplatelet agents forpreventing pre-eclampsia and its complications. Cochrane DatabaseSyst Rev CD004659, 2004

3. Coomarasamy A, Honest H, Papaioannou S, et al: Aspirin for preven-

tion of preeclampsia in women with historical risk factors: a systematicreview. Obstet Gynecol 101:1319-1332, 2003

4. Hofmeyr GJ, Atallah AN, Duley L: Calcium supplementation duringpregnancy for preventing hypertensive disorders and related problems.Cochrane Database Syst Rev 3:CD001059, 2006

5. Spatling L, Spatling G: Magnesium supplementation in pregnancy. Adouble-blind study. Br J Obstet Gynaecol 95:120-125, 1988

6. Sibai BM, Villar MA, Bray E: Magnesium supplementation during preg-nancy: a double-blind randomized controlled clinical trial. Am J ObstetGynecol 161:115-119, 1989

7. Makrides M, Duley L, Olsen SF: Marine oil, and other prostaglandinprecursor, supplementation for pregnancy uncomplicated by pre-eclampsia or intrauterine growth restriction. Cochrane Database SystRev 3:CD003402, 2006

8. Meher S, Duley L: Garlic for preventing pre-eclampsia and its compli-cations. Cochrane Database Syst Rev 3:CD006065, 2006

9. Meher S, Duley L: Progesterone for preventing pre-eclampsia and itscomplications. Cochrane Database Syst Rev CD006175, 2006

0. Rumbold A, Duley L, Crowther C, et al: Antioxidants for preventingpre-eclampsia. Cochrane Database Syst Rev CD004227, 2005

1. Poston L, Briley AL, Seed PT, et al: Vitamin C and vitamin E in pregnantwomen at risk for pre-eclampsia (VIP trial): randomised placebo-con-trolled trial. Lancet 367:1145-1154, 2006

2. Rumbold AR, Crowther CA, Haslam RR, et al: Vitamins C and E and therisks of preeclampsia and perinatal complications. N Engl J Med 354:1796-1806, 2006

3. Riyazi N, Leeda M, de Vries JI, et al: Low-molecular-weight heparincombined with aspirin in pregnant women with thrombophilia and ahistory of preeclampsia or fetal growth restriction: a preliminary study.Eur J Obstet Gynecol Reprod Biol 80:49-54, 1998

4. Kupferminc MJ, Fait G, Many A, et al: Low-molecular-weight heparinfor the prevention of obstetric complications in women with thrombo-philias. Hypertens Pregnancy 20:35-44, 2001

5. Arias F, Romero R, Joist H, et al: Thrombophilia: a mechanism ofdisease in women with adverse pregnancy outcome and thromboticlesions in the placenta. J Matern Fetal Med 7:277-286, 1998

6. Mello G, Parretti E, Fatini C, et al: Low-molecular-weight heparin low-ers the recurrence rate of preeclampsia and restores the physiologicalvascular changes in angiotensin-converting enzyme DD women. Hy-pertension 45:86-91, 2005

7. Widmer M, Villar J, Benigni A, et al: Mapping the theories of pre-eclampsia and the role of angiogenic factors: a systematic review. Ob-stet Gynecol 109:168-180, 2007

8. Duckitt K, Harrington D: Risk factors for pre-eclampsia at antenatalbooking: systematic review of controlled studies. Br Med J 330:565,2005

9. Conde-Agudelo A, Belizan JM: Risk factors for pre-eclampsia in a largecohort of Latin American and Caribbean women. Br J Obstet Gynaecol107:75-83, 2000

0. Sibai BM, Hauth J, Caritis S, et al: Hypertensive disorders in twin versussingleton gestations. National Institute of Child Health and HumanDevelopment Network of Maternal-Fetal Medicine Units. Am J ObstetGynecol 182:938-942, 2000

1. Stamilio DM, Sehdev HM, Morgan MA, et al: Can antenatal clinical andbiochemical markers predict the development of severe preeclampsia?Am J Obstet Gynecol 182:589-594, 2000

2. Abi-Said D, Annegers JF, Combs-Cantrell D, et al: Case-control study ofthe risk factors for eclampsia. Am J Epidemiol 142:437-441, 1995


4. Dugoff L, Hobbins JC, Malone FD, et al: Quad screen as a predictor ofadverse pregnancy outcome. Obstet Gynecol 106:260-267, 2005

5. Tucker MJ, Berg CJ, Callaghan WM, et al: The Black-White disparity inpregnancy-related mortality from 5 conditions: differences in preva-
lence and case-fatality rates. Am J Public Health 97:247-251, 2007

RSON

Pftwbcfpoedoc

bet

*

†

‡

A

1

ecurrent Preterm Birthhali Mazaki-Tovi, MD,*,† Roberto Romero, MD,*,‡ Juan Pedro Kusanovic, MD,*ffer Erez, MD,* Beth L. Pineles, BS,* Francesca Gotsch, MD,* Pooja Mittal, MD,*,†

andor Gabor Than, MD,* Jimmy Espinoza, MD,† and Sonia S. Hassan, MD†

Recurrent preterm birth is frequently defined as two or more deliveries before 37 completedweeks of gestation. The recurrence rate varies as a function of the antecedent for pretermbirth: spontaneous versus indicated. Spontaneous preterm birth is the result of eitherpreterm labor with intact membranes or preterm prelabor rupture of the membranes. Thisarticle reviews the body of literature describing the risk of recurrence of spontaneous andindicated preterm birth. Also discussed are the factors which modify the risk for recurrentspontaneous preterm birth (a short sonographic cervical length and a positive cervicovagi-nal fetal fibronectin test). Patients with a history of an indicated preterm birth are at risk notonly for recurrence of this subtype, but also for spontaneous preterm birth. Individuals ofblack origin have a higher rate of recurrent preterm birth.Semin Perinatol 31:142-158. © 2007 Published by Elsevier Inc.

KEYWORDS recurrent preterm birth, indicated preterm birth, spontaneous preterm birth,rupture of membranes, parturition

nftHl“pp“

ab

DPaetaatpca

reterm birth is the leading cause of perinatal morbidityand mortality worldwide.1 A preterm delivery is a risk

actor for subsequent preterm birth.2-22 Preterm birth can behe result of three obstetrical circumstances: 1) preterm laborith intact membranes; 2) preterm prelabor rupture of mem-ranes (PROM); and 3) “indicated” preterm birth, which oc-urs when maternal or fetal indications require delivery be-ore 37 weeks of gestation. The most common indications arereeclampsia and small for gestational age (SGA). Spontane-us preterm parturition is a syndrome caused by multipletiologies (Fig. 1), which are expressed by synchronous oryssynchronous activation of the common terminal pathwayf parturition. The reader is referred to recent reviews for fullonsideration of this concept.23,24

Although many studies have focused on the rate of pretermirth,25-57 an important consideration is whether these deliv-ries are the result of spontaneous labor (with intact or rup-ured membranes) or “indicated” preterm deliveries. The

Perinatology Research Branch, Intramural Division, NICHD/NIH/DHHS,Hutzel Women’s Hospital, Bethesda, MD, and Detroit, MI.

Department of Obstetrics and Gynecology, Wayne State University/HutzelWomen’s Hospital, Detroit, MI.

Center for Molecular Medicine and Genetics, Wayne State University, De-troit, MI.

ddress correspondence to Roberto Romero, MD, Perinatology ResearchBranch, Intramural Division, NICHD/NIH/DHHS, Hutzel Women’sHospital, 3990 John R St., Box #4, Detroit, MI 48201. E-mail

[email protected]

42 0146-0005/07/$-see front matter © 2007 Published by Elsevier Inc.doi:10.1053/j.semperi.2007.04.001

eed for this distinction is based on the premise that the riskactors for recurrent preterm PROM, preterm labor with in-act membranes, preeclampsia, and/or SGA are different.owever, recent observations suggest that there may be over-

ap among these conditions,18,19 so that a patient with anindicated” preterm birth may also be at risk for spontaneousreterm birth.18,19 The converse may also be true (ie, that aatient with a spontaneous preterm birth is at risk for anindicated” preterm birth in a subsequent pregnancy).

This review will present a summary of the literature andims to clarify the risk of recurrent disease and the biologicalasis for recurrent preterm birth.

efinition of Preterm Birthreterm deliveries are those occurring between fetal viabilitynd 37 completed weeks of gestation (menstrual age). How-ver, the lower limit of gestational age used to define a pre-erm birth has ranged from 13 to 24 weeks of gestationmong various reports.21,58,59 Our view is that the delivery ofpreviable fetus should be considered a spontaneous abor-

ion rather than a spontaneous preterm birth. Otherwise,erinatal and infant mortality estimates in a population orountry will be biased by the frequency of late spontaneousbortions.

The precise definition of viability, however, is subject to
ebate given the increased frequency of survival at very low

gwpdvt

ldttgtsgTcTpiivu

RRbthpd1mab

bc

RSTSRoi

aqrPbttsabt

iacwowp(�

Fpp

Recurrent preterm birth 143

estational ages. Clearly, some infants can survive around 24eeks of gestation, but none at 20 weeks. Therefore, weropose the range of 24 to 36 6/7 weeks of gestation for theefinition of preterm birth. If and when technological ad-ances allow substantial survival below 24 weeks of gesta-ion, this definition should be revised.

A birth weight of 500 g has also been used to define theower limit of viability.11,60 The popularity of this definitionerives from its simplicity. Birthweights can easily be ob-ained anywhere in the world at a very low cost. The limita-ions of this approach are that viable neonates born at viableestational ages and affected by intrauterine growth restric-ion (IUGR) may have birthweights below 500 g, and thatome previable infants may weigh more than 500 g. Ideally,estational age at birth should be used to define viability.here are, however, practical obstacles derived from the un-ertainty of gestational age estimation in many countries.his problem can be overcome in areas where ultrasound iserformed routinely in early pregnancy, but not elsewhere,

ncluding underdeveloped countries. The criteria for the def-nition of viability have implications for the calculation ofital statistics and comparisons of these among different pop-lations.

ecurrent Preterm Birthecurrent preterm birth is defined as two or more deliveriesefore 37 completed gestational weeks.2,9,12,59,61,62 However,he definition among studies is not uniform. Criteria thatave varied and may affect estimation of the rate of recurrentreterm birth include: 1) gestational age thresholds used forefining the upper (ie, 35 or 36 weeks)9,12,63,64 and lower (ie,3 to 28 weeks)10,59 limits of preterm birth; 2) inclusion ofultiple gestations65-67; 3) inclusion of spontaneous, as well

s indicated preterm births11,64; 4) the number of preterm

igure 1 Pathological processes implicated in thereterm parturition syndrome. (Reproduced withermission from Romero and coworkers.24)

irths required to fulfill the definition of recurrent preterm f

irth8; and 5) the requirement that the preterm births beonsecutive.8

ecurrentpontaneous Preterm Birth

he Frequency of Recurrentpontaneous Preterm Birthecurrent spontaneous preterm birth is defined as more thanne preterm birth related to spontaneous onset of labor withntact membranes or preterm PROM.

Several studies have consistently indicated that patients withprevious spontaneous preterm birth are at risk for a subse-uent spontaneous preterm delivery.2-22 Iams et al9 reported theesults of a secondary analysis of the data from the Pretermrediction Study, conducted under the leadership of Golden-erg et al.68 Among 378 patients with a prior spontaneous pre-erm birth or spontaneous abortion in the second trimester (ges-ational age range: 18-36 weeks), the rate of recurrentpontaneous preterm birth (�35 weeks) varied between 14%nd 15%, in contrast to the 3% rate of spontaneous pretermirth among 904 parous women with a previous uncomplicatederm delivery (Table 1).

The rate of recurrent preterm birth was modified accord-ng to the sonographic cervical length in the index pregnancynd the presence of a positive test for fetal fibronectin inervicovaginal fluid at 22 to 24 weeks of gestation.9 Amongomen with a previous spontaneous preterm birth, the ratef recurrence (�35 weeks) was the highest (64%) amongomen with a sonographically short cervix (�25 mm) and aositive fetal fibronectin test. The lowest rate of recurrence7%) occurred in patients with a sonographic cervical length35 mm and a negative fetal fibronectin test.9

Patients with a positive fibronectin test were at higher risk
or spontaneous recurrent preterm birth regardless of cervi-

cogtsratbmrbc

tebhgtsostl9osds

PoPoqePtroP3

ic

dlPat2t

r3wcpPgpwop222

SplaaaPaTPnrwtpwcrlp�

waBr

fitP

TncP

*

144 S. Mazaki-Tovi et al.

al length: 28% for patients with a positive test compared tonly 7% for patients with a negative test. Similarly, a sono-raphic short cervix contributed to the risk of recurrent spon-aneous preterm birth. Among patients with a history ofpontaneous preterm birth and a short cervix, the rate ofecurrence was 25% if the fetal fibronectin test was negativend 64% if the test was positive. This information can be usedo counsel patients about their risk of spontaneous pretermirth. However, further investigation is required in which aathematical model is generated to predict the individual

isk for preterm birth based on clinical, sonographic, andiochemical parameters in which results are expressed asontinuous variables rather than categorical results.

Mercer and coworkers10 performed a secondary analysis ofhe same dataset using different data stratification in order tovaluate the association between prior spontaneous pretermirth and subsequent pregnancy outcome. Women with aistory of spontaneous preterm delivery (before 37 weeks ofestation) had a 2.5-fold increased risk (95% confidence in-erval (CI), 1.9-3.2) of spontaneous preterm delivery in aubsequent pregnancy compared to women with no historyf spontaneous preterm delivery (21.7% versus 8.8%, re-pectively, P � 0.001). This risk was particularly high whenhe analysis focused on recurrent spontaneous preterm de-ivery before 28 weeks of gestation [relative risk (RR) 10.6,5% CI 2.9-38.3].10 Moreover, the earlier the gestational agef the first preterm delivery, the greater the risk for recurrentpontaneous preterm birth (RR 2.4, 2.7, and 3.1 for priorelivery at 35-36, 28-34, and 23-27 weeks of gestation, re-pectively).10

reterm Prelabor Rupturef Membranes and Recurrent Preterm BirthROM is defined as spontaneous rupture of the chorioamni-tic membranes before the onset of labor.69 Since the conse-uences of PROM depend on the gestational age at which thepisode occurs, this condition has been classified as pretermROM or term PROM, depending on whether the rupture ofhe membranes occurs before or after 37 weeks of gestation,espectively.70-76 Term PROM occurs in approximately 10%f pregnancies,70-72,74,77,78 whereas the frequency of pretermROM is 2% to 3.5%.70-72,74,77,78 Preterm PROM accounts for

able 1 Probability and 95% Confidence Intervals of Sponta-eous Recurrent Preterm Birth at Less Than 35 Weeks Ac-ording to the Gestational Age of the Previous Spontaneousreterm Birth*

Gestational Age atDelivery in Previous

SpontaneousPreterm Birth

Probability (95% CI)of Spontaneous

Recurrent PretermBirth at <35 Weeks

18-26 weeks 0.15 (0.05-0.37)27-31 weeks 0.15 (0.05-0.38)32-36 weeks 0.14 (0.05-0.32)>37 weeks 0.03 (0.03-0.03)

Modified from Iams, et al.9 with permission.

0% to 40% of preterm deliveries,70-72,74,77,78 and it is a lead- d

ng clinically identifiable cause of preterm birth and a majorontributor to perinatal morbidity and mortality.21,58,70-78

Analysis of data from the Collaborative Perinatal Projectemonstrated that, among women with a previous term de-

ivery not complicated by PROM, the frequency of pretermROM in a subsequent pregnancy is 4%.79 In contrast,mong patients with two successive singleton pregnancies inhe dataset (n � 5230), the frequency of preterm PROM is1% if the first pregnancy resulted in a preterm delivery dueo preterm PROM.79

Several investigators have confirmed the high recurrenceate of preterm PROM: 1) Asrat and coworkers80 reported a2% (95% CI 23.9-40.5) risk of recurrence in 121 patientsith a previous episode of preterm PROM; 2) Ekwo and

oworkers81 reported that women with preterm PROM in arevious pregnancy had a 5.5-fold higher risk of pretermROM in a subsequent pregnancy than those in the controlroup; and 3) Mercer and coworkers58 reported that, com-ared to women with no history of preterm PROM, womenith previous preterm PROM had a higher risk of spontane-us preterm delivery due to preterm PROM in the indexregnancy (13.5% versus 4.1%, P � 0.001; RR: 3.3, 95% CI.1-5.2) as well as a higher risk of preterm PROM at less than8 weeks (1.8% versus 0.1%, P � 0.01; RR 13.5, 95% CI3-80.3).Mercer and coworkers82 used the Preterm Prediction

tudy population to determine the risk factors for subsequentreterm birth due to preterm PROM alone. Preterm PROM at

ess than 35 weeks of gestation occurred in 2% of patientsnd at less than 37 weeks in 4.5% of patients. Preterm PROMccounted for 32.6% of all preterm deliveries. Clinical char-cteristics associated with preterm birth due to pretermROM, derived from a multivariable analysis, and stratifieds preterm PROM �37 and �35 weeks are displayed inable 2. In nulliparous women, the risk factors for pretermROM were a cervical length �25 mm, working during preg-ancy, and the presence of medical complications [the oddsatios (OR) ranged between 3 and 3.7].82 Among multiparousomen, a previous preterm birth due to preterm PROM was

he primary risk factor for preterm PROM in a subsequentregnancy (OR for preterm PROM at �35 weeks: 4.1; �37eeks: 3.1). Noteworthy is that a previous preterm birth

aused by preterm labor with intact membranes was also aisk factor for preterm PROM, although the odds ratios wereower than if the previous preterm birth was the result ofreterm PROM (OR for preterm PROM at �35 weeks: 2.6;37 weeks: 1.8).82

Interestingly, the only risk factor consistently associatedith preterm PROM at �37 and �35 weeks in both nullip-

rous and multiparous women was a short cervical length.acterial vaginosis was not found to be a risk factor for recur-ent preterm birth.82

Among multiparous women, a short cervix, a positive fetalbronectin test, and a history of preterm birth following pre-erm PROM increased the frequency of recurrent pretermROM at �35 weeks to 25%. If recurrent preterm PROM was
efined as �37 weeks, the combination of these three risk

fm

pbPdsPpmwirsmwptp

Pcaaddacptdi

fitTttaacf

wclbtbpalnmaacbsbpcf

TTcbetcswiwttr(ts2d

T

CPPWMFBFF

* , fetal


actors increased the risk 7.8-fold over the reference group ofultiparous women who had none of these risk factors.82

The mechanisms responsible for the association betweenrevious preterm PROM, short cervix, and positive fetal fi-ronectin and recurrent preterm birth caused by pretermROM have not been elucidated. It is likely that an insulturing gestation (eg, intrauterine infection) would be re-olved by preterm labor with intact membranes or pretermROM. We have proposed83 that selection of the specifichenotype may be determined by genetic and/or environ-ental factors. For example, if patients carry DNA variantshich predispose to an excess production of matrix-degrad-

ng enzymes, such patients will go into premature labor afterupture of membranes. On the other hand, if the genotype isuch that the generation of uterotonic agents rather thanatrix-degrading enzymes is favored, then preterm laborith intact membranes will be the clinical expression of thereterm parturition syndrome. The genotype may explainhe tendency for the same phenotype to occur in subsequentregnancies (ie, preterm PROM).The relationship between a short cervix and preterm

ROM could be due to intrauterine infection.84,85 A longervix with a well-established mucus plug may serve as annatomical and biochemical host defense mechanism againstscending intrauterine infection.86-92 A short cervix may pre-ispose to ascending intrauterine infection by shortening theistance between microorganisms in the lower genital tractnd the chorioamniotic membranes.89,93 In addition, the pro-ess of cervical shortening may lead to the loss of the mucuslug. Cervical mucus contains antimicrobial properties, at-ributed at least in part to antimicrobial peptides such asefensins, lactoferrin, calprotectin, and bacterial permeabil-

ty factor.86-90

The relationship between a positive cervicovaginal fetalbronectin test and subsequent preterm PROM has been at-ributed to the presence of upper genital tract infection.68

his interpretation has been based on the association be-ween a positive fetal fibronectin test in the midtrimester andhe subsequent demonstration of histologic chorioamnionitist the time of preterm delivery.68 However, studies in whichmniocentesis was performed in women with positive cervi-ovaginal fetal fibronectin have found that intraamniotic in-

able 2 Risk Factors Associated with Preterm PROM (at Less

Risk Factor

ervical length <25 mmrevious preterm birth with preterm PROMrevious preterm labor with intact membranesorking during pregnancyedical complications

FN (�)VFN (�) and absent BVFN (�) and present BV

Modified from Mercer, et al.82 with permission. Abbreviations: FFN

ection or inflammation is present in less than 2% of patients 3

ith a positive fibronectin test.94 Because fetal fibronectin is aomponent of the extracellular matrix located in the chorioneave95 and its abundance in cervicovaginal fluid increasesefore both preterm68,96-109 and term labor,110-119 we proposehat detection of this protein is a marker of decidual/mem-rane activation and, therefore, of the common terminalathway of parturition. Thus, a history of preterm PROM andpositive test for fetal fibronectin in the midtrimester are

ikely to reflect activation of the decidua/membrane compo-ent of the pathway. We propose a sequence of events thatay explain the empirical observations reported by Mercer

nd coworkers.82 A patient with a previous preterm PROM ist risk for subsequent PROM.82 If such a patient has a shortervix, the risk of recurrent preterm PROM would increaseecause of ascending intrauterine infection. If the infection isuch that it activates chorion and decidua, then fetal fi-ronectin will be positive. Of course, it is also possible that aatient with preterm PROM would have activation of theommon terminal pathway (and therefore a positive test foretal fibronectin) with a long cervix.

win Gestation and Recurrent Preterm Birthhere are conflicting data as to whether preterm birth in theontext of a multiple gestation is a risk factor for pretermirth in a future singleton pregnancy. Menard and cowork-rs62 were the first to examine the recurrence rate after pre-erm birth of a twin gestation. The authors reported the out-omes of 144 women who first delivered twins, followed by aubsequent singleton gestation. Preterm delivery (before 37eeks of gestation), occurred in 59.7% of twin gestations and

n 14.6% of the subsequent singleton pregnancies. Amongomen who delivered preterm twins, 19.6% delivered pre-

erm in the subsequent singleton pregnancy. Preterm birth inwin gestations was associated with a significantly increasedisk of preterm delivery in a subsequent singleton pregnancyRR 2.87, 95% CI 1.02-8.09). Among the subset of patientshat delivered twins before 30 weeks, 42% of the subsequentingleton pregnancies delivered preterm (RR 6.11, 95% CI.07-18.02). The RR of preterm birth of a singleton afterelivery of twins between 30 and 34 weeks of gestation was

35 weeks) Stratified by Parity*

Nulliparous Multiparous

95% CI OR 95% CI

3.8-25.9 4.2 2.0-8.9– 4.1 2.0-8.7– 2.6 1.2-5.3

1.5-18.7 n.s. n.s.1.1-16.0 n.s. n.s.

n.s. n.s. n.s.n.s. n.s. n.s.n.s. 9.0 3.6-22.5n.s. 2.8 1.2-6.3

fibronectin; BV, bacterial vaginosis; n.s., non-significant.

Than

OR

9.9––

5.34.2n.s.n.s.n.s.n.s.

.63 (95% CI 1.02-12.92). However, women who delivered

ti

lrHrns3Cir

CaMTmswcd“s

dcsamtdwIwdsi1tmwws

sohcawtwoio

iv

ppsetc

rdcgcbancftprvwpobst

gmtLhuactahiapotme

co5dcla


wins between 34 and 37 weeks of gestation did not have anncreased risk for recurrent preterm birth.

In contrast, Rydhstrom59 reported that a preterm twin de-ivery, regardless of etiology, did not increase the risk ofecurrent preterm birth in a subsequent singleton gestation.owever, a prior preterm singleton delivery increased the

isk of preterm birth in subsequent singleton and twin preg-ancies. Bloom and coworkers12 reported that women with aingleton gestation that resulted in preterm birth at less than5 weeks have an increased risk for recurrence (OR 5.6, 95%I 4.5-7.0). However, those whose first pregnancy resulted

n twins delivered at less than 35 weeks did not have a higherisk of recurrent preterm birth (OR 1.9, 95% CI 0.46-8.14).

ervical Insufficiencys a Cause of Recurrentidtrimester Abortion/Preterm Birth

he clinical diagnosis of cervical insufficiency is traditionallyade in patients with a history of recurrent mid-trimester

pontaneous abortions and/or early preterm deliveries inhich “the basic process is thought to be the failure of the

ervix to remain closed during pregnancy.”120 Thus, both byefinition and clinical practice, the condition now termedcervical insufficiency” is recognized as one that recurs inubsequent pregnancies.

Digital examination of the cervix was the method used toetermine cervical status (effacement, dilation, position, andonsistency) before the introduction of ultrasound. Cervicalonography has become an objective and reliable method tossess cervical length, which approximates cervical efface-ent. The shorter the sonographic cervical length in the mid-

rimester, the higher the risk of spontaneous preterm labor/elivery.121-125 However, there is no agreement concerninghat constitutes a sonographic short cervix. For example,

ams et al122 proposed that a cervix of 26 mm or shorter at 24eeks of gestation increases the risk for spontaneous pretermelivery (RR: 6.19, 95% CI 3.84-9.97). The prevalence ofpontaneous preterm delivery (defined as less than 35 weeks)n this study was 4.3%, and the positive predictive value was7.8% for a cervical length �25 mm at 24 weeks of gesta-ion.122 Other investigators have proposed a cut-off of 15m, because a cervical length of 15 mm or less is associatedith nearly a 50% rate of spontaneous preterm delivery at 32eeks of gestation or less, when neonatal morbidity is sub-

tantial.123,125

Sonographic cervical length is not a screening test forpontaneous preterm delivery, because only a small fractionf all patients who will have a spontaneous preterm birthave a short cervix in the mid-trimester. Previous studiesonducted at our institution have indicated that only 8% ofll patients who will have a preterm delivery at less than 32eeks of gestation have a cervical length of 15 mm or less in

he mid-trimester.125 The converse is also true. Amongomen with a short cervix, some have adverse pregnancyutcomes and others have uncomplicated term deliver-es.66,121-123,126-140 Only half of women with a cervical length
f 15 mm or less deliver before 32 weeks of gestation.125 This t
ndicates that a short cervix should not be equated with “cer-ical insufficiency.”

Sonographic cervical length can modify the a priori risk forreterm delivery. For example, in patients with a history ofreterm delivery, a twin gestation, or a triplet gestation, ahort cervix confers an increased risk for preterm deliv-ry.109,141-149 Indeed, among women with a history of spon-aneous preterm birth, the risk of recurrence increases aservical length shortens.9

The hypothesis that cervical competence or sufficiency rep-esents a spectrum was studied by Parikh and Mehta, who usedigital examination of the cervix and concluded that degrees ofervical competence do not exist.150 Iams et al, using sono-raphic examination of the cervix, suggested that cervical suffi-iency/insufficiency is a continuum,66 with a strong relationshipetween cervical length in the index pregnancy and gestationalge at delivery in the first pregnancy. This relationship wasearly linear; patients with a typical history of a cervical insuffi-iency (painless dilation) do not constitute a separate grouprom those with a history of spontaneous preterm delivery (pre-erm labor or preterm PROM).66 Similar results have been re-orted by Guzman et al.151 Collectively, these studies suggest aelationship between a history of preterm delivery and the cer-ical length in a subsequent pregnancy. Inasmuch as patientsith a short cervix are at increased risk for a mid-trimesterregnancy loss (clinically referred to as “cervical insufficiency”)r spontaneous preterm delivery with intact or ruptured mem-ranes,66,121-123,126-131,133-140,151,152 a short cervix could be con-idered the expression of a spectrum of cervical diseases or func-ions.

We have proposed that cervical insufficiency is one of thereat “obstetrical syndromes.”153 Cervical ripening in theid-trimester may be the result of: 1) the loss of connective

issue after a cervical operation such as conization154-156 orEEP procedure156; 2) a congenital disorder such as cervicalypoplasia after diethylstilbestrol exposure157-160; 3) intra-terine infection161,162; and 4) a suspension of progesteronection.163 There is experimental evidence that progesteronean reverse cervical compliance induced by the administra-ion of dexamethasone to pregnant sheep.164 Sherman165 haslso generated evidence that the administration of 17 alphaydroxyprogesterone caproate (17 OH P) may be beneficial

n patients with clinically diagnosed “cervical insufficiency”nd a cervical disorder that manifests itself with the clinicalresentation of “cervical insufficiency.” Each of these causesf the syndrome could be affected by genetic or environmen-al factors. The possibility of novel and yet-to-be-discoveredechanisms of disease playing a role must also be consid-

red.A proportion of patients presenting with asymptomatic

ervical dilation in the mid-trimester have microbial invasionf the amniotic cavity (MIAC)161,162 that can be as high as1.5%.162 Microbial invasion of the amniotic cavity may beue to premature cervical dilation with the exposure of thehorioamniotic membranes to the microbial flora of theower genital tract. Microorganisms may gain access to themniotic cavity by crossing intact membranes.162 Under
hese circumstances, infection would be a secondary phe-

nehgtaacmvwctdcTch

WORTbAt

G2w31

sg

orws

fp31wcaiocTrcwapgi

AfSp

T

M

I

I

GS

M*

†‡§ gh 199


omenon to primary cervical disease. An alternativexplanation is that primary intrauterine infection (ascending,ematogeneous166), or one caused by activation of microor-anisms present within the uterine cavity167 in the secondrimester of pregnancy produces myometrial contractilitynd cervical ripening. Because uterine contractions are usu-lly clinically silent in the mid-trimester of pregnancy, thelinical picture of an infection-induced spontaneous abortionay be indistinguishable from that of an incompetent cer-

ix.65,162 Recently, we have established that 9% (5/57) ofomen with a short endocervix (less than 25 mm) have mi-

robiologically proven intraamniotic infection,168 suggestinghat these infections are subclinical and may precede theevelopment of the clinical picture of acute “cervical insuffi-iency” (dilated and effaced cervix with bulging membranes).he issue of whether subclinical intrauterine infection is aause of recurrent cervical insufficiency and preterm birthas not been answered.

omen of African-Americanrigin Have a Greater Risk ofecurrent Preterm Birth than Caucasianshere is a well-established disparity in the rate of pretermirth among ethnic groups in the U.S.8,11,169-176 Individuals offrican-American origin are at higher risk for recurrent pre-

erm birth.A large population-based cohort study11 in the state of

eorgia found that, among women who delivered between0 and 31 weeks of gestation in their first pregnancy, blackomen had a higher rate of recurrent preterm birth at 20 to1 weeks than did white women [black � 13.4% (95% CI

able 3 Odds Ratios for Recurrence of Preterm Delivery or L

Maternal Characteristic in SecondPregnancy

Delivery

White (n � 84

aternal age (years)10-17 2.3 (0.9-5.6)18-19 1.0 (0.5-2.0)20-49 1.0

nitiation of prenatal care (trimester)First 1.0Second, third, or none 1.2 (0.6-2.2)

nterpregnancy interval, months<6 1.1 (0.5-2.1)6-11 1.6 (0.9-2.9)12-47 1.0>47 1.0 (0.4-2.1)oodness-of-fit P value‡ 0.72moking during the pregnancy§Yes 0.4 (0.2-1.1)No 1.0

odified from Adams, et al.11 with permission.Odds ratio for type of second pregnancy are controlled for all of th

95% confidence intervals.Referent group is delivery in second pregnancy at gestation > 37Goodness-of-fit for model including all variables except smoking.Analysis restricted to second deliveries occurring from 1989 throu

1.4-15.6) versus white � 8.2% (95% CI 6.6-10.1)]. The a

ame was the case for deliveries between 32 and 36 weeks ofestation.

Of interest was that teenagers whose first preterm deliveryccurred between 20 and 31 weeks of gestation had twice theisk of recurrent preterm birth (20-31 weeks) than that ofomen 20 to 49 years of age (Table 3). This observation was

ignificant only among African-American women.Kitska and coworkers64 used a maternally linked database

rom the Missouri Department of Health to study racial dis-arities and recurrent preterm birth. The study focused on68,633 mothers who had two or more deliveries between978 and 1997. The frequency of recurrent preterm birthas 3.1% among African-Americans and 0.6% among Cau-

asians (RR, 5.40; 95% CI 5.06, 5.75). Logistic regressionnalysis indicated that being of African-American descentncreased the risk for recurrent preterm birth independentlyf other factors, such as medical complications and low so-ioeconomic status (adjusted OR, 4.11; 95% CI 3.78, 4.47).wo additional findings of this study were that: 1) the recur-ent preterm birth in women of African-American origin oc-urred at an earlier median gestational age than in Caucasianomen (31 weeks versus 33 weeks); and 2) the gestational

ge of the recurrent preterm birth was similar to that of therevious preterm birth and most likely to occur at the sameestational age (Fig. 2). This finding was consistent amongndividuals in both ethnic groups.

dditional Risk Factorsor Recurrent Preterm Birtheveral environmental factors have been implicated in recurrentreterm birth. Cnattingius and coworkers61 studied the associ-

rthweight Newborn by Race in Georgia, 1980-1995*

0-31 wk† Delivery at 32-36 wk†

lack (n � 145) White (n � 712) Black (n � 1059)

2.0 (1.2-3.5) 1.3 (0.8-2.0) 1.3 (1.1-1.7)1.2 (0.8-2.0) 1.3 (1.0-1.7) 1.2 (1.0-1.4)1.0 1.0 1.0

1.0 1.0 1.01.2 (0.8-1.7) 1.1 (0.9-1.4) 1.1 (1.0-1.3)

1.4 (0.9-2.3) 1.0 (0.7-1.3) 1.2 (1.2-1.5)0.9 (0.5-1.5) 1.2 (0.9-1.5) 1.1 (0.9-1.3)1.0 1.0 1.00.8 (0.4-1.5) 0.9 (0.7-1.2) 0.7 (0.6-0.9)0.33 0.29 0.93

1.7 (0.2-14.5) 0.8 (0.6-1.2) 0.6 (0.3-1.1)1.0 1.0 1.0

r variables in the table except smoking; figures in parentheses are

.

5. Association adjusted for all other variables in the model.

ow-Bi

at 2

) B

e othe

weeks

tion among smoking, previous very early preterm or moderate

pse2spontmsh

t1ct9pnwh

twc

ttvpwswannamabpoo


reterm delivery (before 32 weeks and at 32 to 36 weeks, re-pectively), and the risk of a subsequent very preterm or mod-rately preterm delivery in a population-based cohort of43,858 women in Sweden. The OR for a very early pretermecond delivery among the women who smoked 1 to 9 cigaretteser day was 1.4 (95% CI, 1.1, 1.7) and for those who smoked 10r more cigarettes per day 1.6 (95% CI, 1.3, 2.0), as compared toonsmokers. Furthermore, women who stopped smoking be-ween pregnancies were not at increased risk for very early oroderate preterm delivery, whereas the women who started to

moke in the second pregnancy had the same risk as those whoad continued to smoke.Merlino et al177 investigated the association between ma-

ernal weight loss and recurrent preterm birth in a cohort of241 patients. Women whose body mass index (BMI) de-reased more than 5 kg/m2 had more frequent recurrent pre-erm birth than those whose BMI did not (21.1% versus.3%, P � 0.01). For those with a term birth in the firstregnancy, the rate of preterm birth in the subsequent preg-ancy was not affected by a decline in BMI. In contrast,omen with a preterm birth in the first pregnancy had a

Figure 2 Concordance in timing of preterm (20-34 6/7 wpreterm birth, 1989 to 1997. The line represents the exdistributed event. The bars represent the timing for eachCaucasians (B), or African Americans (C) in correlation wfrom Kistka and coworkers.64)

igher rate of recurrent preterm birth if BMI decreased more w

han 5 kg/m2 (80.0% versus 28.2%, P � 0.01). Hence,omen whose BMI declines between pregnancies are at in-

reased risk for recurrent preterm birth.The effect of sexual behavior on the risk of recurrent pre-

erm birth was the subject of a secondary analysis of a mul-icentric observational study of the association between cer-ical ultrasound at 16 to 18 weeks and the risk for recurrentreterm birth. Women (n � 187) with singleton gestationsho were at high risk for preterm birth because of a prior

pontaneous preterm birth at less than 32 weeks of gestationere included.178 A sexual history was obtained by interview

t the time of enrollment. Information gathered included theumber of sexual partners during the patient’s lifetime, theumber of sexual partners during the patient’s pregnancy,nd the frequency of sexual intercourse in the precedingonth. The greater the number of sexual partners in a wom-

n’s lifetime, the higher the frequency of recurrent pretermirth (1 partner 19%, 2 to 3 partners 29%, more than 4artners 44%, P � 0.007). Of interest, neither the frequencyf sexual intercourse during early pregnancy nor the numberf partners were risk factors for recurrent preterm birth,

f gestation) birth in Missouri to a mother with previousGaussian curve if concordance in timing is a normallybirth after the initial preterm birth for all mothers (A),

e expected normal curve. (Reproduced with permission

eeks opectedpretermith th

hich is consistent with previous reports.179-184

RPIpmtroieddpa

pHm

idwmrb

ae(wtp((wbr

(wiosea

c

IffIPd

bpovbttt

bsrwP“bpp

vpsrbwdwitsbOtc3

cmw9c9hdTwwpl

IawPPP


ecurrent Indicatedreterm Birth

ndicated preterm births are those resulting from delivery ofatients before term, due to complications that place theother and/or fetus at risk. Various authors include among

hose complications hypertension-related disorders, obstet-ical hemorrhage (placenta previa, placental abruption, andther causes of antepartum hemorrhage), and all medicallynduced preterm deliveries.12,40,185 Other investigators18 cat-gorize indicated preterm births into: 1) ischemic placentalisease (ie, preeclampsia, SGA, placental abruption, and fetalistress); and 2) miscellaneous (fetal malformations, placentarevia, unexplained vaginal bleeding, chronic hypertension,nd others).

The incidence of indicated preterm birth has been re-orted to range from 1% to 5.5% of all deliveries.12,176,186,187

owever, indicated preterm birth accounts for approxi-ately one-third of all preterm births.40,185,187

Meis et al185 reported a study examining the risk factors forndicated preterm birth using the Preterm Prediction Studyata set. A history of a previous indicated preterm deliveryas associated with an OR of 2.8 (95% CI 1.5-5.4; P � 0.002;ultivariable analysis by logistic regression including other

isk factors for indicated preterm birth for recurrent pretermirth).Bloom et al12 reported the largest study conducted today in

single unit and concluded that an indicated preterm deliv-ry in singleton gestations is associated with an OR of 5.495% CI 3.1-9.2) for recurrent preterm birth at less than 35eeks of gestation in comparison to patients who delivered at

erm in their first pregnancy. Patients who had an indicatedreterm birth between 24 and 28 weeks had an OR of 12.595% CI 3.8-40.7) for recurrent preterm delivery and 1095% CI 4.8-20.8) if they were delivered between 29 and 32eeks of gestation. In contrast, patients who were deliveredetween 33 and 34 weeks did not have a higher risk forecurrent preterm birth.

Patients with one prior preterm birth had an OR of 2.495% CI 1.5-4.1) for indicated preterm delivery compared toomen with no history of preterm birth. Moreover, the OR

ncreases to 5.2 (95% CI 2.2-11.9) when the patients had twor more previous preterm deliveries.176 Collectively, thesetudies suggest that indicated preterm birth is not an isolatedvent and puts the patient at risk for a subsequent indicated,s well as spontaneous, preterm birth.

The subjects of recurrent preeclampsia and SGA are dis-ussed elsewhere in this issue of the Seminars.

s the Recurrence Riskor Preterm Birth Differentor Spontaneous Versusndicated Preterm Birth?reterm births have been classified as “spontaneous” or “in-
icated” because of the implicit assumption that preterm la- f
or with intact membranes and preterm PROM share patho-hysiologic features and clinical presentation (spontaneousnset of labor). Preterm preeclampsia, fetal distress, and se-ere IUGR—the most common causes of indicated pretermirth186—usually occur in the absence of spontaneous par-urition. Thus, the presence or absence of spontaneous par-urition has been a sharp dividing line between the pheno-ypes of indicated and nonindicated preterm birth.

One can also argue that the mechanisms of disease responsi-le for the phenotypes are shared within the conditions respon-ible for spontaneous preterm birth and within the conditionsesponsible for indicated preterm birth. For example, MIACith bacteria is common in preterm labor188-204 and pretermROM,93,205-211 but rare in preeclampsia and IUGR. In contrast,failure of physiologic transformation of the spiral arteries” cane observed in all of these four conditions,212,213 but is morerevalent and severe in preeclampsia and IUGR214-220 than inreterm labor and preterm PROM.212,213

Ananth and coworkers18 provided evidence in support of thisiew, but also noted that spontaneous preterm birth in the firstregnancy may be followed by an indicated preterm birth in theubsequent pregnancy and vice versa. The observations are de-ived from a population-based retrospective cohort study ofirths in Missouri in which analyses were restricted to womenho delivered their first two consecutive singleton live birthsuring the study period of 1989 to 1997.18 The key observationas that, if the first pregnancy resulted in a spontaneous or

ndicated preterm birth, then women were more likely to havehe same type of preterm birth (spontaneous or indicated) in theecond pregnancy. Indeed, women with spontaneous pretermirth before 35 weeks of gestation in the first pregnancy had anR of 3.6 (95% CI 3.2-4.0) for preterm birth before 35 weeks in

he second pregnancy. However, the risk for a medically indi-ated preterm birth was also increased (OR 2.5, 95% CI 2.1-.0).18

Similarly, women who delivered at less than 35 weeks be-ause of a medical indication in the first pregnancy were muchore likely to have an indicated preterm birth at less than 35eeks of gestation in their subsequent pregnancy (OR 10.6,5% CI 9.1-12.4). However, these patients were also at in-reased risk of having a spontaneous preterm birth (OR 1.6,5% CI 1.3-2.1), although that risk was lower than the risk ofaving an indicated preterm birth.18 Similar findings were evi-ent in pregnancies that ended at less than 32 weeks (Table 4).he greatest risk for recurrence of preterm birth was observedhen women delivered their first preterm birth at less than 28eeks of gestation. The magnitude of the risk for recurrence ofreterm birth decreased progressively as gestational age at de-

ivery of the first preterm birth increased.18

ssues on the Managementnd Prevention of a Patientith a History of Preterm Birth

revention of Recurrent Preterm Birthrogesterone Administrationrogesterone plays a central role in pregnancy. The proposed
unctions of progesterone include maintenance of myome-

ta

mdi

bvbsotvrtt7w(

clplatds3([01cc

ss2

ppw4p2neaeo

dds

ptafRswwa0t1gga

Ta

PP

PP

PP

R*

renata


rial quiescence, downregulation of gap-junction formation,nd inhibition of cervical ripening.221-223

Prevention of recurrent preterm birth by progesterone ad-inistration has been a matter of debate in the literature forecades.224-242 This section will review the results of random-

zed clinical trials and meta-analyses published recently.da Fonseca et al243 recently published a randomized, dou-

le-blind, placebo-controlled trial of vaginal progesteroneersus placebo in decreasing the rate of spontaneous pretermirth. The trial included patients with at least one previouspontaneous preterm birth, a prophylactic cervical cerclage,r a uterine malformation (n � 142). Patients were allocatedo receive either daily progesterone (100 mg) or placebo byaginal suppository from 24 to 34 weeks of gestation. Theates of preterm delivery at both less than 37 weeks and lesshan 34 weeks were lower in the progesterone group than inhe placebo group [for 37 weeks; progesterone: 13.8% (10/2) versus placebo: 28.5% (20/70); P � 0.03 and for 34eeks; progesterone: 2.8% (2/72) versus placebo: 18.6%

13/70); P � 0.002].Meis et al244 reported the results of a double-blind placebo-

ontrolled clinical trial comparing the effects of intramuscu-ar 17-� hydroxyprogesterone caproate (17-OH P) versuslacebo. Patients with a history of spontaneous preterm de-

ivery (n � 463) were enrolled at 16 to 20 weeks of gestationnd randomly assigned in a 2:1 ratio to receive weekly injec-ions of 250 mg of 17-OH P or an inert oil placebo until eitherelivery or 36 weeks of gestation. Treatment with 17-OH Pignificantly reduced the rate of preterm delivery at less than7 weeks [17-OH P 36.3% versus placebo 54.9%, RR 0.6695% CI 0.54-0.81)] and less than 35 weeks of gestation17-OH P 20.6% versus placebo 30.7%, RR 0.67 (95% CI.48-0.93)]. Moreover, neonates born to women treated with7 OH P had significantly lower rates of necrotizing entero-olitis [17 OH P 0% versus placebo 2.6%, RR could not be

able 4 Recurrence of Preterm Birth at <37, <35, and <32 Wt <37, <35, and <32 Weeks in the First Pregnancy, Respe

Preterm Birth in the FirstPregnancy

Preterm

All PretermBirths

reterm birth at <37 wksreterm birth at <37 wks 2.9 (2.8, 3.0)Spontaneous preterm birth 2.8 (2.7, 3.0)Medically indicated preterm birth 3.0 (2.8, 3.3)



eproduced, with permission, from Ananth et al.18

ORs are adjusted for maternal age (second birth), education (sprepregnancy body mass index, and lack of or late initiation of p

alculated], intraventricular hemorrhage [17-OH P 1.3% ver- m

us placebo 5.2%, RR 0.25 (95% CI 0.8-0.82)], and need forupplemental oxygen [17-OH P 14.9% versus placebo3.8%, RR 0.62 (95% CI 0.42-0.92)].Of interest, the effect of 17-OH P in preventing recurrent

reterm delivery was demonstrated only in patients whoserevious preterm delivery had occurred between 20 and 33.9eeks of gestation.245 Moreover, it has been estimated that.7 women would need to be treated to prevent one recurrentreterm delivery among patients who had delivered between0 to 27.9 weeks of gestation in a previous pregnancy. Theumber needed to treat is similar for women who had deliv-red at 28 to 33.9 weeks. Of note, 17-OH P was not associ-ted with a reduction in the rate of recurrent preterm deliv-ries in patients whose previous preterm delivery hadccurred between 34 to 36.9 weeks of gestation.The efficacy of 17-OH P in singleton gestations was not

emonstrated in twin gestations.246 In this trial, no significantifferences were found between the groups in the rates ofpontaneous or indicated preterm birth.

In a 2006 Cochrane review, Dodd and coworkers247 re-orted the results of 6 randomized trials including 988 pa-ients randomized to receive either 17-OH P or placebo. Thedministration of 17-OH P was associated with reduced risksor preterm delivery before 37 weeks of gestation (6 studies,R 0.65, 95% CI 0.54-0.79) and before 34 weeks gestation (1tudy, RR 0.15, 95% CI 0.04-0.64). Moreover, treatmentith progesterone was associated with lower risks for birtheight below 2500 g (4 studies, RR 0.63, 95% CI 0.49-0.81)

nd intraventricular hemorrhage (1 study, RR 0.63, 95% CI.08-0.82). No difference in perinatal death was found be-ween treatment groups (5 studies, RR 0.66, 95% CI 0.37-.19). There were no interactions between the dose of pro-esterone (�500 mg versus �500 mg 17-OH P weekly) orestational age at commencing progesterone administrationnd the reported outcomes (ie, preterm delivery, neonatal

and Subtypes in Second Pregnancy Based on Preterm Birthy: Missouri, 1989 to 1997*

in Second Pregnancy, Adjusted OR (95% CI)

Spontaneous PretermBirths

Medically PretermIndicated Births

2.7 (2.5, 2.9) 3.3 (3.1, 3.6)3.2 (3.1, 3.4) 1.7 (1.5, 1.9)1.0 (0.9, 1.2) 7.7 (7.0, 8.5)

3.1 (2.8, 3.4) 4.8 (4.3, 5.4)3.6 (3.2, 4.0) 2.5 (2.1, 3.0)1.6 (1.3, 2.1) 10.6 (9.1, 12.4)

4.1 (3.4, 4.9) 6.5 (5.2, 8.0)4.6 (3.7, 5.6) 4.3 (3.1, 5.8)2.7 (1.8, 4.1) 11.3 (8.4, 15.1)

birth), marital status, race/ethnicity, smoking and alcohol use,l care and interpregnancy interval.

eeksctivel

Birth

econd

orbidity, and mortality). These results were in accord with

aac

cptrtp0po0w

tbbdeaw�t

twvcboIpdptHtr

optf

TTphtoewmcb

TToasf

Tppauircetfh

STbr“cIert

lp


previous meta-analysis by this group.248 Additionally, theuthors stated that there is currently insufficient informationoncerning the safety of progesterone supplementation.

Sanchez-Ramos et al249 reported another meta-analysis in-luding 10 randomized clinical trails and a total of 1339atients; 8 trials used 17-OH P and 2 used other progesta-ional agents for the prevention of recurrent preterm birth orecurrent abortion. Patients who were treated with progesta-ional agents had a reduced risk of preterm delivery com-ared to patients in the placebo group (OR 0.45, 95% CI.25-0.80). The number needed to treat to prevent a singlereterm delivery was 10 (95% CI 6-24). A similar effect wasbserved when only trials using 17-OH P were included (OR.45, 95% CI 0.25-0.80), and the number needed to treatas 8 (95% CI 5-19).Odibo et al250 performed a cost-effectiveness analysis of

he treatment with 17-OH P for the prevention of pretermirth. The cost savings per quality-adjusted life year gainedy using 17-OH P was $3090 in women with a prior pretermelivery at �32 weeks and $2963 in women who had deliv-red at 32 to 37 weeks of gestation. Moreover, the cost perdditional preterm delivery avoided with the use of 17-OH Pas $35,319 in women with a previous preterm delivery at32 weeks and $36,093 in women who had delivered at 32

o 37 weeks of gestation.In a recent meeting of the Prematurity Interest Group of

he Society for Maternal–Fetal Medicine, da Fonseca and co-orkers reported the results of a randomized clinical trial ofaginal progesterone administration to women with a shortervix (�15 mm). A 40% reduction in the rate of pretermirth by was found in women treated with vaginal progester-ne (da Fonseca E, Nicolaides K, personal communication).n contrast, the largest randomized clinical trial of vaginalrogesterone in women with a history of previous pretermelivery did not demonstrate a beneficial effect of vaginalrogesterone (Lewis D and coworkers, personal communica-ion). The FDA has raised questions about a safety signal.251

owever, this concern was not identified in the trial inwins.246 A review of embryo toxicity in animals has beenecently published.252

In summary, it seems that the administration of progester-ne may be an effective intervention to reduce the rate ofreterm delivery in a subset of women with a previous pre-erm delivery. Women with a short cervix may also benefitrom this intervention.

reatment of Bacterial Vaginosisreatment of patients with bacterial vaginosis and a history ofreterm birth is controversial. Whereas some investigatorsave argued strongly in favor of treatment,253 others believehat this intervention is not justifiable.254-258 Controversyver the choice of antibiotic also exists, with evidence thatarly treatment with clindamycin is preferable to treatmentith metronidazole.259-262 There is no evidence that treat-ent of patients with a previous preterm delivery with inter-

onceptional antibiotics will prevent a subsequent preterm
irth.263,264 a
he Use of Cerclagehe clinical value of cervical cerclage has been subjectf many observational and randomized clinical tri-ls,14,131,133,134,138,265-286 and the studies have been the topic ofeveral systematic reviews.287-289 The evidence suggests theollowing conclusions:

1) Cervical cerclage in women with a sonographic shortcervix (15 mm or less) and a low risk for preterm de-livery (by history) does not reduce the rate of sponta-neous preterm birth.286

2) The effectiveness of cervical cerclage in women with asonographic short cervix and a high risk (by history)for the prevention of preterm delivery is controver-sial.14,136,277,278,290

3) The role of prophylactic cerclage in high-risk patientswithout a sonographic short cervix for the preventionof preterm delivery/midtrimester abortion (by history)is unclear.269-271,278,285 Whereas the largest trial con-ducted before the introduction of ultrasound evalua-tion of the cervix suggested a modest beneficial ef-fect,271 other trials269,270 and systematic reviews120

before the use of ultrasound have indicated that theevidence of effectiveness is either weak or nonexistent.

4) In patients at risk for preterm delivery, serial sono-graphic examination of the cervix followed by cerclagein those who shortened the cervix is a reasonable alter-native to prophylactic placement of a cerclage based onuncontrolled studies.131,282,291

his evidence indicates that only patients with the clinicalresentation of “acute cervical insufficiency” and those with aregnancy history consistent with “cervical insufficiency”nd progressive shortening of the cervix demonstrated withltrasound may benefit from cerclage placement. However,

mportant to consider is that each conclusion is based on theesults of only one randomized clinical trial.278,283 Sakai andoworkers demonstrated that the inflammatory status of thendocervix may be an additional criterion to distinguishhose patients who would benefit from cerclage placementrom those in whom this intervention may be ineffective orarmful.292

ummaryhe evaluation of a patient with a previous preterm birthegins with an examination of the obstetrical circumstancesesponsible for this complication. If the preterm birth wasindicated,” then the risk of recurrence is related to the spe-ific condition, such as preterm preeclampsia, preterm severeUGR, placenta previa, etc. The reader is referred to the rel-vant articles in this volume of the Seminars for details aboutecurrence rate, monitoring of the index pregnancy, and in-erventions.

If the previous preterm birth was the result of spontaneousabor (with intact or ruptured membranes), the informationrovided in this article can be used to counsel the patient
bout the likelihood of recurrence. In general, most patients

wtgotCacat

teAri

rtscwpbabpcgpiear

ATgD

R


ith a previous spontaneous preterm birth will deliver aterm in a subsequent pregnancy.2-22 However, the earlier theestational age of the preterm birth, the higher the likelihoodf recurrence. It is important to be aware that recurrent pre-erm births tend to occur at the same gestational age.10,12,18

ounseling should ideally be conducted before conception,nd efforts should be made to identify potentially treatableauses such as a Mullerian duct abnormalities. However, thettributable risk of these conditions for preterm birth is ex-remely low.

The estimates of risk of recurrence for spontaneous pre-erm birth can be improved by performing a sonographicxamination of the uterine cervix and a fetal fibronectin test.long cervix and a negative fetal fibronectin test reduce the

isk just as a short cervix and a positive fetal fibronectin testncrease the risk.9,122

No intervention has been proven effective in reducing theate of preterm birth in patients with a positive fetal fibronec-in test.293,294 Similarly, the management of patients with ahort cervix remains controversial. Evidence suggests thatervical cerclage does not prevent preterm birth in womenith a short cervix who do not have a history of previousreterm birth.286 Similarly, a prophylactic cerclage has noteen effective in reducing the rate of preterm birth in patientst risk for midtrimester abortion or spontaneous pretermirth.138,277,282 In contrast, one randomized clinical trial ofatients with risk factors or symptoms of cervical insuffi-iency and a shortened cervix (�25 mm before 27 weeks ofestation) in the index pregnancy found a benefit of “thera-eutic cerclage.”278 Although further studies are required to

dentify effective interventions and the patients who will ben-fit from them, monitoring cervical length with ultrasoundnd offering cerclage based on individual risk assessment is aeasonable management strategy.

cknowledgmentshis research was supported by the Intramural Research Pro-ram of the National Institute of Child Health and Humanevelopment, NIH, DHHS.

eferences1. Lawn JE, Cousens S, Zupan J: 4 million neonatal deaths: When?

Where? Why? Lancet 365:891-900, 20052. Carr-Hill RA, Hall MH: The repetition of spontaneous preterm labour.

Br J Obstet Gynaecol 92:921-928, 19853. Ekwo EE, Gosselink CA, Moawad A: Previous pregnancy outcomes

and subsequent risk of preterm rupture of amniotic sac membranes.Br J Obstet Gynaecol 100:536-541, 1993

4. Kristensen J, Langhoff-Roos J, Kristensen FB: Implications of idio-pathic preterm delivery for previous and subsequent pregnancies.Obstet Gynecol 86:800-804, 1995

5. Gibb DM, Salaria DA: Transabdominal cervicoisthmic cerclage in themanagement of recurrent second trimester miscarriage and pretermdelivery. Br J Obstet Gynaecol 102:802-806, 1995

6. Guinn DA, Goldenberg RL, Hauth JC, et al: Risk factors for the devel-opment of preterm premature rupture of the membranes after arrest ofpreterm labor. Am J Obstet Gynecol 173:1310-1315, 1995

7. Kramer MS, Coates AL, Michoud MC, et al: Maternal asthma andidiopathic preterm labor. Am J Epidemiol 142:1078-1088, 1995

8. Ekwo E, Moawad A: The risk for recurrence of premature births to

African-American and white women. J Assoc Acad Minor Phys9:16-21, 1998

9. Iams JD, Goldenberg RL, Mercer BM, et al: The Preterm PredictionStudy: recurrence risk of spontaneous preterm birth. National Insti-tute of Child Health and Human Development Maternal-Fetal Medi-cine Units Network. Am J Obstet Gynecol 178:1035-1040, 1998

10. Mercer BM, Goldenberg RL, Moawad AH, et al: The preterm predic-tion study: effect of gestational age and cause of preterm birth onsubsequent obstetric outcome. National Institute of Child Health andHuman Development Maternal-Fetal Medicine Units Network. Am JObstet Gynecol 181:1216-1221, 1999

11. Adams MM, Elam-Evans LD, Wilson HG, et al: Rates of and factorsassociated with recurrence of preterm delivery. JAMA 283:1591-1596, 2000

12. Bloom SL, Yost NP, McIntire DD, et al: Recurrence of preterm birth insingleton and twin pregnancies. Obstet Gynecol 98:379-385, 2001

13. Koike T, Minakami H, Izumi A, et al: Recurrence risk of preterm birthdue to preeclampsia. Gynecol Obstet Invest 53:22-27, 2002

14. Berghella V, Odibo AO, Tolosa JE: Cerclage for prevention of pretermbirth in women with a short cervix found on transvaginal ultrasoundexamination: a randomized trial. Am J Obstet Gynecol 191:1311-1317,2004

15. Yost NP, Owen J, Berghella V, et al: Number and gestational age ofprior preterm births does not modify the predictive value of a shortcervix. Am J Obstet Gynecol 191:241-246, 2004

16. Tekesin I, Eberhart LH, Schaefer V, et al: Evaluation and validation ofa new risk score (CLEOPATRA score) to predict the probability ofpremature delivery for patients with threatened preterm labor. Ultra-sound Obstet Gynecol 26:699-706, 2005

17. Durnwald CP, Walker H, Lundy JC, et al: Rates of recurrent pretermbirth by obstetrical history and cervical length. Am J Obstet Gynecol193:1170-1174, 2005

18. Ananth CV, Getahun D, Peltier MR, et al: Recurrence of spontaneousversus medically indicated preterm birth. Am J Obstet Gynecol 195:643-650, 2006

19. Ananth CV, Vintzileos AM: Epidemiology of preterm birth and itsclinical subtypes. J Matern Fetal Neonatal Med 19:773-782, 2006

20. Meis PJ, Klebanoff M, Dombrowski MP, et al: Does progesteronetreatment influence risk factors for recurrent preterm delivery? ObstetGynecol 106:557-561, 2005

21. Mercer B, Milluzzi C, Collin M: Periviable birth at 20 to 26 weeks ofgestation: proximate causes, previous obstetric history and recurrencerisk. Am J Obstet Gynecol 193:1175-1180, 2005

22. Mercer BM, Macpherson CA, Goldenberg RL, et al: Are women withrecurrent spontaneous preterm births different from those withoutsuch history? Am J Obstet Gynecol 194:1176-1184, 2006

23. Romero R, Espinoza J, Kusanovic JP, et al: The preterm parturitionsyndrome. Br J Obstet Gynaecol 113:17-42, 2006 (suppl 3)

24. Romero R, Espinoza J, Mazor M, et al: The preterm parturition syn-drome, in Critchely H, Bennett P, Thornton S (eds): Preterm Birth.London, RCOG Press, 2004, pp 28-60

25. Joseph KS, Kramer MS, Marcoux S, et al: Determinants of pretermbirth rates in Canada from 1981 through 1983 and from 1992through 1994. N Engl J Med 339:1434-1439, 1998

26. Preterm singleton births: United States, 1989-1996. MMWR MorbMortal Wkly Rep 48:185-189, 1999

27. Vintzileos AM, Ananth CV, Smulian JC, et al: The impact of prenatalcare in the United States on preterm births in the presence and ab-sence of antenatal high-risk conditions. Am J Obstet Gynecol 187:1254-1257, 2002

28. Little RE, Gladen BC, Birmingham K, et al: Preterm birth rates in AvonCounty, England, and urban Ukraine. Eur J Obstet Gynecol ReprodBiol 113:154-159, 2004

29. Morken NH, Kallen K, Hagberg H, et al: Preterm birth in Sweden1973-2001: rate, subgroups, and effect of changing patterns in mul-tiple births, maternal age, and smoking. Acta Obstet Gynecol Scand84:558-565, 2005

30. Davidoff MJ, Dias T, Damus K, et al: Changes in the gestational age


distribution among U.S. singleton births: impact on rates of latepreterm birth, 1992 to 2002. Semin Perinatol 30:8-15, 2006

31. Papiernik E: Is the high rate of preterm birth in the United Stateslinked to previous induced abortions? Pediatrics 118:795-796, 2006

32. Fuchs K, Wapner R: Elective cesarean section and induction and theirimpact on late preterm births. Clin Perinatol 33:793-801, 2006

33. Hamilton BE, Minino AM, Martin JA, et al: Annual summary of vitalstatistics: 2005. Pediatrics 119:345-360, 2007

34. Collins JW Jr, David RJ, Simon DM, et al: Preterm birth among AfricanAmerican and white women with a lifelong residence in high-incomeChicago neighborhoods: an exploratory study. Ethn Dis 17:113-117,2007

35. Gazaway P, Mullins CL: Prevention of preterm labor and prematurerupture of the membranes. Clin Obstet Gynecol 29:835-849, 1986

36. Nwaesei CG, Young DC, Byrne JM, et al: Preterm birth at 23 to 26weeks’ gestation: is active obstetric management justified? Am J ObstetGynecol 157:890-897, 1987

37. Mueller-Heubach E, Guzick DS: Evaluation of risk scoring in a pre-term birth prevention study of indigent patients. Am J Obstet Gynecol160:829-835, 1989

38. Roberts WE, Morrison JC, Hamer C, et al: The incidence of pretermlabor and specific risk factors. Obstet Gynecol 76:85S-89S, 1990

39. Yan JS, Yin CS: No decline in preterm birth rate over three decades. IntJ Gynaecol Obstet 34:1-5, 1991

40. Tucker JM, Goldenberg RL, Davis RO, et al: Etiologies of preterm birthin an indigent population: is prevention a logical expectation? ObstetGynecol 77:343-347, 1991

41. Kramer MS, McLean FH, Eason EL, et al: Maternal nutrition andspontaneous preterm birth. Am J Epidemiol 136:574-583, 1992

42. Multicenter randomized, controlled trial of a preterm birth preventionprogram. Collaborative Group on Preterm Birth Prevention. Am JObstet Gynecol 169:352-366, 1993

43. Hobel CJ, Ross MG, Bemis RL, et al: The West Los Angeles PretermBirth Prevention Project. I. Program impact on high-risk women. Am JObstet Gynecol 170:54-62, 1994

44. Harbert GM Jr: Efforts to reduce low birth weight and preterm births:a statewide analysis (Virginia). Am J Obstet Gynecol 171:329-338,1994

45. Lubarsky SL, Schiff E, Friedman SA, et al: Obstetric characteristicsamong nulliparas under age 15. Obstet Gynecol 84:365-68, 1994

46. Kristensen J, Langhoff-Roos J, Kristensen FB: Idiopathic preterm de-liveries in Denmark. Obstet Gynecol 85:549-552, 1995

47. Gardner MO, Goldenberg RL, Cliver SP, et al: The origin and outcomeof preterm twin pregnancies. Obstet Gynecol 85:553-557, 1995

48. Graf RA, Perez-Woods R: Trends in preterm labor. J Perinatol 12:51-58, 1992

49. Berkowitz GS, Papiernik E: Epidemiology of preterm birth. EpidemiolRev 15:414-443, 1993

50. Heaman MI, Sprague AE, Stewart PJ: Reducing the preterm birth rate:a population health strategy. J Obstet Gynecol Neonatal Nurs 30:20-29, 2001

51. Mauldin JG, Newman RB: Preterm birth risk assessment. Semin Peri-natol 25:215-222, 2001

52. Kramer MS: The epidemiology of adverse pregnancy outcomes: anoverview. J Nutr 133:1592S-1596S, 2003

53. Wen SW, Smith G, Yang Q, et al: Epidemiology of preterm birth andneonatal outcome. Semin Fetal Neonatal Med 9:429-435, 2004

54. Bibby E, Stewart A: The epidemiology of preterm birth. Neuro Endo-crinol Lett 25:43-47, 2004 (suppl 1)

55. Higgins RD, Delivoria-Papadopoulos M, Raju TN: Executive summaryof the workshop on the border of viability. Pediatrics 115:1392-1396,2005

56. Raju TN: Epidemiology of late preterm (near-term) births. Clin Peri-natol 33:751-763, 2006

57. Steer P: The epidemiology of preterm labor: a global perspective. JPerinat Med 33:273-276, 2005

58. Mercer BM: Preterm premature rupture of the membranes. Obstet
Gynecol 101:178-193, 2003
59. Rydhstroem H: Gestational duration in the pregnancy after a pretermtwin delivery. Am J Obstet Gynecol 178:136-139, 1998

60. Fanaroff AA, Stoll BJ, Wright LL, et al: Trends in neonatal morbidityand mortality for very low birthweight infants. Am J Obstet Gynecol196:147-148, 2007

61. Cnattingius S, Granath F, Petersson G, et al: The influence of gesta-tional age and smoking habits on the risk of subsequent pretermdeliveries. N Engl J Med 341:943-948, 1999

62. Menard MK, Newman RB, Keenan A, et al: Prognostic significance ofprior preterm twin delivery on subsequent singleton pregnancy. Am JObstet Gynecol 174:1429-1432, 1996

63. Krymko H, Bashiri A, Smolin A, et al: Risk factors for recurrent pre-term delivery. Eur J Obstet Gynecol Reprod Biol 113:160-163, 2004

64. Kitska ZA, Palomar L, Lee KA, et al: Racial disparity in the frequencyof recurrence of preterm birth. Am J Obstet Gynecol 196:131, 2007

65. Romero R, Mazor M, Gomez R, et al: Cervix, incompetence and pre-mature labor. The Fetus 3:1-10, 1993

66. Iams JD, Johnson FF, Sonek J, et al: Cervical competence as a contin-uum: a study of ultrasonographic cervical length and obstetric perfor-mance. Am J Obstet Gynecol 172:1097-1103, 1995

67. Rydhstrom H: Twin pregnancy and the effects of prophylactic leave ofabsence on pregnancy duration and birth weight. Acta Obstet GynecolScand 67:81-84, 1988

68. Goldenberg RL, Mercer BM, Meis PJ, et al: The preterm predictionstudy: fetal fibronectin testing and spontaneous preterm birth.NICHD Maternal Fetal Medicine Units Network. Obstet Gynecol 87:643-648, 1996

69. Keirse MJ, Ohlsson A, Treffers PE, et al: Prelabour rupture of themembranes preterm, in Chalmers I, Enkin M, Keirse MJ (eds): Effec-tive Care in Pregnancy and Childbirth. Oxford, Oxford UniversityPress, 1989, p 666

70. Sachs M, Baker TH: Spontaneous premature rupture of the mem-branes. Am J Obstet Gynecol 97:888, 1967

71. Gunn GC, Mishell DR Jr, Morton DG: Premature rupture of the fetalmembranes: a review. Am J Obstet Gynecol 106:469-483, 1970

72. Christensen KK, Christensen P, Ingemarsson I, et al: A study of com-plications in preterm deliveries after prolonged premature rupture ofthe membranes. Obstet Gynecol 48:670-677, 1976

73. Johnson JW, Daikoku NH, Niebyl JR, et al: Premature rupture of themembranes and prolonged latency. Obstet Gynecol 57:547-556, 1981

74. Daikoku NH, Kaltreider DF, Khouzami VA, et al: Premature ruptureof membranes and spontaneous preterm labor: maternal endometritisrisks. Obstet Gynecol 59:13-20, 1982

75. Gibbs RS, Blanco JD: Premature rupture of the membranes. ObstetGynecol 60:671-679, 1982

76. Shubert PJ, Diss E, Iams JD: Etiology of preterm premature rupture ofmembranes. Obstet Gynecol Clin North Am 19:251-263, 1992

77. Lebherz TB, Hellman LP, Madding R, et al: Double-blind study ofpremature rupture of the membranes: A report of 1,896 cases. Am JObstet Gynecol 87:218-225, 1963

78. Fayez JA, Hasan AA, Jonas HS, et al: Management of premature rup-ture of the membranes. Obstet Gynecol 52:17-21, 1978

79. Naeye RL: Factors that predispose to premature rupture of the fetalmembranes. Obstet Gynecol 60:93-98, 1982

80. Asrat T, Lewis DF, Garite TJ, et al: Rate of recurrence of pretermpremature rupture of membranes in consecutive pregnancies. Am JObstet Gynecol 165:1111-1115, 1991

81. Ekwo EE, Gosselink CA, Moawad A: Unfavorable outcome in penul-timate pregnancy and premature rupture of membranes in successivepregnancy. Obstet Gynecol 80:166-172, 1992

82. Mercer BM, Goldenberg RL, Meis PJ, et al: The Preterm PredictionStudy: prediction of preterm premature rupture of membranesthrough clinical findings and ancillary testing. The National Instituteof Child Health and Human Development Maternal-Fetal MedicineUnits Network. Am J Obstet Gynecol 183:738-745, 2000

83. Romero R, Tromp G: High-dimensional biology in obstetrics andgynecology: functional genomics in microarray studies. Am J ObstetGynecol 195:360-363, 2006

84. Rizzo G, Capponi A, Vlachopoulou A, et al: Ultrasonographic assess-

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1


ment of the uterine cervix and interleukin-8 concentrations in cervicalsecretions predict intrauterine infection in patients with preterm laborand intact membranes. Ultrasound Obstet Gynecol 12:86-92, 1998

85. Gomez R, Romero R, Nien JK, et al: A short cervix in women withpreterm labor and intact membranes: a risk factor for microbial inva-sion of the amniotic cavity. Am J Obstet Gynecol 192:678-689, 2005

86. Eggert-Kruse W, Botz I, Pohl S, et al: Antimicrobial activity of humancervical mucus. Hum Reprod 15:778-784, 2000

87. Hein M, Helmig RB, Schonheyder HC, et al: An in vitro study ofantibacterial properties of the cervical mucus plug in pregnancy. Am JObstet Gynecol 185:586-592, 2001

88. Hein M, Valore EV, Helmig RB, et al: Antimicrobial factors in thecervical mucus plug. Am J Obstet Gynecol 187:137-144, 2002

89. Evaldson G, Malmborg AS, Nord CE, et al: Bacteroides fragilis, Strep-tococcus intermedius and group B streptococci in ascending infectionof pregnancy. An animal experimental study. Gynecol Obstet Invest15:230-241, 1983

90. Cole AM: Innate host defense of human vaginal and cervical mucosae.Curr Top Microbiol Immunol 306:199-230, 2006

91. Svinarich DM, Wolf NA, Gomez R, et al: Detection of human defensin5 in reproductive tissues. Am J Obstet Gynecol 176:470-475, 1997

92. Romero R, Gomez R, Araneda H, et al: Cervical mucus inhibits microbialgrowth: a host defense mechanism to prevent ascending infection in preg-nant and non-pregnant women. Am J Obstet Gynecol A57, 1993

93. Espinoza J, Chaiworapongsa T, Romero R, et al: Antimicrobial peptides inamniotic fluid: defensins, calprotectin and bacterial/permeability-increas-ing protein in patients with microbial invasion of the amniotic cavity,intra-amniotic inflammation, preterm labor and premature rupture ofmembranes. J Matern Fetal Neonatal Med 13:2-21, 2003

94. Yoon BH, Romero R, Moon JB, et al: The frequency and clinical sig-nificance of intra-amniotic inflammation in patients with a positivecervical fetal fibronectin. Am J Obstet Gynecol 185:1137-1142, 2001

95. Leitich H, Kaider A: Fetal fibronectin–how useful is it in the prediction ofpreterm birth? Br J Obstet Gynaecol 110:66-70, 2003 (suppl 20)

96. Chang TC, Chew TS, Pang M, et al: Cervicovaginal foetal fibronectinin the prediction of preterm labour in a low-risk population. AnnAcad Med Singapore 26:776-780, 1997

97. Goldenberg RL, Mercer BM, Iams JD, et al: The preterm predictionstudy: patterns of cervicovaginal fetal fibronectin as predictors ofspontaneous preterm delivery. National Institute of Child Health andHuman Development Maternal-Fetal Medicine Units Network. Am JObstet Gynecol 177:8-12, 1997

98. Malak TM, Sizmur F, Bell SC, et al: Fetal fibronectin in cervicovaginalsecretions as a predictor of preterm birth. Br J Obstet Gynaecol 103:648-653, 1996

99. Parker J, Bell R, Brennecke S: Fetal fibronectin in the cervicovaginalfluid of women with threatened preterm labour as a predictor ofdelivery before 34 weeks’ gestation. Aust N Z J Obstet Gynaecol 35:257-261, 1995

00. Burrus DR, Ernest JM, Veille JC: Fetal fibronectin, interleukin-6, andC-reactive protein are useful in establishing prognostic subcategories ofidiopathic preterm labor. Am J Obstet Gynecol 173:1258-1262, 1995

01. Chuileannain FN, Bell R, Brennecke S: Cervicovaginal fetal fibronec-tin testing in threatened preterm labour: translating research findingsinto clinical practice. Aust N Z J Obstet Gynaecol 38:399-402, 1998

02. Goffeng AR, Holst E, Milsom I, et al: Fetal fibronectin and microor-ganisms in vaginal fluid of women with complicated pregnancies.Acta Obstet Gynecol Scand 76:521-527, 1997

03. Lopez RL, Francis JA, Garite TJ, et al: Fetal fibronectin detection as apredictor of preterm birth in actual clinical practice. Am J ObstetGynecol 182:1103-1106, 2000

04. Nageotte MP, Casal D, Senyei AE: Fetal fibronectin in patients atincreased risk for premature birth. Am J Obstet Gynecol 170:20-25,1994

05. Oliveira T, de Souza E, Mariani-Neto C, et al: Fetal fibronectin as apredictor of preterm delivery in twin gestations. Int J Gynaecol Obstet62:135-139, 1998

06. Morrison JC, ?? RW III, Botti JJ, et al: Prediction of spontaneous

preterm birth by fetal fibronectin and uterine activity. Obstet Gynecol87:649-655, 1996

07. Wennerholm UB, Holm B, Mattsby-Baltzer I, et al: Fetal fibronectin,endotoxin, bacterial vaginosis and cervical length as predictors ofpreterm birth and neonatal morbidity in twin pregnancies. Br J ObstetGynaecol 104:1398-1404, 1997

08. Tolino A, Ronsini S, Zullo F, et al: Fetal fibronectin as a screening testfor premature delivery in multiple pregnancies. Int J Gynaecol Obstet52:3-7, 1996

09. Goldenberg RL, Iams JD, Miodovnik M, et al: The preterm predictionstudy: risk factors in twin gestations. National Institute of ChildHealth and Human Development Maternal-Fetal Medicine Units Net-work. Am J Obstet Gynecol 175:1047-1053, 1996

10. Sciscione A, Hoffman MK, DeLuca S, et al: Fetal fibronectin as apredictor of vaginal birth in nulliparas undergoing preinduction cer-vical ripening. Obstet Gynecol 106:980-985, 2005

11. Imai M, Tani A, Saito M, et al: Significance of fetal fibronectin andcytokine measurement in the cervicovaginal secretions of women atterm in predicting term labor and post-term pregnancy. Eur J ObstetGynecol Reprod Biol 97:53-58, 2001

12. Mijovic JE, Demianczuk N, Olson DM, et al: Prostaglandin endoper-oxide H synthase mRNA expression in the fetal membranes correlateswith fetal fibronectin concentration in the cervico-vaginal fluids atterm: evidence of enzyme induction before the onset of labour. Br JObstet Gynaecol 107:267-273, 2000

13. Rozenberg P, Goffinet F, Hessabi M: Comparison of the Bishop score,ultrasonographically measured cervical length, and fetal fibronectinassay in predicting time until delivery and type of delivery at term.Am J Obstet Gynecol 182:108-113, 2000

14. Kiss H, Ahner R, Hohlagschwandtner M, et al: Fetal fibronectin as apredictor of term labor: a literature review. Acta Obstet Gynecol Scand79:3-7, 2000

15. Garite TJ, Casal D, Garcia-Alonso A, et al: Fetal fibronectin: a new toolfor the prediction of successful induction of labor. Am J Obstet Gy-necol 175:1516-1521, 1996

16. Ahner R, Egarter C, Kiss H, et al: Fetal fibronectin as a selectioncriterion for induction of term labor. Am J Obstet Gynecol 173:1513-1517, 1995

17. Ahner R, Kiss H, Egarter C, et al: Fetal fibronectin as a marker topredict the onset of term labor and delivery. Am J Obstet Gynecol172:134-137, 1995

18. Lockwood CJ, Senyei AE, Dische MR, et al: Fetal fibronectin in cervi-cal and vaginal secretions as a predictor of preterm delivery. N EnglJ Med 325:669-674, 1991

19. Brady K, Duff P, Yancey MK: Plasma fibronectin concentrations dur-ing normal term labor. Obstet Gynecol 75:619-621, 1990

20. Grant A: Cervical cerclage to prolong pregnancy, in Chalmers I, EnkinM, Keirse MJNC (eds): Effective Care in Pregnancy and Childbirth(vol. 1). New York, NY, Oxford University Press, 1989, pp 633-646

21. Andersen HF, Nugent CE, Wanty SD, et al: Prediction of risk forpreterm delivery by ultrasonographic measurement of cervical length.Am J Obstet Gynecol 163:859-867, 1990

22. Iams JD, Goldenberg RL, Meis PJ, et al: The length of the cervix andthe risk of spontaneous premature delivery. National Institute ofChild Health and Human Development Maternal Fetal Medicine UnitNetwork. N Engl J Med 334:567-572, 1996

23. Heath VC, Southall TR, Souka AP, et al: Cervical length at 23 weeks ofgestation: prediction of spontaneous preterm delivery. UltrasoundObstet Gynecol 12:312-317, 1998

24. Taipale P, Hiilesmaa V: Sonographic measurement of uterine cervix at18-22 weeks’ gestation and the risk of preterm delivery. Obstet Gy-necol 92:902-907, 1998

25. Hassan SS, Romero R, Berry SM, et al: Patients with an ultrasono-graphic cervical length �15 mm have nearly a 50% risk of earlyspontaneous preterm delivery. Am J Obstet Gynecol 182:1458-1467,2000

26. Kushnir O, Vigil DA, Izquierdo L, et al: Vaginal ultrasonographicassessment of cervical length changes during normal pregnancy. Am J
Obstet Gynecol 162:991-993, 1990

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1


27. Okitsu O, Mimura T, Nakayama T, et al: Early prediction of pretermdelivery by transvaginal ultrasonography. Ultrasound Obstet Gynecol2:402-409, 1992

28. Guzman ER, Rosenberg JC, Houlihan C, et al: A new method usingvaginal ultrasound and transfundal pressure to evaluate the asymp-tomatic incompetent cervix. Obstet Gynecol 83:248-252, 1994

29. Guzman ER, Pisatowski DM, Vintzileos AM, et al: A comparison ofultrasonographically detected cervical changes in response to trans-fundal pressure, coughing, and standing in predicting cervical incom-petence. Am J Obstet Gynecol 177:660-665, 1997

30. Guzman ER, Vintzileos AM, McLean DA, et al: The natural history ofa positive response to transfundal pressure in women at risk for cer-vical incompetence. Am J Obstet Gynecol 176:634-638, 1997

31. Guzman ER, Forster JK, Vintzileos AM, et al: Pregnancy outcomes inwomen treated with elective versus ultrasound-indicated cervical cer-clage. Ultrasound Obstet Gynecol 12:323-327, 1998

32. Guzman ER, Mellon C, Vintzileos AM, et al: Longitudinal assessmentof endocervical canal length between 15 and 24 weeks’ gestation inwomen at risk for pregnancy loss or preterm birth. Obstet Gynecol92:31-37, 1998

33. Berghella V, Daly SF, Tolosa JE, et al: Prediction of preterm deliverywith transvaginal ultrasonography of the cervix in patients with high-risk pregnancies: does cerclage prevent prematurity? Am J ObstetGynecol 181:809-815, 1999

34. Hassan SS, Romero R, Maymon E, et al: Does cervical cerclage preventpreterm delivery in patients with a short cervix? Am J Obstet Gynecol184:1325-1329, 2001

35. Macdonald R, Smith P, Vyas S: Cervical incompetence: the use oftransvaginal sonography to provide an objective diagnosis. Ultra-sound Obstet Gynecol 18:211-216, 2001

36. Rust OA, Atlas RO, Reed J, et al: Revisiting the short cervix detected bytransvaginal ultrasound in the second trimester: why cerclage therapymay not help. Am J Obstet Gynecol 185:1098-1105, 2001

37. To MS, Skentou C, Liao AW, et al: Cervical length and funneling at 23weeks of gestation in the prediction of spontaneous early pretermdelivery. Ultrasound Obstet Gynecol 18:200-203, 2001

38. Berghella V, Haas S, Chervoneva I, et al: Patients with prior second-trimester loss: prophylactic cerclage or serial transvaginal sonograms?Am J Obstet Gynecol 187:747-751, 2002

39. Williams M, Iams JD: Cervical length measurement and cervical cerclageto prevent preterm birth. Clin Obstet Gynecol 47:775-783, 2004

40. Althuisius SM, Dekker GA: A five century evolution of cervical incom-petence as a clinical entity. Curr Pharm Des 11:687-697, 2005

41. Ramin KD, Ogburn PL Jr, Mulholland TA, et al: Ultrasonographicassessment of cervical length in triplet pregnancies. Am J Obstet Gy-necol 180:1442-1445, 1999

42. Souka AP, Heath V, Flint S, et al: Cervical length at 23 weeks in twinsin predicting spontaneous preterm delivery. Obstet Gynecol 94:450-454, 1999

43. Guzman ER, Walters C, O’Reilly-Green C, et al: Use of cervical ultra-sonography in prediction of spontaneous preterm birth in twin ges-tations. Am J Obstet Gynecol 183:1103-1107, 2000

44. Guzman ER, Walters C, O’Reilly-Green C, et al: Use of cervical ultra-sonography in prediction of spontaneous preterm birth in triplet ges-tations. Am J Obstet Gynecol 183:1108-1113, 2000

45. To MS, Skentou C, Cicero S, et al: Cervical length at 23 weeks intriplets: prediction of spontaneous preterm delivery. Ultrasound Ob-stet Gynecol 16:515-518, 2000

46. Yang JH, Kuhlman K, Daly S, et al: Prediction of preterm birth bysecond trimester cervical sonography in twin pregnancies. UltrasoundObstet Gynecol 15:288-291, 2000

47. Maymon R, Herman A, Jauniaux E, et al: Transvaginal sonographicassessment of cervical length changes during triplet gestation. HumReprod 16:956-960, 2001

48. Skentou C, Souka AP, To MS, et al: Prediction of preterm delivery intwins by cervical assessment at 23 weeks. Ultrasound Obstet Gynecol17:7-10, 2001

49. Vayssiere C, Favre R, Audibert F, et al: Cervical length and funneling
at 22 and 27 weeks to predict spontaneous birth before 32 weeks in
twin pregnancies: a French prospective multicenter study. Am JObstet Gynecol 187:1596-1604, 2002

50. Parikh MN, Mehta AC: Internal cervical os during the second half ofpregnancy. J Obstet Gynaecol Br Emp 68:818-821, 1961

51. Guzman ER, Mellon R, Vintzileos AM, et al: Relationship betweenendocervical canal length between 15-24 weeks gestation and obstet-ric history. J Matern Fetal Med 7:269-272, 1998

52. Andersen HF: Transvaginal and transabdominal ultrasonography ofthe uterine cervix during pregnancy. J Clin Ultrasound 19:77-83,1991

53. Romero R: Prenatal medicine: the child is the father of the man.Prenatal Neonatal Med 1:8-11, 1996

54. Moinian M, Andersch B: Does cervix conization increase the risk ofcomplications in subsequent pregnancies? Acta Obstet Gynecol Scand61:101-103, 1982

55. Kristensen J, Langhoff-Roos J, Wittrup M, et al: Cervical conizationand preterm delivery/low birth weight. A systematic review of theliterature. Acta Obstet Gynecol Scand 72:640-644, 1993

56. Raio L, Ghezzi F, Di Naro E, et al: Duration of pregnancy after carbondioxide laser conization of the cervix: influence of cone height. ObstetGynecol 90:978-982, 1997

57. Craig CJ: Congenital abnormalities of the uterus and foetal wastage. SAfr Med J 47:2000-2005, 1973

58. Mangan CE, Borow L, Burtnett-Rubin MM, et al: Pregnancy outcomein 98 women exposed to diethylstilbestrol in utero, their mothers, andunexposed siblings. Obstet Gynecol 59:315-319, 1982

59. Ludmir J, Landon MB, Gabbe SG, et al: Management of the diethyl-stilbestrol-exposed pregnant patient: a prospective study. Am J ObstetGynecol 157:665-669, 1987

60. Levine RU, Berkowitz KM: Conservative management and pregnancyoutcome in diethylstilbestrol-exposed women with and without grossgenital tract abnormalities. Am J Obstet Gynecol 169:1125-1129,1993

61. Mays JK, Figueroa R, Shah J, et al: Amniocentesis for selection beforerescue cerclage. Obstet Gynecol 95:652-655, 2000

62. Romero R, Gonzalez R, Sepulveda W, et al: Infection and labor. VIII.Microbial invasion of the amniotic cavity in patients with suspectedcervical incompetence: prevalence and clinical significance. Am J Ob-stet Gynecol 167:1086-1091, 1992

63. Bengtsson LP: Cervical insufficiency. Acta Obstet Gynecol Scand 47:7-35, 1968 (suppl 1)

64. Stys SJ, Clewell WH, Meschia G: Changes in cervical compliance atparturition independent of uterine activity. Am J Obstet Gynecol 130:414-418, 1978

65. Sherman AI: Hormonal therapy for control of the incompetent os ofpregnancy. Obstet Gynecol 28:198-205, 1966

66. Boggess KA, Madianos PN, Preisser JS, et al: Chronic maternal andfetal Porphyromonas gingivalis exposure during pregnancy in rabbits.Am J Obstet Gynecol 192:554-557, 2005

67. Romero R, Espinoza J, Mazor M: Can endometrial infection/inflam-mation explain implantation failure, spontaneous abortion, and pre-term birth after in vitro fertilization? Fertil Steril 82:799-804, 2004

68. Hassan S, Romero R, Hendler I, et al: A sonographic short cervix as theonly clinical manifestation of intra-amniotic infection. J Perinat Med34:13-19, 2006

69. Blackmore-Prince C, Kieke B Jr, Kugaraj KA, et al: Racial differences inthe patterns of singleton preterm delivery in the 1988 National Ma-ternal and Infant Health Survey. Matern Child Health J 3:189-197,1999

70. Simhan HN, Bodnar LM: Prepregnancy body mass index, vaginalinflammation, and the racial disparity in preterm birth. Am J Epide-miol 163:459-466, 2006

71. Ananth CV, Misra DP, Demissie K, et al: Rates of preterm deliveryamong Black women and White women in the United States over twodecades: an age-period-cohort analysis. Am J Epidemiol 154:657-665, 2001

72. State-specific changes in singleton preterm births among black andwhite women: United States, 1990 and 1997. MMWR Morb Mortal
Wkly Rep 49:837-840, 2000

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

2

2

2


73. Schieve LA, Handler A: Preterm delivery and perinatal death amongblack and white infants in a Chicago-area perinatal registry. ObstetGynecol 88:356-363, 1996

74. Blackmore CA, Savitz DA, Edwards LJ, et al: Racial differences in thepatterns of preterm delivery in central North Carolina, USA. PaediatrPerinat Epidemiol 9:281-295, 1995

75. Schulman ED: Preterm delivery among women in the South CarolinaMedicaid High Risk Channeling Project. J Health Care Poor Under-served 6:352-367, 1995

76. Savitz DA, Dole N, Herring AH, et al: Should spontaneous and med-ically indicated preterm births be separated for studying aetiology?Paediatr Perinat Epidemiol 19:97-105, 2005

77. Merlino A, Laffineuse L, Collin M, et al: Impact of weight loss betweenpregnancies on recurrent preterm birth. Am J Obstet Gynecol 195:818-821, 2006

78. Yost NP, Owen J, Berghella V, et al: Effect of coitus on recurrentpreterm birth. Obstet Gynecol 107:793-797, 2006

79. Klebanoff MA, Nugent RP, Rhoads GG: Coitus during pregnancy: is itsafe? Lancet 2:914-917, 1984

80. Berghella V, Klebanoff M, McPherson C, et al: Sexual intercourseassociation with asymptomatic bacterial vaginosis and Trichomonasvaginalis treatment in relationship to preterm birth. Am J Obstet Gy-necol 187:1277-1282, 2002

81. Kurki T, Ylikorkala O: Coitus during pregnancy is not related tobacterial vaginosis or preterm birth. Am J Obstet Gynecol 169:1130-1134, 1993

82. Read JS, Klebanoff MA: Sexual intercourse during pregnancy andpreterm delivery: effects of vaginal microorganisms. The Vaginal In-fections and Prematurity Study Group. Am J Obstet Gynecol 168:514-519, 1993

83. Brustman LE, Raptoulis M, Langer O, et al: Changes in the pattern ofuterine contractility in relationship to coitus during pregnancies atlow and high risk for preterm labor. Obstet Gynecol 73:166-168,1989

84. Wagner NN, Butler JC, Sanders JP: Prematurity and orgasmic coitusduring pregnancy: data on a small sample. Fertil Steril 27:911-915,1976

85. Meis PJ, Goldenberg RL, Mercer BM, et al: The preterm predictionstudy: risk factors for indicated preterm births. Maternal-Fetal Medi-cine Units Network of the National Institute of Child Health andHuman Development. Am J Obstet Gynecol 178:562-567, 1998

86. Ananth CV, Vintzileos AM: Maternal-fetal conditions necessitating amedical intervention resulting in preterm birth. Am J Obstet Gynecol195:1557-1563, 2006

87. Meis PJ, Goldenberg RL, Mercer B, et al: The preterm predictionstudy: significance of vaginal infections. National Institute of ChildHealth and Human Development Maternal-Fetal Medicine Units Net-work. Am J Obstet Gynecol 173:1231-1235, 1995

88. Romero R, Mazor M, Wu YK, et al: Infection in the pathogenesis ofpreterm labor. Semin Perinatol 12:262-279, 1988

89. Romero R, Sirtori M, Oyarzun E, et al: Infection and labor. V. Preva-lence, microbiology, and clinical significance of intraamniotic infec-tion in women with preterm labor and intact membranes. Am J ObstetGynecol 161:817-824, 1989

90. Romero R, Sepulveda W, Baumann P, et al: The preterm labor syn-drome: Biochemical, cytologic, immunologic, pathologic, microbio-logic, and clinical evidence that preterm labor is a heterogeneousdisease. Am J Obstet Gynecol 288, 1993

91. Romero R, Yoon BH, Mazor M, et al: The diagnostic and prognosticvalue of amniotic fluid white blood cell count, glucose, interleukin-6,and gram stain in patients with preterm labor and intact membranes.Am J Obstet Gynecol 169:805-816, 1993

92. Romero R, Mazor M, Munoz H, et al: The preterm labor syndrome.Ann N Y Acad Sci 734:414-429, 1994

93. Gomez R, Romero R, Nien JK, et al: Idiopathic vaginal bleeding duringpregnancy as the only clinical manifestation of intrauterine infection.J Matern Fetal Neonatal Med 18:31-37, 2005

94. Espinoza J, Goncalves LF, Romero R, et al: The prevalence and clinical

significance of amniotic fluid ‘sludge’ in patients with preterm laborand intact membranes. Ultrasound Obstet Gynecol 25:346-352, 2005

95. Jacobsson B, Mattsby-Baltzer I, Andersch B, et al: Microbial invasionand cytokine response in amniotic fluid in a Swedish population ofwomen in preterm labor. Acta Obstet Gynecol Scand 82:120-128,2003

96. Helmig BR, Romero R, Espinoza J, et al: Neutrophil elastase andsecretory leukocyte protease inhibitor in prelabor rupture of mem-branes, parturition and intra-amniotic infection. J Matern Fetal Neo-natal Med 12:237-246, 2002

97. Edwards RK, Clark P, Locksmith GJ, et al: Performance characteristicsof putative tests for subclinical chorioamnionitis. Infect Dis ObstetGynecol 9:209-214, 2001

98. Yoon BH, Chang JW, Romero R: Isolation of Ureaplasma urealyticumfrom the amniotic cavity and adverse outcome in preterm labor. Ob-stet Gynecol 92:77-82, 1998

99. Rizzo G, Capponi A, Vlachopoulou A, et al: The diagnostic value ofinterleukin-8 and fetal fibronectin concentrations in cervical secre-tions in patients with preterm labor and intact membranes. J PerinatMed 25:461-468, 1997

00. Garry D, Figueroa R, Guero-Rosenfeld M, et al: A comparison of rapidamniotic fluid markers in the prediction of microbial invasion of theuterine cavity and preterm delivery. Am J Obstet Gynecol 175:1336-1341, 1996

01. Rizzo G, Capponi A, Rinaldo D, et al: Interleukin-6 concentrations incervical secretions identify microbial invasion of the amniotic cavity inpatients with preterm labor and intact membranes. Am J Obstet Gy-necol 175:812-817, 1996

02. Mazor M, Hershkovitz R, Ghezzi F, et al: Intraamniotic infection inpatients with preterm labor and twin pregnancies. Acta Obstet Gy-necol Scand 75:624-627, 1996

03. Mazor M, Furman B, Wiznitzer A, et al: Maternal and perinatal out-come of patients with preterm labor and meconium-stained amnioticfluid. Obstet Gynecol 86:830-833, 1995

04. Baumann P, Romero R, Berry S, et al: Evidence of participation of thesoluble tumor necrosis factor receptor I in the host response to intra-uterine infection in preterm labor. Am J Reprod Immunol 30:184-193, 1993

05. Maymon E, Romero R, Pacora P, et al: A role for the 72 kDa gelatinase(MMP-2) and its inhibitor (TIMP-2) in human parturition, prematurerupture of membranes and intraamniotic infection. J Perinat Med29:308-316, 2001

06. Jacobsson B, Mattsby-Baltzer I, Andersch B, et al: Microbial invasionand cytokine response in amniotic fluid in a Swedish population ofwomen with preterm prelabor rupture of membranes. Acta ObstetGynecol Scand 82:423-431, 2003

07. Park KH, Chaiworapongsa T, Kim YM, et al: Matrix metalloproteinase3 in parturition, premature rupture of the membranes, and microbialinvasion of the amniotic cavity. J Perinat Med 31:12-22, 2003

08. Romero R, Yoon BH, Mazor M, et al: A comparative study of thediagnostic performance of amniotic fluid glucose, white blood cellcount, interleukin-6, and gram stain in the detection of microbialinvasion in patients with preterm premature rupture of membranes.Am J Obstet Gynecol 169:839-851, 1993

09. Romero R, Quintero R, Oyarzun E, et al: Intraamniotic infection andthe onset of labor in preterm premature rupture of the membranes.Am J Obstet Gynecol 159:661-666, 1988

10. Romero R, Ghidini A, Mazor M, et al: Microbial invasion of the am-niotic cavity in premature rupture of membranes. Clin Obstet Gy-necol 34:769-778, 1991

11. Romero R, Baumann P, Gomez R, et al: The relationship betweenspontaneous rupture of membranes, labor, and microbial invasion ofthe amniotic cavity and amniotic fluid concentrations of prostaglan-dins and thromboxane B2 in term pregnancy. Am J Obstet Gynecol168:1654-1664, 1993

12. Kim YM, Chaiworapongsa T, Gomez R, et al: Failure of physiologictransformation of the spiral arteries in the placental bed in pretermpremature rupture of membranes. Am J Obstet Gynecol 187:1137-
1142, 2002

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2


13. Kim YM, Bujold E, Chaiworapongsa T, et al: Failure of physiologictransformation of the spiral arteries in patients with preterm labor andintact membranes. Am J Obstet Gynecol 189:1063-1069, 2003

14. Lyall F: Mechanisms regulating cytotrophoblast invasion in normalpregnancy and pre-eclampsia. Aust N Z J Obstet Gynaecol 46:266-273, 2006

15. Robertson WB, Brosens I, Dixon HG: The pathological response of thevessels of the placental bed to hypertensive pregnancy. J Pathol Bac-teriol 93:581-592, 1967

16. Gerretsen G, Huisjes HJ, Elema JD: Morphological changes of thespiral arteries in the placental bed in relation to pre-eclampsia andfetal growth retardation. Br J Obstet Gynaecol 88:876-881, 1981

17. Sheppard BL, Bonnar J: Ultrastructural abnormalities of placental villiin placentae from pregnancies complicated by intrauterine fetalgrowth retardation: their relationship to decidual spiral arterial le-sions. Placenta 1:145-156, 1980

18. Norwitz ER: Defective implantation and placentation: laying the blue-print for pregnancy complications. Reprod Biomed Online 13:591-599, 2006

19. Burton GJ, Jauniaux E: Placental oxidative stress: from miscarriage topreeclampsia. J Soc Gynecol Investig 11:342-352, 2004

20. Redman CW: Immunological aspects of pre-eclampsia. Baillieres ClinObstet Gynaecol 6:601-615, 1992

21. Chwalisz K: The use of progesterone antagonists for cervical ripeningand as an adjunct to labour and delivery. Hum Reprod 9:131-161,1994 (suppl 1)

22. Gorodeski IG, Geier A, Lunenfeld B, et al: Progesterone (P) receptordynamics in estrogen primed normal human cervix following P injec-tion. Fertil Steril 47:108-113, 1987

23. Stjernholm Y, Sahlin L, Akerberg S, et al: Cervical ripening in humans:potential roles of estrogen, progesterone, and insulin-like growth fac-tor-I. Am J Obstet Gynecol 174:1065-1071, 1996

24. Johnson JW, Austin KL, Jones GS, et al: Efficacy of 17alpha-hy-droxyprogesterone caproate in the prevention of premature labor.N Engl J Med 293:675-680, 1975

25. Hill LM, Johnson CE, Lee RA: Prophylactic use of hydroxyprogester-one caproate in abdominal surgery during pregnancy. A retrospectiveevaluation. Obstet Gynecol 46:287-290, 1975

26. Breart G, Lanfranchi M, Chavigny C, et al: A comparative study of theefficiency of hydroxyprogesterone caproate and of chlormadinoneacetate in the prevention of premature labor. Int J Gynaecol Obstet16:381-384, 1979

27. Johnson JW, Lee PA, Zachary AS, et al: High-risk prematurity: pro-gestin treatment and steroid studies. Obstet Gynecol 54:412-418,1979

28. Hartikainen-Sorri AL, Kauppila A, Tuimala R: Inefficacy of 17 alpha-hydroxyprogesterone caproate in the prevention of prematurity intwin pregnancy. Obstet Gynecol 56:692-695, 1980

29. Hauth JC, LC III, Brekken AL, et al: The effect of 17 alpha-hy-droxyprogesterone caproate on pregnancy outcome in an active-dutymilitary population. Am J Obstet Gynecol 146:187-190, 1983

30. Yemini M, Borenstein R, Dreazen E, et al: Prevention of prematurelabor by 17 alpha-hydroxyprogesterone caproate. Am J Obstet Gy-necol 151:574-577, 1985

31. Goldstein P, Berrier J, Rosen S, et al: A meta-analysis of randomizedcontrol trials of progestational agents in pregnancy. Br J Obstet Gynae-col 96:265-274, 1989

32. Daya S: Efficacy of progesterone support for pregnancy in womenwith recurrent miscarriage. A meta-analysis of controlled trials. Br JObstet Gynaecol 96:275-280, 1989

33. Keirse MJ: Progestogen administration in pregnancy may prevent pre-term delivery. Br J Obstet Gynaecol 97:149-154, 1990

34. Meis PJ, Aleman A: Progesterone treatment to prevent preterm birth.Drugs 64:2463-2474, 2004

35. Noblot G, Audra P, Dargent D, et al: The use of micronized proges-terone in the treatment of menace of preterm delivery. Eur J ObstetGynecol Reprod Biol 40:203-209, 1991

36. Iams JD: Supplemental progesterone to prevent preterm birth. Am J
Obstet Gynecol 188:303, 2003
37. Brancazio LR, Murtha AP, Heine RP: Prevention of recurrent pretermdelivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med349:1087-1088, 2003

38. Doggrell SA: Recurrent hope for the treatment of preterm delivery.Expert Opin Pharmacother 4:2363-66, 2003

39. Einstein FH, Bracero LA: Progesterone and preterm birth. Am J ObstetGynecol 190:1798-1799, 2004

40. Keirse MJ: Progesterone and preterm: seventy years of “deja vu” or“still to be seen”? Birth 31:230-235, 2004

41. Di Renzo GC, Mattei A, Gojnic M, et al: Progesterone and pregnancy.Curr Opin Obstet Gynecol 17:598-600, 2005

42. Ness A, Dias T, Damus K, et al: Impact of the recent randomized trialson the use of progesterone to prevent preterm birth: a 2005 follow-upsurvey. Am J Obstet Gynecol 195:1174-1179, 2006

43. da Fonseca EB, Bittar RE, Carvalho MH, et al: Prophylactic adminis-tration of progesterone by vaginal suppository to reduce the incidenceof spontaneous preterm birth in women at increased risk: a random-ized placebo-controlled double-blind study. Am J Obstet Gynecol188:419-424, 2003

44. Meis PJ, Klebanoff M, Thom E, et al: Prevention of recurrent pretermdelivery by 17 alpha-hydroxyprogesterone caproate. N Engl J Med348:2379-2385, 2003

45. Spong CY, Meis PJ, Thom EA, et al: Progesterone for prevention ofrecurrent preterm birth: impact of gestational age at previous delivery.Am J Obstet Gynecol 193:1127-1131, 2005

46. Caritis S, Rouse DJ: A randomized controlled trial of 17-hydroxipro-gesterone caproate (17-OHPC) for the prevention of preterm birth intwins. Am J Obstet Gynecol 6:S2, 2007

47. Dodd JM, Flenady V, Cincotta R, et al: Prenatal administration ofprogesterone for preventing preterm birth. Cochrane Database SystRev CD004947, 2006

48. Dodd JM, Crowther CA, Cincotta R, et al: Progesterone supplemen-tation for preventing preterm birth: a systematic review and meta-analysis. Acta Obstet Gynecol Scand 84:526-533, 2005

49. Sanchez-Ramos L, Kaunitz AM, Delke I: Progestational agents to pre-vent preterm birth: a meta-analysis of randomized controlled trials.Obstet Gynecol 105:273-279, 2005

50. Odibo AO, Stamilio DM, Macones GA, et al: 17alpha-hydroxyproges-terone caproate for the prevention of preterm delivery: a cost-effec-tiveness analysis. Obstet Gynecol 108:492-499, 2006

51. Advisory Committee for Reproductive Health Drugs Meeting. the safetyand efficacy of New Drug Application (NDA 21-945), proposed tradename Gestiva, 17 alpha-hydroxyprogesterone caproate injection, 250mg/mL, Adeza Biomedical, for the proposed indication prevention ofpreterm delivery in women with a history of a prior preterm delivery.Available at: http://www.fda.gov/ohrms/dockets/ac/06/minutes/2006-4227M1.pdf. Last accessed August 29, 2006

52. Christian MS, Brent RL, Calda P: Embryo–fetal toxicity signals for17a-hydroxyprogesterone caproate in high-risk pregnancies: A reviewof the non-clinical literature for embryo–fetal toxicity with progestins.J Matern Fetal Neonatal Med 20:89-112, 2007

53. McDonald H, Brocklehurst P, Gordon A: Antibiotics for treating bac-terial vaginosis in pregnancy. Cochrane Database Syst RevCD000262, 2007

54. Carey JC, Klebanoff MA, Hauth JC, et al: Metronidazole to preventpreterm delivery in pregnant women with asymptomatic bacterialvaginosis. National Institute of Child Health and Human Develop-ment Network of Maternal-Fetal Medicine Units. N Engl J Med 342:534-540, 2000

55. Carey JC, Klebanoff MA: What have we learned about vaginal infec-tions and preterm birth? Semin Perinatol 27:212-216, 2003

56. Riggs MA, Klebanoff MA: Treatment of vaginal infections to preventpreterm birth: a meta-analysis. Clin Obstet Gynecol 47:796-807,2004

57. Odendaal HJ, Popov I, Schoeman J, et al: Preterm labour: is bacterialvaginosis involved? S Afr Med J 92:231-234, 2002

58. Kekki M, Kurki T, Pelkonen J, et al: Vaginal clindamycin in preventing
preterm birth and peripartal infections in asymptomatic women with
http://www.fda.gov/ohrms/dockets/ac/06/minutes/2006-4227M1.pdf

http://www.fda.gov/ohrms/dockets/ac/06/minutes/2006-4227M1.pdf

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2

2


bacterial vaginosis: a randomized, controlled trial. Obstet Gynecol97:643-648, 2001

59. Lamont RF, Jones BM, Mandal D, et al: The efficacy of vaginal clinda-mycin for the treatment of abnormal genital tract flora in pregnancy.Infect Dis Obstet Gynecol 11:181-189, 2003

60. Kekki M, Kurki T, Kotomaki T, et al: Cost-effectiveness of screeningand treatment for bacterial vaginosis in early pregnancy amongwomen at low risk for preterm birth. Acta Obstet Gynecol Scand83:27-36, 2004

61. Larsson PG, Fahraeus L, Carlsson B, et al: Late miscarriage and pre-term birth after treatment with clindamycin: a randomised consentdesign study according to Zelen. Br J Obstet Gynaecol 113:629-637,2006

62. Morency AM, Bujold E: The effect of second-trimester antibiotic ther-apy on the rate of preterm birth. J Obstet Gynaecol Can 29:35-44,2007

63. Andrews WW, Goldenberg RL, Hauth JC, et al: Interconceptionalantibiotics to prevent spontaneous preterm birth: a randomized clin-ical trial. Am J Obstet Gynecol 194:617-623, 2006

64. Espinoza J, Erez O, Romero R: Preconceptional antibiotic treatment toprevent preterm birth in women with a previous preterm delivery.Am J Obstet Gynecol 194:630-637, 2006

65. Briggs RM, Thompson WB Jr: Treatment of the incompetent cervix.Obstet Gynecol 16:414-418, 1960

66. Seppala M, Vara P: Cervical cerclage in the treatment of incompetentcervix. A retrospective analysis of the indications and results of 164operations. Acta Obstet Gynecol Scand 49:343-346, 1970

67. Robboy MS: The management of cervical incompetence. UCLA expe-rience with cerclage procedures. Obstet Gynecol 41:108-112, 1973

68. Crombleholme WR, Minkoff HL, Delke I, et al: Cervical cerclage: anaggressive approach to threatened or recurrent pregnancy wastage.Am J Obstet Gynecol 146:168-174, 1983

69. Lazar P, Gueguen S, Dreyfus J, et al: Multicentred controlled trial ofcervical cerclage in women at moderate risk of preterm delivery. Br JObstet Gynaecol 91:731-735, 1984

70. Rush RW, Isaacs S, McPherson K, et al: A randomized controlled trialof cervical cerclage in women at high risk of spontaneous pretermdelivery. Br J Obstet Gynaecol 91:724-730, 1984

71. Final report of the Medical Research Council/Royal College of Obste-tricians and Gynaecologists multicentre randomised trial of cervicalcerclage. MRC/RCOG Working Party on Cervical Cerclage. Br J ObstetGynaecol 100:516-523, 1993

72. Ayhan A, Mercan R, Tuncer ZS, et al: Postconceptional cervical cer-clage. Int J Gynaecol Obstet 42:243-246, 1993

73. Golan A, Wolman I, Arieli S, et al: Cervical cerclage for the incompe-tent cervical os. Improving the fetal salvage rate. J Reprod Med 40:367-370, 1995

74. Olatunbosun OA, al Nuaim L, Turnell RW: Emergency cerclage com-pared with bed rest for advanced cervical dilatation in pregnancy. IntSurg 80:170-174, 1995

75. Guzman ER, Houlihan C, Vintzileos A, et al: The significance of trans-vaginal ultrasonographic evaluation of the cervix in women treatedwith emergency cerclage. Am J Obstet Gynecol 175:471-476, 1996

76. Kurup M, Goldkrand JW: Cervical incompetence: elective, emergent,or urgent cerclage. Am J Obstet Gynecol 181:240-246, 1999

77. Althuisius SM, Dekker GA, van Geijn HP, et al: Cervical incompetence

prevention randomized cerclage trial (CIPRACT): study design andpreliminary results. Am J Obstet Gynecol 183:823-829, 2000

78. Althuisius SM, Dekker GA, Hummel P, et al: Final results of theCervical Incompetence Prevention Randomized Cerclage Trial(CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.Am J Obstet Gynecol 185:1106-1112, 2001

79. Guzman ER, Ananth CV: Cervical length and spontaneous prematu-rity: laying the foundation for future interventional randomized trialsfor the short cervix. Ultrasound Obstet Gynecol 18:195-199, 2001

80. Novy MJ, Gupta A, Wothe DD, et al: Cervical cerclage in the secondtrimester of pregnancy: a historical cohort study. Am J Obstet Gynecol184:1447-1454, 2001

81. Althuisius SM, Dekker GA, van Geijn HP: Cervical incompetence: areappraisal of an obstetric controversy. Obstet Gynecol Surv 57:377-387, 2002

82. To MS, Palaniappan V, Skentou C, et al: Elective cerclage vs. ultra-sound-indicated cerclage in high-risk pregnancies. Ultrasound ObstetGynecol 19:475-477, 2002

83. Althuisius SM, Dekker GA, Hummel P, et al: Cervical incompetenceprevention randomized cerclage trial: emergency cerclage with bedrest versus bed rest alone. Am J Obstet Gynecol 189:907-910, 2003

84. Berghella V, Bega G, Tolosa JE, et al: Ultrasound assessment of thecervix. Clin Obstet Gynecol 46:947-962, 2003

85. Odibo AO, Elkousy M, Ural SH, et al: Prevention of preterm birth bycervical cerclage compared with expectant management: a systematicreview. Obstet Gynecol Surv 58:130-136, 2003

86. To MS, Alfirevic Z, Heath VC, et al: Cervical cerclage for prevention ofpreterm delivery in women with short cervix: randomised controlledtrial. Lancet 363:1849-1853, 2004

87. Belej-Rak T, Okun N, Windrim R, et al: Effectiveness of cervical cer-clage for a sonographically shortened cervix: a systematic review andmeta-analysis. Am J Obstet Gynecol 189:1679-1687, 2003

88. Drakeley AJ, Roberts D, Alfirevic Z: Cervical stitch (cerclage) for pre-venting pregnancy loss in women. Cochrane Database Syst RevCD003253, 2003

89. Drakeley AJ, Roberts D, Alfirevic Z: Cervical cerclage for prevention ofpreterm delivery: meta-analysis of randomized trials. Obstet Gynecol102:621-627, 2003

90. Rust OA, Atlas RO, Jones KJ, et al: A randomized trial of cerclageversus no cerclage among patients with ultrasonographically detectedsecond-trimester preterm dilatation of the internal os. Am J ObstetGynecol 183:830-835, 2000

91. Higgins SP, Kornman LH, Bell RJ, et al: Cervical surveillance as analternative to elective cervical cerclage for pregnancy management ofsuspected cervical incompetence. Aust N Z J Obstet Gynaecol 44:228-232, 2004

92. Sakai M, Shiozaki A, Tabata M, et al: Evaluation of effectiveness ofprophylactic cerclage of a short cervix according to interleukin-8 incervical mucus. Am J Obstet Gynecol 194:14-19, 2006

93. Goldenberg RL, Klebanoff M, Carey JC, et al: Metronidazole treatmentof women with a positive fetal fibronectin test result. Am J ObstetGynecol 185:485-486, 2001

94. Shennan A, Crawshaw S, Briley A, et al: A randomised controlled trialof metronidazole for the prevention of preterm birth in women posi-tive for cervicovaginal fetal fibronectin: the PREMET Study. Br J Ob-
stet Gynaecol 113:65-74, 2006

PAM

Attc(

amladoprdrt

*

†

A

0d

ostpartum Hemorrhage:Recurring Pregnancy Complication

ichelle A. Kominiarek, MD,* and Sarah J. Kilpatrick, MD, PhD†

Postpartum hemorrhage (PPH) is a potentially life-threatening complication of both vaginaland cesarean deliveries. Although many variables increase the chance for bleeding, a PPHin a previous pregnancy is one of the greatest risk factors for recurrent PPH. A physiologicexplanation for this association is not known, but recurrent risk factors such as a retainedplacenta or underlying medical disorders may account for the majority of recurrent PPHcases. To reduce maternal morbidity and mortality, prevention of PPH in these patients iscritical. Steps to minimize hemorrhagic complications include the identification of high-riskpatients through a complete history, vigilant management of the third stage of labor, andhaving uterotonic medications readily available in the delivery room. Patients with inheritedcoagulopathies require individualized treatment, and their risks for bleeding extend beyondthe first 24 hours after delivery. Further studies are needed to determine whether theadministration of prophylactic measures such as prostaglandins decrease the PPH occur-rence in high-risk patients.Semin Perinatol 31:159-166 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS postpartum hemorrhage, recurrence, risk factors, prevention, retained placenta,coagulation disorders

bc

rqoiaeum

ogupmdtevic

lthough the prevalence of postpartum hemorrhage(PPH) varies between developing and developed coun-

ries, it is the leading cause of maternal morbidity and mor-ality worldwide.1 The etiology of PPH falls into four mainategories: tone, tissue, trauma, and thrombosis disordersTable 1).

An estimated blood loss �500 mL for a vaginal deliverynd �1000 mL for a cesarean delivery defines early or pri-ary PPH, which occurs within the first 24 hours after de-

ivery. This affects 4% to 6% of pregnancies, and uterinetony is the cause in 75% to 90% of cases.2,3 The subjectiveescription of blood loss at delivery limits the interpretationf outcomes in any study of PPH. Comparisons of pre- andostpartum hemoglobin or hematocrit may be more accu-ate, and a decrease in hematocrit levels by 10% has alsoefined PPH.2 Although a patient may meet blood loss crite-ia for a PPH, blood transfusion rates are better estimates ofhe hemorrhage severity. Postpartum anemia resulting in red

Department of Obstetrics and Gynecology, Indiana University School ofMedicine, Indianapolis, IN.

Department of Obstetrics and Gynecology, University of Illinois at Chi-cago, Chicago, IL.

ddress reprint requests to Michelle A. Kominiarek, MD, 550 North Uni-versity Boulevard, Room 2440, Department of Obstetrics and Gynecol-

Pogy, Indianapolis, IN 46202. E-mail: [email protected]


lood cell transfusion occurs in less than 1% and after aesarean delivery 1% to 7%.4

Late or secondary PPH is excessive vaginal bleeding occur-ing between 24 hours and 6 weeks after delivery.5 Unlike theuantitative definition of primary PPH, the definition of sec-ndary PPH is more subjective and requires sufficient bleed-ng to prompt the patient to seek medical attention. Second-ry PPH is also referred to as persistent or delayed PPH and isstimated to occur in 1% to 3% of all deliveries.6 Patientssually present during the second postpartum week, and theost common etiology is retained placenta fragments.Although PPH occurs in women without risk factors, it is

ften a predictable event. There are many risk factors sug-ested for PPH (Table 2). These include an over-distendedterus (multiple gestations, polyhydraminos, macrosomia),rimigravidity, chorioamnionitis, prolonged rupture ofembranes, fibroids, previous cesarean delivery, coagulationisorders, anticoagulant therapy, labor induction and augmen-

ation, prolonged labors, preeclampsia, obesity, and general an-sthetics.7-9 The relationship between grand multiparity (�5aginal births) and PPH has been disputed.2,3,8,10,11 In a univar-ate analysis, delivery by a midwife was associated with a de-reased risk of PPH compared with physician deliveries.2

One of the most important risk factors for PPH is a prior
PH. However, in women without recurrent risk factors,
159

spkrp

ToTwms1Oswethr

la�“teP(ao2Ptas

o

HPrttdpuo3tr9ucfiibposacr

arastlwafga7d

T

T

160 M.A. Kominiarek and S.J. Kilpatrick

uch as a previous twin gestation or chorioamnionitis, thehysiological explanation for the increased recurrence is un-nown. The focus of this article is to review data on theecurrence of either primary or secondary PPH in futureregnancies.

he Epidemiologyf Recurrent PPH

he objective of one of the earliest studies on recurrent PPHas to determine whether a patient with a prior PPH wasore likely to hemorrhage during or after the third stage of a

ubsequent labor. They included deliveries from 1936 to945 and defined PPH as an estimated blood loss �600 mL.3

f interest, cesareans, which were excluded from the analy-is, accounted for only 1.7% of the deliveries. Of the 12,243omen with vaginal births, the rate of PPH was 5.2%. How-

ver, the recurrence rate of PPH was 8.1%. In 68% (21/31) ofhe repeated PPH cases, the cause was the same as the initialemorrhage, suggesting that recurrent risk factors may play aole in PPH recurrence.3

In a study of recurring abnormalities of the third stage ofabor, 132 pregnancies with a known prior PPH were evalu-ted.12 Twenty-three percent had a recurrent PPH (blood loss20 oz or required placenta extraction), and 22% had a

potentially” abnormal third stage but precautions wereaken to prevent hemorrhage (placenta removed or givenrgometrine even though blood loss was �20 oz). After onePH, there was a recurrence in 20/100 (20%) and in 7/2528%) after more than one PPH. Even if a normal third stageccompanied a second pregnancy following a prior PPH, theccurrence of PPH in the third pregnancy was still high, at1%. There was a subset of 10 women whose risk factors forPH in the first pregnancy included nonrecurring complica-ions such as general anesthesia, antepartum hemorrhage,nd multiple gestations. These women had a total of 17 sub-equent pregnancies without PPH.12

Changes in obstetrical practice, including routine use of

able 1 Etiology of Postpartum Hemorrhage

ToneUterine atonyUterine inversion

TissueRetained placenta or blood clotsAbnormal placentation (previa, accreta)Connective tissue disorders (Ehlers-Danlos, Marfans)

TraumaLower genital tract lacerationsUterine rupture

Thrombosis disordersCoagulopathies

Inherited coagulation disordersHELLPDIC

Anticoagulant use

xytocics and decreasing parity among women, prompted

all and coworkers to study the recurrence risk for eitherPH (�500 mL blood loss at a singleton vaginal delivery) oretained placenta (requiring manual removal) using a longi-udinal analysis which controlled for the patient’s reproduc-ive history.11 The population included 6615 women whoelivered between 1967 and 1981. Although the originalaper did not reports statistics, for patients with two consec-tive live births and a history of PPH in the first birth, thedds ratio (OR) for a patient to have a subsequent PPH was.68 (95% CI 2.49-5.45). Interestingly, after a labor induc-ion for the second birth, the risk of either recurrent PPH oretained placenta was also significant at 19.6% (OR 2.89,5% CI 2.89-5.28). Of 1106 women who had three consec-tive live births, only one had a PPH three times. PPH oc-urred in 13.5% in the second live birth in 375 women whoserst birth was preceded by an abortion. In summary, the

ncidence of PPH and/or retained placenta in the next twoirths was �12% in the presence of a PPH and/or retainedlacenta in the first birth. If there was no prior history of PPHr retained placenta, the incidence of either one in the sub-equent pregnancy was �6% unless there was an interveningbortion.11 Although this study did not address pregnancyomplications or risk factors, patients with a prior PPH oretained placenta were more likely to have a recurrence.

In a more recent case-control study of risk factors for PPH,previous PPH had an OR of 3.55 (95% CI 1.24-10.19) for

ecurrent PPH (a hematocrit decrease of �10 points betweendmission and postdelivery or by the need for blood transfu-ion) in a multivariate analysis.2 A previous PPH was one ofhe strongest predictors of recurrent PPH in their 17-factorogistic regression model. An interesting aspect of the studyas their analysis of which hemorrhages might be predict-

ble. A regression model of variables potentially known be-ore the onset of labor, such as preeclampsia, parity, multipleestation, prior PPH, or previous cesarean delivery, derivedn adjusted OR for a previous PPH of 2.85 (95% CI 1.07-.60). Similar results were reported in a review of 19,476eliveries at a tertiary care center. PPH (blood loss �1000 mL

able 2 Risk Factors for Postpartum Hemorrhage

Non-recurringPrimigravidityMacrosomiaPolyhydraminosMultiple gestationsProlonged or augmented laborsProlonged third stageChorioamnionitisAntepartum hemorrhageOperative deliveries

RecurringFibroidsMaternal obesityCoagulation disordersPrevious cesareanSpecific medical/genetic disorders�Grand multiparity
Previous postpartum hemorrhage

av1

bfo(

psm

CRPPTusateovrpsm

owvc“if

ell

afaa

eppcartwfcB

saurtctdma

FFmlctlar

PsebUIomtfifsp

To

PSP

P

M

U

V

�

†

Postpartum hemorrhage 161

nd/or need for blood transfusion) occurred in 5.15% ofaginal deliveries and a prior PPH had an OR of 2.2 (95% CI.7-2.9) for recurrent PPH.13

Although data are more limited, recurrent PPH has alsoeen described for secondary PPH. A retrospective studyrom two hospitals in the United Kingdom identified 48 casesf secondary PPH (0.73% incidence). Of the 32 multiparas, 618.8%) also had a secondary PPH in a previous pregnancy.14

Over several decades, despite changing obstetrical care, arevious PPH remains an important predictor of PPH in sub-equent deliveries. Clinicians should consider it as one of theost important risk factors for PPH (Table 3).

ondition-Specificecurrence Risks forostpartum Hemorrhagerevious Invasive Treatment for PPHhe primary management of PPH is medical: treatment withterotonic drugs. In most cases, this is successful; however,urgical techniques to control bleeding, such as hypogastricrtery ligation, bilateral uterine artery ligation (O’Leary su-ures), and the B-Lynch technique, are alternatives to hyster-ctomy in cases of persistent bleeding. Arterial embolizationr balloon occlusion of radiographically identified bleedingessels is another nonsurgical option for continued hemor-hage. Patients requiring any of these treatments for either arimary or secondary PPH were very likely to have suffered aignificant hemorrhage. The risk for recurrent PPH is of ut-ost concern in this group of patients.Nizard and coworkers reported subsequent pregnancy

utcomes after bilateral hypogastric artery ligation in 43omen.15 Information was obtained either through chart re-iew or telephone interview. Of the 13 completed pregnan-ies, two had a “threatened” PPH which was described aseasily managed by manual evacuation of the placenta andntravenous oxytocin,” and one patient had a PPH success-

able 3 Recurrence Risk for Postpartum Hemorrhage Basedn Prior History and Risk Factors

Risk FactorRecurrence

Risk for PPH

rimary PPH3,11,12 8-28%econdary PPH14 19%†rior PPH treated with hypogastric arteryligation15

27%

rior PPH treated with uterine arteryembolization16-18

0-100%

yomectomy as treatment forfibroids29,31

0-1%

terine artery ligation as treatment forfibroids21,29,30,32,33

6-19%

on Willebrand disease46 80%

Baseline risks: primary PPH 4-6%, secondary PPH 1-3%.Risk is for recurrent secondary PPH.

ully treated with prostaglandins during a cesarean deliv- c

ry.15 Although the numbers in this report are small, it is theargest series of pregnancies reported after hypogastric arteryigation, and a recurrence of 3/13 (27%) is significant.

Several studies have addressed the long-term effect of uterinertery embolization (UAE) for the control of PPH on menses,ertility, and future pregnancies. Thirteen full-term deliveriesfter a previous bilateral UAE for a PPH were uncomplicated,nd no instances of recurrent PPH were reported.16,17

In both of these studies, the authors do not mention thetiology or any high-risk factors for hemorrhage in the priorregnancy or if any prophylactic measures were taken torevent a recurrence. In contrast to these reassuring out-omes, a separate study describes four patients who deliveredt term after a UAE for PPH in a prior pregnancy. Hemor-hage recurred in all of them, and two required a hysterec-omy for placenta accreta.18 The etiology for the prior PPHas uterine atony in three cases and a placenta accreta in the

ourth. There are case reports of successful pregnancy out-omes without hemorrhage after a B-lynch suture19 and a-lynch suture combined with hypogastric artery ligation.20

The small number of cases limits the interpretation of thesetudies. The exact cause for the recurrent PPH is unclear, butdevascularized myometrium after UAE may not allow theterus to adequately contract after a delivery.21 Other theo-ies propose that a prior UAE may lead to abnormal implan-ation and excessive trophoblast invasion in future pregnan-ies.18 It may be reasonable to consider antenatal imaging inhese cases to assess the placental implantation site for evi-ence of invasive placental disease. Although outcomes areixed, preparation for the recurrence of PPH and placenta

ccreta is the best approach in these patients.

ibroidsibroids may increase the chance for PPH. The mechanismay be mediated through abnormal muscle contractility

eading to uterine atony. Although they have been cited as theause for many complications in pregnancy, including pre-erm labor, abruption, fetal growth restriction, dysfunctionalabor, and cesarean delivery, the studies regarding fibroidsnd PPH are conflicting, where some show an increasedisk22-25 and others show no association.26-28

Although no investigations have directly evaluated thePH recurrence in the setting of uterine fibroids, a number oftudies have described PPH outcomes in patients treated withither an interval myomectomy or UAE for symptomatic fi-roids. In an analysis of 50 published cases of pregnancy afterAE for fibroids, there were 2/23 (9%) occurrences of PPH.21

n a subsequent study, these authors compared pregnancyutcomes in women treated with UAE to laparoscopic myo-ectomy. The difference in the PPH rates did not reach sta-

istical significance: 6% versus 1% (OR 6.3, 95% CI 0.6-71.8)or the UAE and myomectomy groups, respectively.29 A sim-lar study surveyed women who became pregnant after a UAEor fibroids. Of the 26 pregnancies, 16 continued beyond theecond trimester with a PPH rate of 19%.30 A review of 158regnancies after laparoscopic myomectomy reported no
ases of PPH, even though the cesarean delivery rate was

7nstut

mrwemoti

RTrppofoImwsc1c8oai

UUoctcunpnhPKDsiaioir

ife

VAVbrcWpbRPrcl

hprsc

l1wetns

PoaPl1wmahfldfvfimntpe

tV


5%.31 Walker and coworkers described 56 completed preg-ancies after UAE with 6 cases (18.2%) of PPH, which wasimilar to the 3/18 (17%) rate of PPH described in the On-ario Multicenter Trial, a large prospective study of womenndergoing UAE as an alternative to hysterectomy for symp-omatic fibroids.32,33

The recurrence of PPH in patients treated with either myo-ectomy or UAE is difficult to interpret from these small

etrospective studies as the previous pregnancy outcomesere not compared and PPH was not always defined. How-

ver, the available data suggest that the risk for PPH afteryomectomy is less than it is after UAE. Clearly, the removal

r treatment of fibroids before pregnancy does not eliminatehe risk of PPH, and in some instances, it appears to bencreased.

etained Placentahe incidence of retained placenta is 1% to 5.5%. The occur-ence of PPH is significantly higher when there is a retainedlacenta: 9.6% to 21.3% versus 2.3% to 3.5%.2,11 Retainedlacenta is essentially another complication of the third stagef labor, and many studies suggest that it is likely to recur inuture pregnancies. The medical records of 24,750 deliveriesver an 8-year period at a hospital in Norway were reviewed.n summary, 165 women (0.6%) had a manual placenta re-oval, and PPH occurred in 16 (10%).34 Of the 74 parousomen, 12 (16%) had a prior retained placenta. A similar

tudy of 134 patients with a retained placenta reported re-urrences in 23% to 32% of cases.35 In a case-control study of13 women with a retained placenta, a prior retained pla-enta was a risk factor for recurrence, OR 9.8 (95% CI 1.1-5.5) in a logistic regression analysis.36 Although recurrencef PPH has not been included in these reports, it is appropri-te to conclude that a history of retained placenta is a risk forts recurrence and, therefore, PPH.

terine Inversionterine inversion, a potentially life-threatening complicationf the third stage of labor, is associated with PPH in 94% ofases.37,38 In most situations, appropriate management of thehird stage can prevent this occurrence. Because it compli-ates few deliveries (about 1 in 2,500), the recurrence ofterine inversion is difficult to quantify. A review of 56 preg-ancies after a previous uterine inversion from various re-orts suggested that it recurred in 33% of subsequent preg-ancies.39 All recurrences were in those patients who did notave surgery (Spinelli operation, Piccoli’s incision, combinediccoli-Borelius-Westermann method, Küstner operation,üstner-Borelius-Westermann operation, Hehrer’s method,uret’s Method, or Aveling repositor) to correct the inver-

ion. In contrast, a more recent review of 40 cases of uterinenversion where PPH and blood transfusion rates were 65%nd 47.5%, respectively, reported no recurrences of uterinenversion in 26 subsequent pregnancies.40 Although the rec-mmended delivery route for patients with a prior uterinenversion has never been described in the literature, recur-
ent uterine inversion is a rare event. If there are no contra- h
ndications to vaginal delivery, then this should be the pre-erred route, but the placenta should be delivered withoutxcessive cord traction.

on Willebrand Diseasematernal coagulation disorder increases the risk for PPH.

on Willebrand disease (VWD), the most common inheritedleeding disorder, is found in 1% of the general populationegardless of ethnicity. Of the three categories, type 1 ac-ounts for 70% of all cases. Although factor VIII and vonillebrand factor antigen (vWF:Ag) levels often improve in

regnancy, women with VWD are at increased risk for PPHecause of the rapid decline in these levels after delivery.ates have been estimated as high as 16% to 29% for primaryPH and 20% to 28% for secondary PPH,41-43 and transfusionates as high as 25% have been reported.44 PPH occurs moreommonly in type 2 and 3 patients and those with vWF:Agevels �50% of normal at term.41,45

Of 75 women with type I VWD followed at 46 differentemophilia treatment centers in the United States, 32% re-orted PPH a total of 54 times, suggesting a high recurrenceate in subsequent pregnancies.44 PPH recurred with succes-ive pregnancies in 80% of patients who had type I VWD inombination with PPH during the first birth.46

Kadir and coworkers described 67 women with VWD fol-owed at the Royal Free Hospital Hemophilia Centre from980 to 1996.44 The diagnosis of VWD was unknown in sixomen with 11 pregnancies, and the diagnosis was later

stablished in two women when PPH occurred after an abor-ion. In an analysis of 38 women with type I VWD, the diag-osis of VWD was unknown before presentation in a sub-tantial number of cases, including 5 (13.1%) with PPH.47

This information leads one to question whether a patient’sPH is the consequence of an undiagnosed coagulation dis-rder such as VWD. The purpose of the study by Hundeggernd coworkers was to retrospectively evaluate patients withPH, defined as a drop in hemoglobin concentration of at

east 8% between delivery day and postpartum day 3, for typeVWD.48 Of the 3565 women who delivered in 1997, 14omen were eligible after excluding cesarean deliveries andajor birth injuries. These patients were contacted by phone

nd returned for testing after stopping estrogen-containingormones for 2 months and aspirin and nonsteroidal antiin-ammatory drugs for 1 week. Blood samples, collected onays 5 to 7 of the menstrual cycle, were tested for vWF:Ag,actors II, V, VII, VIII, IX, X, XI, and XII, prothrombin, acti-ated partial thromboplastin time (aPTT), thrombin time,brinogen, and for a complete blood count. None had abnor-al levels of vWF:Ag or other factor levels.48 Although theumbers are small, this is the only study that evaluated pa-ients for a coagulation disorder after having a PPH. Therevalence of undiagnosed VWD in women with PPH is pres-ntly unknown.

It is critical that the obstetrician/gynecologist is aware ofhe prevalence and clinical presentation of patients withWD. At the first prenatal visit or after a pregnancy-related
emorrhage, it is necessary to obtain a bleeding history, es-

poowfhssodarwwaFa

OCttrftaJir(rds

ttco

tdnco

CCnhDttbsvmf

edO2taso

PPtshavasdpeudeo0pcmmas

bdauptgl(mvt

T

IPH

AVH


ecially with respect to menorrhagia and bleeding at the timef surgical or dental procedures and injuries. A family historyf these occurrences is important to ascertain as well. Womenith unexplained significant menorrhagia should be tested

or VWD.49,50 In addition, acute postpartum or postoperativeemorrhages that do not respond to customary treatmenthould prompt an investigation of the patient’s coagulationtatus (aPTT and PT). Prolonged values may be an indicationf an underlying bleeding disorder. Identifying a coagulationisorder can improve morbidity from bleeding events andvoid unnecessary surgery. Given the limited available data,outine testing for VWD is currently not indicated in patientshose bleeding history is only a primary PPH. It is not clearhether VWD testing is of benefit for women having second-

ry PPH without evidence for retained placental fragments.uture studies will hopefully provide more guidance in thisrea.

ther Coagulation Disordersongenital coagulation disorders such as deficiencies in fac-

ors II, VII, and X are rare, but the obstetric experience ofhese patients has been described in case series and literatureeviews. The majority of patients with deficiencies in eitheractor II, VII, or X have excessive bleeding at delivery.51 Fac-or XI (Hemophilia C) deficiency is transmitted autosomally,nd the carrier rate is as high as 9% to 13% in the Ashkenaziewish population.52,53 In 1 report, there were 3/14 (21%)ncidences of PPH in severely deficient women.54 In a sepa-ate study, primary and secondary PPH complicated 4/2516%) and 6/25 (24%) of deliveries, respectively.42 Even car-iers of hemophilia A (factor VIII deficiency) and B (factor IXeficiency), have high rates of both primary (21.7%) andecondary PPH (10.9%).55

Primary PPH has been reported in 2.1% to 7.7% of patientsreated with anticoagulants for either a thrombophilia orhromboembolism during pregnancy.56,57 Immune thrombo-ytopenic purpura is also associated with PPH in 24% to 27%f cases.58,59

No study has addressed repeated PPH in these popula-ions, including those with rare inherited coagulation disor-ers. Assuming the conditions persist in subsequent preg-ancies, recurrence is likely. Increased awareness amonglinicians of these less common causes is essential for optimalutcome.

onnective Tissue Disordersompared with the general population, patients with con-ective tissue disorders have higher occurrences of PPH. PPHas been reported in up to 19% of patients with Ehlers-anlos syndrome, a disease characterized by abnormalities in

he metabolism and synthesis of collagen.60 The manifesta-ions of Ehlers-Danlos syndrome, such as easy bruising andleeding tendencies, are attributed to poor connective tissueupport of the blood vessels and direct involvement of theessel wall, rather than defective platelet function or abnor-al coagulation.61 Consequently, uterine atony from inef-

ective contractions or lacerations of the perineum likely A

xplain the increased frequency of PPH. Other maternaliseases associated with PPH include Gaucher (21-77%),steogenesis Imperfecta (11%), and Marfans Syndrome (8-1%).62-65 In patients with connective tissue disorders, thereatment approach to PPH is similar in that uterotonicgents are the first-line therapy. Uterine ruptures and inver-ions in these patients may account for some of the instancesf PPH.61,64

revention of PPH RecurrencePH prevention remains an important issue in obstetrics. Thehird stage of labor in women with a previous PPH historyhould be managed vigilantly, with anticipation of a potentialemorrhage. Patients should be appropriately counseledbout the potential for recurrence and its implications. Intra-enous access during labor and sending a type and screen areppropriate prophylactic measures. Management of the thirdtage of labor includes administration of oxytocin after theelivery of the fetus or placenta (the latter is a more commonractice in the United States) and expeditious placenta deliv-ry. A study comparing the administration of oxytocin (20nits in 500 mL crystalloid intravenous bolus) either after theelivery of the fetal shoulder or placenta showed no differ-nce in PPH (5.4% versus 5.8%, OR 0.92, 95% CI 0.59-1.43)r retained placenta (2.4% versus 1.6%, OR 1.49, 95% CI.72-3.08).66 At a cesarean, spontaneous delivery of thelacenta has been shown to decrease blood loss by 31%ompared with manual removal.67 Additional uterotonicedications, such as misoprostol, 15-methyl PGF2�,ethylergonovine, or dinoprostone, should be readily avail-

ble in the delivery room.68 These recommendations areummarized in Table 4.

One study investigated prostaglandin use in a double-linded randomized trial for women with a history of PPH,efined as a blood loss �1000 mL due to uterine atony aftervaginal delivery.69 Exclusion criteria were preexisting coag-lation disorders, anticoagulation treatment, fibroids, multi-le pregnancies, hypertension, and an induction or augmen-ation of labor. Patients were randomized either to the studyroup (0.5 mg sulprostone after delivery of the shoulder fol-owed by placebo after placenta delivery) or control group5 U oxytocin after delivery of the shoulder followed by 0.2g ergometrine after placenta delivery). An additional inter-

ention in the third stage was “fundal pressure while holdinghe lower segment of the uterus, after the first signs of pla-

able 4 Steps to Prevent a Recurrent Postpartum Hemorrhage

dentify risk factors including bleeding history.rovide intravenous access in labor.ave blood products available (type and screen as aninitial measure).

void an induction or augmentation of labor.igilant management of the third stage of labor.ave uterotonic medications readily available in thedelivery room.

llow spontaneous delivery of the placenta at cesareans.

ctbnt(ibss

nbaw

ihisvlus

ticocltDrsppdp

FApsmowscpmiitsco

qpr

CTmapiebtfumoc

R

1

1

1

1

1

1

1

1

1


ental detachment.” The trial was stopped prematurely dueo concerns of the cardiovascular side effects of sulprostone,ut 69 cases were available for analysis. Although there waso statistically significant difference between the 2 groups,he rates of PPH (�500 mL blood loss) were high: 16/3644%) in the oxytocin/ergometrine group and 16/33 (47%)n the sulprostone group.69 Although this study found noenefit to sulprostone in a high-risk population, the conclu-ion may have been different if enrollment had not beentopped.

Autologous blood donation has been described for preg-ant patients at high risk for hemorrhage.70-72 However, mostlood donated for these purposes is not transfused, and thispproach has not been shown to be cost effective in patientsithout a placenta previa.73,74

There are no known preventive strategies for patients atncreased risk for a secondary PPH. Counseling patients onow to recognize the signs and symptoms of excessive bleed-

ng may allow early diagnosis, but is unlikely to preventecondary PPH. Complete placental removal is the best pre-ention strategy. Antenatal findings of an accessory placentalobe may identify those patients who would benefit fromterine exploration after delivery to confirm complete expul-ion of placental tissue.

Patients with inherited coagulopathies require specializedreatment during the peripartum. Bleeding risks in affectedndividuals are difficult to predict given the wide variations inlinical presentation. However, preparing for the recurrencef PPH in these patients is the cornerstone of their medicalare. The risk for PPH is often directly correlated to the factorevel activity. Replacement of the specific deficient factor andransfusion of other products, such as fresh frozen plasma orDAVP, can correct the problem, but transfusion of packed

ed blood cells is still the treatment for acute anemia. Con-ultation or comanagement with a hematologist or hemo-hilia treatment center and the hospital’s blood bank is ap-ropriate in these instances. In the event of an operativeelivery, meticulous hemostasis during surgery will also im-rove outcomes.

uture Researchll further research on the topic of bleeding complications inregnancy should define PPH in a consistent manner so thattudies can be compared and generalized. A more objectiveeasurement of estimated blood loss, such as a comparison

f pre- and postpartum blood counts or calibrated drapes,ould be beneficial in prospective studies. Previous retro-

pective studies have identified the risk factors for PPH andonfirmed a significantly increased risk for recurrent PPH inatients with a prior history. Determining an underlyingechanism for recurrent PPH, such as a molecular or biolog-

cal marker, could identify those at risk and potentially givensight to the prevention and treatment of PPH. It is possiblehat a coagulation disorder, inherited or acquired, is respon-ible for many cases of PPH without a clear cause. A moreomprehensive understanding of the frequency of these dis-
rders in cases of PPH is needed since these conditions re-
uire special therapies. Randomized controlled trials of pro-hylactic medications such as misoprostol for PPH in high-isk patients are needed as well.

onclusionshe key to management of PPH is early recognition and treat-ent. The goals are to ultimately reduce maternal morbidity

nd mortality. Predicting the risk of PPH is an essential part ofrenatal screening and assessment on labor and delivery and

s heavily dependent on an accurate history. During thevent, a cause either related to tone, tissue, trauma, or throm-osis should be identified and documented. Evidence showshat a prior PPH increases the risk for recurrence, 8% to 28%or primary PPH and 19% for secondary PPH. Although thenderlying pathophysiology is often unclear, repeated PPHay be related to recurring identifiable risk factors. Prevention

f bleeding complications in the third stage centers on identifi-ation of these risk factors and preparation for recurrence.

eferences1. AbouZahr C: Global burden of maternal death and disability. Br Med

Bull 67:1-11, 20032. Combs CA, Murphy EL, Laros RK: Factors associated with postpartum

hemorrhage with vaginal birth. Obstet Gynecol 77:69-76, 19913. Doran JR, O’Brien SA, Randall JH: Repeated postpartum hemorrhage.

Obstet Gynecol 5:186-192, 19554. Jansen AJG, van Rhenen DJ, Steegers EAP, et al: Postpartum hemor-

rhage and transfusion of blood and blood components. Obstet GynecolSurv 60:663-671, 2005

5. Hoveyda F, MacKenzie IZ: Secondary postpartum haemorrhage: inci-dence, morbidity and current management. Br J Obstet Gynaecol 108:927-930, 2002

6. King PA, Duthie SJ, Dong ZG, et al: Secondary postpartum hemor-rhage. Aust NZ J Obstet Gynaecol 29:394-398, 1989

7. Sheiner E, Sarid L, Levy A, et al: Obstetric risk factors and outcome ofpregnancies complicated with early postpartum hemorrhage: a popu-lation-based study. J Matern Fetal Neo Med 18:149-154, 2005

8. Stones RW, Paterson CM, Saunders NJ: Risk factors for major obstetrichemorrhage. Eur J Obstet Gynecol Reprod Biol 48:15-18, 1993

9. Brinsden PR, Clark AD: Postpartum haemorrhage after induced andspontaneous labour. Br Med J 2:855-856, 1978

0. Tsu VD: Postpartum haemorrhage in Zimbabwe: a risk factor analysis.Br J Obstet Gynaecol 100:327-333, 1993

1. Hall MH, Halliwell R, Carr-Hill R: Concomitant and repeated happen-ings of complications of the third stage of labour. Br J Obstet Gynaecol92:732-738, 1985

2. Dewhurst CJ, Dutton WAW: Recurrent abnormalities of the third stageof labour. Lancet 273:764-767, 1957

3. Magann EF, Evans S, Hutchinson M, et al: Postpartum hemorrhageafter vaginal birth: an analysis of risk factors. Southern Med J 98:419-422, 2005

4. Mongelli M: Secondary post-partum haemorrhage: a recurrent condi-tion? Clin Exp Obstet Gynecol 19:97, 1992

5. Nizard J, Barrinque L, Frydman R, et al: Fertility and pregnancy out-comes following hypogastric artery ligation for severe post-partumhaemorrhage. Hum Reprod 18:844-848, 2003

6. Descargues G, Tinlot FM, Douvrin F, et al: Menses, fertility and preg-nancy after arterial embolization for the control of postpartum haem-orrhage. Hum Reprod 19:339-343, 2004

7. Stancato-Pasik A, Mitty H, Richard HM, et al: Obstetric embolotherapy:effect on menses and pregnancy. Radiology 204:791-793, 1997

8. Salomon LJ, De Tayrac R, Castaigne-Meary V, et al: Fertility and preg-nancy outcome following pelvic arterial embolization for severe post-
partum haemorrhage. A cohort study. Hum Reprod 18:849-852, 2003

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6

7


9. B-Lynch C, Coker A, Lawal AH, et al: The B-Lynch surgical techniquefor the control of massive postpartum haemorrhage: an alternative tohysterectomy? Five cases reported. Br J Obstet Gynaecol 104:372-375,1997

0. Api M, Api O, Yayla M: Fertility after B-lynch suture and hypogastricartery ligation. Fertil Steril 84:509, 2005

1. Goldberg J, Pereira L, Berghella V: Pregnancy after uterine artery em-bolization. Obstet Gynecol 100:869-872, 2002

2. Qidwai GI, Caughey AB, Jacoby AF: Obstetric outcomes in women withsonographically identified uterine leiomyomata. Obstet Gynecol 107:376-382, 2006

3. Hasan F, Arumugam K, Sivanesaratnam V: Uterine leiomyomata inpregnancy. Int J Gynecol Obstet 34:45-48, 1990

4. Winer-Muram HT, Muram D, Gilleison MS: Uterine myomas in preg-nancy. J Can Assoc Radiol 35:168-170, 1984

5. Muram D, Gillieson MS, Walters JH, et al: Myomas of the uterus inpregnancy: ultrasonographic follow-up. Am J Obstet Gynecol 138:16-19, 1980

6. Roberts WE, Fulp KS, Morrison JC, et al: The impact of leiomyomas onpregnancy. Aust NZ J Obstet Gynaecol 39:43-47, 1999

7. Vergani P, Ghidini A, Stobelt N, et al: Do uterine leiomyomas influencepregnancy outcome. Am J Perinatol 11:356-358, 1994

8. Lev-Toaff AS, Coleman BG, Arger PH, et al: Leiomyomas in pregnancy:a sonographic study. Radiology 164:375-380, 1987

9. Goldberg J, Pereira L, Berghella V, et al: Pregnancy outcomes aftertreatment for fibromyomata: uterine artery embolization versus laparo-scopic myomectomy. Am J Obstet Gynecol 191:18-21, 2004

0. Carpenter TT, Walker WJ: Pregnancy following uterine artery embo-lisation for symptomatic fibroids: a series of 26 completed pregnancies.Br J Obstet Gynaecol 112:321-325, 2005

1. Seracchioli R, Manuzzi L, Vianello F, et al: Obstetric and delivery out-come of pregnancies achieved after laparoscopic myomectomy. FertilSteril 86:159-165, 2006

2. Walker WJ, McDowell SJ: Pregnancy after uterine artery emoblizationfor leiomyomata: a series of 56 completed pregnancies. Am J ObstetGynecol 195:1266-1271, 2006

3. Pron G, Mocarski E, Bennett J, et al: Pregnancy after uterine arteryembolization for leiomyomata: the Ontario multicenter trial. ObstetGynecol 105:67-76, 2005

4. Tandberg A, Albrechtsen S, Iversen OE: Manual removal of the pla-centa: incidence and clinical significance. Acta Obstet Gynecol Scand78:33-36, 1999

5. VanBeekhuizen ME, Vierhout ME: Risk of recurrence of retained pla-centa. Med Tijdschr Geneeskd 138:2149-2152, 1994

6. Titiz H, Wallace A, Voaklander DC: Manual removal of the placenta: acase control study. Aust NZ J Obstet Gynaecol 41:41-44, 2001

7. Platt LD, Druzin ML: Acute puerperal inversion of the uterus. Am JObstet Gynecol 141:187-190, 1981

8. Watson P, Besch N, Bowes WA: Management of acute and subacutepuerperal inversion of the uterus. Obstet Gynecol 55:12-16, 1980

9. Miller NF: Pregnancy following inversion of the uterus. Am J ObstetGynecol 13:307-322, 1927

0. Baskett TF: Acute uterine inversion: a review of 40 cases. J ObstetGynaecol Can 24:953-956, 2002

1. Ramsahoye BH, Davies SV, Dasani H, et al: Obstetric management invon Willebrand’s disease: a report of 24 pregnancies and a review of theliterature. Haemophilia 1:140-144, 1995

2. Kadir RA, Lee CA, Sabin CA, et al: Pregnancy in women with vonWillebrand’s disease or factor XI deficiency. Br J Obstet Gynaecol 105:314-321, 1998

3. Greer IA, Lowe GDO, Walker JJ, et al: Haemorrhagic problems inobstetrics and gynaecology in patients with congenital coagulopathies.Br J Obstet Gynecol 98:909-918, 1991

4. Kirtava A, Crudder S, Dilley A, et al: Trends in clinical management ofwomen with von Willebrand disease: a survey of 75 women enrolled inhaemophilia treatment centres in the United States. Haemophilia 10:158-161, 2004

5. Kujovich JL: Von Willebrand disease and pregnancy. J Thromb Hae-
most 3:246-253, 2005
6. Kouides PA, Phatak PD, Burkart P, et al: Gynaecological and obstetricalmorbidity in women with type I von Willebrand disease: results of apatient survey. Haemophilia 6:643-648, 2000

7. Ragni MV, Bontempo FA, Hassett AC: von Willebrand disease andbleeding in women. Haemophilia 5:313-317, 1999

8. Hundegger R, Husslein P, Berghammer P, et al: Postpartum bleedingand von Willebrand’s disease. Arch Gynecol Obstet 266:160-162,2002

9. American College of Obstetricians and Gynecologists CommitteeOpinion #263. Von Willebrand’s disease in gynecologic practice, 2001

0. American College of Obstetricians and Gynecologists CommitteeOpinion #329. Menstruation in girls and adolescents. Using the men-strual cycle as a vital sign, 2006

1. Girolami A, Randi ML, Ruzzon E, et al: Pregnancy and oral contracep-tives in congenital bleeding disorders of the vitamin K-dependent co-agulation factors. Acta Haematol 115:58-63, 2006

2. Kadir RA, O’Connell NM, Economides DL, et al: Screening for factor XIdeficiency amongst pregnant women of Ashkenazi Jewish origin. Hae-mophilia 12:625-628, 2006

3. Shpilherg O, Peretz H, Zivelin R, et al: One of the two common muta-tions causing factor XI deficiency in Ashkenazi Jews (type 11) is alsoprevalent in Iraqi Jews, who represent the ancient gene pool of Jews.Blood 85:429-432, 1995

4. Bolton-Maggs PHB, Wan-Yin BY, McCraw AH, et al: Inheritance andbleeding in factor XI deficiency. Br J Haemotol 69:521-528, 1988

5. Kadir RA, Economides DL, Braithwaite J, et al: The obstetric expe-rience of carriers of haemophilia. Br J Obstet Gynaecol 104:803-810, 1997

6. Maslovitz S, Many A, Landsberg JA, et al: The safety of low molecularweight heparin therapy during labor. J Matern Fetal Neonat Med 17:39-43, 2005

7. Rowan JA, McLintock C, Taylor RS, et al: Prophylactic and therapeuticenoxaparin during pregnancy: Indications, outcomes, and monitoring.Aust NZ J Obstet Gynecol 43:123-128, 2003

8. Borna S, Borna H, Zhazardoost S: Maternal and neonatal outcomes inpregnant women with immune thrombocytopenic purpura. Arch Ira-nian Med 9:115-118, 2006

9. Hwa HL, Wang TR, Huang SC, et al: Maternal and fetal outcome ofpregnant women with idiopathic thrombocytopenic purpura: retro-spective analysis of 25 cases. J Formos Med Assoc 92:957-961, 1993

0. Lind J, Wallenburg HCS: Pregnancy and the Ehlers-Danlos syndrome:a retrospective study in a Dutch population. Acta Obstet Gynecol Scand81:293-300, 2002

1. Volkov N, Nisenblat V, Ohel G, et al: Ehlers-Danlos syndrome: insightson obstetric aspects. Obstet Gynecol Surv 62:51-57, 2007

2. Rosnes JS, Sharkey MF, Veille JC, et al: Gaucher’s disease in pregnancy.Obstet Gynecol Surv 51:549-558, 1996

3. Young BK, Gorstein F: Maternal osteogenesis imperfecta. Obstet Gy-necol 31:461-470, 1968

4. Rahman J, al-Suleiman SA, Rahman MS, et al: Obstetric and gyneco-logic complications in women with Marfan syndrome. J Reprod Med48:723-728, 2004

5. Meijboom L, Drenthen W, Pieper PG, et al: Obstetric complications inMarfan syndrome. Int J Card 110:53-59, 2006

6. Jackson KW, Allbert JR, Schemmer GK, et al: A randomized controlledtrial comparing oxytocin administration before and after placental de-livery in the prevention of postpartum hemorrhage. Am J Obstet Gy-necol 185:873-877, 2001

7. McCurdy C, Magann E, McCurdy C, et al: The effect of placental man-agement at cesarean delivery on operative blood loss. Am J ObstetGynecol 167:1363, 1992

8. American College of Obstetricians and Gynecologists Practice Bulletin#76. Postpartum hemorrhage, 2006

9. VanSelm M, Kanhai HHH, Keirse M: Preventing the recurrence ofatonic postpartum hemorrhage: a double-blind trial. Acta Obstet Gy-necol Scand 74:270-274, 1995

0. Newton M, Mosey LM, Egli GE, et al: Blood loss during and immedi-
ately after delivery. Obstet Gynecol 17:9-18, 1961

7

7

7

7


1. McVay PA, Hoag RW, Hoag MS, et al: Safety and use of autologousblood donation during the third trimester of pregnancy. Am J ObstetGynecol 160:1479-1488, 1989

2. Herbert WN, Owen HG, Collins ML: Autologous blood storage in
obstetrics. Obstet Gynecol 72:166-170, 1988
3. Andres AL, Piacquadio KM, Resnik R, et al: A reappraisal of the need forautologous blood donation in the obstetric patient. Am J Obstet Gy-necol 163:1551-1552, 1990

4. Combs CA, Murphy EL, Laros RK: Cost-benefit analysis of autologous
blood donation in obstetrics. Obstet Gynecol 80:621-625, 1992

TPAH

Nlowrhcp4dPhs

D

A

0d

hromboembolism inregnancy: Recurrence and Its Prevention

ndra H. James, MD, MPH, Chad A. Grotegut, MD, Leo R. Brancazio, MD, andaywood Brown, MD

Fifteen to 25% of thromboembolic events in pregnancy are recurrent events. Women witha history of thrombosis have a three- to fourfold increased risk of recurrence when they arepregnant compared with when they are not. The risks are even higher postpartum. The rateof recurrent venous thromboembolic events without anticoagulation is 2.4% to 12.2%,whereas the rate with anticoagulation is 0% to 2.4%. Because the rates of recurrentthromboembolism can be reduced with anticoagulation, women with a history of thrombo-sis who are not on lifelong anticoagulation will likely require anticoagulation duringpregnancy, or at least during the postpartum period. Women who are already on lifelongwarfarin for the prevention of recurrent venous thromboembolism should be counseledabout the teratogenic effects of warfarin and offered the opportunity to be converted toheparin before conception. During pregnancy, low-molecular-weight heparin, with fewerside effects and a longer half-life, is generally preferred over unfractionated heparin.Unfractionated heparin with its shorter half-life is generally preferred around the time ofdelivery. Women on antiplatelet medication for prevention of arterial thromboembolismmay be converted to low-dose aspirin after conception and supplemented with low-doseheparin or low-molecular-weight heparin during pregnancy. Because current recommen-dations rely on case series and expert opinion, additional studies including randomizedtrials might enhance our ability to prevent recurrent thromboembolism in pregnancy.Semin Perinatol 31:167-175 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS recurrent thromboembolism, venous thromboembolism, arterial thromboembo-lism, deep vein thrombosis, pulmonary embolus, pregnancy

bT1emivih

ERDTTt

ormal pregnancy is accompanied by an increase in clot-ting factors.1,2 The resulting hypercoagulable state has

ikely evolved to protect women from hemorrhage at the timef miscarriage and childbirth. Indeed, in the developingorld, the leading cause of maternal death is still hemor-

hage.3 In Western Europe and the United States, whereemorrhage is successfully treated or prevented, the leadingause of maternal death is thromboembolic disease.4 Duringregnancy, the risk of venous thromboembolism is increased-fold,5 and the risk of arterial thromboembolism, myocar-ial infarction, and stroke is also increased 3- to 4-fold.6,7

ostpartum, the risk of venous thromboembolism is 20-foldigher,5 and the risk of arterial thromboembolism (stroke) isimilarly elevated.8

ivision of Maternal–Fetal Medicine, Department of Obstetrics and Gyne-cology, Duke University Medical Center, Durham, NC.

ddress reprint requests to Andra H. James, MD, MPH, Box 3967 DUMC,
(Durham, NC 27710. E-mail: [email protected]

The overall number of maternal deaths from thromboem-olism in pregnancy is approximately 3 per 1000 deliveries.wo per 100,000 are due to arterial thrombosis and 1 per00,000 are due to venous thrombosis.6,7,9 Although arterialvents account for more deaths, venous events are more com-on, accounting for 4 out of 5 thromboembolic events dur-

ng pregnancy.6,7,9 Eighty percent of these events are deepein thrombosis,9 and 20% are pulmonary emboli.9 The mostmportant risk factor for thrombosis during pregnancy is aistory of thrombosis.6,7,9,10

pidemiology ofecurrent Thrombosis

uring pregnancy, the risk of thromboembolism is increased.his is especially true for women with a history of thrombosis.he risk of recurrent venous thromboembolism in pregnancy is

hree- to fourfold higher compared with women not pregnant
RR 3.5; 95% CI 1.6, 7.8),11 and the risk of a recurrent arterial
167

tf1

rahai(4wmveiw

err1or

RTAtta

vaSnpsnscecraT(wel8ip(

wattdrip

TA

BT

SB

LP

D

*†‡§

¶#

168 A.H. James et al.

hromboembolism is probably similarly increased. One studyound the relative risk of recurrent stroke to be 2.2 (95% CI 0.5,7.5) during pregnancy and 9.7 (95% CI 1.2, 78.9) postpartum.12

Fifteen to 25% of thromboembolic events in pregnancy areecurrent events. There are few data about the risk of recurrentrterial events in pregnancy, but a study of 115 women with aistory of stroke (89 with a history of arterial ischemic strokend 26 with a history of cerebral vein thrombosis) found thencidence of recurrent stroke during pregnancy to be 1.8%0.5%, 7.5%) and during the first 6 weeks postpartum to be.8% (0.6%, 35.9%). During pregnancy, most of the 115omen with a history of arterial thrombosis received antiplateletedication, and most of the women with a history of cerebral

ein thrombosis received no prophylaxis.12 Sanson and cowork-rs found no recurrence of arterial thromboembolic events dur-ng pregnancy among 8 women who received anticoagulationith low-molecular-weight heparin.There are several studies about the risk of recurrent venous

vents in pregnancy (Table 1). In recent studies, the rate ofecurrent venous thromboembolism in women who did noteceive anticoagulation has been reported to range from 2.4% to2.2%.13-15 In women who did receive anticoagulation, the ratef recurrent venous thromboembolism has been reported toange from 0% to 2.4%.13,16,17

isk Factors for Recurrenthrombosis in Pregnancy

lthough there are no data on the risk factors for recurrenthrombosis in pregnancy, the risk factors are thought to behe same as the overall risk factors for thrombosis. Aside from

able 1 Rates of Recurrent Venous Thromboembolism (VTE)nticoagulation*

StudyRate of Recwith Antic

adaracco and Vessey, 197491 6/30 (2engborn et al., 198992

Pregnancy 3/20 (1Delivery and puerperium 2/57 (5

anson et al., 199916 3/149 (rill-Edwards et al., 200013

PregnancyPostpartum

epercq et al., 200117 3/125 (abinger et al., 200514 7/574 (

e Stefano et al., 200615 0/87 (0

When available, the rate in pregnancy compared to postpartum isAnticoagulation consisted of heparin 5000 units subcutaneously twCompiled from case series. Anticoagulation consisted of varying dTwo of these events occurred in women who were not taking w

anticoagulation was discontinued at 6 weeks postpartum.357/574 (65%) received 20 mg of enoxaparin once daily and 217/Anticoagulation consisted of heparin 5000 units subcutaneously

anti-factor X units/day.

history of thrombosis, the most important risk factor for i

enous thromboembolism is thrombophilia.9,10 From annalysis of 14,335 records from the Nationwide Inpatientample,9 other medical conditions that were statistically sig-ificant risk factors for venous thromboembolism duringregnancy were heart disease, sickle cell disease, lupus, obe-ity, anemia, diabetes, hypertension, and smoking. Preg-ancy and delivery complications that were associated with aignificantly increased risk of venous thromboembolism in-luded multiple gestation, hyperemesis, disorders of fluid,lectrolyte and acid-base balance, antepartum hemorrhage,esarean delivery, postpartum infection, postpartum hemor-hage, and transfusion (Table 2). In the same analysis, agend race were also risk factors for venous thromboembolism.he odds ratio (OR) for women age 35 and older was 2.12.0, 2.3).9 When controlled for age, the OR for black womenas still 1.4 (1.2, 1.6).9 The risk factors for arterial thrombo-

mbolism were similar to those for venous thromboembo-ism, except for stroke, where hypertension OR � 6.1 (4.5,.1) and thrombocytopenia OR � 6.0 (1.5, 24.1) were more

mportant risk factors and myocardial infarction, where hy-ertension OR � 11.7 (6.9, 21.2) and smoking OR � 6.24.1, 9.5) were more important risk factors.

Thrombophilia is present in 20%18 to 40%15,19 of womenho experience venous thromboembolism during pregnancy

nd the postpartum period. Both acquired and inheritedhrombophilia increase the risk, but it is not clear whetherhrombophilia increases the risk of recurrent venous eventsuring pregnancy and the postpartum period. In a systematiceview of the risk of recurrent venous thromboembolism inndividuals who were not pregnant, factor V Leiden and therothrombin gene mutation were found to confer a slightly

g Pregnancy and the Postpartum period with and without

t VTElation

Rate of Recurrent VTE withoutAnticoagulation

-

8/67 (12%)3/34 (11%)

‡ -

3/125 (2.4%)(95% CI: 0.2%, 6.9%)

§ -¶ 8/197 (6.9%)

(95% CI: 1.6%, 10.6%)19/155 pregnancies (12.2%)

(95% CI: 7.9%, 18.3%)

ed.ily.f varying low-molecular-weight heparins.as prescribed and the third occurred in a woman 2 weeks after

5%) received 40 mg of enoxaparin once daily.or three times daily or low-molecular-weight heparin 4000-5000

durin

urrenoagu

0%)†

5%)%)1.5%)

-

2.4%)1.2%)

%)#

specifiice da

oses oarfarin

574 (3two

ncreased risk of recurrence with an OR for factor V Leiden of

11drdcfifbfmanr

wradcbnobe4tabenii

EwIbpomabiapcbmnvsa

cocgccsWwtet

TT

M

C

Thromboembolism in pregnancy 169

.41 (1.14, 1.75) and for the prothrombin gene mutation of

.72 (1.27, 2.31).20 In a separate study, protein C and Seficiencies were found to confer only a slightly increasedisk of recurrence [OR 1.4 (0.9, 2.2)],21 but antithrombineficiency was found to confer approximately a twofold in-reased risk of recurrence [OR 1.9 (1.0, 3.9)].21 This con-rmed the findings of Christiansen and coworkers who alsoound a twofold increased risk of recurrence with antithrom-in deficiency [OR 1.8 (0.9, 3.7)].22 Although homozygosityor factor V Leiden, homozygosity for the prothrombin gene

utation, and combined heterozygosity for both factors aressociated with a high risk of venous thromboembolism, it isot clear whether they also confer an increased risk of recur-ent thrombosis. Results are conflicting.22-24

Brill-Edwards and coworkers13 and Pabinger and co-orkers found that thrombophilia conferred an increased

isk of recurrent venous thromboembolism in pregnancynd the postpartum period, but De Stefano and coworkersid not.15 Nonetheless, any thrombophilic condition thatonfers an increased risk of primary venous thromboem-olism above the background risk of recurrence in preg-ancy of 2.4% to 12.2% will likely confer an increased riskf recurrent thrombosis. This would be true for antithrom-in deficiency where the absolute risk of venous thrombo-mbolism in pregnancy has been reported to be as high as0% to 68%,25-27 the antiphospholipid syndrome wherehe absolute risk has been reported to be as high as 30%,28

nd homozygosity for factor V Leiden where the risk haseen estimated to be 32%.29 Homozygosity, let alone het-rozygosity, for the MTHFR C677T polymorphism doesot confer any increased risk of venous thromboembolism

n pregnancy29 and, therefore, would not be expected to

able 2 Medical Conditions and Complications of Pregnanhromboembolism in Pregnancy9

Risk Factor

edical conditionsHeart diseaseSickle cell diseaseLupusObesityAnemiaDiabetesHypertensionSmokingomplications of pregnancy and deliveryMultiple gestationHyperemesisFluid and electrolyte imbalanceAntepartum hemorrhageCesarean deliveryPostpartum infectionPostpartum hemorrhageTransfusion

ncrease the risk of recurrent events. p

valuation of the Womanith a History of Thrombosis

deally, evaluation of the woman with a history of throm-osis should occur before conception, or at least early inregnancy. At the time of the evaluation, the clinicianbtains a history, reviewing any pertinent discharge sum-aries, reports of imaging studies and laboratory reports,

nd makes recommendations. In most cases, women cane counseled that, although their history puts them at an

ncreased risk for recurrent thrombosis and, if they haven underlying hypercoagulable state, an increased risk ofoor pregnancy outcome, these risks are manageable andan be reduced with anticoagulation. Women with throm-oembolic conditions that place them at a high risk ofaternal mortality may be best served by avoiding preg-ancy. These conditions include mechanical heartalves,30 chronic thromboembolic pulmonary hyperten-ion, a history of recurrent thrombosis while fully antico-gulated, and a history of myocardial infarction.

Usually, women with a history of thrombosis will re-eive anticoagulation during pregnancy. Women who aren lifelong anticoagulation will be continued on full anti-oagulation. They should be counseled about the terato-enic effects of warfarin and offered the opportunity to beonverted to low-molecular-weight heparin before con-eption. This can be done in collaboration with the phy-ician who prescribes the patient’s anticoagulation.

omen who have not had a complete thrombophiliaorkup should be offered appropriate testing. Whereas

he results of thrombophilia testing will not alter the gen-ral recommendation for anticoagulation in pregnancy,he results may alter the intensity of anticoagulation that is

d Delivery Associated with an Increased Risk of Venous

dds Ratio Confidence Interval

7.1 6.2, 8.36.7 4.4, 10.18.7 5.8, 13.04.4 3.4, 5.72.6 2.2, 2.92.0 1.4, 2.71.8 1.4, 2.31.7 1.4, 2.1

1.6 1.2, 2.12.5 2.0, 3.24.9 4.1, 5.92.3 1.8, 2.82.1 1.8, 2.44.1 2.9, 5.71.3 1.1, 1.67.6 6.2, 9.4

cy an

O

rescribed.

TTttpLpilompwodvCmsnptmtc4ns

apatmo

RTwPiaapo

ret

bfu

OAdUmcttbbti8rcapbrr

hlcttraAplk

nbhqrparedwcfw


esting for Thrombophiliaesting should include assays for both antiphospholipid an-

ibodies and inherited thrombophilia. Tests for inheritedhrombophilia should include assays for antithrombin (III),rotein C and protein S deficiency, as well as tests for factor Veiden and the prothrombin gene G20210A mutation. Sincerotein S activity levels fall during pregnancy, a cutoff of 35%

s suggested.31,32 Antithrombin (III)33-38 and protein C39-42

evels do not change significantly during pregnancy, so lab-ratory reference values can be used. Hyperhomocysteine-ia is associated with thrombosis and possibly with poorregnancy outcome. Treatment of hyperhomocysteinemia isith folic acid, vitamin B12, and vitamin B6.32 In the absencef adequate folate levels, the methylene tetrahydrofolate re-uctase (MTHFR) C677T polymorphism can result in ele-ated homocysteine levels,43 but testing for the MTHFR677T polymorphism is not recommended, since even ho-ozygosity for the polymorphism does not appear to be as-

ociated with an increased risk of thrombosis or poor preg-ancy outcome.29 There is insufficient information on otherolymorphisms for genes in the folate metabolism pathwayo make specific testing recommendations. Even if the poly-orphism were present, the treatment for hyperhomocys-

einemia is folic acid, vitamin B12, and vitamin B6,32 whichan be adequately supplemented with prenatal vitamins. TheG/5G PAI-1 polymorphism was associated with poor preg-ancy outcome in one retrospective study,44 but has not beentudied sufficiently to be included in testing.

Tests for antiphospholipid antibodies include the lupusnticoagulant, anticardiolipin antibodies, and anti-�2 glyco-rotein I antibodies. The addition of anti-�2 glycoprotein Intibodies to the lupus anticoagulant and anticardiolipin an-ibodies is recommended in an international consensus state-ent to increase the sensitivity of the assays for the diagnosis

f the antiphospholipid syndrome.45

In summary, the recommended tests are:

● Lupus anticoagulant,● Anticardiolipin antibodies,● Anti-�2 glycoprotein I antibodies,● Activated protein C resistance with factor V Leiden if

abnormal,● Prothrombin G20210A polymorphism,● Protein C,● Protein S, and● Antithrombin (III).

ationale forhromboprophylaxis in Womenith a History of Thrombosis

regnancy increases the risk of thrombosis fourfold,5 includ-ng women who have a history of thrombosis.11 Women whore not on lifelong anticoagulation will likely require antico-gulation during pregnancy, or, at a minimum, during theostpartum period. Although there are no randomized trials
f anticoagulation versus placebo for the prevention of recur- b
ent thrombosis in pregnancy, the risk of recurrent thrombo-mbolism is significantly reduced in women who receivehromboprophylaxis.13,14,16,17

Women with a history of recurrent thrombosis will likelye on lifelong anticoagulation and, because of warfarin’s ef-ects on the fetus, will require conversion from warfarin tonfractionated or low-molecular-weight heparin.

ptions fornticoagulationuring Pregnancynique aspects of anticoagulation in pregnancy include bothaternal and fetal issues. During pregnancy, there is an in-

rease in blood volume of 40% to 50%46 and an increase inhe volume of distribution. An increase in glomerular filtra-ion46 results in increased renal excretion of drugs eliminatedy this route. Additionally, there is an increase in proteininding of heparin. Since 45% of pregnancies are unin-ended,47 many women do not realize they are pregnant dur-ng the critical period for organogenesis, which is the 4th toth weeks after conception.48 Warfarin taken during this pe-iod is associated with a 14.6% to 56% reported risk of mis-arriage,49-55 and carries up to a 30% risk of congenital anom-lies.49-51,53,55-58 Placental transfer of warfarin later inregnancy can result in fetal bleeding58,59 or still-irth.49,51,53,54 Long-term sequelae include a 14% reportedisk of adverse neurological outcome60 and a 4% reportedisk of low intelligence quotient (IQ).60

The preferred agents for anticoagulation in pregnancy areeparin compounds.61 Neither heparin62-64 nor low-molecu-

ar-weight heparin63,64 crosses the placenta, and both areonsidered safe in pregnancy.17,65 Disadvantages of unfrac-ionated heparin include the necessity of parenteral adminis-ration, a 2% risk of major bleeding,65 a 2% to 36% risk ofeduced bone density,66-68 a 2% risk of vertebral fracture,69

nd a risk of heparin-induced thrombocytopenia (HIT).61

lthough the risk of HIT, a life- and limb-threatening com-lication of heparin therapy, is low in pregnancy, and may be

ower than in nonpregnant patients,70 the actual risk is un-nown.61

There are few comparative studies in pregnancy, but inonpregnant patients, low-molecular-weight heparin haseen associated with fewer side effects than unfractionatedeparin.61 Although parenteral administration is still re-uired, potential advantages of low-molecular-weight hepa-in over unfractionated heparin are less bleeding, a moreredictable response, a lower risk of HIT, a longer half-life,nd maybe less bone loss.61 However, in a recently completedandomized trial of low-dose unfractionated heparin versusnoxaparin for thromboprophylaxis in pregnancy, the inci-ence of clinically significant bone loss was 2% to 2.5% andas no different in women who took unfractionated heparin

ompared with those who took enoxaparin.68 Another studyound that bone loss in women who took low-molecular-eight heparin was approximately 4%, no different than
one loss in controls.71 A definite advantage of enoxaparin,

afdmt

tuottmsdboctdbit

PiwPltnddmcper

trwsptircw

timwc8mi

IdNtprflCacTtt

f

T

U

L

g qd


nd probably other low-molecular-weight heparins, over un-ractionated heparin is less bruising at injection sites.72 Aisadvantage of low-molecular-weight heparin is that it isore expensive. Also, its longer half-life may be a problem at

he time of delivery.Long-term therapy with fondaparinux, a new selective fac-

or Xa inhibitor, may result in less bone loss than eithernfractionated or low-molecular-weight heparin,73 but datan the use of fondaparinux in pregnancy are limited. Al-hough Lagrange and coworkers74 observed no transplacen-al passage of fondaparinux using a perfused cotyledonodel, Dempfle and coworkers75 found transplacental pas-

age of fondaparinux in five women who took it for 1 to 101ays because of heparin allergy. Antifactor Xa levels in um-ilical cord plasma of newborns, however, were found to benly one-tenth the concentration of maternal plasma, a con-entration well below that required for effective anticoagula-ion.75 Nonetheless, at the present time, there are insufficientata to justify the routine use of fondaparinux in pregnancy,ut fondaparinux is probably still the anticoagulant of choice

n cases of severe cutaneous allergies or heparin-inducedhrombocytopenia in pregnancy.75,76

rotocols for Anticoagulationn Pregnancy for Womenith a History of Thrombosis

rotocols for prevention of recurrent venous thromboembo-ism are presented in Table 3. Full-dose (adjusted dose) an-icoagulation is recommended61,77 for women with either aeed for life-long anticoagulation or antiphospholipid syn-rome with a history of thrombosis. Full-dose (adjustedose) or an intermediate or moderate dose is recom-ended61,77 for women with either antithrombin (III) defi-

iency or homozygosity for the factor V Leiden mutation, therothrombin gene G20210A mutation, or compound het-rozygosity for both mutations. Low-dose anticoagulation is

able 3 Protocols for Thromboprophylaxis in Pregnancy61,93

Bates et al

nfractionated HeparinMini-dose “low dose” 5000 U sc q 12 hrs

Moderate-dose “low dose” q 12 hrs to target anti-0.1-0.3 U/ml

Adjusted dose “full dose” q12 hrs to target mid-itherapeutic range

ow-Molecular-Weight HeparinProphylactic dose “low dose” Enoxaparin 40 mg qd

Dalteparin 5000 U qdTinzaparin 4500 U qd

Weight-adjusted dose “full dose” Enoxaparin 1 mg/kg biDalteparin 100 U/kg q

1 24 hrsTinzaparin 175 mg U/k

ecommended32,61 for women with a history of unprovoked n

hrombosis. Although in one series women with transientisk factors had a rate of recurrence similar to that of otheromen with a history of thrombosis,14 close observation (as-

essment of signs and symptoms of thrombosis at routinerenatal visits) may be an option for women with a history ofhrombosis in the setting of transient risk factors such asnjury or immobility.61 Nonetheless, if these women do noteceive anticoagulation during pregnancy, they should beonsidered for thromboprophylaxis during the first 2 to 6eeks postpartum.There are few protocols for thromboprophylaxis of arterial

hromboembolism. Currently, there is no consensus regard-ng the best protocol for women with a history of stroke78 or

yocardial infarction. Outside of pregnancy, most of theseomen are on an antiplatelet agent. It is our practice to dis-

ontinue full-dose antiplatelet agents and prescribe aspirin1 mg per day along with low-dose unfractionated or low-olecular-weight heparin to counteract the hypercoagulabil-

ty of pregnancy.

nitiating Anticoagulationuring Pregnancyormal pregnancy is accompanied by increased concentra-

ions of factors VII, VIII, X, and von Willebrand factor and byronounced increases in fibrinogen.1 Factors II, V, and IX areelatively unchanged.1 Free protein S, the active, unboundorm, is decreased during pregnancy secondary to increasedevels of its binding protein, the complement component4b.1 Plasminogen activator inhibitor type 1 (PAI-1) levelsre increased fivefold.1 Levels of PAI-2, produced by the pla-enta, increase dramatically during the third trimester.2

hese changes, which may not return to baseline until morehan 8 weeks postpartum,1 begin with conception. So doeshe risk of thrombosis.10,14,79

Women on lifelong anticoagulation should be convertedrom warfarin to low-molecular-weight heparin before preg-

461 Duke Protocol93

5000 U sc q 12hrs <8 weeks7500 U sc q 12 hrs 8-28 weeks10,000 U sc q 12 hrs >28 weeks

Xa level of q 8 or12 hrs to target mid-interval aPTT intherapeutic range

l aPTT in

Enoxaparin 40 mg qd or 30 mg bid before28 weeks then

Enoxaparin 40 mg bid after 28 weeks.5 mg/kg qds or 200 U/kg

Enoxaparin 1 mg/kg bid with target ofantifactor Xa level of 0.5-1.0

., 200

factor

nterva

d or 112 hr

ancy or as soon as possible after conception. The problem

wdrc3tcwio

csbast

etptsta

MaWaoudwtrwow1aawwpswthhfbpfbg

iiib

PPpatahhumfwnmwcdimow

fufiuharo

lctpcha

FHelstaTf


ith conversion before pregnancy is the inconvenience andiscomfort of parenteral administration of heparins and theisks associated with their long-term use. The problem withonversion after conception is that the half-life of warfarin is6 to 42 hours80 and it may remain in the maternal circula-ion for several days, increasing the risk of miscarriage andongenital anomalies. Only a few women are candidates forarfarin rather than heparins during pregnancy. Candidates

nclude women with mechanical heart valves30 and certainther unusual conditions.Women who are not on lifelong anticoagulation, but are

andidates for thromboprophylaxis in pregnancy, shouldtart soon after conception. An exception is women who wille undergoing ovulation induction. Because hormone ther-py, including clomiphene, increases the risk of thrombo-is,81 these women should begin anticoagulation at the timehey start ovulation induction.

Because HIT manifests within the first 5 to 15 days ofxposure to heparins,82 platelet counts may be monitored forhe first 2 to 3 weeks after initiation of therapy. Althoughlatelet counts usually drop by 10% in pregnancy83 andhrombocytopenia affects up to 10% of all pregnancies,83 HITtill must be considered in women who develop thrombocy-openia after starting heparin or low-molecular-weight hep-rin.

anagementt the Time of Deliveryomen are often converted from low-molecular-weight hep-

rin to unfractionated heparin at 36 to 37 weeks of gestationr sooner if there is preterm labor, preeclampsia, fetal (intra-terine) growth restriction, oligohydramnios, or other evi-ence of imminent delivery. The purpose of convertingomen to the shorter-acting unfractionated heparin has less

o do with any risk of bleeding at the time of delivery, butather the rare possibility of an epidural or spinal hematomaith regional anesthesia.84 Due to this possibility, anesthesi-logists usually will not place a regional anesthetic if aoman has received low-molecular-weight heparin within2 to 24 hours. Because of the benefits of regional analgesiand anesthesia over other analgesia and anesthesia for labornd delivery,84 every effort should be made to ensure that aoman has not received low-molecular-weight heparinithin 12 to 24 hours of needing a regional anesthetic. De-ending on the risk of thrombosis, unfractionated heparinhould be held for 6 to 24 hours before delivery. Should aoman go into labor while taking unfractionated heparin,

he heparin will usually clear within 6 hours. Reversal ofeparin is rarely required and is not indicated for low-doseeparin. Although the use of pneumatic compression devicesor the prevention of pregnancy-related thrombosis has noteen studied, extrapolating from perioperative data,85 thelacement of pneumatic compression devices in labor or be-ore cesarean delivery is recommended. In the woman who iseing aggressively anticoagulated, consideration should be
iven to entering the abdomen through a vertical midline s
ncision at the time of cesarean delivery. A vertical midlinencision, which is made in a watershed area and does notnvolve dissection of the rectus muscles off of the fascia, maye associated with less bleeding.

ostpartum Managementneumatic compression devices are left in place until theatient is ambulatory and until anticoagulation is restartedfter delivery. To minimize bleeding complications, resump-ion of anticoagulation should be postponed until 12 hoursfter vaginal delivery, 12 hours after epidural removal, or 24ours after cesarean delivery. After the risk of postpartumemorrhage has lessened (2 or more weeks after delivery),nless a woman prefers to remain on unfractionated or low-olecular-weight heparin, she may be bridged to warfarin

or the remainder of the 6-week postpartum period. Womenho have had a thrombotic event during the current preg-ancy should remain on warfarin for at least another 3 to 6onths after delivery. Women on lifelong anticoagulationill remain on warfarin indefinitely. Although warfarin is

ontraindicated during pregnancy, it is not contraindicateduring breastfeeding. In a study of the transfer of warfarin

nto breast milk, less than 25 ng of warfarin was detected perL.86 The American Academy of Pediatrics Committee

n Drugs supports breastfeeding for women who takearfarin.87

At the present time, since there are no data about whetherondaparinux enters breast milk, fondaparinux should not besed routinely in women who are breastfeeding, butondaparinux is an option for women who are not breastfeed-ng. Fondaparinux, a selective factor Xa inhibitor, may besed until a woman is bridged to warfarin. For women whoave to be fully anticoagulated for only 6 weeks postpartum,n advantage of fondaparinux over warfarin is that it does notequire monitoring. In addition, it requires only daily, aspposed to twice-daily, dosing.Women who are not breastfeeding and who will be on

ifelong anticoagulation should be allowed to take estrogen-ontaining contraceptives, but these are contraindicated forhe woman who will discontinue anticoagulation at 6 weeksostpartum. Although progestins may increase the levels ofertain coagulation factors, progestin-only contraceptivesave not been found to increase the risk of thrombosis andre, therefore, generally allowed.88

etal Surveillanceypercoagulability, as manifested by a history of thrombo-

mbolism, may alter normal hemostasis and increase the like-ihood of poor pregnancy outcome, including abruption,tillbirth, fetal growth restriction, and preeclampsia.29 Al-hough anticoagulation may improve pregnancy outcome,89

plan for fetal surveillance should still be implemented.90

his would include periodic ultrasounds for fetal growth,etal testing in the last month or two of pregnancy, and con-
ideration of delivery by 39 weeks gestation.

IFFtfwdgfbfw

KaAirlhlaaf

STithrthstaeWncthhcomb

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2


mplications foramily Members

amily members should be informed about any inheritedhrombophilia and what the specific condition is (such asactor V Leiden or protein S deficiency). Family membersho inherit the same genes may be at an increased risk ofeveloping venous thromboembolism. Obviously, first-de-ree relatives are at greater risk compared with extendedamily members, but extended family members may alsoenefit from the information. Family members should in-orm their health care providers and together they can decidehether testing for thrombophilia is indicated.

nowledge Gapsnd Future Researchlthough there are several case series, there are no random-

zed trials of thromboprophylaxis for the prevention of recur-ent venous thromboembolism in pregnant women. Guide-ines rely on expert opinion. Randomized trials would beelpful. As pointed out, even though arterial thromboembo-

ism may result in more deaths, there is even less informationbout thromboprophylaxis for the prevention of recurrentrterial thromboembolism. Even case series would be help-ul. Since these events are rare, registries may be required.

ummaryhe risk of recurrent venous and arterial thromboembolism

s high during pregnancy. The rate of recurrent venoushromboembolism in women not receiving anticoagulationas been reported to range from 2.4% to 12.2%. In womeneceiving anticoagulation, the rate of recurrent venoushromboembolism ranges from 0% to 2.4%. Women with aistory of arterial thromboembolism probably experienceimilar reductions in the rate of recurrence with anticoagula-ion. Therefore, anticoagulation with one of the various hep-rin compounds is generally recommended in pregnancyven for women who are not on lifelong anticoagulation.

omen who are taking warfarin can be offered the opportu-ity to be converted to low-molecular-weight heparin beforeonception. To minimize bleeding complications, resump-ion of anticoagulation should be postponed until 12 to 24ours after delivery. After the risk of postpartum hemorrhageas subsided, women may be bridged to warfarin. Becauseurrent recommendations rely on case series and expertpinion, additional studies including randomized trialsight enhance our ability to prevent recurrent thromboem-

olism in pregnancy.

eferences1. Bremme KA: Haemostatic changes in pregnancy. Best Pract Res Clin

Haematol 16:153-168, 20032. Medcalf RL, Stasinopoulos SJ: The undecided serpin. The ins and outs

of plasminogen activator inhibitor type 2. FEBS J 272:4858-4867, 20053. Postpartum hemorrhage. From the Prevention of Postpartum Hemor-

rhage Initiative Web Site. Office of Health, Infectious Diseases and

Nutrition, Bureau for Global Health, USAID. Available at:http://www.pphprevention.org/pph.php. Last accessed January 28,2007.

4. Chang J, Elam-Evans LD, Berg CJ, et al: Pregnancy-related mortalitysurveillance: United States, 1991–1999. MMWR Surv Summ 52:1-8,2003

5. Heit JA, Kobbervig CE, James AH, et al: Trends in the incidence ofvenous thromboembolism during pregnancy or postpartum: a 30-yearpopulation-based study. Ann Intern Med 143:697-706, 2005

6. James AH, Bushnell CD, Jamison MG, et al: Incidence and risk factorsfor stroke in pregnancy and the puerperium. Obstet Gynecol 106:509-516, 2005

7. James AH, Jamison MG, Biswas MS, et al: Acute myocardial infarctionin pregnancy: a United States population-based study. Circulation 113:1564-1571, 2006

8. Kittner SJ, Stern BJ, Feeser BR, et al: Pregnancy and the risk of stroke.N Engl J Med 335:768-774, 1996

9. James AH, Jamison MG, Brancazio LR, et al: Venous thromboembolismduring pregnancy and the postpartum period: incidence, risk factors,and mortality. Am J Obstet Gynecol 194:1311-1315, 2006

0. James AH, Tapson VF, Goldhaber SZ: Thrombosis during pregnancyand the postpartum period. Am J Obstet Gynecol 193:216-219, 2005

1. Pabinger I, Grafenhofer H, Kyrle PA, et al: Temporary increase in therisk for recurrence during pregnancy in women with a history of ve-nous thromboembolism. Blood 100:1060-1062, 2002

2. Lamy C, Hamon JB, Coste J, et al: Ischemic stroke in young women: riskof recurrence during subsequent pregnancies. French Study Group onStroke in Pregnancy. Neurology 55:269-274, 2000

3. Brill-Edwards P, Ginsberg JS, Gent M, et al: Safety of withholdingheparin in pregnant women with a history of venous thromboembo-lism. Recurrence of Clot in This Pregnancy Study Group. N Engl J Med343:1439-1444, 2000

4. Pabinger I, Grafenhofer H, Kaider A, et al: Risk of pregnancy-associatedrecurrent venous thromboembolism in women with a history of venousthrombosis. J Thromb Haemost 3:949-954, 2005

5. De Stefano V, Martinelli I, Rossi E, et al: The risk of recurrent venousthromboembolism in pregnancy and puerperium without antithrom-botic prophylaxis. Br J Haematol 135:386-391, 2006

6. Sanson BJ, Lensing AW, Prins MH, et al: Safety of low-molecular-weight heparin in pregnancy: a systematic review. Thromb Haemost81:668-672, 1999

7. Lepercq J, Conard J, Borel-Derlon A, et al: Venous thromboembolismduring pregnancy: a retrospective study of enoxaparin safety in 624pregnancies. Br J Obstet Gynaecol 108:1134-1140, 2001

8. Dilley A, Austin H, El-Jamil M, et al: Genetic factors associated withthrombosis in pregnancy in a United States population. Am J ObstetGynecol 183:1271-1277, 2000

9. Gerhardt A, Scharf RE, Beckmann MW, et al: Prothrombin and factor Vmutations in women with a history of thrombosis during pregnancyand the puerperium. N Engl J Med 342:374-380, 2000

0. Ho WK, Hankey GJ, Quinlan DJ, et al: Risk of recurrent venous throm-boembolism in patients with common thrombophilia: a systematic re-view. Arch Intern Med 166:729-736, 2006

1. De Stefano V, Simioni P, Rossi E, et al: The risk of recurrent venousthromboembolism in patients with inherited deficiency of natural an-ticoagulants antithrombin, protein C and protein S. Haematologica91:695-698, 2006

2. Christiansen SC, Cannegieter SC, Koster T, et al: Thrombophilia, clin-ical factors, and recurrent venous thrombotic events. J Am Med Assoc293:2352-2361, 2005

3. De StefanoV, Martinelli I, Mannucci PM, et al: The risk of recurrentdeep venous thrombosis among heterozygous carriers of both factor VLeiden and the G20210A prothrombin mutation. N Engl J Med 341:801-806, 1999

4. Lindmarker P, Schulman S, Sten-Linder M, et al: The risk of recurrentvenous thromboembolism in carriers and non-carriers of the G1691Aallele in the coagulation factor V gene and the G20210A allele in theprothrombin gene. DURAC Trial Study Group. Duration of Anticoag-
ulation. Thromb Haemost 81:684-689, 1999

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6

7


5. Pabinger I, Schneider B: Thrombotic risk in hereditary antithrombinIII, protein C, or protein S deficiency. A cooperative, retrospectivestudy. Gesellschaft fur Thrombose- und Hamostaseforschung (GTH)Study Group on Natural Inhibitors. Arterioscler Thromb Vasc Biol16:742-748, 1996

6. Conard J, Horellou MH, Van Dreden P, et al: Thrombosis and preg-nancy in congenital deficiencies in AT III, protein C or protein S: studyof 78 women. Thromb Haemost 63:319-320, 1990

7. Hellgren M, Tengborn L, Abildgaard U: Pregnancy in women withcongenital antithrombin III deficiency: experience of treatment withheparin and antithrombin. Gynecol Obstet Invest 14:127-141, 1982

8. Krnic-Barrie S, O’Connor CR, Looney SW, et al: A retrospective reviewof 61 patients with antiphospholipid syndrome. Analysis of factorsinfluencing recurrent thrombosis. Arch Intern Med 157:2101-2108,1997


0. James AH, Brancazio LR, Gehrig TR, et al: Low-molecular-weight hep-arin for thromboprophylaxis in pregnant women with mechanicalheart valves. J Matern Fetal Neonatal Med 19:543-549, 2006

1. Rhee E, Beckman M, Dowling N, et al: Protein S levels in normalpregnancy. Am J Obstet Gynecol 193:S174, 2005

2. Lockwood CJ: Inherited thrombophilias in pregnant patients: detectionand treatment paradigm. Obstet Gynecol 99:333-341, 2002

3. Hellgren M, Blomback M: Studies on blood coagulation and fibrinolysisin pregnancy, during delivery and in the puerperium. I. Normal con-dition. Gynecol Obstet Invest 12:141-154, 1981

4. Stirling Y, Woolf L, North WR, et al: Haemostasis in normal pregnancy.Thromb Haemost 52:176-182, 1984

5. Dahlman T, Hellgren M, Blomback M: Changes in blood coagulationand fibrinolysis in the normal puerperium. Gynecol Obstet Invest 20:37-44, 1985

6. Gerbasi FR, Bottoms S, Farag A, et al: Increased intravascular coagula-tion associated with pregnancy. Obstet Gynecol 75:385-389, 1990

7. Sanchez-Luceros A, Meschengieser SS, Marchese C, et al: Factor VIIIand von Willebrand factor changes during normal pregnancy and pu-erperium. Blood Coagul Fibrinolysis 14:647-651, 2003

8. Wickstrom K, Edelstam G, Lowbeer CH, et al: Reference intervals forplasma levels of fibronectin, von Willebrand factor, free protein S andantithrombin during third-trimester pregnancy. Scand J Clin Lab In-vest 64:31-40, 2004

9. Faught W, Garner P, Jones G, et al: Changes in protein C and protein Slevels in normal pregnancy. Am J Obstet Gynecol 172:147-150, 1995

0. Cerneca F, Ricci G, Simeone R, et al: Coagulation and fibrinolysischanges in normal pregnancy. Increased levels of procoagulants andreduced levels of inhibitors during pregnancy induce a hypercoagula-ble state, combined with a reactive fibrinolysis. Eur J Obstet GynecolReprod Biol 73:31-36, 1997

1. Clark P, Brennand J, Conkie JA, et al: Activated protein C sensitivity,protein C, protein S and coagulation in normal pregnancy. ThrombHaemost 79:1166-1170, 1998

2. Kjellberg U, Andersson NE, Rosen S, et al: APC resistance and otherhaemostatic variables during pregnancy and puerperium. ThrombHaemost 81:527-531, 1999

3. Jacques PF, Bostom AG, Williams RR, et al: Relation between folatestatus, a common mutation in methylenetetrahydrofolate reductase,and plasma homocysteine concentrations. Circulation 93:7-9, 1996

4. Glueck CJ, Phillips H, Cameron D, et al: The 4G/4G polymorphism ofthe hypofibrinolytic plasminogen activator inhibitor type 1 gene: anindependent risk factor for serious pregnancy complications. Metabo-lism 49:845-852, 2000

5. Miyakis S, Lockshin MD, Atsumi T, et al: International consensus state-ment on an update of the classification criteria for definite antiphos-pholipid syndrome (APS). J Thromb Haemost 4:295-306, 2006

6. Gordon M: Maternal physiology in pregnancy, in Gabbe S, Niebyl J,Simpson J (eds): Normal and Problem Pregnancies (ed 4). New York,NY, Churchill Livingstone, 2002, pp 63-92

7. Naimi TS, Lipscomb LE, Brewer RD, et al: Binge drinking in the pre-

conception period and the risk of unintended pregnancy: implicationsfor women and their children. Pediatrics 111:1136-1141, 2003

8. Moore K: Organogenetic period: the fourth to eighth weeks, in TheDeveloping Human: Clinically Oriented Embryology (ed 6). Philadel-phia, PA, WB Saunders Company, 1998, pp 83-106

9. Nassar AH, Hobeika EM, Abd Essamad HM, et al: Pregnancy outcomein women with prosthetic heart valves. Am J Obstet Gynecol 191:1009-1013, 2004

0. Chan WS, Anand S, Ginsberg JS: Anticoagulation of pregnant womenwith mechanical heart valves: a systematic review of the literature. ArchIntern Med 160:191-196, 2000

1. Blickstein D, Blickstein I: The risk of fetal loss associated with Warfarinanticoagulation. Int J Gynaecol Obstet 78:221-225, 2002

2. Vural KM, Ozatik MA, Uncu H, et al: Pregnancy after mechanical mitralvalve replacement. J Heart Valve Dis 12:370-376, 2003

3. Sadler L, McCowan L, White H, et al: Pregnancy outcomes and cardiaccomplications in women with mechanical, bioprosthetic and ho-mograft valves. Br J Obstet Gynaecol 107:245-253, 2000

4. Cotrufo M, De Feo M, De Santo LS, et al: Risk of warfarin duringpregnancy with mechanical valve prostheses. Obstet Gynecol 99:35-40, 2002

5. Schaefer C, Hannemann D, Meister R, et al: Vitamin K antagonists andpregnancy outcome. A multi-centre prospective study. Thromb Hae-most 95:949-957, 2006

6. Iturbe-Alessio I, Fonseca MC, Mutchinik O, et al: Risks of anticoagulanttherapy in pregnant women with artificial heart valves. N Engl J Med315:1390-1393, 1986

7. Srivastava AK, Gupta AK, Singh AV, et al: Effect of oral anticoagulantduring pregnancy with prosthetic heart valve. Asian Cardiovasc ThoracAnn 10:306-309, 2002

8. Meschengieser SS, Fondevila CG, Santarelli MT, et al: Anticoagulationin pregnant women with mechanical heart valve prostheses. Heart 82:23-26, 1999

9. Chen WW, Chan CS, Lee PK, et al: Pregnancy in patients with pros-thetic heart valves: an experience with 45 pregnancies. Q J Med 51:358-365, 1982

0. Wesseling J, Van Driel D, Heymans HS, et al: Coumarins during preg-nancy: long-term effects on growth and development of school-agechildren. Thromb Haemost 85:609-613, 2001

1. Bates SM, Greer IA, Hirsh J, et al: Use of antithrombotic agents duringpregnancy: the Seventh ACCP Conference on Antithrombotic andThrombolytic Therapy. Chest 126:627S-644S, 2004 (suppl)

2. Flessa HC, Kapstrom AB, Glueck HI, et al: Placental transport of hep-arin. Am J Obstet Gynecol 93:570-573, 1965

3. Harenberg J, Schneider D, Heilmann L, et al: Lack of anti-factor Xaactivity in umbilical cord vein samples after subcutaneous administra-tion of heparin or low molecular mass heparin in pregnant women.Haemostasis 23:314-320, 1993

4. Schneider D, Heilmann L, Harenberg J: [Placental transfer of low-mo-lecular weight heparin]. Geburtshilfe Frauenheilkd 55:93-98, 1995

5. Ginsberg JS, Kowalchuk G, Hirsh J, et al: Heparin therapy duringpregnancy. Risks to the fetus and mother. Arch Intern Med 149:2233-2236, 1989

6. Dahlman TC, Sjoberg HE, Ringertz H: Bone mineral density duringlong-term prophylaxis with heparin in pregnancy. Am J Obstet Gy-necol 170:1315-1320, 1994

7. Barbour LA, Kick SD, Steiner JF, et al: A prospective study of heparin-induced osteoporosis in pregnancy using bone densitometry. Am JObstet Gynecol 170:862-869, 1994

8. Casele H, Haney EI, James A, et al: Bone density changes in women whoreceive thromboprophylaxis in pregnancy. Am J Obstet Gynecol 195:1109-1113, 2006

9. Dahlman TC: Osteoporotic fractures and the recurrence of thrombo-embolism during pregnancy and the puerperium in 184 women un-dergoing thromboprophylaxis with heparin. Am J Obstet Gynecol 168:1265-1270, 1993

0. Fausett MB, Vogtlander M, Lee RM, et al: Heparin-induced thrombo-cytopenia is rare in pregnancy. Am J Obstet Gynecol 185:148-152,
2001

7

7

7

7

7

7

7

7

7

8

8

8

8

8

8

8

8

8

8

9

9

9

9


1. Carlin AJ, Farquharson RG, Quenby SM, et al: Prospective observa-tional study of bone mineral density during pregnancy: low molecularweight heparin versus control. Hum Reprod 19:1211-1214, 2004

2. Casele H, Haney E: Bruising in women undergoing thromboprophy-laxis in pregnancy. Am J Obstet Gynecol 193:S81, 2006

3. Matziolis G, Perka C, Disch A, et al: Effects of fondaparinux comparedwith dalteparin, enoxaparin and unfractionated heparin on human os-teoblasts. Calcif Tissue Int 73:370-379, 2003

4. Lagrange F, Brun JL, Vergnes MC, et al: Fondaparinux sodium does notcross the placental barrier: study using the in-vitro human dually per-fused cotyledon model. Clin Pharmacokinet 41:47-49, 2002 (suppl 2)

5. Dempfle CE: Minor transplacental passage of fondaparinux in vivo.N Engl J Med 350:1914-1915, 2004

6. Mazzolai L, Hohlfeld P, Spertini F, et al: Fondaparinux is a safe alter-native in case of heparin intolerance during pregnancy. Blood 108:1569-1570, 2006

7. ACOG. Thromboembolism in Pregnancy. American College of Obste-trician and Gynecologists, 2000:1-9

8. Coppage KH, Hinton AC, Moldenhauer J, et al: Maternal and perinataloutcome in women with a history of stroke. Am J Obstet Gynecol190:1331-1334, 2004

9. Ray JG, Chan WS: Deep vein thrombosis during pregnancy and thepuerperium: a meta-analysis of the period of risk and the leg of presen-tation. Obstet Gynecol Surv 54:265-271, 1999

0. Ansell J, Hirsh J, Poller L, et al: The pharmacology and management ofthe vitamin K antagonists: the Seventh ACCP Conference on Anti-thrombotic and Thrombolytic Therapy. Chest 126:204S-233S, 2004(suppl)

1. Stewart JA, Hamilton PJ, Murdoch AP: Thromboembolic disease asso-ciated with ovarian stimulation and assisted conception techniques.Hum Reprod 12:2167-2173, 1997

2. Warkentin TE, Levine MN, Hirsh J, et al: Heparin-induced thrombo-

cytopenia in patients treated with low-molecular-weight heparin orunfractionated heparin. N Engl J Med 332:1330-1335, 1995

3. McCrae KR, Bussel JB, Mannucci PM, et al: Platelets: an update ondiagnosis and management of thrombocytopenic disorders. Hematol-ogy (Am Soc Hematol Educ Program) 282-305, 2001

4. Horlocker TT: Low molecular weight heparin and neuraxial anesthesia.Thromb Res 101:141-154, 2001

5. Baker WH, Mahler DK, Foldes MS, et al: Pneumatic compression de-vices for prophylaxis of deep venous thrombosis (DVT). Am Surg 52:371-373, 1986

6. Orme ML, Lewis PJ, de Swiet M, et al: May mothers given warfarinbreast-feed their infants? Br Med J 1:1564-1565, 1977

7. American Academy of Pediatrics Committee on Drugs: Transfer ofdrugs and other chemicals into human milk. Pediatrics 108:776-789,2001

8. Cardiovascular disease and use of oral and injectable progestogen-onlycontraceptives and combined injectable contraceptives. Results of aninternational, multicenter, case-control study. World Health Organiza-tion Collaborative Study of Cardiovascular Disease and Steroid Hor-mone Contraception. Contraception 57:315-324, 1998

9. Kupferminc MJ, Eldor A, Steinman N, et al: Increased frequency ofgenetic thrombophilia in women with complications of pregnancy.N Engl J Med 340:9-13, 1999

0. James A, Brancazio L, Ortel T: Thrombophilia in pregnancy: maternaland fetal implications. Curr Womens Health Rev 2:51-59, 2006

1. Badaracco MA, Vessey MP: Recurrence of venous thromboembolic dis-ease and use of oral contraceptives. Br Med J 1:215-217, 1974

2. Tengborn L, Bergqvist D, Matzsch T, et al: Recurrent thromboembo-lism in pregnancy and puerperium. Is there a need for thromboprophy-laxis? Am J Obstet Gynecol 160:90-94, 1989

3. James AH, Brancazio LR, Ortel TL: Thrombosis, thrombophilia, andthromboprophylaxis in pregnancy. Clin Adv Hematol Oncol 3:187-
197, 2005

RFJ

Gnpateogapd

bmaf

D

A

1

ecurrent Gestational Diabetes: Riskactors, Diagnosis, Management, and Implications

oseph N. Bottalico, DO

Gestational diabetes mellitus (GDM) should be regarded as a sentinel event in a woman’slife that presents challenges and disease prevention opportunities to all providers of healthcare for women of reproductive age. Prediabetic risk factors are rising in prevalence andinclude dietary and lifestyle habits, which when superimposed on genetic predispositioncontribute to the rising prevalence of type 2 diabetes and GDM. There is growing evidencethat treatment of GDM matters, with a continuum of adverse pregnancy outcome risksproportional to degrees of maternal glucose intolerance. GDM in an index pregnancyincreases the risk of recurrent GDM in subsequent pregnancies, and recurrence rates of upto 70% have been reported. GDM recurrence rates are influenced by maternal healthcharacteristics and past pregnancy history. The risk of later metabolic syndrome and type2 diabetes is increased in women with a history of GDM and women should be screenedfor postpartum glucose intolerance. Opportunities to prevent recurrent GDM and later type2 diabetes require attention to risk factors and plasma glucose status with identification ofimpaired fasting glucose or impaired glucose tolerance.Semin Perinatol 31:176-184 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS gestational diabetes, recurrence, diabetes mellitus-type 2, diabetes risk factors,prevention, metabolic syndrome, preconception care

acn

TGGwGtvsoteaTG

icw

estational diabetes mellitus (GDM) is defined as anydegree of glucose intolerance with onset or first recog-

ition during pregnancy. The definition of GDM does notreclude the possibility that unrecognized glucose intoler-nce may have existed before the pregnancy, and the defini-ion applies whether insulin, oral antidiabetic agents, or di-tary modification is used for treatment. Approximately 7%f all pregnancies in the United States are complicated byestational diabetes resulting in more than 200,000 casesnnually,1 but the prevalence ranges from 1% to 14% of allregnancies depending on the population studied and theiagnostic tests employed.Gestational diabetes is often considered to be type 2 dia-

etes unmasked by pregnancy. The two entities share com-on risk factors, a number of which are increasing in prev-

lence, such as obesity and advancing maternal age. Theseactors, combined with more universal screening, have led to

ivision of Maternal Fetal Medicine, Department of Obstetrics and Gyne-cology, UMDNJ-School of Osteopathic Medicine, Stratford, NJ.

ddress reprint requests to Joseph N. Bottalico, DO, Department of Obstet-rics and Gynecology, Division of Maternal Fetal Medicine, UMDNJ-School of Osteopathic Medicine, University Doctors Pavilion-Suite3500, 42 East Laurel Road, Stratford, NJ 08084-1504. E-mail: bottaljn

[email protected]


n increased incidence of GDM. It is thus becoming moreommon to care for women who have had a previous preg-ancy complicated by GDM.

he Epidemiology ofestational Diabetes Recurrence

iven that most of the risk factors for GDM persist or becomeorse in subsequent pregnancies, it is not surprising thatDM has a high recurrence rate of 35.6% to 70%. The fact

hat a wide range of recurrence rates have been reported inarious studies is, in part, due to the variability of GDMcreening methods and the use of different diagnostic thresh-ld values for various glucose tolerance tests. It is clear fromhe literature, however, that a diagnosis of GDM confers anlevated risk of recurrent GDM in subsequent pregnancies inddition to an increased risk of later type 2 diabetes mellitus.able 1 summarizes studies reporting rates of recurrence forDM.Philipson and Super2 examined the recurrence of glucose

ntolerance in 36 women with an index pregnancy compli-ated by GDM. Oral or intravenous glucose tolerance testsere used to document glucose intolerance or gestational
iabetes. They reported that 55% (20 of 36) of women de-

vwwwcwwiGaasdtcBLn

adct1ad8irhv

iumtt

T

P

G

F

S

M

M

Recurrent gestational diabetes 177

eloped gestational diabetes in a subsequent pregnancy. Theomen with consecutive pregnancies complicated by GDMere heavier and delivered larger neonates than the womenho did not develop recurrent GDM. In 1992, Gaudier and

oworkers3 conducted a retrospective review of 90 womenhose index pregnancies were complicated by GDM andhose subsequent pregnancies were managed at the same

nstitution. There were 52% (47) who had a recurrence ofDM in their subsequent pregnancy. These 47 women hadn increased BMI (32.8 kg/m2 versus 28.9 kg/m2, P � 0.03)nd more large for gestational age (LGA) neonates (38% ver-us 14%; P � 0.05). A greater number also required insulinuring their index pregnancy (38% versus 19%; P � 0.05)han those who did not have a recurrence of GDM. Theyoncluded that women with a history of GDM who have aMI greater than 35 kg/m2, whose previous newborn wasGA, and who required insulin during the previous preg-ancy are at increased risk for recurrence of GDM.

able 1 Summary of Literature on Recurrence of Gestational D

Author(Country-

Year)DataYears Diagnostic Criteria

hilipson2

USA-19991979-1987 100 g OGTT (C/C criteria) or

IVGTT (25-28 weeks)

audier3

USA-19921983-1990 50 g screen (>135 mg/dl) @

24-28 weeks. If screen �,100 g OGTT (105/190/165/145 mg/dl)

oster-Powell9

Australia-1998

1990-1996 75 g 2-hour OGTT (100/180/145 mg/dl)

pong4

USA-19981988-1992 50 g screen (>140 mg/dl) @

24-26 weeks. If screen �,100g OGTT (105/190/165/145 mg/dl)

ajor5

USA-19981992-1996 50g screen (>140 mg/dl) @

24-28 weeks. If screen �,100g OGTT (105/190/165/145 mg/dl)

acNeill6

Canada-2001

1980-1996 50g screen (>140 mg/dl). Ifscreen �, 100g OGTT(105/190/165/145 mg/dl)

Spong and coworkers4 evaluated the influence of maternal 1

nd neonatal factors on the recurrence of GDM in 164 pre-ominantly Hispanic women whose index pregnancy wasomplicated by GDM. The diagnosis of GDM was based onhe National Diabetes Data Group (NDDG) criteria using a00-g oral glucose tolerance test. There were 68% (111) withrecurrence of GDM. Women with recurrence had GDM

iagnosed earlier, frequently required insulin (25% versus%, P � 0.05), and had more hospital admissions in their

ndex pregnancy compared with women without a recur-ence of GDM. Infants of women with recurrence wereeavier and had an increased incidence of macrosomia (26%ersus 10%, P � 0.05).

Major and coworkers5 reported on 78 women with GDMn a previous pregnancy. Universal screening was performedsing a 50-g oral glucose load. Those with a value of 140g/dL or greater were then given the 100-g oral glucose

olerance test (OGTT) using the threshold criteria of the Na-ional Diabetes Data Group (fasting of 105 mg/dL, 1 hour of

s Including Diagnostic Criteria, Risk Factors, and Outcomes

currenceRate

Risk Factors forRecurrence

Complicationswith

Recurrence

% (20/36) -BMI > 30 kg/m2

-Obesity-More large for

gestational agebirths

-More macrosomia% (47/90) -BMI > 35

-Insulin required inindex pregnancy

-Previous baby > 4kg% (82/117) -Maternal age


-Pre-pregnancy BMI-Weight gain between

pregnancies% (111/164) -Earlier GDM in index

pregnancy-Insulin required in

index pregnancy-Hospital admission in

index pregnancy

-Higher birthweight

-More infants>4kg

% (54/78) -BMI > 30 in indexpregnancy

-GDM Dx < 24 weeksin index pregnancy


-Weight gain > 15 lbsbetween pregnancies

-Pregnancy interval <24 months

% (232/651) -Birth weight >4 kgindex pregnancy

-Pre-pregnancy weightin subsequentpregnancy

iabete

Re

55

52

70

68

69

35.6

90 mg/dL, 2 hours of 165 mg/dL, and 3 hours of 145 mg/

dRngmig

twnlsAw1thmAt�a�aaGMwwr

drt7Wlwtpd

sstu

IoFSemwc

tcATstwwp34oalmoct

mcsf�dsmipwcSbt2ooai

UDDNl(mSaa2wHA

178 J.N. Bottalico

L). GDM recurrence was found in 69% (54) of women.ecurrence of GDM was more common with an index preg-ancy parity �1, BMI �30 kg/m2, a GDM diagnosis at �24estational weeks in the first pregnancy, and insulin require-ent in the first pregnancy. A weight gain of �15 lbs and an

nterval between pregnancies of �24 months had the stron-est association with recurrence of GDM.

MacNeill and coworkers6 published a retrospective longi-udinal study in 2001 of a predominantly white cohort of 651omen in Nova Scotia diagnosed with GDM during a preg-ancy that occurred between 1980 and 1996 and who had at

east one subsequent pregnancy during that time period. Thecreening method employed was a 50-g 1-hour glucose load.value of �140 mg/dL was considered a positive screen andas followed by a 3-hour oral glucose tolerance test with00 g of glucose that required more than one glucose over thehreshold values for a positive test (fasting �95mg/dL, 1our �190 mg/dL, 2 hours �165 mg/dL, and 3 hours 145g/dL). These criteria combine elements from the currentDA endorsed Carpenter and Coustan7 threshold values for

he 100-g glucose 3-hour OGTT (fasting �95mg/dL, 1-hour180 mg/dL, 2-hour �155 mg/dL, and 3-hour 140 mg/dL)

nd the National Diabetes Data Group8 values (fasting105mg/dL, 1 hour �190 mg/dL, 2 hours �165 mg/dL,

nd 3 hours 145 mg/dL), which also require more than onebnormal value for a positive test. The rate of recurrence ofDM in the subsequent pregnancy was found to be 35.6%.ultivariate regression models showed that infant birtheight in the index pregnancy and maternal prepregnancyeight before the subsequent pregnancy were predictive of

ecurrent GDM.A 1998 study by Foster-Powell and Cheung9 through a

iabetes service in New South Wales, Australia reported aetrospective review of 540 women with GDM managed be-ween 1990 and 1996. There was a GDM recurrence rate of0% using a 2-hour 75-g OGTT for diagnosis with modifiedHO threshold values. Finally, Danilenko-Dixon and col-

eagues10 in 2000 reported that perinatal outcomes in womenith previous gestational diabetes but with normal glucose

olerance tests during a subsequent pregnancy were not im-roved with regard to birth weight, macrosomia, route ofelivery, and neonatal complications.Unfortunately, these studies on the recurrence of GDM

uffer from the limitations of having no formal proof thatome of the women did not have undiagnosed type 2 diabe-es, impaired fasting glucose, or impaired glucose tolerancesing current ADA criteria before the subsequent pregnancy.

ncreasing Prevalencef Pre-Diabetic Riskactors in the United States

ince GDM may very well be a “tip of the iceberg” phenom-non for many women, it is important to consider the epide-iologic landscape affecting the first occurrence of GDM asell as the risks for recurrent GDM in subsequent pregnan-

ies and the longer term risk of type 2 diabetes. P

Excessive weight is an important public health concern inhe United States, as well as other affluent societies. Pleis andoworkers11 reported that, by the end of 2000, 34% of adultmericans were overweight and another 27% were obese.his represented an increase of 75% compared with 1980tatistics. These data showed that, at the end of 2000, morehan 50% of the adults in the United States were either over-eight or obese. United States data from 1997 and 1998 foromen aged 18 to 24 and 25 to 44 revealed an estimatedrevalence of overweight (BMI � 25 kg/m2) or obese (BMI �0 kg/m2) of over 30% for the 18 to 24 age group and over0% for the 25 to 44 group.12 An increased prevalence ofbesity in U.S. adolescents has also been documented andssociated with declining levels of physical activity.13 Theinks between obesity, insulin resistance, and type 2 diabetes

ellitus are well known, and the long-term complications ofbesity are numerous. The strong association between in-reased BMI and higher risks for GDM has been clearly es-ablished.

There is also growing recognition of the importance of theetabolic syndrome as a multiplex risk factor for cardiovas-

ular disease and type 2 diabetes. In women, the metabolicyndrome is defined by the presence of three or more of theollowing criteria: abdominal obesity (waist circumference

88 cm or 34.7 inches), elevated triglycerides (�150 mg/L), high-density lipoprotein (�50 mg/dL), high blood pres-ure (�130/85 mm Hg), and elevated fasting glucose (110g/dL).14 It should be noted that the current definition of

mpaired fasting glucose is 100 to 125 mg/dL.1 The estimatedrevalence of the metabolic syndrome in women is 24%,ith prevalence increasing with age as reported by Ford and

oworkers using data from the National Health and Nutritionurvey (NHANES III).15 For 4549 female subjects reportedy Ford, metabolic syndrome prevalence was about 6% inhose 20 to 29 years of age, 14% in those 30 to 39 years of age,0% in those 40 to 49 years of age, and �30% for womenlder than 50 years of age. As with obesity, the pathophysi-logic features of the metabolic syndrome are likely to bessociated with a variety of pregnancy complications, includ-ng GDM.

nited Statesiabetes Prevalence:iagnosed and Undiagnosed

ational estimates on diabetes prevalence have been pub-ished by the Centers for Disease Control and PreventionCDC) in Atlanta.16 The total prevalence of frank diabetesellitus among women age 20 years or older in the United

tates in 2005 was 9.7 million, or 8.8% of all women in thisge group. The prevalence of diabetes by race and ethnicitymong people 20 years of age or older in the United States in005 was an estimated 8.7% prevalence for all non-Hispanichites, 13.3% for non-Hispanic blacks, and about 9.5% ofispanic/Latino Americans. The rates are also higher formerican-Indians, Alaskan natives, Asian Americans, and
acific Islanders. Many of these women will have DM iden-

tHbttn

tig1fiaaerwct

DGctfstucrdGalscow

m1“bclaCasmdn2thet

aNsd

thrn6swhnersw0hwrw

GDGD

TR

PPIP

P

P

P

E

FAOGMI


ified for the first time in pregnancy and be classified as GDM.owever, if not tested after delivery, a second pregnancy maye incorrectly classified as recurrent GDM instead of preges-ational DM. These epidemiologic trends suggest that diabe-es, GDM, and recurrent GDM will likely complicate a greaterumber of pregnancies in the next few decades.In a cross-sectional sample of U.S. adults over the age of 40

ested between 1988 and 1994, 33.8% had an impaired fast-ng glucose (IFG; which is currently defined as a plasmalucose of 100-125 mg/dL after an overnight fast). A total of5.4% had impaired glucose tolerance (IGT), currently de-ned as 140 to 199 mg/dL after a 2-hour oral glucose toler-nce test with 75 g of oral glucose solution.16 When percent-ges were applied to the entire U.S. population in 2000, anstimated 41 million adults had prediabetes. Data are noteadily available for the reproductive age female populationith regard to the prevalence of prediabetes, although in-

reasing overweight, obesity, and ethnicity trends suggesthat a rising prevalence may be occurring.

iagnosis of GDMDM is a condition that lacks international consensus con-erning the method of choice for screening and specific cri-eria for diagnosis. The Fourth International Workshop-Con-erence on Gestational Diabetes in 1998 recommended acreening strategy based on risk assessment for detecting ges-ational diabetes and that risk assessment for GDM should bendertaken at the first prenatal visit.7 Women can then belassified as either low-risk, average-risk, or high-risk. High-isk women should be screened for GDM as soon as feasibleuring pregnancy. If high-risk women are found not to haveDM at initial screening, they should be retested between 24nd 28 weeks of gestation.1 Risk factors most consistentlyinked with GDM include a previous macrosomic infant, obe-ity, suspected glucose intolerance when not pregnant, glu-osuria, a strong immediate family history of type 2 diabetesr GDM, history of unexplained fetal demise and, especially,omen with a history of GDM in a prior pregnancy (Table 2).A diagnosis of GDM is based on a positive OGTT (�2 abnor-al values) most commonly performed after an abnormal

-hour plasma glucose screen with a 50-g oral glucose load (thetwo-step” approach). Diabetes mellitus in pregnancy can alsoe diagnosed if the fasting plasma glucose is �126 mg/dL or aasual plasma glucose is �200 mg/dL, but these criteria may beess reliable since they are mostly applicable to nonpregnantdults. Guidelines issued by the Fourth International Workshoponference on Gestational Diabetes Mellitus,7 the American Di-betes Association (ADA-2004),1 and American College of Ob-tetricians and Gynecologists17 in 2001 lean toward recom-ending the use of the Carpenter-Coustan criteria for theiagnostic 3-hour, 100-g glucose OGTT. However, these orga-izations also recognize the alternative use of the single-step,-hour, 75-g glucose OGTT in certain high prevalence popula-ions (threshold values of 95 mg/dL fasting, 180 mg/dL at 1our, and 155 mg/dL at 2 hours). ACOG also acknowledges thatxpert panels have supported both the Carpenter-Coustan and
he National Diabetes Data Group criteria and report that there a
re “no data from clinical trials to determine which is superior.”evertheless, it is important to remember that, similar to other

creening tests, those for GDM may have problems with repro-ucibility.18

As with other high-risk groups, ADA guidelines for the detec-ion of GDM indicate that glucose testing for women with aistory of prior GDM should occur “as soon as feasible,” withe-screening, between 24 and 28 weeks if the initial screen isegative. Maser and coworkers19 evaluated medical records for0 women enrolled in a GDM management program that pre-ented with 2 subsequent pregnancies with GDM to determinehether more specific guidelines for detection are needed. Overalf (55%) of these women required insulin during both preg-ancies, and 16.7% (10) required insulin during the secondnrollment for GDM, but not the first. For those subjects whoequired insulin during both pregnancies, 88% (29 of 33) weretarted earlier during the subsequent pregnancy (21.6 � 8.4eeks of gestation versus 31.5 � 2.7 weeks of gestation, P �.001). During the subsequent pregnancy, approximately one-alf of the women requiring insulin needed it before 24 weeks,hereas a third required it by week 15. Therefore, it seems

easonable to perform the initial screen by 15 weeks in womenith a previous history of GDM.

estationaliabetes Management,eneral and Recurrent

espite the debate over screening strategies, there is a clear

able 2 Risk Factors for Gestational Diabetes Mellitus andecurrence

revious history of gestational diabetesrevious diagnosis of GDM before 24 weeks

nsulin requirement in previous pregnancyrevious GDM and weight gain of >15 lbs between

pregnanciesrevious GDM and interval between pregnancies <24

monthsrevious history of macrosomia (birth weight >9 lbs or

4100 grams)revious history of adverse pregnancy outcome (not

clearly attributable to a condition other than diabetesduring pregnancy)

thnic group with an elevated risk of gestational and type2 diabetes

HispanicAfricanNative AmericanSouth or East AsianPacific Island Ancestry

irst degree relative with diabetesdvanced maternal agebesity (BMI >30-35 kg/m2)lucosuriaetabolic syndrome

nsulin resistance (eg, PCOS)

ssociation between maternal carbohydrate intolerance and

papPmtcntiwpwsdtwrsftidlfswtfgofcttugsl

wgda(1rgh5cissntuh

wtws3wqi

arnsGgtt

PGTrGsowGdem3no3fW7vftiribatvrt

CtTl

180 J.N. Bottalico

erinatal complications. The rates of excessive fetal growthnd neonatal morbidity rise with the degree of maternal hy-erglycemia. The Toronto Tri-Hospital Gestational Diabetesroject was a prospective study on the impact of increasingaternal carbohydrate intolerance on both maternal and fe-

al outcomes.20 Women with GDM were compared with aohort of 3637 women with singleton pregnancies that didot have GDM. The study found a direct relationship be-ween OGTT values and multiple adverse outcomes, includ-ng preeclampsia, macrosomia, and operative delivery inomen without frank GDM. Saks and coworkers21 found aositive relationship between OGTT results and fetal birtheight in approximately 3500 pregnant women undergoing

creening. Langer and Mazze,22 in a large prospective studyesigned to determine optimal levels of glycemic control forreatment of women with GDM, demonstrated in 246omen with GDM that the women who achieved the lowest

ange of blood glucose values throughout pregnancy had aignificantly lower incidence of LGA and macrosomic in-ants. In 1994, Langer and coworkers23 compared intensiveo conventional therapy in 2500 women with GDM. Womenn the intensive therapy group had a significantly lower inci-ence of macrosomic and LGA infants. These infants also had

ower rates of admission to the neonatal intensive care unit,ewer metabolic complications, and less need for respiratoryupport. Langer and coworkers24 compared 555 gravidasith GDM diagnosed after 37 weeks with 1110 subjects

reated for GDM and 1110 nondiabetic subjects matchedrom the same delivery year for obesity, parity, ethnicity, andestational age at delivery. They found a composite adverseutcome rate of 59% for untreated, 18% for treated, and 11%or nondiabetic subjects. A 2- to 4-fold increase in metabolicomplications and macrosomia/LGA was found in the un-reated group with no difference between nondiabetic andreated subjects. A 2- to 3-fold higher morbidity rate for thentreated groups was found compared with the otherroups. They also concluded that untreated GDM confersignificant risks of perinatal morbidity at all disease severityevels.

The Australian Carbohydrate Intolerance Study (ACHOIS)as a large randomized multicenter trial for the treatment ofestational diabetes.25 The criteria used in this study for theiagnosis of GDM were based on the 75-g OGTT between 24nd 34 weeks gestation. Those with values below 140 mg/dL7.8 mmol/L) after an overnight fast and between 140 and98 mg/dL (7.8-11.0 mmol/L) at 2 hours were eligible forandomization. The 490 women assigned to the interventionroup had premeal glucose levels below 99 mg/dL and 2-our postprandial levels that did not exceed 126 mg/dL. The10 women assigned to the control group received routineare that was consistent with care provided in other facilitiesn which screening for GDM was not standard. The rate oferious perinatal complications (which was defined as death,houlder dystocia, bone fracture, and nerve palsy) was sig-ificantly lower among infants of the women in the interven-ion group (self-monitoring of blood glucose levels, individ-alized medical nutrition therapy counseling and insulin);
owever, more infants of women in the intervention group d
ere admitted to the neonatal ICU. Women in the interven-ion group had a higher rate of induction of labor thanomen in the routine care group, although the rates of ce-

arean delivery were similar at 31% and 32%, respectively. Atmonths postpartum, follow-up data were available for 573omen and revealed lower rates of depression and betteruality of life scores, consistent with improved health status

n the intervention group.The above data suggest that making a diagnosis of GDM

nd identifying effective interventions have the potential toeduce morbidities associated with GDM and improve preg-ancy outcomes. However, it is not clear which managementtrategies might be most effective for women with recurrentDM. It seems likely given the probable greater degree oflucose intolerance and onset or diagnosis earlier in gestationhat a need for more aggressive use of medications (oral an-idiabetic agents or insulin) should be anticipated.

revention ofestational Diabetes Recurrence

here are limited data examining interventions to preventecurrence of GDM. Medication interventions to preventDM have been studied in women with polycystic ovary

yndrome (PCOS). Affected women are characteristicallybese, have insulin resistance and hyperinsulinemia, all ofhich are risk factors for primary (and probably recurrent)DM.26,27 Metformin increases insulin sensitivity and re-uces insulin resistance. It has been found to be a safe andffective treatment for the metabolic and endocrine abnor-alities in PCOS. Glueck and coworkers28 reported that, in

3 women with PCOS who received metformin during preg-ancy, GDM developed in 1 of 33 (3%) versus 8 of 12 (67%)f their previous pregnancies without metformin. Among the9 women with a history of PCOS who did not take met-ormin, GDM developed in 14 of 60 (23%) of pregnancies.

hen all live births were combined, GDM occurred in 22 of0 pregnancies (31%) in women who did not take metforminersus 1 of 33 pregnancies (3%) in those who did take met-ormin. The authors concluded that, in women with PCOS,he use of metformin was associated with a 10-fold reductionn the rate of GDM. The effect was probably achieved byeducing insulin resistance and insulin secretion, thus reduc-ng the secretory demands imposed on pancreatic beta cellsy insulin resistance in pregnancy. In light of the limitedvailable information, it is not clear whether medication in-erventions after a pregnancy complicated by GDM will pre-ent recurrence, but at the minimum, it seems reasonable toecommend lifestyle changes such as increased physical ac-ivity and weight reduction if overweight or obese.

ontribution of GDMo Glucose Tolerance Statushere is epidemiologic evidence associating GDM with insu-

in resistance, glucose intolerance, and type 2 diabetes.29 The
evelopment of overt diabetes mellitus appears to require an

alwf

itypycramGiLmNlGyspatSi

rocvwaiafmcorIsst

PLaTooipo

lhnwsgnwp2

fbfiGtredpmscam3(pssbdwn

tncwowohiastptc

prttt


ssociated defect in insulin secretion. There may be a finiteevel of pancreatic “beta cell reserve” that is further depletedith recurring GDM. Insulin resistance is another significant

actor for many women who eventually develop GDM.Verma and coworkers30 examined the prevalence of the

nsulin resistance syndrome (IRS) among women with a his-ory of GDM compared with controls over a period of 11ears post delivery. They reported that 27.2% of women withrior GDM versus only 8.2% of controls developed IRS by 11ears after delivery. Prepregnancy obesity was an importantontributor to this increased risk. Kousta and coworkers31

eported on normoglycemic women (34 with previous GDMnd 44 with no previous GDM) who were given insulin-odified IV glucose tolerance tests. They found that post-DM women were more obese than controls, had evidence of

nsulin resistance, and had higher fasting triglycerides.auenborg and colleagues32 estimated the prevalence of theetabolic syndrome by three different criteria (WHO 1999,CEP 2001, and the European Group for the Study of Insu-

in Resistance 2002) among Danish women with previousDM. A cohort of women were followed for a median of 9.8ears (6.4-17.2 years) after their pregnancy. Outcome mea-ures included BMI, glucose tolerance, blood pressure, lipidrofile, and insulin resistance, and they found that the prev-lence of the metabolic syndrome was three times higher inhe prior GDM group in comparison to the control group.imilar findings were reported by Albareda and coworkers33

n 2005 based on their study in a Spanish population.It is possible that recurrent episodes of heightened insulin

esistance, such as with recurrent GDM, place high demandsn the pancreas and contribute to an eventual decline in betaell function that leads to type 2 diabetes in high-risk indi-iduals. Peters and coworkers34 studied 666 Latino womenith a history of GDM. Among the 87 (13%) who completed

n additional pregnancy, the rate ratio of type 2 diabetesncreased significantly in comparison to women without andditional pregnancy. Although the longevity of beta cellunction may be genetically determined, some individuals

ay respond to repeated demands on pancreatic insulin se-retion (as seen in normal pregnancy, obesity, and polycysticvarian disease) in a way that eventually exhausts beta celleserve leading to the development of frank diabetes mellitus.nterestingly, for the nonpregnant population, decreasing in-ulin resistance through diet and exercise or medicationsuch as metformin or rosiglitazone has been shown to reducehe risk of subsequent type 2 diabetes or delay its onset.35-37

ostpartum andong-Term Managementfter Gestational Diabetes

he definition of GDM encompasses a fairly wide spectrumf disease. The degree of glucose intolerance and the time ofnset varies, ranging from transient mild hyperglycemia lim-ted to pregnancy to unrecognized pregestational diabetes toregestational metabolic syndrome (insulin resistance) with
r without features of PCOS. Although data are lacking, it is p
ikely that women with GDM who were most likely to havead unrecognized pregestational diabetes in their index preg-ancy include: those in whom GDM was diagnosed before 24eeks (especially the first trimester), those who required in-

ulin before 20 weeks, and those with initial fasting plasmalucose levels of �126 mg/dL (fasting glucose criterion foronpregnant diabetes mellitus). Other risk factors associatedith an increased risk of resistant glucose intolerance afterregnancy include obesity, high-risk ethnic group, and typediabetes in a first-degree relative.In a systematic literature review, Kim and coworkers38

ound that, after a pregnancy with GDM, rates of overt dia-etes ranged from 2.6% to 70% in studies with follow uprom 6 weeks to 28 years. The incidence of type 2 diabetesncreased most in the first 5 years after the pregnancy withDM and seemed to plateau after 10 years. The Fourth In-

ernational Workshop-Conference on Gestational Diabetesecommended that women diagnosed with GDM undergovaluation with the 75-g OGTT test at 6 to 12 weeks afterelivery.7 The American Diabetes Association in their 2004osition statement on gestational diabetes mellitus recom-ends that “reclassification of maternal glycemic status

hould be performed at least 6 weeks after delivery and ac-ording to the guidelines of the expert committee” referencedbove.1 If glucose levels are normal postpartum, reassess-ent of glycemia should be undertaken at a minimum of

-year intervals. All women with impaired fasting glucoseIFG) or impaired glucose tolerance (IGT) in the postpartumeriod should be tested for diabetes annually. These patientshould receive intensive medical nutrition therapy andhould be placed on an individualized exercise programecause of their very high risk for the development ofiabetes.1 A summary of these recommendations togetherith the author’s suggested evaluation plan after a preg-ancy with GDM is provided in Figure 1.Unfortunately, studies have shown that rates of postpar-

um glucose tolerance testing compliance in women diag-osed with GDM have been less than optimal. Russel andoworkers39 retrospectively studied a cohort of 344 womenith GDM from 2001 to 2004. They found that less thanne-half (45%) of women in the cohort with GDM under-ent postpartum glucose testing. More than one-third (36%)f women who did have postpartum testing were found toave persistent abnormal glucose tolerance. They found no

ndependent relationship between most demographic char-cteristics and postpartum testing. Postpartum testing wastrongly associated only with the attendance at the postpar-um visit as shown by the 54% postpartum testing rate com-ared with 17% of women who did not attend their postpar-um visit. Therefore, postpartum visits are a critical link toompliance with follow-up testing.

These data highlight the importance of postpartum andreconception care to allow further assessment of modifiableisk factors for women with previous GDM, whether or nothey are considering another pregnancy. If not feasible forhose with subsequent pregnancies, then early prenatal careo allow timely GDM screening and management is appro-
riate.

182 J.N. Bottalico

Figure 1 Suggested evaluation plan for women with a previous history of gestational diabetes.

DiTiwtcimwadt

ieagof4mvmwayocsit

acwg9oai0rr

twootl

FTo

wGuciocat

dlompfh

tiGpPdsssOkbmpos

tttmttsatit

ATfab

R


iabetes Preventionn Women with Previous GDMhe TRIPOD study demonstrated that the drug troglitazone

mproved glucose tolerance after GDM in high-risk Hispanicomen.36 Treatment with troglitizone delayed or prevented

he onset of type 2 diabetes in treated women compared withontrols, with a 55% reduction in the average annual diabetesncidence rates over a median follow-up period of 30

onths. The proposed mechanism of this protective effectas the preservation of pancreatic beta cell function which

ppeared to be mediated by a reduction in the secretoryemands placed on the beta cells by chronic insulin resis-ance.

The Diabetes Prevention Program (DPP) study publishedn 200235 randomly assigned 3234 nondiabetic adults withlevated fasting and postload plasma glucose concentrationsnd randomized groups to either a lifestyle modification pro-ram with at least 150 minutes of physical activity per weekr metformin (850 mg twice daily) or placebo. The averageollow up was 2.8 years, and the incidence of diabetes was.8, 7.8, and 11.0 cases per 100 person years in the lifestyle,etformin, and placebo groups, respectively. Lifestyle inter-

ention reduced the incidence of type 2 diabetes by 58% andetformin by 31% as compared with placebo. Of the 2191omen randomized into the 3-arm trial of the DPP, 353 gavehistory of GDM. The women with a history of GDM wereounger but had no other differences in parity, weight, levelf carbohydrate tolerance, or degree of insulin resistanceompared with those without such a history. Despite theseimilarities in cohorts, a history of GDM conferred a 74%ncreased hazard rate to develop diabetes compared withhose without such history.

More recently, the DREAM study37 included 5269 adultsged 30 years or more who had either impaired fasting glu-ose or impaired glucose tolerance. This trial included 1536omen treated with rosiglitazone with 1584 female controlsiven a placebo. Histories of GDM were positive in 9.1% and.3% of these two subgroups of women, respectively. A totalf 11.6% of all persons in the rosiglitizone group (306/2635)nd 26.0% given placebo (686/2634) developed the compos-te primary outcome that included diabetes (hazard ratio.40, 95% CI 0.35-0.46, P � 0.0001). The mean ages in theisiglitazone and placebo groups were 54.6 and 54.8 years,espectively.

These landmark trials illustrate that persons at high risk forhe development of type 2 diabetes, which includes womenith histories of GDM, may benefit from some combinationf lifestyle interventions and drug therapy to reduce the riskf eventual type 2 diabetes. Potential problems with medica-ion side effects and lack of confirmatory studies, however,imit full consideration of their use at the current time.

uture Directionhe concept of gestational diabetes mellitus being an issue
nly for obstetricians and gynecologists needs to be replaced
ith the idea that GDM is a “tip of the iceberg” phenomenon.DM is often the culmination of years of unrecognized andnmodified diabetes risk factors that lead to overt and occultlinical manifestations during pregnancy. It is clear that anndex case of gestational diabetes increases the risks of notnly recurrent gestational diabetes in subsequent pregnan-ies, but also type 2 diabetes later in life for many women. Welso know that recurrent episodes of GDM further increasehe risk of eventual type 2 diabetes.

Although lifestyle interventions and insulin sensitizingrugs can prevent or delay type 2 diabetes mellitus, there are

imited data on prevention of recurrent GDM and adversebstetric outcomes. Considering the possible fetal program-ing effects of GDM on the offspring (albeit through incom-letely understood mechanisms and with a clear need forurther research), the concept of “diabetes begets diabetes”as also emerged,40 inviting strategies to break this cycle.Challenges for the future include development of strategies

o prevent the first episode of gestational diabetes. After anndex pregnancy with GDM, the prevention of recurrentDM may be possible through interventions that addressotentially modifiable risk factors, such as weight and BMI,COS, and insulin resistance. Small studies involving therug metformin have suggested a reduction in the risk ofubsequent GDM, but data are insufficient to make wide-pread recommendations endorsing this strategy and furthertudies will be necessary to address safety and efficacy issues.ther candidate medications deserve further study. Othernowledge gaps include the lack of a definition of the meta-olic syndrome in pregnancy leading to the notion that GDMay be a manifestation of the metabolic syndrome duringregnancy. Further studies are also needed to address theptimal screening strategy to detect recurrent GDM in sub-equent pregnancies.

Finally, given the rising prevalence of prediabetic risk fac-ors, concerns are mounting that, without effective interven-ion strategies, the incidence of gestational diabetes as well asype 2 diabetes will continue to increase. Gestational diabetesellitus for many women represents only one stage in a con-

inuum of risks that presents many challenges and preventa-ive opportunities to health care systems. Further under-tanding of specific risks as well as their complex interactionsnd rates of progression may allow the identification of effec-ive strategies to break the cycle with the goals of not onlymproving pregnancy outcomes but also reducing longererm health risks for women and their children.

cknowledgmentshe author would like to express his gratitude to Naomi Spina

or her expert assistance in the preparation of this manuscriptnd John Smulian, MD, MPH, for his valuable editorial contri-utions.

eferences1. American Diabetes Association. Position Statement, Gestational Diabe-

tes Mellitus. Diabetes Care, Vol. 27, Supplement I, January 2004.
2. Philipson EH, Super DM: Gestational diabetes mellitus: does it recur in

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

184 J.N. Bottalico

subsequent pregnancy? Am J Obstet Gynecol 160:1324-1329,discussion 1329-1331, 1989

3. Gaudier FL, Hauth JC, Poist M, et al: Recurrence of gestational diabetesmellitus. Obstet Gynecol 80:755-758, 1992

4. Spong CY, Guillermo L, Kuboshige J, et al: Recurrence of gestationaldiabetes: identification of risk factors. Am J Perinatol 15:29-33, 1998

5. Major CA, DeVeciana M, Weeks J, et al: Recurrence of gestationaldiabetes: who is at risk? Am J Obstet Gynecol 179:1038-1042, 1998

6. MacNeill S, Dodds L, Hamilton D, et al: Rates and risk factors forrecurrence of gestational diabetes. Diabetes Care 24:659-662, 2001

7. Metzger BE, Coutsan DR: Summary and recommendations of theFourth International Workshop-Conference on Gestational DiabetesMellitus. The Organizing Committee. Diabetes Care 21:B161-B167,1998 (suppl 2)

8. National Diabetes Data Group: Classification and diagnosis of diabetesmellitus and other categories of glucose intolerance. Diabetes 28:1039-1057, 1979

9. Foster-Powell KA, Cheung NW: Recurrence of gestational diabetes.Austr NZ J Obstet Gynecol 38: 384-387, 1998

0. Danilenko-Dixon D, Annamalai A, Mattson L, et al: Perinatal outcomesin consecutive pregnancies discordant for gestational diabetes. Am JObstet Gynecol 182:80, 2000

1. Pleis JR, Senson V, Schiller JS: Summary Health Statistics for US Adults:National Health Interview Survey, 2000. National Center for HealthStatistics. Vital Health STAT 10, 2003

2. Schoenborn CA, Adams PF, Barnes PM: Body weight status of adults.United States, 1997-98. Advance data from vital and health statistics,no. 330, National Center for Health Statistics, 2002

3. Kimm S, Glynn NW, Kriska AM, et al: Decline in physical activity inblack girls and white girls during adolescence. N Engl J Med. 347:709,2002

4. Third report of the national cholesterol education program (NCEP)Expert panel on detection, evaluation, and treatment of high bloodcholesterol adults (Adult Treatment Panel III) final report. Circulation106: 3143-3421, 2002

5. Ford E, Giles W, Dietz W: Prevalence of the metabolic syndromeamong US adults: findings from the third National Health and Nutri-tion Examination Survey. J Am Med Assoc 287:356-359, 2002

6. Centers for Disease Control and Prevention. National Diabetes FactSheet. United States, 2005

7. American College of Obstetricians and Gynecologists. ACOG PracticeBulletin, No. 30, Gestational diabetes, September 2001

8. Catalano PM, Avallone D, Drago N, et al: Reproducibility of the oralglucose tolerance test in pregnant women. Am J Obstet Gynecol 169:874-881, 1993

9. Maser R, Lenhard MJ, Henderson BC, et al: Detection of subsequentepisodes of gestational diabetes mellitus, a need for specific guidelines.J Diabetes Complications 18:86-90, 2004

0. Sermer M, Naylor CD, Gare DJ, et al: Impact of increasing carbohydrateintolerance on maternal-fetal outcomes in 3637 women without GDM.The Toronto Tri-Hospital Gestational Diabetes Project. Am J ObstetGynecol 173:146-156, 1995

1. Sacks DA, Greenspoon JS, Abu-Fadil S, et al: Towards universal criteriafor gestational diabetes: the 75 gram glucose tolerance test in preg-
nancy. Am J Obstet Gynecol 172:607-614, 1995
2. Langer O, Mazze R: The relationship between large-for-gestational-ageinfants and glycemic control in women with gestational diabetes. Am JObstet Gynecol 159:1478-1483, 1988

3. Langer O, Rodriguez DA, Xenakis EM, et al: Intensified versus conven-tional management of gestational diabetes. Am J Obstet Gynecol 170:1036-1047, 1994

4. Langer O, Yogev Y, Most O, et al: Gestational diabetes: the conse-quences of not treating. Am J Obstet Gynecol 192:989-997, 2005

5. Crowther C, Hiller J, Moss J, et al: Effect of treatment of gestationaldiabetes mellitus on pregnancy outcomes. N Engl J Med 352:2477-2486, 2005

6. Radon P, McMahon MJ, Meyer WR: Impaired glucose tolerance inpregnant women with polycystic ovary syndrome. Diabetologia 94:194-197, 1999

7. Mikola M, Hiilesmaa V, Halttunen M, et al: Obstetric outcome in womenwith polycystic ovarian syndrome. Hum Reprod 16:226-229, 2001

8. Glueck M, Wang P, Kobayashi S, et al: Metformin therapy throughoutpregnancy reduces the development of gestational diabetes in womenwith polycystic ovary syndrome. Fertil Steril 77:520-525, 2002

9. Hanna F, Peters JR: Screening for gestational diabetes; past, present andfuture. Diabet Med 19:351-358, 2002

0. Verma A, Boney C, Tucker R, et al: Insulin resistance syndrome inwomen with prior history of gestational diabetes mellitus. J Clin Endo-crinol Metab 87:3227-3235, 2002

1. Kousta E, Lawrence N, Godsla I, et al: Insulin resistance and beta cellfunction in normoglycemic European women with a history of GDM.Clin Endocrinol 59:289-229, 2003

2. Lauenborg J, Mathiesen E, Hansen T, et al: The prevalence of themetabolic syndrome in a Danish population of women with previousgestational diabetes is 3-fold higher than in the general population.J Clin Endocrinol Metab 90:4004-4010, 2005

3. Albareda M, Caballero A, Badell G, et al: Metabolic syndrome at followup in women with and without gestational diabetes in index pregnancy.J Metab Clin Exp 54:1115-1121, 2005

4. Peters RK, Kjos SL, Xiang A, et al: Long term diabetogenic effect ofsingle pregnancy in women with previous gestational diabetes mellitus.Lancet 347:227-230, 1996

5. Diabetes Prevention Program Research Group: Reduction in the inci-dence of type 2 diabetes with lifestyle intervention or metformin.N Engl J Med 346:393-403, 2002

6. Buchanan T, Xiang A, Peters RK, et al: Preservation of pancreatic beta-cell function and prevention of type 2 diabetes by pharmacologicaltreatment of insulin resistance in high-risk Hispanic women. Diabetes51:2796-2803, 2002

7. Gerstein H, Yusuf S, Bosch J, et al: Effect of rosiglitazone on the fre-quency of diabetes in patients with impaired glucose tolerance or im-paired fasting glucose: a randomized controlled trial. Lancet 368:1096-1105, 2006

8. Kim C, Newton K, Knopp RH: Gestational diabetes and the Incidence oftype 2 diabetes: a systematic review. Diabetes Care 25:1862-1868, 2002

9. Russell M, Phipps M, Olson C, et al: Rates of postpartum glucose testingafter gestational diabetes mellitus. Obstet Gynecol 8:1456-1462, 2006

0. Van Assche F, Holemas K, Aerts L: Long term consequences for off-
spring of diabetes during pregnancy. Br Med Bull 60:173-182, 2001

AME

Tiaeewfkanead1tpd

*

†

S

A

0d

fter Shoulder Dystocia:anaging the Subsequent Pregnancy and Delivery

dith D. Gurewitsch, MD,*,† Tara L. Johnson, BS, BM,† and Robert H. Allen, PhD†

Among risk factors for shoulder dystocia, a prior history of delivery complicated byshoulder dystocia is the single greatest risk factor for shoulder dystocia occurrence, withodds ratios 7 to 10 times that of the general population. Recurrence rates have beenreported to be as high as 16%. Whereas prevention of shoulder dystocia in the generalpopulation is neither feasible nor cost-effective, intervention efforts directed at the partic-ular subgroup of women with a prior history of shoulder dystocia can concentrate onpotentially modifiable risk factors and individualized management strategies that canminimize recurrence and the associated significant morbidities and mortality.Semin Perinatol 31:185-195 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS prior shoulder dystocia, history of shoulder dystocia, recurrence, brachial plexuspalsy, injury prevention, risk management

dts

mpcfttshiecmvdufawcp

pcao

he occurrence of shoulder dystocia is largely consideredunpredictable. Epidemiologically, although diabetics’

nfants weighing more than 4 kg at birth are disproportion-tely at higher risk for shoulder dystocia compared with av-rage weight newborns,1-5 nearly half of shoulder dystociavents occur in nonmacrosomic infants of otherwise healthyomen.2,6 Although both antepartum and intrapartum risk

actors associated with occurrence of shoulder dystocia arenown,1,2 a substantial number of mother–infant pairs whoctually experience shoulder dystocia during delivery haveone of these risk factors, whereas many more deliver un-ventfully despite having several risk factors.7 However,mong risk factors for shoulder dystocia, a history of shoul-er dystocia in a prior delivery carries a recurrence risk of0% to 16%.8-10 Akin to the single greatest predictor of pre-erm delivery being a history of preterm delivery in a previousregnancy, a prior history of delivery complicated by shoul-er dystocia is indeed the single greatest risk factor for shoul-

Department of Gynecology and Obstetrics, Division of Maternal Fetal Med-icine, The Johns Hopkins University School of Medicine, Baltimore, MD.

Department of Biomedical Engineering, The Johns Hopkins UniversitySchool of Medicine, Baltimore, MD.

ome of the research reported in this review was funded by a grant from theCenters for Disease Control’s National Center for Injury Prevention andControl: Grants for Traumatic Injury Biomechanics Research Program04047: #CE00433-03. The contents of the article are the sole opinions ofthe authors and do not represent the opinions of the NCIPC.

ddress reprint requests to Edith D. Gurewitsch, MD, The Johns HopkinsHospital, 600 North Wolfe Street, Phipps 217, Baltimore, MD 21287.

iE-mail: [email protected]


er dystocia recurrence, with odds ratios 7 to 10 times that ofhe general population.8-10 Table 1 provides a summary oftudies describing recurrence risks.8-11

Unfortunately, the pathogenesis of shoulder dystocia isultivariate, screening for “at-risk” individual pregnancies isoorly predictive (positively or negatively) of its actual oc-urrence, and thus effective interventions for prevention areew. However, although avoidance of shoulder dystocia inhe general population is neither feasible nor cost-effec-ive,12-14 concentrating intervention efforts on the particularubgroup of women with a prior history of shoulder dystociaas distinct merit. First, the occurrence of shoulder dystocia

n a previous pregnancy indicates the need for a targetedvaluation. Second, review of the details of a woman’s spe-ific shoulder dystocia experience can elucidate potentiallyodifiable conditions and circumstances amenable to inter-

ention in a subsequent pregnancy. Third, even if shoulderystocia recurs, complications may be minimized by individ-alized management aimed at reducing or controlling thoseactors that predispose to significant morbidity and mortalityrising from shoulder dystocia. In focusing on the womenith a prior history of shoulder dystocia, this review will

oncentrate on those aspects of shoulder dystocia and injuryathophysiology that are knowable and hence modifiable.It is axiomatic that birth remains a moderately hazardous

rocess and that certain unintended and unavoidable out-omes will occur even when complications are anticipatednd managed appropriately. Importantly, it is the untowardutcomes of the condition and not necessarily the condition
tself that we wish to prevent. Indeed, avoidance of shoulder
185

dappefoUpndowboppot

RStPomidwaoubsnede

DRmaodepii

ydstrle

1S

umm

ary

ofth

eLi

tera

ture

that

Inve

stig

ates

Rec

urre

nce

ofS

houl

der

Dys

toci

aA

mon

gC

ohor

tsof

Wom

enw

itha

Rec

orde

dH

isto

ryof

Sho

ulde

rD

ysto

cia

ina

Pri

orliv

ery

Stu

dy

(Tim

eP

erio

d)

Bac

kgro

und

Rat

eof

SD

Pro

port

ion

Hav

ing

Pre

gnan

cyaf

ter

SD

N(%

)

Pro

port

ion

wit

hC

esar

ean

for

His

tory

of

SD

N(%

)

Rec

urr

ence

Rat

eof

SD

Am

ong

All

Sub

ject

sw

ith

His

tory

of

SD

(RR

)

SD

Rec

urr

ence

Rat

e/C

esar

ean

Rat

efo

rFai

led

Tri

alof

Lab

or

ith,

etal

.8(1

980-

1985

)0.

85%

51/2

03(2

5.1%

)5/

51(9

.8%

)9.

8%(R

R-1

7)11

.9%

/9.5

%w

is,

etal

.9(1

983-

1992

)2.

0%10

1/74

7(1

3.5%

)—

—(R

R-7

)13

.8%

/unk

now

nns

berg

,et

al.10

(199

3-19

99)

1.5%

73/6

02(1

2.1%

)7/

73(9

.6%

)15

.1%

(RR

-10)

16.7

%/9

.6%

rew

itsch

,et

al.11

(199

3-20

04)

1.9%

78/3

85(2

0.2%

)4/

78(5

.1%

)11

.5%

(RR

-6.3

)12

.2%

/12.

8%

brev

iatio

ns:

SD

,sh

ould

erdy

stoc

ia;

RR

,re

lativ

eri

sk.

186 E.D. Gurewitsch, T.L. Johnson, and R.H. Allen

ystocia recurrence per se should neither be expected nor sets a goal. However, its outcome may be significantly im-roved if those factors that increase the risk for harm arerepared for and managed. As with any emergency, logicalvidence-tested response to shoulder dystocia—either be-ore it occurs or once in progress—is directly correlated withutcome, whether it occurs for the first time or is recurrent.nfortunately, fully tested and evidence-validated ap-roaches to management specifically of the subsequent preg-ancy and delivery of a woman with a prior history of shoul-er dystocia do not exist. Nevertheless, empiric modificationf delivery method or of intrapartum management foromen with a history of prior shoulder dystocia appears toe commonplace8,10,11; however, without targeted emphasisn modifiable risk factors, such strategies have had little im-act on either recurrence or morbidity.10,11 Our goal is toropose a reasonable and logical approach to the managementf a subsequent pregnancy and delivery after a shoulder dys-ocia based on interpretation of the available literature.

ecurrence Preventiontarts Immediately After

he Initial Shoulder Dystociareventive strategies for the next possible shoulder dystociaccurrence begin with the proper management of the after-ath of the index event. The incidence of shoulder dystocia

s increasing.1,15,16 This may be due to better diagnoses, betterocumentation, or physical factors such as increasing birtheight and obesity. The same is true for shoulder dystocia-

ssociated brachial plexus injury.17 Fortunately, the majorityf shoulder dystocia events are inconsequential in terms ofntoward outcome. Still, up to 27% of shoulder dystocia wille complicated by some maternal and/or fetal injury, withignificant proportion (up to 10%) having attendant perma-ent sequelae.18 Whether the original shoulder dystociavent was mild, moderate, or severe, the occurrence of shoul-er dystocia at any delivery is worthy of documentation andvaluation.

ocumentationegardless of outcome, details of the shoulder dystocia—evenerely that it occurred—must be communicated to the mother

nd, by way of documentation in the medical record, to the nextbstetrician who will care for her. Ideally, each participant in theelivery’s emergency response should author a note about thevent itself and their role. However, there is no substitute for therimary delivering clinician’s giving a comprehensive account-

ng of each aspect of the management and its results. Itemsmportant to document are listed in Table 2.

Acker proposed a shoulder dystocia intervention form 15ears ago that was in keeping with the idea that this degree ofetail and documentation is not only important, but should betandardized.19 The merits of documenting all of this informa-ion from risk management and medicolegal perspectives areeviewed elsewhere.20-22 However, the relevance of such detail

to assessment of recurrence risk should not be underestimated.Tab

De

Sm

Le Gi

Gu

Ab

Epatpf

RDSp

ftfpdbtbsdosvienr

ttstpttsrtdr(aosp

mtTttptifitcdbtitnmgor

Tcm

After shoulder dystocia 187

ach aspect provides potential clues to the pathogenesis of thatarticular shoulder dystocia and, regardless of any retrospectivessessment of the specific management’s propriety or prudence,horough review of the precise sequence of events may identifyotentially modifiable contributors to recurrence and outcomeor that particular patient.

elating Deliveryetails to Pathophysiology

houlder dystocia results from heterogeneous processes that

able 2 Documentation Suggested after a Delivery Compli-ated by a Shoulder Dystocia That May Help with Manage-ent of a Subsequent Pregnancy

Second stage of labor detailsDurationTime spent pushingTime the head was on the perineum prior to initiating

tractionTotal head to body delivery time

Details of any operative interventionsType (forceps or vacuum)IndicationsStation initiatedAsynclitism (if present)Rotation of the head (if performed)

Anesthesia detailsUse of any prophylactic positioning or maneuversPosition of head and side of restitution (manual or

spontaneous)Timing and type of episiotomy with extensionsHow and when shoulder dystocia was recognized (eg,

turtle sign, with first attempt at traction)Maneuvers (eg, prophylactic McRoberts’) and traction

prior to diagnosisEvidence for compound presentationOther procedures

Nuchal cord managementSuctioning of oropharynxMeconium management

Personnel presentDescription of each maneuver performedManeuver documentation

Type and sequence of maneuversWho performed which maneuversTraction direction and orientation of head and thoraxEffectiveness of each maneuverFetal response to each maneuverFinal successful maneuverTiming of management choicesClear descriptions of right versus left shoulder and

anterior versus posterior shoulder (eg, which wasinitially anterior, which was manipulated directly,which delivered first)

Birth weightApgar scoresArterial and venous cord gasInfant injury (if any)Maternal injury (if any)

roduce the same clinical effect. Among impaired or dys- g

unctional births characterized by fetopelvic disproportion,here is a continuum of obstruction wherein delivery of theetal trunk is impeded following completed (or near com-leted) delivery of the fetal head. Whereas the most commonefinition of shoulder dystocia requires the presence of aony impaction preventing spontaneous delivery of the fetalrunk, its pathogenesis derives from variable contributions ofoth size and positional incompatibilities between fetalhoulder and maternal pelvic dimensions. Some shoulderystocia events are caused mainly by positional misalignmentf only marginally incompatible shoulder and pelvic dimen-ions (usually of a normal-sized infant and normal-sized pel-is). Others arise from a significant shoulder-to-pelvis sizencompatibility necessarily producing a “tight fit” betweenither a large-for-gestational age infant (most common) and aormal-to-large pelvic cavity or a normal-sized infant and aeduced-capacity pelvis (least common).

In a normal delivery, a winding, forward-progressing mo-ion occurs as the head and body traverse the birth canal. It ishis rotational-type motion that accomplishes delivery of thehoulders following emergence of the head through the in-roitus. This is effectively an anatomical requirement im-osed by the typical arrangement of the bony boundaries ofhe pelvic inlet, midpelvis, and pelvic outlet, which is akin tohreads of a screw. Thus, just after delivery of the head, thehoulders often must come to lie in an oblique orientationelative to the maternal pelvis before they can to descend intohe midpelvis and deliver through the pelvic outlet. Shoulderystocia occurs when the shoulders retain the anteroposte-ior orientation they had assumed on the head’s initial entryusually in occiput transverse position) into the pelvic inletnd are unable to (or have yet to) rotate spontaneously toccupy the oblique diameter of the pelvis. Subsequently, thehoulders fail to deliver from behind the maternal pubic sym-hysis.Positional misalignments resulting in shoulder dystociaay also be produced by incomplete entry to the hollow of

he sacrum by the posterior shoulder as the head delivers.his may occur when there is a compound presentation with

he posterior arm beneath the posterior shoulder, from rela-ive lordosis of the maternal lumbosacral spine or from arominent sacral promontory impeding descent of the pos-erior shoulder. The anterior shoulder then “rides high” andmpacts behind the pubic symphysis. Another cause is insuf-cient time allotted to permit spontaneous shoulder rotationo the normal oblique orientation relative to the pelvis. Thisan occur following precipitous delivery of the head pro-uced either by increased compliance of pelvic tissues or elsey instrumented deliveries. This can even occur simply byoo-hasty application of traction to the head immediately onts emergence from the birth canal (eg, “to keep the momen-um going”) without awaiting the next contraction. It is alsooteworthy that shoulder dystocia is disproportionatelyore common among anomalous or stillborn fetuses, sug-

esting that physiologic negotiation of the pelvic inlet andutlet is indeed a dynamic process, facilitated by normal fetalesponsiveness.16 It is likely that any of these antecedents to a
iven shoulder dystocia can lead to an “accidental” or dynam-

istpwitt

spsmldmvgppoemuciwapcv

ptotgetitaoTascstotawbTcs

dn

RoTstwatfhaamd(asomsd

ldafsse

ehewspSwdjffcetPhaeppop


cally determined misalignment of otherwise compatible fetalhoulder and maternal pelvic dimensions. This latter situa-ion is probably associated with a lower recurrence risk com-ared with shoulder dystocias from other causes. For theseomen, a recurrence may best be prevented by avoiding

nstrumented deliveries in the future or by patiently allowinghe fetal head to spontaneously rotate to the oblique diame-er, before traction is applied to the head and neck.

More severe shoulder dystocia occurs when the anteriorhoulder becomes tightly impacted behind the pubic sym-hysis. In the most severe cases, the posterior shoulder isimultaneously impacted on the sacral promontory. Theost common cause for this type of shoulder dystocia is a

arge-for-gestational age or macrosomic fetus, whose shoul-er width is typically 14 cm or more.23 Less commonly, theaternal pelvis may be contracted or have a greater trans-

erse dimension at the pelvic outlet than the obstetric conju-ate, as in the platypelloid pelvis. Acquired deformities of theelvis, such as coccygeal fracture from either trauma or arevious delivery, may also cause such difficulty in deliveryf the shoulders. It is these types of shoulder dystociavents—those encountered in significantly size-discrepantother–infant dyads—that are associated more often withntoward outcome.24 It is also these types of shoulder dysto-ia that would be more likely to recur in subsequent deliver-es. The approach to such women in the next pregnancyould focus on careful monitoring of fetal growth antenatally

nd timely delivery for some; others with narrow or reducedelvic capacity may be appropriately managed with plannedesarean birth. We return to these strategies later in this re-iew.

The apparent simplicity of the above distinction betweenositionally determined and size-determined pathogenesis ofhe index shoulder dystocia is misleading. The contributionsf positional misalignment and true fetopelvic disproportiono the occurrence of shoulder dystocia at a given delivery in aiven woman are variable and are by no means mutuallyxclusive. A grand multipara who experiences shoulder dys-ocia after a precipitous second stage labor and delivery of annfant comparably sized to those she has delivered severalimes before without incident may have been caused by ancquired pelvic deformity that prevented normal occupancyf the hollow of the sacrum by the infant’s posterior shoulder.hus, an apparent dynamically determined positional mis-lignment was also impacted on by narrowed pelvic dimen-ions. Similarly, an unanticipated finding of a randomizedontrolled trial of forceps versus vacuum delivery was thathoulder dystocia followed the vacuum deliveries more oftenhan it followed forceps deliveries.25 One explanation for thisutcome was that the vacuum may be more apt than forcepso be successfully applied to the head of a macrosomic infantlready maximally occupying the pelvic cavity since, unlikeith forceps, the vacuum instrument itself does not have toe accommodated between the head and the pelvic sidewalls.hus, shoulder dystocias seemingly attributable to insuffi-ient time for shoulder rotation following instrumented as-
istance with delivery of the head often also involve true size p
iscrepancy between a large-for-gestational-age fetus andormal maternal pelvic capacity.

etrospective Evaluationf Pelvic Capacity and Infant Biometryhe contour of the pelvis undergoes modification and expan-ion during pregnancy and delivery by a progressive laxityhat develops in the cartilage between the pubic rami andithin the sacroiliac joints. Thus, the prospective de novo

ssessment of the likelihood of shoulder dystocia by tradi-ional clinical or even medical image-derived pelvimetry per-ormed in either a nullipara or someone who previously hasad uneventful deliveries is bound to have a low yield. It islso probable that the simple assessment of pelvic diameterss done in traditional pelvimetry may be too crude an assess-ent of pelvic capacity where propensity toward shoulderystocia is concerned. It is unknown whether volumetricthree-dimensional) or surface-rendered (topographical)nalysis of bony contours rather than geometric (two-dimen-ional) assessment could better differentiate women withtherwise “normal” or “adequate” pelves who nonethelessay have a predilection for shoulder dystocia from similar-

ized pelves of women who are unlikely to have shoulderystocia.Similarly, not all fetuses of the same weight, even when

arge-for-gestational age, are equally predisposed to shoulderystocia. The notion of macrosomia refers not simply to anrbitrary weight cut-off, but rather a disproportionately largeetal trunk relative to the head. Thus, anthropomorphic mea-urements of an infant whose delivery was complicated byhoulder dystocia may also yield some insight into pathogen-sis.

Table 3 depicts the results of traditional and novel pelvim-try and corresponding infant biometry performed within 24ours of delivery on three mother–infant pairs who experi-nced shoulder dystocia.26 Based on clinical history aloneithout benefit of postpartum pelvimetry, it would be rea-

onable to surmise that patient 1 must have a “borderline”elvis where large-for-gestational-age fetuses are concerned.he was unable to deliver a 10-lb baby at all, and a babyeighing just less than 9 lbs experienced shoulder dystociauring delivery. Patient 2, on the other hand, by clinical

udgment and without benefit of retrospective pelvic andetal biometry, could have been judged to be at reduced riskor developing shoulder dystocia based on her documentedompliance with blood sugar management and ultrasound-stimated fetal weight that was in a reasonable range.27 Yet, ofhe three shoulder dystocia cases, hers was the most severe.atient 3 was only modestly compliant with management ofer blood sugar, as demonstrated by sonographic evidence ofccelerated fetal growth by the time she reached term. How-ver, given the extensive history of her “tested pelvis” forresumably similarly “macrosomic” (trunk-to-head dispro-ortionate) infants in the past and the relatively mild naturef the shoulder dystocia she experienced in the context of herrecipitous delivery, it might be reasonable to suspect that
ositional misalignment resulting from precipitous delivery

wt

mlhocsomsd

waor

npidrncdmssao

atb

TD

C

S

I

BBL

S

I

ISR

VS

*


as a greater factor in her shoulder dystocia pathogenesishan true fetopelvic disproportion.

It is noteworthy that all three infants had similar bisacro-ial widths. Babies 1 and 3 had proportionate upper and

ower body anthropomorphic dimensions, whereas Baby 2ad markedly asymmetric shoulder girth despite normalverall body weight and ponderal index, the latter of which isonsistent with well-controlled gestational diabetes.28 Retro-pectively, the traditionally derived sonographic estimationf fetal weight of Baby 2 obtained before delivery overesti-ated the actual birth weight, yet did not fully detect this

houlder width asymmetry that perhaps might have beenetectable by additional nonstandardized means.29-32

Anecdotally, patient 1 had a subsequent pregnancy thatas scheduled for a repeat cesarean section at 39 weeks to

void recurrent shoulder dystocia. However, she presentednly a few hours before her scheduled operation already in

able 3 Clinical and Computerized Tomographic Pelvimetricelivery Complicated by Shoulder Dystocia

Case 1

linical presentation VBAC, prior c-sec for CPDPrior infant 4318 gBMI 46.9, non-diabetic,

28-lb wt gain, EFW3800 g, 39-wk NSVDafter 27-min secondstage

houlder dystocia severity Head-to-body interval 70seconds

Required McRoberts’,posterior Rubins’ andModified Woods’ Screw

mmediate neonatal outcome Apgars 9/9, cord pH 7.3,facial bruising, nofractures or neurapraxia

irth weight (g) 3933isacromial width (cm) 14.33argest transverse diameter(normal > 12 cm)

15.76

acral promontory to top ofpubic symphysis (normal>11cm)

11.63

nterspinous distance (normal>10 cm)

12.47

ntertuberous distance (cm) 14.71ymphyseal separation (cm) 1.82ight/left sacroiliac jointspace (cm)

0.49/0.42

olume (cm3)* 1205urface area (cm2)* 2141

Novel pelvimetry was derived using 3D Doctor™ and reconstructinsurface-fitting of a mesh to the images. Results were validatedradiologist interpreting the CT scans.

apidly progressive labor. She successfully delivered vagi- p

ally without shoulder dystocia a 3945-g infant, similar to therevious child who developed shoulder dystocia. Interest-

ngly, the pelvimetric assessment following her initial shoul-er dystocia delivery was consistent with a low recurrenceisk since it suggested that a true fetopelvic disproportion wasot evident. In light of the retrospective evaluation of pelvicapacity and infant biometry following this patient’s shoulderystocia delivery, it is more probable that dynamic positionalisalignment was more contributory to the pathogenesis of

houlder dystocia and her 27-minute second stage was moreignificant in producing this than was any presumed discrep-ncy in fetal shoulder and maternal pelvic dimensions basedn her prior cesarean delivery indication.By contrast, the shoulder dystocia deliveries of patients 2

nd 3 were more likely to have resulted from size discrepancyhan from positional misalignment of otherwise compatibleony dimensions between mother and fetus. In the case of

from Mother–Infant Dyads Examined Within 24 Hours of

Case 2 Case 3

rimigravida, insulin-requiringGDM, euglycemic duringpregnancy, BMI 34.5, 33-lbwt gain, EFW 3900 g,induction at 39 wk, NSVDafter 70-min second stage

Grand grand multipara(10th delivery),recurrent insulinrequiring GDM, largestprior infant 3973 g, noprior history shoulderdystocia, 3476-g sonoEFW @ 37 wk,precipitous delivery onarrival to hospital at 38wks

ead-to-body interval 2minutes

equired McRoberts’,suprapubic pressure, andWoods’ Screw

Head-to-body interval 45seconds

Required McRoberts’ andsuprapubic pressure

pgars 7/9, no neonatalcomplications

Apgars 8/9, temporaryErb’s palsy

613 4,8804.83 14.672.09 11.72

0.41 11.21

0.57 11.18

3.36 13.53.62 0.64.34/0.48 0.58/0.62

14 1463650 2690

mages in Matlab™ via intensity-based segmentation, stacking, andparing 3D Doctor™ measurements against those derived by the

Data

P

H

R

A

311

1

1

100

71

g CT iby com

atient 2, postpartum CT pelvimetry obtained after her index

ssGtpe

rGtut“ePtoracm

mdt

d(tTdtacttpnfattfitteaw

PItjaoewiotfiwdsfe

ipcscpmpstehd

Fiwstawddr


houlder dystocia event revealed a reduced pelvic capacity,uggesting that she and her infant were size incompatible.iven these findings, it would seem reasonable to propose

hat recurrence of shoulder dystocia would be likely in thisatient, and she would be a good candidate for a scheduledlective cesarean delivery in her next pregnancy.

Patient 3’s postpartum CT evaluation (Fig. 1) shows aather capacious pelvic cavity by standard measurements.iven the clinical history alone, it would appear as though

he shoulder dystocia was more likely “sporadic” and attrib-table to positional misalignment in this patient. Evenhough her infant was large-for-gestational age, it was notmacrosomic” as defined by trunk-to-head disproportion (orven by criteria for diabetic gravida given in the 2002 ACOGractice Bulletin on Shoulder Dystocia). Indeed, her antena-al estimated fetal weight was at the 72nd percentile. Basedn this, it might have been reasonable to surmise that recur-ence of shoulder dystocia based on fetopelvic size discrep-ncy would be relatively low in this patient, and potentiallyould be modified further with improved diabetic manage-ent to avoid excess neonatal adiposity.However, a surprise finding of an acquired pelvic defor-ity from a healed coccygeal fracture (likely sustained at theelivery that immediately preceded the index shoulder dys-

igure 1 Postpartum CT pelvimetry following shoulder dystocia. CTmages were obtained from a grand grand multiparous (para 9)oman immediately following delivery, which was complicated by

houlder dystocia. (A) Largest transverse diameter. (B) Anteropos-erior diameter (sacral promontory to top of pubic symphysis); im-ge shows marked angulation of sacrococcygeal joint consistentith healed fracture. (C) Interspinous diameter. (D) Intertuberousiameter. Acquired pelvic deformity, likely sustained during earlierelivery, probably contributed to “high-riding” anterior shoulderesulting in shoulder dystocia.

ocia delivery) suggests a different pathogenesis for shoulder c

ystocia. The sharp angulation of the sacrococcygeal jointFig. 1b) likely impeded normal occupancy of the hollow ofhe sacrum by the posterior shoulder in the index delivery.his, in turn, would have caused the anterior shoulder to beisplaced forward and cephalad and become lodged behindhe pubic symphysis in an anteroposterior orientation. Suchsacrococcygeal deformity could potentially reduce the effi-acy of first-line shoulder dystocia maneuvers that still main-ain the shoulders in the anteroposterior orientation.33 Addi-ionally, given the borderline small transverse diameter of herelvic inlet, the anteroposterior orientation is perhaps moreaturally predisposed to being the “path of least resistance”or fetal descent in this patient. Thus, the combination of hercquired deformity and her anthropoid-type pelvic architec-ure makes her likely to experience recurrent shoulder dys-ocia in future deliveries. A planned cesarean delivery foruture pregnancy should be considered. If a vaginal deliverys planned, then antenatal management should focus on bet-er antenatal control of fetal growth and, with awareness ofhe potential interference by the pelvic deformity, the deliv-ring clinician can opt to prioritize rotational type maneuverss initial management for any recurrent shoulder dystocia,hich could reduce the risk of injury.34

atient Counselingt is often counterintuitive or seemingly premature to attempto counsel a patient regarding the possible pathogeneses of aust-experienced shoulder dystocia or to prognosticate eitherbout its possible recurrence or about the long-term outcomef any associated injuries. Nevertheless, it is important toxplain, in a candid and honest manner, what transpired,hat are the possible causes, and what might be anticipated

n the future and to begin to involve the patient actively in herwn and her child’s future care. Perhaps precisely becausehere are so many unknowns about exact pathogeneses, riskactors, prognostic indicators, and future management plans,t is critically important to educate the woman whose deliveryas complicated by shoulder dystocia about the often unpre-ictable and unpreventable nature of the complication, en-ure her understanding that it is a significant issue as far asuture childbearing is concerned, and elicit the patient’s pref-rences and priorities when planning future deliveries.

Until more information is amassed through epidemiolog-cal and clinical studies, it is reasonable to encourage theatient’s cooperation in investigating possible causes andommitment to compliance with proposed managementchema that would aim to ameliorate potentially modifiableontributors to risk. However, the inability to guarantee aarticular outcome or to entirely avoid risk in any chosenode of delivery should always be explained clearly. De-ending on a particular patient’s degree of risk aversion, shehould be encouraged to consider how willing she would beo maintain a flexible, dynamic, and evolving plan for deliv-ry throughout the next pregnancy and perhaps even duringer next trial of labor. Knowledge of the occurrence of shoul-er dystocia, even if otherwise uneventful in terms of diffi-
ulty of management or associated morbidity, is important

fw

TBFIdpsasnfqi

iacratrusotnrotSwasl

wr

PMAvnwfswsfwbp

dAngdpcapcaodw

wtsa

T

MMWS

O

*

*


or the empowerment of both the patient and the obstetricianho will deliver her next child.

he Interval Period:efore the Next Pregnancy

ollow-Up of Infant Statusf the infant had sustained an injury at the index shoulderystocia delivery, whether skeletal or neurologic, the im-ortance of follow up of the infant’s condition must betressed to the patient (to optimize long-term outcome18),s well as to the delivering clinician. Not only is ongoingensitive and caring communication with the family sig-ificant from a risk management perspective,35,36 the in-ormation is also invaluable to management of the subse-uent pregnancy and delivery. Parental input is extremely

mportant in this process.For the delivering clinician, knowledge of whether an

nfant that was discharged from the newborn nursery withn as-yet unresolved brachial plexus palsy eventually re-overs completely or requires surgical intervention andemains with a permanent deficit is critical to retrospectivessessment of the management of the index shoulder dys-ocia. Whereas many brachial plexus injuries of a tempo-ary or even mild permanent nature (eg, restricted to thepper plexus with only mild functional deficit in thehoulder’s active range of motion) may have myriad etiol-gies (including malpositioning in utero),37 shoulder dys-ocia-associated brachial plexus injuries involving allerve roots from C5 to C8/T1 and/or avulsion of any nerveoot from the spinal cord must have involved some degreef externally applied lateral traction of sufficient magni-ude and rate to produce an injury of such an extent.38-41

etting aside whether the degree of traction applied wasithin the standard of care, the contribution of externally

pplied traction to the eventual outcome cannot, andhould not, be denied. Rather, it behooves the clinician to

able 4 Comparison of Risk Factors Between Mild and Sever

Mild ShouldeDystocia

N Val

aternal height (cm) 266 162.9aternal weight (kg) 274 87.4eight gain (>15.9 kg) 250 100 (

econd stage abnormality 279 125 (Prolonged second stage 279 43 (Precipitous second stage 279 82 (perative vaginal delivery 290 93 (Forceps 290 36 (Vacuum 290 47 (Combined 290 10 (

Severe shoulder dystocia defined as have met at least one of the folldeliberate proctoepisiotomy or direct manipulation of the fetus (eof neonatal depression at birth (5-minute Apgar <7 and/or arter

*Mean � SD or N (%).

earn from the experience, and familiarize him/herself h

ith shoulder dystocia management techniques that couldeduce necessary traction at any delivery.34,42,43

reconceptualanagement of Maternal Risk Factors

mong women with a history of shoulder dystocia in a pre-ious pregnancy, predictors of recurrence seem to vary. Sig-ificant differences, as well as lack thereof between womenith and without recurrent shoulder dystocia, have been

ound for such variables as parity at index delivery (eg, prioruccessful delivery without shoulder dystocia), maternaleight or diabetic status, operative delivery, and length of

econd stage.8-10 However, a consistent correlation has beenound between subsequent shoulder dystocia and high birtheight in the subsequent pregnancy, as well as comparativeirth weight with the index pregnancy, especially in diabeticregnancies.44

The most recognized antepartum risk factors for shoulderystocia are also the same risk factors for fetal macrosomia.side from a history of shoulder dystocia in a prior preg-ancy, other risk factors include maternal obesity, weightain during pregnancy of more than 35 pounds, gestationaliabetes and pregestational diabetes without vascular com-lications, and postdatism.21 As shown in Table 4, severeompared with mild shoulder dystocia is more likely to bessociated with obese gravidas weighing more than 200ounds or who gain excessive weight during pregnancy. If alinician is fortunate enough to have a motivated patient withhistory of shoulder dystocia who wishes to reduce the riskf fetal macrosomia before her next pregnancy, recommen-ation of weight reduction may be the single interventionith the greatest impact.Screening for occult diabetes is also reasonable. Some

omen with a history of a false-positive glucose challengeest may well be predisposed to macrosomia, suggestingome degree of impaired glucose tolerance not yet manifesteds overt diabetes.45 In most instances, these women likely

ulder Dystocia56

Severe ShoulderDystocia*

P valueN Value**

143 162.7 � 6.8 0.82148 93.7 � 22.0 0.002138 81 (58.7) 0.0004154 61 (39.6) 0.30154 23 (14.9)154 38 (24.6)170 62 (36.4) 0.34170 13 (7.6)170 38 (22.3)170 11 (6.5)

three criteria: 1) head-to-body interval >90 seconds; 2) use of eithertional maneuvers, delivery of the posterior arm); and/or 3) evidence

pH <7.1).

e Sho

r

ue**

� 6.7� 18.440.0)44.8)15.4)29.3)32.1)12.4)16.2)3.4)

owingg, rotaial cord

ave some degree of a metabolic syndrome or excess adipose

tolsoo

AoEAJamttdasttgwpwtgtpigaae

i(ggs1fwc

2erdws3cot

cg

anlipdo

AIeftwtava

gedf

dieampv

PThcpepsoPosd

wfpstwu


issue that contributes to hyperinsulinemia, a precursor tovert diabetes. Proper nutrition and exercise are well estab-ished as effective modifiers of this effect. A commitment toustained lifestyle changes may improve future pregnancyutcome, as well as the woman’s long-term health statusverall.

ntenatal Managementf the Subsequent Pregnancyliciting the History andssessing Patient-Specific Recurrence Risk

ust as it is important for the delivering clinician who man-ged a given patient’s shoulder dystocia to carefully docu-ent and communicate to the patient the details of the event, so

oo should any practitioner meeting an obstetric patient forhe first time specifically attempt to elicit a history of shoulderystocia in a prior pregnancy. The major antepartum (listedbove) and intrapartum risk factors for shoulder dystociahould be assessed. Intrapartum risk factors include opera-ive vaginal delivery and abnormal second-stage length, ei-her prolonged or precipitous. Even if there is no history ofestational diabetes, the patient should be asked to recallhether she had a false-positive glucose challenge test45 orerhaps had not been screened. If the infant whose deliveryas complicated by shoulder dystocia had weighed more

han 4 kg at birth, this is reason enough to perform an earlylucose screen at registration. A specific discussion of nutri-ion and monitored weight gain is warranted, especially if therevious infant was large-for-gestational age. Vigilance for

mpaired glucose tolerance leading to accelerated fetalrowth, with dietary modification and flexible medical ther-py as needed,46,47 should be a mainstay of antenatal man-gement throughout the next pregnancy of any such patient,specially one with a history of prior shoulder dystocia.

For those women whose early glucose screen is negative, its important to rescreen at the appropriate gestational age26-28 weeks). For those with a false-positive 1-hour 50-glucose screen (normal 3-hour glucose tolerance test) early inestation, repeat “screening” in the early third trimesterhould be by the diagnostic 3-hour test rather than the-hour screening test. This is based on the principle that aalse-positive screening test may not be sensitive enough,hen repeated in a given patient, to consistently detect the

ondition being screened for.Patients who have a normal glucose screening test at 26 to

8 weeks who later develop either clinical or sonographicvidence of accelerated fetal growth should be considered foretesting again at 30 to 34 weeks since there may be delayedetection of up to 15% of gestational diabetics. For thoseomen with a history of shoulder dystocia who, in the sub-

equent pregnancy, manifest one abnormal value on the-hour glucose tolerance test obtained after an abnormal glu-ose screen at 26 to 28 weeks, retesting should probablyccur at 30 to 34 weeks, regardless of growth parameters by
hat stage. If there is any maternal or fetal suggestion of glu- i
ose intolerance, aggressive management of diet and bloodlucose is warranted.

For those women with a history of shoulder dystocia whore diagnosed with gestational diabetes in a subsequent preg-ancy, it is especially important to monitor blood glucose

evels at several time points each day. A low threshold fornitiating oral hypoglycemic or insulin therapy may be ap-ropriate in these gravidas as well, especially if there is evi-ence of accelerated fetal growth while on diet despite reportf normal blood glucose levels.47

ssessment of Fetal Growtht is well established that the margin for error for ultrasoundstimation of fetal weight is too large to be relied on heavilyor predicting delivery complications.48 However, whenhere has been a previous shoulder dystocia (similar to casesith gestational diabetes), serial ultrasound measurements of

he trajectory of fetal growth are often more informative thansingle growth estimation for assessing the need for inter-

ention. Particular attention should be given to evidence ofsymmetrical growth using ponderal indices.

The most common way to assess for accelerated truncalrowth is the head circumference-to-abdominal circumfer-nce ratio.49 If below 0.9 near term, the risk for shoulderystocia may be increased, even when the overall estimatedetal weight has not surpassed the 90th percentile.

For the gestational diabetic apparently well managed oniet alone, there is some evidence to suggest that empiric

nitiation of insulin therapy once the abdominal circumfer-nce exceeds the 75th percentile may curtail further acceler-ted fetal growth.46 Although not studied prospectively, thisay be reasonable in a gestational diabetic with a history of arevious shoulder dystocia who desires to have a subsequentaginal delivery.

lanning the Delivery: Mode and Timinghe practice of inducing labor for “impending macrosomia”as no proven benefit. Such practices increase the rate ofesarean delivery performed for failed induction without im-acting the incidence of shoulder dystocia.50-52 This lack ofvidentiary support for early “elective” inductions and theersistent inability to shrink the margin of error in ultra-ound estimation of fetal weight14 led the American Collegef Obstetricians and Gynecologists in 2002 to revise theirractice Bulletin entitled “Shoulder Dystocia,” raising the cut-ffs for estimated fetal weight above which primary cesareanection without trial of labor may be offered, to 4500 g for theiabetic gravida and to 5000 g for the nondiabetic gravida.However, these recommendations may not apply for some

omen. There is still some evidence that lower estimatedetal weight thresholds for cesarean birth may be appropriate,articularly in women with “preexisting” predilection forhoulder dystocia.53,54 Although this evidence is not specifico the woman with a history of shoulder dystocia, theseomen may benefit the most from lower thresholds. Contin-ed research is needed to more appropriately individualize

ntrapartum management for such women.

uabafsagscStergtpltwvwttdp

ASSSsmtflgmdrtts

csstfvfstaa

ptItstwnwbs

ihSsbtcaccnmtccds

CCaascpdrnsdtBatmottndrso


Regardless of the specific estimated fetal weight cut-offsed to offer a primary cesarean section, only a small percent-ge of parturients will ever meet these criteria. What then cane said about the strategy of elective induction of labor in anttempt to ensure a lower birth weight than would resultrom expectant management and, thereby, potentially avoidhoulder dystocia? Whereas ineffective as a strategy whenpplied to the general population, this may have merit in theravida with a history of prior shoulder dystocia. First, sincehe is already parous, her risk of failed induction is signifi-antly lower compared with her nulliparous counterpart.econd, postdatism is among the potentially modifiable an-epartum risk factors for shoulder dystocia that would beliminated by elective induction at term. Third, the risk ofecurrence of shoulder dystocia correlates with similar orreater birth weight at subsequent delivery compared withhe index shoulder dystocia delivery.10 Finally, and most im-ortantly, it is the severe shoulder dystocia, which is more

ikely to occur in obese gravida with macrosomic infants,55

hat is most predictive of subsequent injury,56 and that weould most wish to prevent or mitigate. However, these se-ere shoulder dystocias occur in the same type of patient inhom complications of cesarean delivery are also more likely

o occur. Therefore, a strategy of early induction of labor aterm, before development of substantial trunk-to-head sizeiscrepancy, may present a balanced alternative to electiverimary cesarean delivery in these women.

nticipatinghoulder Dystocia Recurrenceimulation-Based Training and Rehearsal

ince the exact threshold for permanent injury in in vivohoulder dystocia is not known, the goal in shoulder dystociaanagement should be to reduce uterine force and clinician

raction as much as possible. Maternal pushing and uterineorces can only be controlled in a limited way; these also haveimited contribution to injury.57 Therefore, training and on-oing research must focus on how clinician-applied tractionight be reduced.43 This is especially important for shoulder

ystocia management because the very natural, unconsciousesponse when attempting to deliver the fetus once first at-empts have failed is to increase traction on subsequent at-empts.33,43,58 This training can be accomplished throughimulation-based training and with drills.

Already addressed in other fields of medicine, obstacles toomparative research and provider training within the clinicaletting are best overcome by use of medical simulation.59,60 Forurgically oriented skills, including vaginal delivery techniques,here is no substitute for mechanical simulation to allow hapticeedback and biofidelic learning for the student obstetric pro-ider.61,62 Real-time measurement of actual clinician-appliedorces during simulated shoulder dystocia deliveries—where as-ociated mechanical fetal response can be measured prospec-ively in a fetal model—may enable clinicians to self-assess moreccurately the magnitude, direction, and rate of traction they
pply during delivery.63 i
Simulation-based experimentation has also proven the hy-othesis that rotational maneuvers to resolve shoulder dys-ocia require less applied force than McRoberts’ positioning.ndeed, brachial plexus stretch is also reduced by a factor ofhree with use of Rubin’s maneuver as an initial approach tohoulder dystocia management.34 The comparative advan-age of fetal maneuvers for reducing brachial plexus strainas also predicted computationally.64 Clinicians are enjoinedot to fear fetal manipulation, but to familiarize themselvesith the techniques by availing themselves of simulation-ased training and even to practice routine assessment of fetalhoulder position by direct palpation at every delivery.65

The likelihood of favorable outcome of shoulder dystocias maximized by preparation and coordination of a well-re-earsed response by all members of the health care team.uch resource utilization and management is best ensured byystematic rehearsal.66 Communication and continuous feed-ack to team members regarding effectiveness of interven-ions or lack thereof is essential to this effort. Otherwise,ross-purposeful actions of two or more team members can,lbeit unintentionally, increase the risk for injury. Importantomponents of shoulder dystocia management include aontinuous assessment of the success or failure of each ma-euver, employing an alternative maneuver within approxi-ately 30 seconds of a previous failed maneuver,67 and main-

aining a calm and unhurried approach by the primarylinician. Each of these elements, as well as how and when toommunicate with the family during and after a shoulderystocia, can be rehearsed during drills until they progressmoothly.

onclusiononsidered one of the greatest fears of obstetric providers,ssociated with considerable risk for injury to both mothernd fetus and fraught with potential liability for the clinician,evere shoulder dystocia is an emergency that no one wouldare to relive. Thus, after a woman has experienced the com-lication, managing the risk of its recurrence in a subsequentelivery is desirable and prudent. Indeed, unlike other spo-adically occurring and unpredictable complications of preg-ancy, recurrence of shoulder dystocia is not infrequent. Yethoulder dystocia’s definitive prevention, namely cesareanelivery, while expedient and facile is also costly and poten-ially risky,54,68 especially in the obese and diabetic gravida.ecause many women choose to attempt a vaginal birth aftershoulder dystocia, it is important to use what information

here is in the literature to form a rational and reasonableanagement plan for risk modification and for management

f a shoulder dystocia event should it recur. At least some ofhe risk of shoulder dystocia recurrence indeed may be quan-ifiable and potentially modifiable during subsequent preg-ancies of individual women with a history of shoulderystocia. Importantly, avoidance of shoulder dystocia recur-ence per se should neither be expected nor used as a mea-ure of success in management. As with other medical andbstetric conditions at higher risk for morbidity and mortal-
ty, it is the untoward outcomes of shoulder dystocia and not

nTiit

ATHdv

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4


ecessarily shoulder dystocia itself that we wish to prevent.hus, circumstances and conditions that increase the risk of

njury from shoulder dystocia should be targeted, and if theres recurrence, the goal becomes the atraumatic resolution ofhe shoulder dystocia.69

cknowledgmentshe authors wish to thank Dr. Elliot Fishman of The Johnsopkins University School of Medicine’s Department of Ra-iology for his assistance with interpretation of the CT pel-imetry.

eferences1. Acker DB, Sachs BP, Friedman EA: Risk factors for shoulder dystocia.

Obstet Gynecol 66:762-768, 19852. Benedetti TJ, Gabbe SG: Shoulder dystocia: a complication of fetal

macrosomia and prolonged second stage of labor with midpelvic deliv-ery. Obstet Gynecol 52:526-529, 1978

3. Berard J, Dufour P, Vinatier D, et al: Fetal macrosomia:risk factors andoutcome: a study of the outcome concerning 100 cases �4500 g. Eur JObstet Gynecol Reprod Biol 77:51-59, 1998

4. Hassan AA: Shoulder dystocia: risk factors and prevention. Aust NZ JObstet Gyn 28:107-109, 1988

5. Nesbitt T, Gilbert W, Herrchen B: Shoulder dystocia and associated riskfactors with macrosomic infants born in California. Am J Obstet Gy-necol 179:476-480, 1998

6. Acker DB, Sachs BP, Friedman EA: Risk factors for shoulder dystocia inthe average-weight infant. Obstet Gynecol 67:614-618, 1986

7. Geary M: Risk factors and fetal outcome in cases of shoulder dystociacompared with normal deliveries of a similar birthweight. Br J ObstetGynaecol 104:121-122, 1997

8. Smith RB, Lane C, Pearson JF: Shoulder dystocia: what happens at thenext delivery? Br J Obstet Gynecol 101:713-715, 1994

9. Lewis D, Raymond RC, Perkins MB, et al: Recurrence rate of shoulderdystocia. Am J Obstet Gynecol 172:1369-1371, 1995

0. Ginsberg NA, Moisidis C: How to predict recurrent shoulder dystocia.Am J Obstet Gynecol 184:1427-1430, 2001

1. Gurewitsch E, Landsberger E, Jain A, et al: Does knowledge of priorshoulder dystocia affect management and outcome of subsequent de-liveries? Am J Obstet Gynecol 193:S42, 2005

2. Blickstein I, Ben-Arie A, Hagay ZJ: Antepartum risks of shoulder dys-tocia and brachial plexus injury for infants weighing 4200 g or more.Gynecol Obstet Invest 45:77-80, 1998

3. Nagey DA: Can shoulder dystocia be prevented? Am J Obstet Gynecol163:1095-1096, 1990

4. Rouse DJ, Owen J, Goldenberg RL, et al: The effectiveness and costs ofelective cesarean delivery for fetal macrosomia diagnosed by ultra-sound. J Am Med Assoc 276:1480-1486, 1996

5. Spong CY, Beall M, Rodrigues D, et al: An objective definition of shoul-der dystocia: prolonged head to body delivery intervals and/or the useof ancillary obstetric maneuvers. Obstet Gynecol 86:433-440, 1995

6. Gurewitsch ED, Johnson E, Hamzehzadeh S, et al: Risk factors forbrachial plexus injury with and without shoulder dystocia. Am J ObstetGynecol 194:486-492, 2006

7. Bager B: Perinatally acquired brachial plexus palsy: A persisting chal-lenge. Acta Paediatr 86:1214-1219, 1997

8. Waters PM: Comparison of the natural history, the outcome of micro-surgical repair, and the outcome of operative reconstruction in brachialplexus palsy. J Bone Jt Surg 81A:649-659, 1999

9. Acker DB: A shoulder dystocia intervention form. Obstet Gynecol 78:150-151, 1991

0. Peleg D, Powell S: Shoulder dystocia: prediction, prevention, manage-ment, and defense. Postgraduate obstetrics and gynecology 18:1-6,1998

1. Dildy GA, Clark SL: Shoulder dystocia: risk identification. Clin Obstet
Gynecol 43:265-282, 2000 4
2. Gross TL, Sokol RJ, Williams T, et al: Shoulder dystocia: a fetal-physi-cian risk. Am J Obstet Gynecol 15:1408-1414, 1987

3. Verspyck E, Goffinet F, Hellot MF, et al: Newborn shoulder width: aprospective study of 2222 consecutive measurements. Br J Obstet Gy-necol 106:589-593, 1999

4. Gurewitsch ED, Donithan M, Stallings S, et al: Episiotomy versus fetalmanipulation in managing severe shoulder dystocia: a comparison ofoutcomes. Am J Obstet Gynecol 191:911-916, 2004

5. Bofill JA, Rust OA, Devidas M, et al: Shoulder dystocia and operativevaginal delivery. J Matern Fetal Med 6:220-224, 1997

6. Johnson T, Allen R, Fishman E, et al: Use of traditional and novel CTpelvimetry distinguish between different types of shoulder dystocia. JSoc Gynecol Invest 13:117A, 2006

7. Sokol RJ, Blackwell SC: ACOG practice bulletin: shoulder dystocia.Number 40, November 2002. Int J Gynecol Obstet 80:87-92, 2003

8. Rosenn B, Miodovnik M, Combs CA, et al: Preconception managementof insulin-dependent diabetes: improvement of perinatal outcome. Ob-stet Gynecol 1991:846-849, 1991

9. Klaij FA, Geirsson RT, Nielsen H, et al: Humerospinous distance mea-surements: accuracy and usefulness for predicting shoulder dystocia indelivery at term. Ultrasound Obstet Gynecol 12:115-119, 1998

0. Cohen BF, Penning S, Ansley D, et al: The incidence and severity ofshoulder dystocia correlates with a sonographic measurement of asym-metry in patients with diabetes. Am J Perinatol 16:197-201, 1999

1. Riska A, Lain H, Voutilainen P, et al: Estimation of fetal shoulder widthby measurement of the humerospinous distance by ultrasound. Ultra-sound Obstet Gynecol 7:272-274, 1996

2. Mintz MC, Landon MB, Gabbe SG, et al: Shoulder soft tissue width as apredictor of macrosomia in diabetic pregnancies. Am J Perinatol 6:240-243, 1989

3. Gurewitsch ED, Allen RH: Fetal manipulation for managment of shoul-der dystocia. Fetal Matern Med Rev 17:239-280, 2006

4. Gurewitsch E, Kim E, Yang JH, et al: Comparing McRoberts’ and Ru-bin’s maneuvers for initial management of shoulder dystocia: an objec-tive evaluation. Am J Obstet Gynecol 192:153-160, 2005

5. Bellew M, Kay SP: Early parental experiences of obstetric brachialplexus palsy. J Hand Surg 28B:339-346, 2003

6. Queenan JT: Professional liability: some solutions. Obstet Gynecol 98:3658, 2001

7. Papazian O, Alfonso I, Yaylali I, et al: Neruophysiological evaluation ofchildren with traumatci radiculopathy, plexopathy, and peripheralneuropathy. Semin Pediatr Neurol 7:26-35, 2000

8. Metaizeau JP, Gayet C, Plenat F: [Brachial plexus injuries: an experi-mental study]. Chirug Ped 20:159-163, 1979

9. Sunderland S, Bradley KC: Stress-strain phenomena in human periph-eral nerve trunks. Brain 84:102-119, 1961

0. Slooff ACJ, Ubachs JMH: Brachial plexus impairment: a birth trauma?Am J Obstet Gynecol 169:230, 1993

1. Ubachs JMH, Slooff ACJ, Peeters LLH: Obstetric antecedents of surgi-cally treated obstetric brachial-plexus injuries. Br J Obstet Gynecol102:813-817, 1995

2. Gurewitsch ED, Johnson E, Hamzehzadeh S, et al: Risk factors forbrachial plexus injury with and without shoulder dystocia. Am J ObstetGynecol 194:486-492, 2006

3. Allen RH, Sorab J, Gonik B: Risk factors for shoulder dystocia: anengineering study of clinician-applied forces. Obstet Gynecol 77:352-355, 1991

4. Yogev Y, Langer O: Recurrence of gestational diabetes: pregnancy out-come and birth weight diversity. J Matern Fetal Neonatal Med 15:56-60, 2004

5. Stamilio DM, Olsen T, Ratcliffe S, et al: False-positive 1-hour glucosechallenge test and adverse perinatal outcomes. Obstet Gynecol 103:148-156, 2004

6. Raychaudhuri K, Maresh MJ: Glycemic control throughout pregnancyand fetal growth in insulin-dependent diabetes. Obstet Gynecol 95:190-194, 2000

7. Schwartz R, Teramo KA: Effects of diabetic pregnancy on the fetus andnewborn. Semin Perinatol 24:120-135, 2000

8. Rouse DJ, Owen J: Sonography, suspected macrosomia, and prophy-

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6


lactic cesarean: a limited partnership. Clin Obstet Gynecol 43:326-334,2000

9. Spellacy WN, Miller S, Winegar A, et al: Macrosomia: maternal char-acteristics and infant complications. Obstet Gynecol 66:158-161, 1985

0. Combs CA, Singh NB, Khoury JC: Elective induction versus spontane-ous labor after sonographic diagnosis of fetal macrosomia. Obstet Gy-necol 81:491-496, 1993

1. Gonen O, Rosen DJ, Dolfinn Z, et al: Induction of labor versus expect-ant management in macrosomia: a randomized study. Obstet Gynecol89:913-917, 1997

2. Sanchez-Ramos L, Bernstein S, Kaunitz AM: Expectant managementversus labor induction for suspected fetal macrosomia: a systematicreview. Obstet Gynecol 100:997-1002, 2002

3. Conway DL: Choosing route of delivery for the macrosomic infant of adiabetic mother: cesarean section versus vaginal delivery. J Matern FetalNeonatal Med 12:442-448, 2002

4. Conway DL: Delivery of the macrosomic infant: cesarean section versusvaginal delivery. Semin Perinatol 26:225-231, 2002

5. Allen R, Petersen S, Moore P, et al: Do antepartum and intrapartum riskfactors differ between mild and severe shoulder dystocia? Am J ObstetGynecol 189:S208, 2003

6. Dyachenko A, Ciampi A, Fahey J, et al: Prediction of risk for shoulderdystocia with neonatal injury. Am J Obstet Gynecol 195:1544-1549,2006

7. Allen RH, Cha SL, Kranker LM, et al: Comparing mechanical fetalresponse during descent, crowning and restitution among deliverieswith and without shoulder dystocia. Am J Obstet Gynecol 2007, inpress

8. Allen RH, Bankoski BR, Butzin CA, et al: Comparing clinician-appliedloads for routine, difficult and shoulder dystocia deliveries. Am J ObstetGynecol 171:1621-1627, 1994

9. Kim J, Neilipovitz D, Cardinal P, et al: A pilot study using high-fidelity
simulation to formally evaluate performance in the resuscitation of
critically ill patients: The University of Ottawa Critical Care Medicine,High-Fidelity Simulation, and Crisis Resource Management I Study.Crit Care Med 34:2167-2174, 2006

0. Rosenthal ME, Adachi M, Ribaudo V, et al: Achieving housestaff com-petence in emergency airway management using scenario based simu-lation training: comparison of attending vs housestaff trainers. Chest129:1453-1458, 2006

1. Strom P, Hedman L, Sarna L, et al: Early exposure to haptic feedbackenhances performance in surgical simulator training: a prospective ran-domized crossover study in surgical residents. Surg Endosc 20:1383-1388, 2006

2. Jude DC, Gilbert GG, Magrane D: Simulation training in the obstetricsand gynecology clerkship. Am J Obstet Gynecol 195:1489-1492, 2006

3. Gurewitsch E, Cha S, Johnson T, et al: Traction training for routine andshoulder dystocia delivers: an experimental study. Am J Obstet Gy-necol 193:S41, 2005

4. Gonik B, Costello R, Zhang N, et al: Effect of clinician applied maneu-vers on fetal brachial plexus strain during shoulder dystocia delivery.Am J Obstet Gynecol 189(6 Suppl 1):S200, 2003

5. Reid DE: Conduct of normal labor and the puerperium, in Reid DE(ed): Textbook of Obstetrics. Philadelphia, PA, W.B. Saunders Com-pany, 1962, pp 448-489

6. Freeth D, Ayida G, Berridge EJ, et al: MOSES: Multidisciplinary Ob-stetric Simulated Emergency Scenarios. J Interprof Care 20:552-554,2006

7. Gurewitsch E, Stallings S, Tam W, et al: Does maneuver rate affectshoulder dystocia outcome? Am J Obstet Gynecol 191:S66, 2004

8. Langer O, Berkus MD, Huff RW, et al: Shoulder dystocia: should thefetus weighing �4000 grams be delivered by cesarean section? Am JObstet Gynecol 165:831-837, 1991

9. Mollberg M, Hagberg H, Bager B, et al: High birthweight and shoulderdystocia: the strongest risk factors for obstetrical brachial plexus palsyin a Swedish population-based study. Acta Obstet Gynecol Scand 84:
654-659, 2005

ESCC

Mpopssagt

tassd

D

A

1

pidemiologic Approaches fortudying Recurrent Pregnancy Outcomes:hallenges and Implications for Research

ande V. Ananth, PhD, MPH

The study of recurrence of pregnancy-related complications and outcomes can offerpowerful insights to understanding patient-related risks for subsequent pregnancies. Suchstudies, when designed, analyzed, and interpreted correctly, can help distinguish geneticfrom environmental causes that portend increased recurrence of a particular pregnancycomplication (eg, recurrence of gestational diabetes) or a perinatal outcome (eg, recur-rence of preterm birth or preeclampsia). Recurrence risk studies can be challenging inother dimensions, including inherent biases, generalizability of findings, inadequate studysize, and inappropriate use of analytic models to study recurrence. Other common mis-perceptions in studies of recurrence risk are highlighted, including issues with terminologyand interpretation of recurrence risks. A review of available epidemiologic study designs ispresented and the usefulness and applicability of each design for addressing specificetiologic questions as they relate to recurrence risks are contrasted.Semin Perinatol 31:196-201 © 2007 Elsevier Inc. All rights reserved.

KEYWORDS recurrence risk, heterogeneity, epidemiology, clustered data, case-crossoverdesign, prospective cohort, cohort studies

repocap(s

prfnBsi

WFpO

ost diseases in medicine do not tend to recur. Preg-nancy is perhaps the only state that provides the op-

ortunity to study recurrence risks. The study of recurrencef pregnancy-related complications and outcomes can offerowerful insights to understanding patient-related risks forubsequent pregnancies. Whereas recurrent outcome re-earch can help guide clinical care for future pregnancies, itlso affords a unique opportunity to explore disease etiolo-ies through the study of heterogeneity in risk factors orhose that change across successive pregnancies.

This paper provides an overview of some of the fundamen-al concepts in studying recurrence, focusing largely on avail-ble epidemiologic and statistical approaches. In addition,ome of the common misperceptions about recurrence re-earch in human reproduction (broadly defined) are ad-ressed. In particular, several challenges in studying recur-

ivision of Epidemiology and Biostatistics, Department of Obstetrics, Gy-necology, and Reproductive Sciences, UMDNJ-Robert Wood JohnsonMedical School, New Brunswick, NJ.

ddress reprint requests to Cande V. Ananth, PhD, MPH, Division of Epidemi-ology and Biostatistics, Department of Obstetrics, Gynecology, and Repro-ductive Sciences, UMDNJ-Robert Wood Johnson Medical School, 125Paterson Street, New Brunswick, NJ 08901-1977. E-mail: cande.

[email protected]


ent events in pregnancy are highlighted with particularmphasis on: (1) challenges in accurately defining the end-oint being observed; (2) choosing an appropriate epidemi-logic design for studying patterns of recurrence risks; (3)onfounding; (4) limitations contributed through various bi-ses in epidemiologic studies; (5) sample size, statisticalower, and analytic models; (6) a note on terminology; and7) interpretation of recurrence risks. Finally, an outline ofome unifying concepts in studies on recurrence is presented.

The fundamental ideas and concepts espoused in this pa-er regarding epidemiologic approaches to studying recur-ent pregnancy outcomes stem, in part, from discussionsrom the “Second International Symposium on Successive Preg-ancy Outcomes: A Decade of Progress”, which was held in Newrunswick, NJ, in August 2005, and the proceedings of thisymposium that are due to be published in a forthcomingssue of Paediatric and Perinatal Epidemiology.1

hy Study Recurrence?or epidemiologists as well as clinicians, studying recurrencerovides an opportunity to examine etiologic heterogeneity.ne fundamental, yet much underappreciated, advantage to

tudying recurrence is the ability to separate “low-risk”

wpcmrcaaa

ipbacgfdstptberotc

vetctvewgdertpfieiot(nt

(epsc

bovvpsidpCqcmacepet

CoEOdhcaMbassb

cgapsiTttptrhc

CfSr

Studying recurrent events in pregnancy 197

omen from those at “high risk” of recurrence of a particularregnancy complication or outcome. This concept was suc-inctly illustrated by Wilcox,2 who made a compelling argu-ent that what separates low risk from high risk of recur-

ence is driven by the etiology of the initial clinical event oromplication. When characteristics of persons in the low-nd high-risk groups differ, then depending on the strengthnd magnitude, these differences may provide some cluesbout disease causation.

As an illustration, consider a study to examine differencesn risk factors associated with spontaneous and indicatedreterm births. Some studies have shown a few differencesetween these two clinical subtypes of preterm birth.3-5 Thebility to separate preterm birth based on the underlyinglinical subtypes addresses issues related to etiologic hetero-eneity in studying preterm birth.4,6 The similarities in riskactors between the two subtypes suggest that these two con-itions may overlap in terms of similar etiologic profiles.7 Yet,tudying the recurrence of preterm birth provides one oppor-unity to examine separately and in greater detail the riskrofiles between low-risk women (women with a previouserm birth) and high-risk women (those with a prior pretermirth). Whereas the data regarding increased recurrence (ofither spontaneous or indicated preterm birth) suggest a di-ect biologic influence of the previous preterm birth, otherbserved (both fixed and transient) and unobserved factorshat are shared between the two pregnancies also probablyontribute to this increased recurrence.

The heterogeneity in recurrence risks is derived throughariable contributions of genetic effects, environmental influ-nces, their interaction, and “time.” A genetic contribution tohe recurrence of an adverse reproductive endpoint can beontributed by the mother such as is seen with inheritedhrombophilias and venous thromboembolism risk. An en-ironmental effect (eg, smoking during pregnancy) can op-rate through either, or both, of the pregnancies withinhich the recurrence is studied. An interaction betweenenes and environment is one when the effect of genotype onisease risk depends on the type and level of exposure to annvironmental factor, or vice versa. A good example is theisk of recurrence for neural tube defects in siblings in rela-ion to a genetic polymorphism (in the folate metabolismathway) and the variability in environmental exposure toolate supplementation. The “time” factor in recurrence stud-es, although important to understand heterogeneity in dis-ase etiologies, is highly complex.8 Exposures or risk factorsnevitably operate through time with varying windows ofpportunity: a short exposure window (eg, exposure to aeratogen during organogenesis), throughout the pregnancyeg, cumulative exposure to smoking throughout preg-ancy), or a lifetime of exposure (eg, inadequate nutritionhrough harsh social conditions9 or across generations).

For example, attempts to separate biologic aging effectssuch as maternal age) from those of aging of the uterinenvironment contributed through repeated pregnancies (aarity effect) on the risk of placental abruption,10 althoughtraightforward in theory, are challenging to separate in
ross-sectional studies. Consider the task of separating the c
iologic aging effects from other influences on the recurrencef a pregnancy complication using the interpregnancy inter-al as the important “time” factor. Both short and long inter-als between pregnancies have been associated with adverseregnancy outcomes.11-14 Although short intervals can be as-ociated with the biologic effect of possible inadequate uter-ne recovery between pregnancies, there may also be a ten-ency for couples with an adverse outcome to have anotherregnancy sooner, something known as selective fertility.onversely, a long interpregnancy interval may allow ade-uate replenishment of nutrients to ensure a subsequent suc-essful pregnancy, but may be associated with the develop-ent of a medical disease (such as diabetes or hypertension)

nd possibly a change in partner between the two pregnan-ies. It is extremely difficult to separate the relative effects ofach of these components on the recurrence of a particularregnancy complication. Taken together, all these factors in-vitably contribute to recurrence risks, are highly inter-wined, and complex to disentangle.

hallenge 1: The Importancef Accurately Definingxposures and End Pointsne of the fundamental tenets of any research is to be able toefine both the exposure and the outcome of interest withigh degree of precision. The challenges in defining out-omes accurately have serious implications both in data an-lytic approach and in the interpretation of study findings.ost of the outcomes in perinatal epidemiology and, more

roadly in obstetrics and gynecology, are fairly well definednd can be measured accurately. Whereas some outcomes,uch as preterm birth, gestational diabetes, and preeclamp-ia, are easy to define accurately, others are not (eg, pretermirth clinical subtypes).Several conditions and reproductive end points present a

hallenge regarding an acceptable definition. Intrauterinerowth restriction refers to a fetus that has not been growingdequately, and is thought to be the result of an underlyingathological process.15 Because true intrauterine growth re-triction is very difficult to assess in population-based stud-es, a proxy, small-for-gestational age, has instead been used.he latter refers to infants (at birth) that weigh below a cer-

ain threshold (usually the 10th centile) for a given gesta-ional age. Although small-for-gestational age is used as aroxy of “true” in utero fetal growth restriction, not all babieshat are “small” for their gestational age are truly growthestricted. Thus, inaccuracies in defining an end point canave serious implications as to how data on recurrence risksan be interpreted.

hallenge 2: Study Designsor Studies of Recurrenceeveral epidemiologic study designs are available to studyecurrence of a specific disease or outcome. A prospective
ohort study is one that has been a commonly applied design

tdgpsttn

nMoeasoiimgtst“csttpthtvo

se“tedtbntstecde

CAaas

ittattbecijCffipo

ifsvwpdeetr

vpnr(npisbn(ttaft

CiAatcrn

198 C.V. Ananth

o study recurrence of pregnancy complications. With thisesign, women having an index pregnancy are followed lon-itudinally over time to record their outcome in subsequentregnancies. Existing registry-based data can be used to de-ign a prospective cohort study assuming an ability to iden-ify and link women having more than one pregnancy. This ishe most commonly used design in studies of recurrent preg-ancy outcomes.For some types of recurrent pregnancy complications,

ewer hybrid study designs may offer some advantages.ost notably, the case-crossover study design is one that

ffers great promise for studying the effect of transientxposures on the risk of outcomes.16 This design helpsnswer the question: “Was an outcome triggered by somepecific exposure that occurred immediately preceding theutcome (ie, an antecedent exposure)?” The distinguish-ng characteristic of this design is that each case serves asts own control, and is analogous to a crossover experi-

ent viewed retrospectively. This means that the investi-ator does not control when a patient starts being exposedo a potential trigger.17 The case-crossover design hasome similarities to a traditional matched-pair case-con-rol study.18 In both types of designs, each case has amatched” control. However, in a traditional matchedase-control study, the control is a different individual at aimilar time. As in the case in the case-crossover design,he control is the same person, but observed at a differentime. The case-crossover design applies best when the ex-osure is intermittent, the effect of risk is immediate andransient, and the outcome is abrupt. Therefore, it may beelpful to examine the effects of an abdominal trauma onhe risk of placental abruption among women with a pre-ious abruption and those with prior normal pregnancyutcomes.One of the inherent strengths of the case-crossover de-

ign is the implicit control for confounding bias. Sincevery subject in a case-crossover study serves as their owncontrol,” confounding is less of an issue relative to studieshat incorporate other designs. This phenomenon was el-gantly illustrated by Warner and colleagues19 in a con-om effectiveness study in preventing infections, wherehey demonstrated that epidemiologic studies confoundedy unmeasured differences between condom users andonusers underestimate condom effectiveness against cer-ain infections. For instance, prospective cohort studiesuffer from unmeasured confounding (a term also referredo as “residual confounding”), often leading to distortedstimates of the exposure– disease relationship. The case-rossover method, however, provided a technique for re-ucing unmeasured confounding in studies of condomffectiveness.19

hallenge 3: Confoundingconfounder is, by definition, a factor that distorts the

ssociation between an exposure and the outcome unlessdjusted. Three essential criteria for a variable to be clas-
ified as a confounder include: (1) the variable in question a
s a risk factor for the outcome being examined amonghose unexposed; (2) the factor must be associated withhe exposure in the population from which the subjectsrose; and (3) the variable is not in the causal pathway ofhe exposure– outcome relationship.18 Adjustment for po-ential confounders in any analysis eliminates bias (causedy those confounders) that could otherwise distort thexposure– disease association. In a cross-sectional or aase-control study, adjustment for available confounderss straightforward; however, the issue of confounder ad-ustment gets more complicated in studies of recurrence.onsider, as an illustration, if one is interested in quanti-

ying the extent to which preeclampsia recurs between therst and the second pregnancy. Should one adjust forotential confounders present in the first pregnancy, sec-nd pregnancy, or both?Confounders largely fall into two broad types: time-

ndependent and time-varying (or time-dependent). In theormer, when the value of a variable for subjects undertudy does not change over time (eg, race/ethnicity), theariable is said to be time-independent. On the contrary,hen the value of the variable changes with time (eg,arity, maternal smoking or prepregnancy body mass in-ex across successive pregnancies), the variable is consid-red time-dependent. The type of variable being consid-red for adjustment during statistical analysis, presumablyo minimize confounding, has implications in studies ofecurrence.

Although sophisticated statistical models have been de-eloped for correcting time-varying confounders,20,21 sim-ler approaches can also be undertaken. Consider a sce-ario where one is interested in estimating the risk ofecurrence of placental abruption, and maternal smokingas a binary factor denoting if the women was a smoker oronsmoker) is a potential confounder. One efficient ap-roach to fully adjust for the confounding effects of smok-

ng is to construct a four-level factor as follows: non-moker in both pregnancies (coded 0), smoker in the firstut not the second (coded 1), smoker in the second, butot the first (coded 2), and smoker in both pregnanciescoded 3). The construction of the new variable is similaro allowing an interaction effect of smoking status betweenhe first and second pregnancies. Adjustment for this vari-ble in the regression model will ensure control for con-ounding as opposed to adjusting for smoking effects inhe first or in the second pregnancy alone.

hallenge 4: Biasesn Studies of Recurrencecouple’s decision to achieve a desired family size introducesbias in studies of recurrence. Selective fertility, the tendency

o control fertility on the basis of previous pregnancy out-omes,22 is one such phenomenon that can affect studies ofecurrence risks.23 Specifically, couples that experience peri-atal losses in the first pregnancy tend to go on to have
dditional pregnancies to achieve a desired family size. This

plotlh

bbcablciite

CtSMcmwSeiUaIsi

tclaaspoipdtsttptotcs

lot

trirprewea“itmoeiotTdacois

CAUhHpttnsriaipBtphcio


henomenon operates across all parity levels, but is particu-arly a concern in higher-order births, and by analogy, tolder women. Skjaerven and coworkers23 suggest that, al-hough women differ in their inherent risks for perinatalosses, selective fertility leads to an overrepresentation ofigh-risk women at higher birth orders.Other biases often operate in studies of recurrence, with

ias due to selection being one of the more important ofiases. When a study of recurrence of a reproductive out-ome is restricted to successive singleton live births, for ex-mple, women with pregnancy losses or stillbirths, eitherefore the first singleton live birth or between two successive

ive births, are inevitably excluded. These inclusion and ex-lusion criteria have implications to studies on recurrence,ncluding generalizability or clinical relevance of the find-ngs. Although this is not a critical limitation, careful atten-ion to such biases must be considered while drawing infer-nces from such studies.

hallenge 5: Choosinghe Appropriate Studyize and Analytic Approachost published research on recurrent pregnancy outcomes

ome from large, population-based data. These studiesostly come from the Scandinavian countries, notably, Nor-ay, Sweden, and Denmark, and others in Europe such ascotland. Data collection from these countries date back sev-ral decades with perhaps the Norwegian birth registry dat-ng back to the mid-1960s. Similar data from within thenited States are few, and are largely based on identifyingnd linking biologic mothers to their successive pregnancies.n the United States, these data come primarily from vitaltatistics registers that are based on birth and death (fetal andnfant) certificates.

The advantages of using large, population-based registrieso address specific research on recurrent pregnancy out-omes is the large sample size which, in turn, affords excel-ent statistical power to discern patterns of recurrence risk. Inddition, such studies offer greater generalizability since theyre population-based. One of the inherent difficulties withmaller studies on recurrence of a particular pregnancy com-lication is adequate study size. As an illustration, consider ifne is interested in estimating the risk of recurrence of uter-ne rupture. Uterine rupture has a reported incidence of 1.6er 1000 pregnancies in women with a previous cesareanelivery attempting a subsequent vaginal birth.24 If the goal iso estimate the recurrence risk of uterine rupture, then thetudy should exclude women who have undergone hysterec-omy and include only women with a repaired uterine rup-ure in the first pregnancy who then go on to have a secondregnancy that again results in a uterine rupture. The rarity ofhis sequence of events would require hundreds of thousandsf women to adequately study. To be successful, a study ofhis magnitude would need several large registries, perhaps aollaborative effort of all population-based registries. Smaller
tudies will inevitably lead to conclusions being affected by i
ack of statistical power (type II error). Therefore, picking anutcome that is observable with an available study popula-ion is paramount to a successful study.

One of the important assumptions of any statistical test ishe “independence” of observations. Thus, studying a recur-ent pregnancy-related event entails evaluation of the eventn the same woman twice. Studies dealing with the recur-ence of an outcome clearly violate this assumption of inde-endence. Consider, for example, studies dealing with theisk of recurrence of fetal growth restriction, a topic reviewedxtensively by Kinzler and Kaminsky in this issue.25 Since aoman contributes data on two successive pregnancies, with

ach pregnancy resulting in a growth-restricted or appropri-tely grown infant, the responses (fetal growth) tend to becorrelated.” This tendency of “clustering” gives rise to anntracluster correlation (the woman, in this instance, is saido constitute a cluster), which will produce incorrect esti-ates of variance parameters when left unadjusted.26-28 In

ther words, although the odds ratio for the exposure–dis-ase relationship will remain unaffected due to ignoring thentracluster correlation, the 95% confidence interval for thedds ratio will be biased on either direction (depending onhe strength and direction of the intracluster correlation).his, in turn, will affect significance tests, and eventuallyiminish the scientific merit of the intended research. Severalnalytic models and methods have been developed to ac-ount for this clustering phenomenon. Although descriptionf these methods fall well beyond the scope of this paper,nterested readers are referred to papers on this topic withpecific applications to studies on human reproduction.5,29

hallenge 6:Note on Terminology

nfortunately, the terms “multivariable” and “multivariate”ave been used interchangeably in the medical literature.owever, these terms have distinct implications that are im-ortant for recurrence research. Consider, as an illustration,he setting of a cross-sectional (or case-control) study, wherehe goal is to estimate the risk of preeclampsia in relation toulliparity (with parous women as the reference). A regres-ion model to derive an estimate of the relative risk (or oddsatio) for preeclampsia in relation to nulliparity after adjust-ng for potential confounders is, by definition, a multivari-ble model. On the other hand, if one is interested in estimat-ng the effect of change in paternity on the recurrence risk ofreeclampsia, then a study of successive births is required.ecause preeclampsia is repeatedly assessed (ie, in each of thewo pregnancies), a regression model that is adjusted forotential confounders is a “multivariate” model. Researchersave used the two terms “multivariable” and “multivariate”asually, and the medical literature has paid little attention tots correct usage. Correct use and understanding of terminol-gy will facilitate communication among all parties involved
n recurrence research.

CoOrbecrwlcttpr

raovawcnmrilpmr

CNrOpaiccrmdg

teItfitai

raiiocp

ATMVctt

R

1

1

1

1

1

1

1

1

200 C.V. Ananth

hallenge 7: Interpretationf Recurrence Risksne of the fundamental issues common to studies on recur-

ence of pregnancy outcomes pertains to interpretation. Theiggest challenge is to successfully translate the results ofpidemiologic observations on recurrence risk to clinical de-ision making for individual patients. Most studies on recur-ence are population-based, which is a strength in manyays, but population-based recurrence risks do not readily

ead to customized risks for an individual patient. Aside fromonceptual differences in the regression models for popula-ion-based versus subject-specific approaches to risk estima-ion,30,31 the rich set of factors that shape risk in an individualatient limits the direct translation of population-derivedisks to individuals.

As an illustration, consider the following scenario. Theecurrence risk for stillbirth in a large study is fourfold highermong women with a previous stillbirth. Let us assume thatne is interested in applying this recurrence risk to an indi-idual patient with gestational diabetes that had experiencedprevious stillbirth. But this particular patient may have losteight, maintained extremely good glycemic control inter-

onceptionally, or was managed differently in the next preg-ancy. This is in addition to other contributing factors thatay have been altered, but are difficult to ascertain. That the

ecurrence risk for stillbirth for this patient is fourfold highers clearly misleading. This illustration highlights the chal-enge in applying population-based risk profiles to individualatient settings. Smulian32 provides a clinical perspective onany of the issues surrounding interpretation of research in

ecurrent pregnancy complications.

onclusionso research can be accomplished in isolation, and a study on

ecurrence of pregnancy-related conditions is no exception.ptimizing study design using the best available data, appro-riately adjusting for confounding and other biases, choosingppropriate statistical models and, most importantly, carefulnterpretation of findings are extremely important. Given theomplexities in all these facets in studies of recurrence, thehallenges are quite daunting. It is clear that as recurrenceesearch becomes more common and as the methodologyatures, active collaborations will be necessary among epi-emiologists, biostatisticians, and the clinical community toet the most out of our efforts.

Observational studies of recurrent pregnancy complica-ions are difficult to perform well, and cohort studies thatntail a recurrence component are much more complicated.n addition to the limitations afforded in following prospec-ive cohorts (including resources, time and personnel ef-orts), studies of recurrence need to carefully consider thessues highlighted in this review. Such studies require atten-ion to inherent biases (eg, selective fertility, selective man-gement), and failure to recognize and account for such lim-
tations will invariably result in distorted findings of 1
ecurrence risks. More importantly, incorrect models to an-lyze data arising from correlated responses will likely yieldncorrect statistical and biologic inferences. Finally, correctnterpretation of recurrence risks will greatly enhance ourngoing pursuits for studying recurrence of adverse out-omes in pregnancy-related conditions, which is best ap-roached through a collaborative environment.

cknowledgmentshe author extends special thanks to Russell Kirby, PhD,organ Peltier, PhD, John Smulian, MD, MPH, and Anthonyintzileos, MD, for their thoughtful suggestions and valuableomments that helped improve the paper. Dr. Ananth is par-ially supported through a grant (HD038902) from the Na-ional Institutes of Health awarded to him.

eferences1. Ananth CV: Second International Symposium on Successive Pregnancy

Outcomes: a decade of progress. Paediatr Perinat Epidemiol 21(Suppl)2007, in press

2. Wilcox AJ: The analysis of recurrence risks as an epidemiologic tool.Paediatr Perinat Epidemiol 21(Suppl) 2007, in press

3. Berkowitz GS, Blackmore-Prince C, Lapinski RH, et al: Risk factors forpreterm birth subtypes. Epidemiology 9:279-285, 1998

4. Savitz DA, Blackmore CA, Thorp JM: Epidemiologic characteristics ofpreterm delivery: etiologic heterogeneity. Am J Obstet Gynecol 164:467-471, 1991

5. Ananth CV, Platt RW, Savitz DA: Regression models for clustered bi-nary responses: implications of ignoring the intracluster correlation inan analysis of perinatal mortality in twin gestations. Ann Epidemiol15:293-301, 2005

6. Ananth CV, Vintzileos AM: Maternal-fetal conditions necessitating amedical intervention resulting in preterm birth. Am J Obstet Gynecol195:1557-1563, 2006

7. Ananth CV, Getahun D, Peltier MR, et al: Recurrence of spontaneousversus medically indicated preterm birth. Am J Obstet Gynecol 195:643-650, 2006

8. Basso O: Options and limitations in studies of successive pregnancyoutcomes: an overview. Paediatr Perinat Epidemiol 21(Suppl) 2007, inpress

9. Wilcox AJ, Skjaerven R, Irgens LM: Harsh social conditions and peri-natal survival: an age-period-cohort analysis of the World War II occu-pation of Norway. Am J Public Health 84:1463-1467, 1994

0. Ananth CV, Wilcox AJ, Savitz DA, et al: Effect of maternal age andparity on the risk of uteroplacental bleeding disorders in pregnancy.Obstet Gynecol 88:511-516, 1996

1. Basso O, Olsen J, Knudsen LB, et al: Low birth weight and preterm birthafter short interpregnancy intervals. Am J Obstet Gynecol 178:259-263, 1998

2. Smith GC, Pell JP, Dobbie R: Interpregnancy interval and risk of pre-term birth and neonatal death: retrospective cohort study. Br Med J327:313, 2003

3. Zhu BP, Rolfs RT, Nangle BE, et al: Effect of the interval betweenpregnancies on perinatal outcomes. N Engl J Med 340:589-594, 1999

4. Conde-Agudelo A, Rosas-Bermudez A, Kafury-Goeta AC: Birth spacingand risk of adverse perinatal outcomes: a meta-analysis. J Am MedAssoc 295:1809-1823, 2006

5. Resnik R: Intrauterine growth restriction. Obstet Gynecol 99:490-496,2002

6. Maclure M: The case-crossover design: a method for studying transienteffects on the risk of acute events. Am J Epidemiol 133:144-153, 1991

7. Maclure M, Mittleman MA: Should we use a case-crossover design?Annu Rev Public Health 21:193-221, 2000

8. Greenland S, Rothman KJ: Measures of effect and measures of associa-

1

2

2

2

2

2

2

2

2

2

2

3

3

3


tion, in Rothman KJ, Greenland S (eds): Modern Epidemiology (ed 2).Philadelphia, PA, Lippincott Williams & Wilkins, 1998, pp 47-64

9. Warner L, Macaluso M, Austin HD, et al: Application of the case-crossover design to reduce unmeasured confounding in studies of con-dom effectiveness. Am J Epidemiol 161:765-773, 2005

0. Bodnar LM, Davidian M, Siega-Riz AM, et al: Marginal structural mod-els for analyzing causal effects of time-dependent treatments: an appli-cation in perinatal epidemiology. Am J Epidemiol 159:926-934, 2004

1. Robins JM, Hernan MA, Brumback B: Marginal structural models andcausal inference in epidemiology. Epidemiology 11:550-560, 2000

2. Wilcox AJ, Gladen BC: Spontaneous abortion: the role of heteroge-neous risk and selective fertility. Early Hum Dev 7:165-178, 1982

3. Skjaerven R, Wilcox AJ, Lie RT, et al: Selective fertility and the distor-tion of perinatal mortality. Am J Epidemiol 128:1352-1363, 1988

4. Lydon-Rochelle M, Holt VL, Easterling TR, et al: Risk of uterine ruptureduring labor among women with a prior cesarean delivery. N Engl
J Med 345:3-8, 2001
5. Kinzler WL, Kaminsky L: Fetal growth restriction and subsequent preg-nancy risks. Semin Perinatol 2007, in press

6. Liang KY, Zeger SL: Regression analysis for correlated data. Annu RevPublic Health 14:43-68, 1993

7. Zeger SL, Liang KY: Longitudinal data analysis for discrete and contin-uous outcomes. Biometrics 42:121-130, 1986

8. Liang K-Y, Zeger SL: Longitudinal data analysis using generalized linearmodels. Biometrika 73:13-22, 1986

9. Louis GB, Dukic V, Heagerty PJ, et al: Analysis of repeated pregnancyoutcomes. Stat Methods Med Res 15:103-126, 2006

0. Zeger SL, Liang KY: An overview of methods for the analysis of longi-tudinal data. Stat Med 11:1825-1839, 1992

1. Zeger SL, Liang KY, Albert PS: Models for longitudinal data: a general-ized estimating equation approach. Biometrics 44:1049-1060, 1988

2. Smulian JC: Research on recurrent pregnancy complications: a clini-
cian’s perspective. Paediatr Perinat Epidemiol Suppl 2007, in press

seminars in logy june2007 recurring complications of pregnancy

Documents

recurrence of complications

md topicsneonatal

md volume

recurrent pregnancy

recurrence research

future pregnancy compl

prediction of complications

pregnancy track record