obstet gynecol 2007; p153

Clinical Expert Series

Fetal DeathRobert M. Silver, MD

The death of a formed fetus is one of the most emotionally devastating events for parents andclinicians. With improved care for conditions such as RhD alloimmunization, diabetes, andpreeclampsia, the rate of fetal death in the United States decreased substantially in the midtwentieth century. However, the past several decades have seen much greater reductions inneonatal death rates than in fetal death rates. As such, fetal death remains a significant andunderstudied problem that now accounts for almost 50% of all perinatal deaths. The availabilityof prostaglandins has greatly facilitated delivery options for patients with fetal death. Risk factorsfor fetal death include African American race, advanced maternal age, obesity, smoking, priorfetal death, maternal diseases, and fetal growth impairment. There are numerous causes of fetaldeath, including genetic conditions, infections, placental abnormalities, and fetal–maternalhemorrhage. Many cases of fetal death do not undergo adequate evaluation for possible causes.Perinatal autopsy and placental examination are perhaps the most valuable tests for theevaluation of fetal death. Antenatal surveillance and emotional support are the mainstays ofsubsequent pregnancy management. Outcomes may be improved in women with diabetes,hypertension, red cell alloimmunization, and antiphospholipid syndrome. However, there isconsiderable room for further reduction in the fetal death rate.(Obstet Gynecol 2007;109:153–67)

Pregnancy loss is one of the most common obstetriccomplications, affecting over 30% of conceptions.1

Most of these occur early in gestation, are due toproblems with implantation and may not be clinicallyapparent. However, 12–15% of conceptions result inclinically recognized pregnancy loss. The majority ofthese are first trimester miscarriages and fewer thanfive percent of pregnancies are lost after 10 weeks ofgestation. These later losses (fetal deaths) are particu-larly emotionally devastating for families and clini-cians, yet relatively little is known about second andthird trimester pregnancy loss. This article will reviewthe epidemiology, causes, management and evalua-tion of fetal death.

OVERVIEWNomenclature of Pregnancy LossThe terminology of pregnancy loss is confusing andcould potentially benefit from revision. Historically,pregnancy losses before 20 weeks of gestation arereferred to as abortions, whereas those after 20 weeksof gestation are termed fetal deaths or stillbirths.These definitions are somewhat arbitrary, inconsistentwith advances in our understanding of reproductivebiology, and not clinically helpful. Instead, it may bemore useful to classify pregnancy losses in terms ofstages of gestational development. Pregnancy lossescould be defined in terms of developmental biology,as preembryonic (anembryonic), embryonic, or fetal.The expression “blighted ovum” should be aban-doned and replaced with anembryonic or preembry-onic pregnancy loss. The preembryonic period beginsfrom conception and lasts through 5 weeks of gesta-tion (based upon menstrual dating). The embryonicperiod lasts from 6 through 9 weeks of gestation. At 10weeks of gestation, the fetal period begins, extendingthrough delivery. Alternatively, losses less than 20weeks gestation could be described as early (eg, lessthan 10 weeks gestation) compared with late (morethan 10 weeks gestation) abortions.

From the Department of Obstetrics and Gynecology, University of Utah, SaltLake City, Utah.

Partially supported by HD-045944, National Institutes of Health, NationalInstitute of Child Health and Human Development.

The author thanks Janice L. B. Byrne, MD, for the provision of figures.

Corresponding author: Robert M. Silver, MD, Department of Obstetrics andGynecology, University of Utah School of Medicine, 30 North 1900 East, Room2B308, Salt Lake City, UT 84132; e-mail: [email protected].

© 2006 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/06

VOL. 109, NO. 1, JANUARY 2007 OBSTETRICS & GYNECOLOGY 153

Increased specificity regarding the timing in ges-tation of pregnancy loss has important clinical impli-cations. First, the causes of losses are different acrossgestational ages. For example, losses before the de-velopment of an embryo (anembryonic losses) aremore likely to be associated with genetic problemsthan those later in gestation.2 In contrast, losses after10 weeks of gestation are more strongly associatedwith disorders that may affect placental blood flowsuch as antiphospholipid syndrome or heritablethrombophilias, when compared with early preg-nancy losses.3,4 The timing in gestation of pregnancylosses also has considerable influence on the recur-rence risk and timing in gestation of subsequentlosses.5 Too often, details regarding the timing ingestation of pregnancy losses are lacking, with pa-tients and physicians simply reporting a “miscarriage”based on the interval between menses and the onsetof vaginal bleeding. However, failure of growth ordeath of the conceptus often precedes clinical symp-toms of miscarriage, sometimes by several weeks.Clinicians are strongly encouraged to document ul-trasound findings, pathologic examination, and otherdata pertinent to distinguishing among types of preg-nancy losses.

Nomenclature regarding stillbirth also is contro-versial. The World Health Organization defines still-birth as pregnancy lost after 20 completed weeks ofgestation. If gestational age is unknown, a birth weightof 500 g or more is considered to be a stillbirth.6

However, others advocate the use of 24 weeks or 28weeks of gestation to define stillbirth.7 The rationalebehind the use of these latter definitions is to focus onfetal deaths after viability outside of the womb and islimited in clinical relevance by changes in the limit ofviability over time and according to survival potentialin different countries and regions.

A distinction also must be made between spo-radic and recurrent early pregnancy loss and fetaldeath. Sporadic pregnancy loss is common in nor-mal couples and is usually due to de novo nondis-junctional events. Recurrent pregnancy loss is vari-ably defined, most often as three or more losseswith no more than one live birth. Up to one percentof couples suffer recurrent pregnancy loss. This ismore common than would be expected by chancealone, suggesting that some couples have underly-ing conditions increasing the probability of preg-nancy loss. Both with early miscarriage and fetaldeath, recurrent cases substantially increase theodds of an underlying predisposing medical orgenetic condition. In turn, this influences the prog-nosis for subsequent pregnancies.

EpidemiologyThe rate of stillbirth has decreased substantially fromthe 1950s (20 per 1,000 births) through the 1980s withimproved care for conditions such as diabetes, red cellalloimmunization, and preeclampsia. However, still-birth rates have been relatively stable over the past 20years, reaching a plateau in the United States ofapproximately 6.4 per 1,000 births in 2002. In con-trast, infant mortality has decreased by more than30% in the last 20 years. In the United States in 2001,26,373 fetal deaths were recorded as compared with27,568 infant deaths. Thus, fetal death currently ac-counts for close to 50% of all perinatal deaths. As withother perinatal morbidity, there is considerable racialdisparity in fetal death rates. In 2001, African Amer-icans suffered a stillbirth rate of 12.1 per 1,000 birthscompared with 5.5 per 1,000 for whites.

DiagnosisFetal death may be associated with a cessation ofpreviously perceived fetal movements or a decreasein pregnancy-related symptoms such as nausea. Insome cases, women will present with bleeding,cramping, or labor. However, many patients withfetal death have no bleeding or contractions, and fetaldeath may precede clinical symptoms by a variableand often extended period of time. A definitivediagnosis is made by real-time ultrasonography con-firming the presence of a fetus and the absence of fetalheart pulsations. If the ultrasonographer is inexperi-enced, the diagnosis should be confirmed by someonewith appropriate expertise.

Classification of Fetal DeathIt is often difficult to determine a “certain” cause offetal death. First, many risk factors that are associatedwith fetal death in epidemiologic studies are presentin numerous apparently normal women with uncom-plicated pregnancies. Second, most studies of fetaldeath do not include controls, making it difficult toascertain the contribution of a potential abnormalityto the stillbirth. For example, heritable thrombo-philias often are present in women delivering livebirths. Accordingly, a positive test for thrombophiliain a case of fetal death, especially without evidence ofplacental insufficiency, does not prove causality.Third, several conditions may be present simulta-neously. If a stillborn fetus with trisomy 13 hasevidence of group B streptococcal infection, is thedeath due to infection or fetal aneuploidy? Sometimesfetal death may be due to the interaction or additiveeffect of two or more disorders. Finally, even after an

154 Silver Fetal Death OBSTETRICS & GYNECOLOGY

extensive evaluation, it may not be possible to deter-mine a cause of fetal death. Such unexplained lossesare common, especially in third trimester stillbirth.

There have been numerous attempts to catalogcauses of fetal death, typically greater than 20 weeksof gestation, using classification systems. None havebeen universally accepted, and all have advantagesand disadvantages. Further confusion arises from theuse of different definitions of fetal death amongsystems and the inclusion of neonatal deaths in somebut not all classification schemes. Popular classifica-tions schemes include the Aberdeen clinicopathologicclassification8 and the Wigglesworth classification9

scheme that is probably most commonly used today.Recently, Gardosi and colleagues developed a newsystem that substantially decreased the proportion ofunexplained stillbirths compared with traditional clas-sification schemes.10 However, this system ascribed avery large proportion (43%) of deaths to fetal growthrestriction, which may be an association rather than acause of fetal death (see below). There is ongoingdialogue among investigators throughout the world toagree on a uniform system to facilitate comparison offetal death rates and research into causes and preven-tion of fetal death.

It is important to distinguish between conditionsthat clearly and unequivocally cause fetal death andthose that are associated with the condition. Theselatter conditions are present in many cases of livebirths and do not always cause the unavoidable deathof the fetus. This distinction is not merely academic; ithas important implications for clinical practice andcounseling of couples with fetal death.

RISK FACTORS AND CAUSESMaternal ConditionsDemographicsConsistent demographic factors for fetal death in-clude race, low socioeconomic status, inadequateprenatal care, less education, and advanced maternalage.11,12 African-American women have rates of fetaldeath that are more than twice the rate for whitemothers. In part, this may be due to secondary riskfactors such as socioeconomic status and a lack ofprenatal care. However, African Americans havehigher fetal death rates than whites even amongwomen receiving prenatal care.13 This may be due tohigher rates of medical and obstetric complications inAfrican Americans.13 It is unclear whether improveduse of obstetric care would reduce the fetal death ratein African-American women, but it seems likely.

Maternal AgeFretts and colleagues demonstrated that increasingmaternal age after 35 years is associated with anincreased risk for fetal death.14 These findings havebeen confirmed in numerous studies, and the associ-ation persists when adjusting for potential confound-ing variables such as genetic problems, birth defects,medical problems, and maternal weight. A largeinpatient-based study in the United States estimatedthe odds ratio for stillbirth to be 1.28 (95% confidenceinterval [CI], 1.24–1.32) in women aged 35–39 yearsand 1.72 (95% CI, 1.6–1.81) in women aged 40 yearsor older compared with 20–34-year-old women.15

ObesityThe rate of fetal death also is increased in obesewomen. Numerous studies have shown a consistentdoubling in the risk for fetal death in cases of maternalobesity (body mass index of 30 or more).16 Increasedbody mass index increases the risk for several condi-tions known to increase the risk of stillbirth, such asdiabetes, hypertensive disorders including pre-eclampsia, socioeconomic status, and smoking. None-theless, obesity remains associated with fetal deathafter controlling for these confounders. The associa-tion between obesity and fetal death is of particularconcern given the dramatic and persistent increase inthe rate of maternal obesity.

Medical DisordersSeveral maternal medical disorders are associated withan increased risk for fetal death. It is debatable as towhether these conditions are causal or risk factorsbecause most affected women deliver liveborn infants.Perinatal outcome is influenced by obstetric manage-ment and decreased morbidity, and mortality frommaternal diseases such as diabetes and hypertension areresponsible for much of the improvement in fetal deathrates over the past half century. It is estimated thatmaternal diseases play a role in 10% of fetal deaths.

Despite improved care, women with diabetesmellitus (type 1 and 2) have a 2.5-fold increase in therisk for fetal death.17 Conversely, true gestationaldiabetes (type 2 diabetes may be first recognizedduring pregnancy) is not associated with an increasedrisk for fetal death. In part, fetal death in diabeticwomen is due to increased fetal anomalies and co-morbid conditions such as high blood pressure andobesity. The increased risk of fetal death persists aftercontrolling for these factors. Maternal hyperglycemiaand disorders of fetal growth, metabolism, and possi-ble acidosis contribute to fetal death, and treating

VOL. 109, NO. 1, JANUARY 2007 Silver Fetal Death 155

diabetics with insulin during pregnancy decreases therisk.18

Fetal death has been attributed to numerous othermaternal diseases, including hypertension, thyroiddisease, kidney disease, asthma, cardiovascular dis-ease, and systemic lupus erythematosus. Pregnancyloss in the setting of these conditions typically occursin women with clinically apparent and severe disease.Asymptomatic disease (such as mild glucose intoler-ance or abnormal thyroid function) has been pro-posed as a cause of fetal death. Although of interest,the theory remains unproven.

Antiphospholipid syndrome is an autoimmunedisorder characterized by the presence of specifiedlevels of antiphospholipid antibodies and one or moreclinical features, including pregnancy loss, thrombo-sis, or autoimmune thrombocytopenia.19 The histo-logic findings of placental infarction, necrosis, andvascular thrombosis in some cases of pregnancy lossassociated with antiphospholipid antibodies have ledto the hypothesis that thrombosis in the uteroplacen-tal circulation may lead to placental infarction andultimately, pregnancy loss (Fig. 1). Numerous retro-spective and prospective studies have linked recur-rent pregnancy loss, especially fetal death, with an-tiphospholipid syndrome.3 The two bestcharacterized antiphospholipid antibodies are lupusanticoagulant and anticardiolipin antibodies.

ThrombophiliasFetal death also has been associated with heritablethrombophilias. These disorders typically involve de-ficiencies or abnormalities in anticoagulant proteinsor an increase in procoagulant proteins, and likeantiphospholipid syndrome, have been associatedwith a risk for vascular thrombosis and pregnancyloss. Several case series and retrospective studiesreported an association between the factor V Leidenmutation (associated with abnormal factor V resis-tance to the anticoagulant effects of protein C), theG20210A mutation in the promoter of the prothrom-bin gene, deficiencies of the anticoagulant proteinsantithrombin III, protein C, and protein S and fetaldeath.19

In most studies, thrombophilias are more stronglyassociated with losses after 10 weeks of gestation asopposed to early pregnancy losses. A recent meta-analysis indicated an odds ratio of 2 for “early” and7.8 for “late” recurrent pregnancy loss for womenwith the factor V Leiden mutation, and an odds ratioof 2.6 for “early” recurrent fetal loss in those with theprothrombin gene mutation.4 Protein S deficiency,but not the methylenetetrahydrofolate mutation asso-ciated with hyperhomocysteinemia, protein C defi-ciency, or antithrombin III deficiency were associatedwith pregnancy loss in this meta-analysis.4 Fetalthrombophilias also have been associated with fetal

Fig. 1. Placenta demonstrating vil-lous infarction (arrow) from a caseof second trimester fetal death in apatent with antiphospholipid syn-drome (A). B. Normal placenta isshown for comparison. A. x 40,original magnification. B. x 100,original magnification.)Silver. Fetal Death. Obstet Gynecol2007.


death but data are inconsistent and should be consid-ered preliminary.

It is important to be careful when attributing fetaldeath to thrombophilias in women testing positive.These conditions are extremely common in normalindividuals,19 and prospective studies have failed todemonstrate an association between the factor VLeiden mutation and fetal death (or any adverseobstetric outcome).20 Most pregnancies in womenwith heritable thrombophilias result in healthy in-fants, and women with thrombophilias but no priorobstetric complications should be reassured.

As with all potential causes, context is important.Thrombophilia is more likely to contribute to a fetaldeath if there is objective evidence of placental insuf-ficiency such as intrauterine growth restriction(IUGR), placental infarction, or abnormal Dopplervelocimetry. It is less plausible as a cause of death ina term fetal demise weighing 8.5 pounds with normalamniotic fluid volume and a normal placenta.

ExposuresSmoking is the most common exposure that has beenassociated with fetal death. Although the vast majorityof women who smoke deliver liveborn infants, nu-merous studies identify smoking as a risk factor forfetal death. The risk is typically 1.5-fold over non-smokers; the risk decreases to background rate inwomen who stop smoking after the first trimester. Thecause is uncertain but may involve an increase in fetalcarboxyhemoglobin and vascular resistance, causingimpaired growth and hypoxia. Smoking also in-creases the risk for potentially catastrophic conditionssuch as abruption.

Other recreational drugs have been associatedwith fetal death, although substance abuse is associ-ated with many other risk factors. Cocaine has con-vincingly been associated with an increased risk forfetal death as well as abruption and IUGR. Althoughsimilar pathophysiology should apply to metham-phetamines, it has not been linked to fetal death atpresent. Data regarding alcohol and fetal death aremixed; some studies show an association and some donot.21,22 Marijuana has not been associated with fetaldeath. Abrupt narcotic withdrawal (eg, heroin) isanother potential cause of fetal death.

Fetal exposure to medications and environmentaltoxins such as pesticides or radiation has been pro-posed as a risk for fetal death. This subject is of greatinterest to communities but is very hard to study.Exposures likely contribute to a very small proportionof fetal deaths.

Systemic Maternal InfectionsSevere maternal infection with any type of organismmay result in fetal death. Examples include appendi-citis, pneumonia, pyelonephritis, and viruses such asinfluenza. The pathophysiology of fetal loss mayinclude hypoxia due to respiratory distress, pooruterine perfusion related to factors such as sepsis anddehydration, the metabolic effects of high fever, andthe initiation of a cascade of toxic inflammatorymediators. Systemic infection (as well as intra-amni-otic infection) also may lead to fetal death by initiatingpreterm labor, resulting in intrapartum death, espe-cially at previable gestations.

Fetal ConditionsGenetic ConditionsThe best studied genetic cause of fetal death ischromosomal abnormalities (Fig. 2). These have beenreported in 6–12 % of stillbirths. The proportion ishigher in first trimester losses and is likely intermedi-ate for losses between 10 and 20 weeks of gestation.2

These numbers may underestimate the true percent-age, because karyotype is not always assessed orsuccessfully obtained in all cases of fetal death. Thechances of fetal aneuploidy are increased in thesetting of fetal abnormalities noted on the antenatalultrasonogram or postmortem examination (up to

Fig. 2. Second-trimester fetal death with cystic hygroma(arrows) and nonimmune hydrops. The fetus had Turner’sSyndrome. Copyright 1994, J. L. B. Byrne. Used withpermission.Silver. Fetal Death. Obstet Gynecol 2007.


25%) and if the fetus is small for gestational age(SGA). Conversely, karyotype is more likely to benormal in the absence of these findings. The mostcommon abnormalities are monosomy X (23%), tri-somy 21 (23%), trisomy 18 (21%), and trisomy 13(8%).23

Many fetal deaths have genetic abnormalities thatare not detected by conventional cytogenetic analysis.Malformations, deformations, syndromes, or dyspla-sias have been reported in up to 35% of fetal lossesundergoing perinatal autopsy.23,24 Although up to 25%of these fetuses are aneuploid, most will have normalkaryotypes. However, it is probable that many ofthese fetuses have genetic abnormalities that are notidentified by traditional cytogenetic analysis.

Genetic abnormalities may contribute to fetaldeath in cases without obvious malformations. Singlegene disorders such as autosomal recessive condi-tions, including glycogen storage diseases, other met-abolic disorders, and hemoglobinopathies, may causefetal death. X-linked conditions may cause death inmale fetuses. There are likely a host of other singlegene disorders that contribute to some cases of fetaldeath, especially early in gestation. Supporting evi-dence comes from experiments in transgenic micewherein single gene mutations cause embryonicdeath. Typically, mice with these mutations sufferabnormal angiogenesis, placental, cardiac, or neuro-logic development and die at mid gestation.

Other types of genetic abnormalities also con-tribute to some cases of fetal death. Confinedplacental mosaicism refers to the presence of ab-normal chromosomes in some placental tissue inthe setting of normal fetal karyotype. This leads toabnormal placental development and function andhas been associated with fetal death and otherobstetric abnormalities such as IUGR. Other fetaldeaths may have very small deletions or additionsof chromosomes that are too small to detect byconventional karyotype. This type of abnormalityhas been found in some cases of unexplainedmental retardation. Newer molecular genetic tech-niques such as comparative genomic hybridizationmay allow detection of such “microdeletions.”

The identification of genetic causes of fetaldeath has been somewhat hampered by the fact thatthere are few syndromes or mutations that accountfor a large proportion of losses. Instead, manydifferent abnormalities each contribute to a smallproportion of cases. Recent developments in mo-lecular genetic technology should greatly facilitateour ability to determine previously unrecognizedgenetic conditions associated with fetal death.

InfectionInfections have been reported to account for 10–25% offetal deaths in developed countries.25 The proportion ishigher in developing countries. In addition to the pop-ulation studied, the percentage of losses due to infectionis influenced by gestational age and the thoroughness (orlack) of investigation into infectious causes of fetal death.For example, in developed countries, bacterial infec-tions are more common in fetal deaths before 28 weeksof gestation than later in pregnancy.

As with all causes of fetal death, it is importantto distinguish causality from association. One maybe confident (although not 100% certain) that infec-tion is the cause of death in cases of positivecultures and histologic evidence of inflammationand infection in fetal tissues. It is a different storywhen there are positive placental or vaginal cul-tures or maternal serology without histologic evi-dence of fetal infection.

Viral InfectionThe proportion of fetal deaths due to viruses isuncertain because of a lack of systematic evalua-tion. This is especially important for viral infectionsbecause they are often hard to culture. Perhaps themost common viral infection associated with preg-nancy loss is parvovirus B19 (Fig. 3). The virus isthought to cause fetal death through fetal anemialeading to hydrops, direct myocardial toxicity, orother mechanisms. This organism has been re-ported in 7.5% of fetal deaths in a Swedish studythat used polymerase chain reaction to determineinfection.26 The proportion of fetal deaths associ-ated with parvovirus has been reported to be ashigh as 15% when polymerase chain reaction isused to detect parvovirus B19, but is considerablylower (less than 1%) in studies that did not system-atically assess for the virus.25 Parvovirus B19 ismore likely to cause fetal death in the first or secondtrimester; deaths late in gestation are rarely causedby the virus.

The most common fetal or neonatal viral infec-tion is cytomegalovirus. Most fetal infections occurafter primary maternal infection—which occurs inabout 1% of women in the United States. Placentaland fetal infection with cytomegalovirus is well doc-umented, as are adverse neonatal sequelae such asIUGR and damage to organ systems including thebrain. Fetal death is rare but has been described.

Coxsackie viruses (A and B) also have beenreported to cause fetal death. The organism can causeplacental inflammation, myocarditis, and hydrops.


Other viruses sporadically linked to fetal death in-clude echoviruses, enteroviruses, chickenpox, mea-sles, rubella, and mumps. For those viruses amenableto vaccine prevention, fetal death is rare in countrieswith routine vaccination. Human immunodeficiencyvirus (HIV) may cross the placenta and cause fetalinfection. Although fetal death has occasionally beenattributed to HIV, HIV-positive women usually haveother risk factors for fetal death, making it difficult todocument an independent association. Herpes sim-plex virus rarely causes fetal death because it is rarelytransmitted to the fetus in utero.

Bacterial InfectionBacterial infections are generally accepted as a causeof some cases of fetal death throughout gestation. Indeveloped countries, a higher proportion of losses inthe second trimester are due to infection comparedwith term fetal deaths.25 In contrast, fetal deaths dueto bacterial infection persist through term in develop-ing countries. This may be due to increased burden ofexposure to infectious agents or a decreased immuneresponse associated with low socioeconomic status indeveloping nations.25

Most bacterial infections associated with fetaldeath are organisms that reach the fetus by ascending

from the lower genital tract into the decidua andchorion and occasionally the amniotic fluid. The fetusmay then swallow fluid, leading to fetal infection.Examples include group B streptococcus, Escherichiacoli, Klebsiella, Ureaplasma urealyticum, Mycoplasmahominis, and Bacteroides. Virulent organisms (eg, groupB streptococcus) may be associated with clinical evi-dence of intra-amniotic infection, whereas those re-sponsible for bacterial vaginosis (eg, Mycoplasma andUreaplasma) often do not cause clinically apparentdisease. Bacteria such as Listeria monocytogenes mayreach the fetus by hematogenous transmission.

Other InfectionsSpirochete, protozoal, and fungal infections may oc-casionally cause fetal death. Treponema pallidum, theorganism responsible for syphilis, may cross the pla-centa in the second and third trimesters and directlyinfect the fetus. This risk increases with advancinggestation. Fetal death may occur due to direct infec-tion or due to placental vasculopathy associated withplacental infection. Syphilis was a common cause offetal death at the beginning of the twentieth century.Although less common today, syphilis remains a rarecause of fetal death in developed countries and acommon one in places with higher prevalence, suchas Africa. Other spirochetes that may cause fetaldeath include Borrelia burgdorferi, which causes Lymedisease, Leptospirosis, and African tick borne relaps-ing fever.

The parasite Toxoplasma gondii may cross theplacenta in association with acute maternal infection.The organism may directly infect the fetus and hasbeen linked to sporadic fetal death, which may occurin up to 5% of pregnancies after first trimester infec-tion. However, the rate of primary infection is about1 per 1,000 in the United States, making it unlikely tocause a substantial proportion of fetal death in theUnited States.25 Other infectious diseases that havebeen associated with sporadic fetal death includemalaria and Q fever. The relative importance of all ofthese infections is influenced by the local prevalenceof the infectious agent.

Small for Gestational Age FetusA major obstetric risk factor for fetal death is thepresence of an SGA fetus. This is a complicatedsubject for several reasons. First, the risk factor ofinterest is fetal IUGR rather than SGA. However,IUGR implies a downward deflection on a growthcurve, requiring several measurements over time,which is unavailable in many populations. In fact,precise knowledge of gestational age is often missing.

Fig. 3. Fetal spleen from a case of parvovirus B19–associ-ated fetal death in the second trimester. Erythroblasts showmarginated chromatin and typical amphophilic intra-nu-clear inclusions (arrows). Hematoxylin and eosin stain. (x1,000, original magnification.) Copyright 1994, J. L. B.Byrne. Used with permission.Silver. Fetal Death. Obstet Gynecol 2007.


Second, by definition, many SGA fetuses will beentirely normal, often termed “constitutionally small.”Third, the use of population-based tables for weightpercentiles does not account for an individual’s inher-ent growth potential. A 6-pound infant may be appro-priate for families with constitutionally small childrenbut represent lagging fetal growth in families destinedto have larger infants. Recent investigations havefocused on developing methods for tracking individ-ual growth potential for fetuses.27 Finally, SGA is nota diagnosis unto itself. Rather it is a sign or a clue fora variety of other conditions that may lead to fetaldeath.

Despite these concerns, it is increasingly apparentthat the presence of an SGA fetus is strongly associ-ated with fetal death. Moreover, there seems to be a“dose–response curve”; the more profound the SGA,the greater the risk for fetal death. This observationstrongly supports biologic plausibility. A populationbased study from Sweden illustrates the use of cus-tomized, rather than population-based growth curves.The odds ratio for stillbirth was 6.1 (95% CI, 5.0–7.5)for SGA fetuses using customized growth charts,compared with 1.2 (95% CI, 0.8–1.9) for SGA fetusesdetermined by population-based curves.27 Antenatalsurveillance including Doppler velocimetry may beuseful in distinguishing which SGA fetuses are at riskfor fetal death. Absent end diastolic velocity or re-verse flow in the umbilical artery is a particularlyworrisome sign and should prompt consideration ofdelivery.

Obstetric ConditionsFetal–Maternal HemorrhageFetal–maternal hemorrhage is one of the most com-mon single disorders responsible for fetal death. Thecondition has been reported in 5–14% of cases.28

Fetal-maternal hemorrhage may be associated withvaginal bleeding or abdominal pain due to abruptionbut also may occur in the absence of symptoms.Because labor and delivery cause fetal–maternal hem-orrhage, ideally assessment of fetal blood in thematernal circulation should be done before delivery.However, it is probably useful to assess for thecondition after delivery if not done previously. Smallamounts of fetal blood routinely enter the maternalcirculation. Accordingly, only large amounts of fetal–maternal hemorrhage, ideally in association with au-topsy confirmation of fetal anemia and hypoxia,should be considered causal for fetal death. In rarecases of vasa previa, fetal blood passes per vaginarather than entering the maternal circulation. Histo-

logic evaluation of the placenta and cord may confirmthe diagnosis.

Multiple GestationThe risk of fetal death is substantially increased inmultiple gestations. Although they account for about3% of births in the United States, multiple gestationscontribute to 10% of fetal deaths. The proportion islikely higher if second trimester losses are included.This is of concern given the continued increase in therate of multiple gestations due to increased use ofassisted reproductive technology. The potentialcauses of fetal death in multiple gestations are numer-ous and include virtually every obstetric complica-tion, including placental insufficiency, abruption, pre-eclampsia, and preterm labor. Other problems areunique to multiple gestation, especially in cases ofmonochorionic placentation, such as twin–twin trans-fusion syndrome (Fig. 4), cord entanglement withmonoamniotic gestations, and twin-reverse arterialperfusion sequence.

Placental AbnormalitiesConditions specific to gestational tissues that maycause fetal death include umbilical cord thrombosis,velamentous cord insertion or vasa previa, and amni-otic band syndrome (Fig. 5). Placental abnormalities

Fig. 4. Placenta demonstrating arterial-to-venous anastomo-ses after injection of milk (arrows) in a pregnancy compli-cated by twin–twin transfusion syndrome. Copyright 1994,J. L. B. Byrne. Used with permission.Silver. Fetal Death. Obstet Gynecol 2007.


also can provide clues regarding other mechanisms ofdeath, such as infection, thrombosis, inflammation,and vascular abnormalities.

Cord AccidentsMany cases of fetal death, especially at term, areattributed to umbilical cord accidents. This is thoughtto occur due to cord occlusion in the presence ofnuchal or body cords and true knots in the cord.However, because cord entanglement occurs in up to30% of uncomplicated pregnancies, and becausethese may be transient findings, caution should beused in attributing fetal death to the presence of thesefindings. Similarly, true knots are usually associatedwith live births. Thus, the presence of a true knot ornuchal cord is insufficient evidence that cord accidentis the cause of death. Ideally, the demonstration ofcord occlusion, fetal hypoxia, and the exclusion ofother causes is required to confirm the diagnosis.

Other Obstetric DisordersNumerous obstetric disorders may directly or indi-rectly cause fetal death. Examples include abruption,preeclampsia, cord prolapse, cervical insufficiency,preterm labor, and preterm premature rupture ofmembranes. These conditions often lead to intrapar-tum, or early neonatal, rather than antepartum death.

Nonetheless, taken together they account for a mean-ingful proportion of fetal deaths.

Other ConditionsA variety of other disorders such as red blood cellalloimmunization may contribute to some cases offetal death. Although fetal death from this conditionhas decreased dramatically due to the use of RhDimmune globulin and improved obstetric care, thecondition continues to be a cause of fetal death.Uterine malformations have been associated withfetal death and should be considered in cases ofrecurrent losses and very early preterm labor orpreterm premature rupture of membranes. Maternaltrauma through motor vehicle accident or violence isa rare but important cause of loss, especially inteenagers.

UnexplainedIn the majority of studies of fetal death, many casesare unexplained, even after extensive evaluation. Inmany cases this is due to inadequate attempts todetermine a cause of death. The proportion of unex-plained fetal deaths also are influenced by whetherconditions that are associated with, but may not bedirectly causal, are accepted as a cause. Losses later ingestation (third trimester) are more likely to be unex-plained than losses earlier in gestation. Such losses arestrongly associated with IUGR as well as most of thepreviously described risk factors for fetal death. Un-doubtedly, continued investigation will identify pre-viously unrecognized causes of fetal death so thatfewer cases remain unexplained.

“Workup” of Fetal DeathThe value of an investigation into potential causes offetal death cannot be overemphasized. First, deter-mining a cause of death helps bring emotional closureto the event. Second, most families at least considertrying to have another child. Invariably, they are quiteinterested in whether there is a chance for recurrence.Finally, in some cases, medical intervention mayreduce the risk of recurrence and improve outcome insubsequent pregnancies. Couples should be coun-seled regarding these issues in a supportive mannerand encouraged to permit an evaluation of theirpregnancy loss with sensitivity to their needs andconcerns.

The optimal evaluation for potential causes offetal death is uncertain. It is necessary to balance costand yield when considering which testing to perform.Thus, it is appropriate to focus on the most commoncauses of fetal death. It also is desirable to emphasize

Fig. 5. Third-trimester fetal death with acalvarium. Onultrasonography, there was suspicion of possible neuraltube defect. However, autopsy demonstrated amnioticband syndrome (arrow points to amniotic band). Copyright1994, J. L. B. Byrne. Used with permission.Silver. Fetal Death. Obstet Gynecol 2007.


conditions with recurrence risk, especially those ame-nable to effective therapies. There is still value, how-ever, in the identification of sporadic conditions. Thismay allow women to avoid unnecessary tests andinterventions in subsequent pregnancies and facilitateemotional healing from the loss.

In the majority of cases, the most valuable test isperinatal autopsy because it provides information thatis pertinent to nearly every potential cause of fetaldeath. Autopsy can identify intrinsic abnormalitiessuch as malformations and metabolic abnormalities,as well as extrinsic problems including hypoxia andinfection. Importantly, perinatal autopsy providesnew information that influences counseling and recur-rence risk in 26–51% of cases.29,30 The rate of perina-tal autopsy varies widely throughout the United Statesand the world, but in all but a few centers withdedicated programs is typically less than 50%. Ifpatients are reluctant to proceed with autopsy, it isworthwhile to ask about their reservations. It often ispossible to work with individuals so that they arecomfortable with the procedure. If families are stilluncomfortable with autopsy, partial autopsy, X-raysor postmortem magnetic resonance imaging mayprovide valuable information. Other barriers to au-topsy are cost (although families are rarely charged), alack of adequately trained pathologists, and miscon-ceptions on the part of clinicians about its value.

After autopsy, placental evaluation is perhaps thenext most valuable test. Gross and histologic exami-nation of gestational tissues is pertinent to a widevariety of conditions, including infection, anemia,hypoxia, and thrombophilias. Patients rarely object tothis procedure. Placental examination is considerablymore informative in the hands of a trained andinterested pathologist.

Karyotype is recommended in all cases of fetaldeath. However, cost may be an issue, dependingupon payor status. In such cases, careful externalevaluation by a pathologist or geneticist is invaluablein deciding whether to obtain a karyotype. The risk ofchromosomal abnormalities in stillbirths with no dys-morphic features (especially if no abnormalities werenoted on antenatal ultrasonography) is probably lessthan 2%.

In some cases of fetal death it is not possible tosuccessfully culture fetal cells to assess karyotype.This is particularly true if there was a long intervalbetween death and the delivery of the fetus. Onestrategy to obtain a reliable karyotype is to do anamniocentesis before delivery. Other approaches areto attempt to culture cells that may survive afterdemise or under low oxygen tension, such as placenta

(especially chorionic plate), fascia latta, tendons, andskin from the nape of the neck. Blood is an excellentcell source if available. In cases of autopsy, tissueshould be sent to the cytogenetic laboratory by thepathologist so that the gross examination is not com-promised. If autopsy is not obtained, it is importantfor the clinician to avoid placing the placenta or fetaltissues in formalin so that cells may be grown inculture. If attempts to culture fetal cells are unsuccess-ful, comparative genomic hybridization has beenused to successfully evaluate fetal chromosomes.31

This technique is increasingly available in cytogeneticlaboratories.

Authorities uniformly agree that autopsy, placen-tal evaluation, and karyotype are worthwhile for mostcases of fetal death. Additional testing is controversialand the cost/benefit ratio is uncertain. Screening forfetal–maternal hemorrhage is advised, because thetest is inexpensive and noninvasive, and the conditionis common. This is typically done with a Kleihauer-Betke test, although some laboratories are using alter-native methods such as flow cytometry to screen forfetal cells in the maternal circulation.

Testing for infection is probably best accom-plished with autopsy and placental histology. Basedon histologic findings, the pathologist may choose toculture fetal tissues or to assess for bacterial or viralnucleic acids. Placental cultures should be consideredexperimental, because positive cultures are commonin association with live births. Serologic testing forsyphilis is advised. In cases of negative first trimestertesting, repeat testing in the third trimester likely canbe limited to high-risk populations. It is reasonable toassess parvovirus serology because this organism ac-counts for a substantial proportion of fetal deaths, andtesting is reliable. Although traditionally recom-mended, the usefulness of “TORCH titers,” (serologyfor toxoplasmosis, rubella, cytomegalovirus, and her-pes simplex) remains uncertain.

Routine testing for antiphospholipid syndromeand heritable thrombophilias also is of questionableusefulness. Most authorities advise testing for lupusanticoagulant and anticardiolipin antibodies in allcases of fetal death. However, the condition is rareunless there is clear evidence of placental insuffi-ciency or other features of antiphospholipid syn-drome such as thrombosis. Testing for heritablethrombophilias is controversial because positive re-sults in healthy women are common. As with an-tiphospholipid syndrome, testing cases associatedwith placental insufficiency or recurrent cases seemsmost indicated. Testing for antiphospholipid syn-drome and heritable thrombophilias is attractive be-


cause treatment may improve outcome in subsequentpregnancies (see below).

An antibody screen (indirect Coomb’s test) ishelpful to exclude red cell alloimmunization. A toxi-cology screen should always be considered. This isusually accomplished with maternal urine, but mea-surement of stable metabolites in fetal tissues such ashair or meconium is gaining popularity. Testing forthyroid disease or diabetes in asymptomatic womenhas been suggested by many authorities, but subclin-ical thyroid disease and diabetes have not beenproven to be associated with fetal death. Uterineimaging studies should be considered in cases of

recurrent loss, preterm premature rupture of mem-branes, and preterm labor.

A summary of recommended tests for the evalu-ation of fetal death is shown in the box, “Recom-mended Evaluation for Stillbirth.” It seems reason-able to limit testing for rare conditions to caseswherein clinical history or other testing raises suspi-cion for a particular disorder. Ideally, the clinicianshould discuss clinical details as well as physical andlaboratory findings with the pathologist so that theworkup is tailored for each individual loss. Ongoingpopulation based studies such as that being conductedby the Stillbirth Collaborative Research Network ofthe National Institute of Child Health and HumanDevelopment (scrn.rti.org) may clarify the optimalevaluation of fetal death.

Delivery of the FetusMost women prefer to proceed with delivery of thefetus after diagnosis of fetal death. It is emotionallystressful to carry a nonviable fetus, especially late ingestation. Nonetheless, there is no medical urgency toeffect delivery and for some women a delay is emo-tionally desirable. For example, some women preferto grieve with their families and do not feel “up to” amedical procedure immediately after diagnosis.Other couples may even prefer prolonged expectantmanagement, usually prompted by a desire to avoidinduction of labor. It is important to offer both theoptions of delivery and expectant management towomen experiencing fetal death.

Risks of expectant management include intrauter-ine infection and maternal coagulopathy. These risksare poorly characterized due to the relative infre-quency of expectant management. Older reports statethat 80–90% of women will spontaneously laborwithin two weeks of fetal death.32 However, this“latency” period may be substantially longer. It seemsprudent to perform surveillance for infection andcoagulopathy in women undergoing expectant care.Examples include serial assessment of maternal tem-perature, abdominal pain, bleeding, and labor. Reg-ular office visits (eg, weekly) may be useful for emo-tional support and medical surveillance. Someauthorities advise serial (eg, weekly) determination ofcomplete blood count, platelet count, and fibrinogenlevel. The usefulness of this is uncertain. A consump-tive coagulopathy has been reported in 25% of pa-tients who retain a dead fetus for more than 4 weeks,33

but the condition is rare in clinical practice and is notusually associated with clinical sequelae. A fibrinogenlevel of less than 100 mg/dL is considered evidence ofcoagulopathy. Patients should be advised to immedi-

Recommended Evaluation for StillbirthRecommended in most cases:• Perinatal autopsy• Placental evaluation• Karyotype• Antibody screen1

• Serologic test for syphilis2

• Screen for fetal–maternal hemorrhage(Kleihauer-Betke or other)

• Urine toxicology screen• Parvovirus serologyRecommended if clinical suspicion:

• Lupus anticoagulant screen3

• Anticardiolipin antibodies3

• factor V Leiden mutation3

• Prothrombin G20210A mutation3

• Screen for protein C, protein S, and antithrom-bin III deficiency4

• Uterine imaging study5

Not recommended at present:• Thyroid-stimulating hormone• Glycohemoglobin• TORCH titers6

• Placental cultures• Testing for other thrombophilias

1. Negative first-trimester screen does not require repeat testing.

2. Repeat testing in cases of negative first trimester screen ifhigh-risk population.

3. Test in cases of thrombosis, placental insufficiency, and recur-rent fetal death.

4. These thrombophilias are rare in the absence of personal orfamily history of thrombosis.

5. Test in cases of unexplained recurrent loss, preterm prematurerupture of membranes, and preterm labor.

6. TORCH titers, serology for toxoplasmosis, rubella, cytomega-lovirus, and herpes simplex.


ately report symptoms such as fever, pain, bleeding,contractions, leaking fluid, or foul discharge.

Depending upon gestational age, evacuation ofthe uterus may be accomplished medically or surgi-cally. Ideally, the choice of delivery method shouldbe made by the parents based on personal preference.Some women prefer surgical evacuation of the uterusbecause the procedure is rapid and they may receivea general anesthetic. Others choose induction of laborso that they may experience labor and deliver anintact fetus.

In experienced hands, dilation and evacuation inthe early second trimester is as safe as medicalinduction of labor. This procedure becomes techni-cally more difficult in the later second and thirdtrimesters. At more advanced gestational ages, laborinduction is safer and dilation and evacuation shouldbe reserved for physicians with more extensive expe-rience and skills in this procedure. Induction of laborhas been greatly aided by the availability of prosta-glandin preparations. There is considerable experi-ence with the use of prostaglandin E2� (PGE2�) toinduce labor in women with fetal demise. During thepast decade, misoprostol has largely replaced PGE2

for induction of labor in cases of fetal death due tosimilar efficacy with fewer side effects.34 Adverseeffects of prostaglandins include fever, nausea, eme-sis, and diarrhea, particularly if a PGE2 preparation isused. Pretreatment with antiemetics, antipyretics, andantidiarrheals may reduce symptoms.

Misoprostol may be administered orally as a loz-enge or vaginally in the posterior fornix. Dosing ofmisoprostol is influenced by the size of the uterus andseveral approaches have been published. If the uterus isless than 28 weeks size, our approach is to place 200 mgof misoprostol in the posterior fornix every 4 hours untildelivery of the fetus and placenta. The dose could beincreased to 400 mg every 2 hours, but delivery is nothastened compared with 200 mg every 4 hours. Theoral dose (taken as a lozenge) is 200–400 mg every 2–4hours. The interval to delivery is less when the drug isadministered vaginally compared with orally, but somewomen may prefer to take the drug by mouth. If theuterus is greater than 28 weeks size, we administer aninitial dose of 25 mg of misoprostol in the posteriorfornix, followed by 25–50 mg every 4 hours. Alterna-tively, it may be given orally at a dosage of 25 mg every4 hours. Prostaglandin E2 should not be used in womenwith active cardiac, pulmonary or renal disease, andglaucoma. All prostaglandins for medical induction oflabor should be avoided in cases of prior cesarean ifuterine size is greater than 26 weeks of gestation at thetime of induction. In such cases we use oxytocin (low

dose if the cervix is unfavorable). Risks of uterinerupture must be weighed against the desire to avoidhysterotomy in women with fetal death. Although uter-ine rupture may occur, misoprostol has been used safelyfor second trimester induction of labor in women withprior cesarean delivery.35

In cases of fetal death in the second trimester,especially at less than 20 weeks of gestation, there isan increased risk for retained placenta. Allowing theplacenta to deliver spontaneously without “pulling onthe umbilical cord” can greatly reduce this risk.Additional doses of misoprostol may be administered(at appropriate intervals) to promote uterine contrac-tility between delivery of the fetus and placenta. Inmy practice I do not use a “time limit” for the deliveryof the placenta in the absence of bleeding or emo-tional duress. Placental delivery rarely takes morethan 2 hours. The use of misoprostol may reduce theincidence of retained placenta to less than 5%, whichseems to be lower than seen with oxytocin or PGE2.34

BereavementThe facilitation of bereavement is an important op-portunity for clinicians to help families. Many practi-tioners, especially obstetric providers, are uncomfort-able with death and avoid frank discussions withpatients. This is especially true in cases wherein theclinician is worried about being at fault. It is importantto overcome these fears and to directly address all ofthe patient’s questions. It is helpful to develop astandard bereavement protocol, particularly in unitsthat rarely deliver fetal deaths. Patients should beoffered the opportunity to hold their infants and tokeep mementos such as pictures, foot and handprints,and plaster casts. Visits with clergy and supportgroups and psychological counseling should be of-fered. Patients should be allowed to make as manychoices as possible regarding their experience. Pro-longed hospitalizations are unnecessary and recoveryon postpartum wards should be avoided.

Subsequent Pregnancy ManagementThe risk for virtually all adverse pregnancy outcomesare influenced by prior obstetric history, and fetal deathis no exception. The recurrence risk for fetal death is notwell studied and reliable numbers for individual patientsare often unavailable. A recent population-based studyfrom Missouri noted a stillbirth rate of 22.7 of 1,000 inwomen with prior stillbirth, representing an odds ratioof 4.7 (95% CI 1.2–5.7) compared with women withoutprior stillbirth.36 Increased recurrence risks were notedin African Americans (35.9/1,000) compared withwhites.36 Recurrence risk may be stratified by cause of


stillbirth. For example, losses associated with placentalinsufficiency, prematurity, or some genetic conditionsare more likely to recur, whereas those due to infectionor abnormalities of twinning are less likely. Also, fetaldeath earlier in gestation is more likely to recur thanlosses at term. Finally, patients with recurrent fetal deathare at much higher risk than those with sporadic loss.5

Strategies to prevent recurrence depend upon thecause of the prior loss(es). Families with identifiedgenetic conditions may be counseled about reproduc-tive options including antenatal and preimplantationgenetic diagnosis. Improved medical care for mater-nal disorders such as diabetes and hypertension cansubstantially improve outcome in subsequent preg-nancies. The same is true for women with red cellalloimmunization. Although not universally accepted,there is evidence that treatment with thromboprophy-laxis can improve the live birth rate in women withantiphospholipid syndrome.37,38

Data are less clear for heritable thrombophiliasbecause they are common in healthy women. Onewell-designed prospective randomized trial com-pared low molecular weight heparin and low-doseaspirin to low-dose aspirin alone in 160 womenwith prior fetal death and thrombophilia.39 Preg-nancy outcome was dramatically improved in thelow molecular weight heparin group with a livebirth rate of 71% compared with 14% for low-doseaspirin alone.39 These data are promising but mustbe interpreted with caution. First, results have notbeen confirmed in other trials. Second, the rate ofpregnancy loss in the control group was extremelyhigh (86%) and much higher than anticipated basedon risk factors. Thus, current data are insufficient torecommend routine thromboprophylaxis forwomen with thrombophilias.

Counseling regarding smoking cessation, weightloss in obese women and the proper use of seat beltsduring pregnancy also may reduce the rate of stillbirth.Although of unproven efficacy these public health mea-sures make good common sense for all women.

Antenatal surveillance is widely recommendedin subsequent pregnancies for patients with priorfetal death. Clinical usefulness has been suggestedby older studies and the test is likely to benefit thesubset of pregnancies at risk for placental insuffi-ciency. It is noteworthy that in addition to recurrentpregnancy loss, prior fetal death increases the riskfor many obstetric complications, including IUGR,abruption, and preterm birth. The most commonlyemployed surveillance method is the nonstress test.Although some authorities advise testing 2– 4 weeksbefore the gestational age of the fetal death, initiat-

ing testing at 32 weeks of gestation likely works aswell and may reduce the chance of false-positiveresults.40 Alternatively, Doppler velocimetry, amni-otic fluid indexes, and serial ultrasonograms toassess growth may be used to assess placentalfunction. Induction of labor is another commonstrategy used in women with prior fetal death. Aswith antenatal surveillance, many clinicians advisedelivery at a gestational age 2 weeks before theprevious loss. This recommendation should beviewed with caution because of unproven efficacy(with regard to stillbirth prevention) and the poten-tial for clinically relevant prematurity. However,induction has tremendous emotional benefit formany couples with prior fetal death. Accordingly,elective induction in the setting of pulmonary ma-turity and a favorable cervix may be appropriate inwell-selected cases. Indeed, a large component ofproviding good care in subsequent pregnancies inwomen with prior fetal death is to tend to thepatient’s emotional needs. Frequent visits, docu-mentation of fetal heart tones and well being andlots of positive reinforcement are invaluable.

CONCLUSIONFetal death remains a common, traumatic and insome cases preventable complication of pregnancy.Delivery may be safely accomplished either medi-cally or surgically and expectant management is asafe alternative for interested patients. The stron-gest risk factors for fetal death are African-Ameri-can race, prior fetal death, obesity, small for gesta-tional age fetus, and advanced maternal age.Common causes and risk factors for fetal deathinclude chromosomal abnormalities, genetic syn-dromes, infections, placental abnormalities, fetal–maternal hemorrhage, maternal diseases such asdiabetes and hypertension, antiphospholipid syn-drome, thrombophilias, and abnormalities of mul-tiple gestation. Clinicians should encourage fami-lies to allow a thorough investigation of potentialcauses of fetal death to facilitate emotional closure,to assess recurrence risk, and in some cases toreduce recurrence risk. The optimal “workup” forfetal death is uncertain. Recommended tests in-clude perinatal autopsy, placental evaluation, fetalkaryotype, Kliehauer-Betke, antibody screen, and aserologic test for syphilis. Other tests to considerinclude anticardiolipin antibodies, lupus anticoagu-lant screen, testing for heritable thrombophilias,urine toxicology screen, and parvovirus serology.Subsequent pregnancies may be at increased riskfor fetal death and obstetric complications. Treat-


ment of underlying medical or obstetric conditions,antenatal surveillance, and induction of labor withfetal maturity may improve outcome. One hopesthat ongoing research will elucidate causes forpreviously unexplained fetal death and focus effortson effective prevention.

REFERENCES1. Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer

JP, Canfield RE, et al. Incidence of early loss of pregnancy.N Engl J Med 1988;319:189–94.

2. Kline J, Stein Z. Epidemiology of chromosomal abnormalitiesin spontaneous abortion: prevalence, manifestation and deter-minants. In: Spontaneous and recurrent abortion. Bennett MJ,Edmonds DK, editors. Oxford (United Kingdom): BlackwellScientific Publications; 1987. p. 29–50.

3. Oshiro BT, Silver RM, Scott JR, Yu H, Branch DW. Antiphos-pholipid antibodies and fetal death. Obstet Gynecol 1996;87:489–93.

4. Rey E, Kahn SR, David M, Shrier I. Thrombophilic disordersand fetal loss: a meta-analysis. Lancet 2003;361:901–8.

5. Frias AE Jr, Luikenaar RA, Sullivan AE, Lee RM, Porter TF,Branch DW, et al. Poor obstetric outcome in subsequentpregnancies in women with prior fetal death. Obstet Gynecol2004;104:521–6.

6. Johansen KS, Hod M. Quality development in perinatal care: theOBSQID project. OBStetrical Quality Indicators and Data. Int JGynaecol Obstet 1999;64:167–72.

7. Cartlidge PH, Stewart JH. Effect of changing the stillbirthdefinition on evaluation of perinatal mortality rates. Lancet1995;346:486–8.

8. Baird D, Walker J, Thomson AM. The causes and preventionof stillbirths and first week deaths. III. A classification of deathsby clinical cause; the effect of age, parity and length ofgestation on death rates by cause. J Obstet Gynaecol Br Emp1954;61:433–48.

9. Hey EN, Lloyd DJ, Wigglesworth JS. Classifying perinataldeath: fetal and neonatal factors. Br J Obstet Gynaecol 1986;93:1213–23.

10. Gardosi J, Kady SM, McGeown P, Francis A, Tonks A. Classifi-cation of stillbirth by relevant condition at death (ReCoDe):population based cohort study. BMJ 2005;331:1113–7.

11. Fretts RC. Etiology and prevention of stillbirth. Am J ObstetGynecol 2005;193:1923–35.

12. Froen JF, Arnestad M, Frey K, Vege A, Saugstad OD, Stray-Pedersen B. Risk factors for sudden intrauterine unexplaineddeath: epidemiologic characteristics of singleton cases in Oslo,Norway, 1986-1995. Am J Obstet Gynecol 2001;184:694–702.

13. Vintzileos AM, Ananth CV, Smulian JC, Scorza WE, KnuppelRA. Prenatal care and black-white fetal death disparity in theUnited States: heterogeneity by high-risk conditions. ObstetGynecol 2002;99:483–9.

14. Fretts RC, Schmittdiel J, McLean FH, Usher RH, GoldmanMB. Increased maternal age and the risk of fetal death. N EnglJ Med 1995;333:953–7.

15. Bateman BT, Simpson LL. Higher rate of stillbirth at the extremesof reproductive age: a large nationwide sample of deliveries in theUnited States. Am J Obstet Gynecol 2006;194:840–5.

16. Kristensen J, Vestergaard M, Wisborg K, Kesmodel U, SecherNJ. Pre-pregnancy weight and the risk of stillbirth and neonataldeath. BJOG 2005;112:403–8.

17. Cundy T, Gamble G, Townend K, Henley PG, MacPherson P,Roberts AB. Perinatal mortality in Type 2 diabetes mellitus.Diabet Med 2000;17:33–9.

18. Beischer NA, Wein P, Sheedy MT, Steffen B. Identificationand treatment of women with hyperglycaemia diagnosedduring pregnancy can significantly reduce perinatal mortalityrates. Aust N Z J Obstet Gynaecol 1996;36:239–47.

19. Lockwood C, Silver R. Thrombophilias in pregnancy. In:Creasy R, Resnick R, Iams J. Maternal-fetal medicine: princi-ples and practice. 5th ed. Philadelphia (PA): WB SaundersCompany; 2003. p. 1005–22.

20. Dizon-Townson D, Miller C, Sibai B, Spong CY, Thom E,Wendel G Jr, et al. The relationship of the factor V Leidenmutation and pregnancy outcomes for mother and fetus.Obstet Gynecol 2005;106:517–24.

21. Kesmodel U, Wisborg K, Olsen SF, Henriksen TB, Secher NJ.Moderate alcohol intake during pregnancy and the risk ofstillbirth and death in the first year of life. Am J Epidemiol2002;155:305–12.

22. Whitehead N, Lipscomb L. Patterns of alcohol use before andduring pregnancy and the risk of small-for-gestational-agebirth. Am J Epidemiol 2003;158:654–62.

23. Wapner RJ, Lewis D. Genetics and metabolic causes ofstillbirth. Semin Perinatol 2002;26:70–4.

24. Pauli RM, Reiser CA. Wisconsin Stillbirth Service Program: II.Analysis of diagnoses and diagnostic categories in the first1,000 referrals. Am J Med Genet 1994;50:135–53.

25. Goldenberg RL, Thompson C. The infectious origins of still-birth. Am J Obstet Gynecol 2003;189:861–73.

26. Skjoldebrand-Sparre L, Tolfvenstam T, Papadogiannakis N,Wahren B, Broliden K, Nyman M. Parvovirus B19 infection:association with third-trimester intrauterine fetal death. BJOG2000;107:476–80.

27. Clausson B, Gardosi J, Francis A, Cnattingius S. Perinataloutcome in SGA births defined by customised versus popula-tion-based birthweight standards. BJOG 2001;108:830–4.

28. Laube DW, Schauberger CW. Fetomaternal bleeding as acause for “unexplained” fetal death. Obstet Gynecol 1982;60:649–51.

29. Faye-Petersen OM, Guinn DA, Wenstrom KD. Value ofperinatal autopsy. Obstet Gynecol 1999;94:915–20.

30. Michalski ST, Porter J, Pauli RM. Costs and consequences ofcomprehensive stillbirth assessment. Am J Obstet Gynecol2002;186:1027–34.

31. Christiaens GC, Vissers J, Poddighe PJ, de Pater JM. Compar-ative genomic hybridization for cytogenetic evaluation ofstillbirth. Obstet Gynecol 2000;96:281–6.

32. Goldstein DP, Reid DE. Circulating fibrinolytic activity—aprecursor of hypofibrinogenemia following fetal death in utero.Obstet Gynecol 1963;22:174–80.

33. Diagnosis and management of fetal death. ACOG TechnicalBulletin Number 176—January 1993. Int J Gynaecol ObstetJanuary 1993;42:291–9.

34. Ramsey PS, Savage K, Lincoln T, Owen J. Vaginal misoprostolversus concentrated oxytocin and vaginal PGE2 for second-trimester labor induction. Obstet Gynecol 2004;104:138–45.

35. Dickinson JE. Misoprostol for second-trimester pregnancytermination in women with a prior cesarean delivery. ObstetGynecol 2005;105:352–6.

36. Sharma PP, Salihu HM, Oyelese Y, Ananth CV, Kirby RS. Israce a determinant of stillbirth recurrence? Obstet Gynecol2006;107:391–7.


37. Kutteh WH. Antiphospholipid antibody-associated recurrentpregnancy loss: treatment with heparin and low-dose aspirin issuperior to low-dose aspirin alone. Am J Obstet Gynecol1996;174:1584–9.

38. Rai R, Cohen H, Dave M, Regan L. Randomised controlledtrial of aspirin and aspirin plus heparin in pregnant womenwith recurrent miscarriage associated with phospholipid anti-bodies (or antiphospholipid antibodies). BMJ 1997;314:253–7.

39. Gris JC, Mercier E, Quere I, Lavigne-Lissalde G, Cochery-Nouvellon E, Hoffet M, et al. Low-molecular-weight heparinversus low-dose aspirin in women with one fetal loss and aconstitutional thrombophilic disorder. Blood 2004;103:3695–9.

40. Weeks JW, Asrat T, Morgan MA, Nageotte M, Thomas SJ,Freeman RK. Antepartum surveillance for a history of still-birth: when to begin? Am J Obstet Gynecol 1995;172:486–92.

ACOG members can access many of the world's premier medical journals online throughOvid. In these highly cited journals, you're sure to find the content resources you need toanswer your clinical and research questions and advance your medical knowledge.

Ovid offers you:· Full-text online access to selected Lippincott Williams & Wilkins publications· Ability to search across the Ovid platform· Access to Ovid Medline

To access Ovid online:· Enter the member section at www.acog.org · Under the “Information” tab, click on the link to “Search Ovid”

Search the World's Premier Medical Research!Special Member-Only Access to OVID Online


obstet gynecol 2007; p153

Documents