Infertility trial outcomes: healthymoms and babies

Robert Silver, M.D.

Division of Maternal–Fetal Medicine, Department of Obstetrics and Gynecology, University of Utah, Salt Lake City, Utah

Traditionally, the primary outcome of infertility trials has been a positive pregnancy test or a clinically recognized pregnancy. However,parents desire a healthy baby that grows up to be a healthy adult, rather than a positive pregnancy test. Too often results of infertilitytrials are lacking in crucial obstetric details. This is problematic because treatments for infertility have the capacity to increase the riskfor a variety of adverse obstetric outcomes. This reviewwill outline important obstetric variables that should be included when reporting

Use your smartphone

infertility research. The rationale for including these data, precise definitions of the variables,and cost-effective strategies for obtaining these obstetric details will be highlighted. (Fertil Ster-il� 2014;101:1209–16. �2014 by American Society for Reproductive Medicine.)Key Words: Obstetric complications, preterm birth, pregnancy loss, infertility treatment,clinical trials

Discuss: You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/silver-infertility-trial-outcomes-healthy-moms-babies/

to scan this QR codeand connect to thediscussion forum forthis article now.*

* Download a free QR code scanner by searching for “QRscanner” in your smartphone’s app store or app marketplace.

T he primary outcome for manyinfertility trials is a positivepregnancy test or ‘‘ongoing’’

live pregnancy. This is understandablebecause the goal of most infertilitytreatment is to facilitate conception.However, parents view this quite differ-ently. They desire a live baby that theycan take home with them. Indeed, theyhave no interest in a positive pregnancytest that does not result in a ‘‘take-home’’ baby. Moreover, they are inter-ested in having healthy infants withno handicaps and normal lifespans.Avoiding serious maternal morbidityduring pregnancy also is desirable. Ofcourse it is expensive and impracticalto follow children for many years toassess long-term developmental out-comes and whether they gain admis-sion to the college of their choice.Nonetheless, many importantmaternal, fetal, and neonatal outcomescan be easily and efficiently assessed,

Received February 12, 2014; revised March 12, 2014;R.S. has nothing to disclose.Reprint requests: Robert Silver, M.D., Division of Mat

and Gynecology, University of Utah, 30 N. 190bob.silver@hsc.utah.edu).

Fertility and Sterility® Vol. 101, No. 5, May 2014 001Copyright ©2014 American Society for Reproductivehttp://dx.doi.org/10.1016/j.fertnstert.2014.03.027

VOL. 101 NO. 5 / MAY 2014

and such outcomes should routinelybe reported in infertility trials.

When reporting pregnancy out-comes it is important to preciselycommunicate using standardized defi-nitions. Unfortunately, some of thepertinent outcomes have numerousdefinitions in common use that varyamong countries and occasionally pro-viders. Whenever possible, this articlewill use evidence-based, standard, andgenerally accepted definitions ofadverse outcomes. If there are insuffi-cient data available to yield generallyaccepted definitions, the rationale forthose used will be presented.

There are numerous examples ofpotential interactions between infer-tility treatment and obstetric outcomes.For example, IVF and/or intracytoplas-mic sperm injection (ICSI) seem to beassociated with a slight increase in therisk for birth defects. Although the as-sociation remains controversial and

accepted March 13, 2014.

ernal–Fetal Medicine, Department of Obstetrics0 E., 2B200, Salt Lake City, Utah 84132 (E-mail:

5-0282/$36.00Medicine, Published by Elsevier Inc.

may be related to infertility ratherthan treatment for infertility, a recentmeta-analysis including 46 studiesand 124,468 infants noted a pooledrisk estimation for birth defects of1.37 (95% confidence interval [CI]1.26–1.48) for IVF/ICSI (1). In contrast,a recent population-based study fromAustralia noted an increased risk ofbirth defects after ICSI but not IVF (2).Additionally, IVF and ICSI have beenassociated with an increased risk forimprinting disorders. A systematic re-view of eight studies reported that therelative risk for having a child withBeckwith–Wiedemann Syndrome was5.2 (95% CI 1.6–7.4) after IVF/ICSI (3).However, many of the studies did notadequately correct for the effects ofinfertility itself, and although the au-thors acknowledged an increase in therisk of imprinting disorders after IVF/ICSI, they also state that proof of acausal relationship is lacking (3).

In addition to fetal abnormalities,obstetric complications may be affectedby treatments for infertility. A recentlarge, systematic review and meta-analysis found increased risks of ante-partum hemorrhage (relative risk [RR]2.49; 95% CI 2.30–2.69), hypertensivedisorders of pregnancy (RR 1.49; 95%

1209

VIEWS AND REVIEWS

CI 1.39–1.59), preterm premature rupture of membranes(PPROM) (RR 1.16; 95% CI 1.07–1.26), preterm birth (RR1.54; 95% CI 1.47–1.62), small for gestational age infant(RR 1.39; 95% CI 1.27–1.53), perinatal mortality (RR 1.87;95% CI 1.48–2.37), and gestational diabetes (RR 1.48; 95%CI 1.33–1.66) (4). It is still unclear whether these risks arerelated to the fertility treatments themselves or to risks asso-ciated with conditions linked to infertility (5). Nonetheless,some effects may be related to the specific method of treat-ment. In pregnancies achieved with IVF, the use of frozen em-bryos resulted in decreased perinatal mortality, small forgestational age infant, preterm birth, and antepartum hemor-rhage compared with those using fresh embryos (6). Anotherreview noted a higher risk of preterm delivery but a lowerchance of small for gestational age infant in pregnancies re-sulting from blastocyst (rather than cleavage-stage embryo)transfer (7).

There are numerous other examples of the potential ef-fects of fertility treatments on a variety of obstetric outcomes.Any medical exposure or technical manipulation of gametesor embryos could potentially affect clinically relevant preg-nancy outcomes. Consequently, accurate reporting of theseoutcomes should be mandatory for clinical trials of infertilitytherapies.

GESTATIONAL AGEWhen assessing most adverse perinatal outcomes it is impor-tant to have accurate information regarding gestational ageat delivery. This is crucial for conditions such as preterm birth(PTB) and fetal growth restriction (FGR). Fortunately, mostinfertility trials should capture high-quality data regardinggestational age. Numerous algorithms are available to deter-mine gestational age. Most are based on a hierarchical schemebased on last menstrual period (if known and reliable) and ob-stetric sonograms performed early in gestation. It should bevery easy to accurately determine gestational age in pregnan-cies conceived with IVF and ET or after well-documentedovulation by objective means. For other circumstances, closeattention to last menstrual period and early sonogram isadvised.

PREGNANCY LOSSThe terminology used to describe both pregnancy loss and livebirth is confusing for patients and physicians. Traditional def-initions do not reflect our current understanding of reproduc-tive biology, and many of the terms do not make sense givenour current knowledge base (8). For example, traditionally allpregnancy losses before 20 weeks' gestational age are termed‘‘spontaneous abortions.’’ Fetal death in utero after 20 weeks'gestation are termed stillbirths, and live births between 20 and37 weeks' gestation are referred to as preterm live births. Thedownside of this approach is that it lumps together manydisparate conditions with different causes and prognoses.For example, spontaneous abortions include early lossesdue to aneuploidy, second-trimester fetal deaths due toabnormal placentation, and second-trimester preterm birthsof live fetuses.

1210

It would be preferable to use terminology that more accu-rately reflects our current knowledge of developmentalbiology. Pregnancy losses can be stratified according to thedevelopmental stage during which they happen. Theseinclude pre-embryonic, embryonic, and fetal losses, occurringbefore 6, between 6 and 10, and after 10 weeks' gestation,respectively. This is important for several reasons. First, thereis considerable overlap between infertility (or subfertility) andpregnancy loss. Although few data are available, it is likelythat the overlap is most profound in cases of early pregnancyloss. Thus, such early losses are potentially linked to infertilitytreatments. Second, the causes of pregnancy loss vary acrossdevelopmental epochs. Early losses are most likely to be asso-ciated with aneuploidy, whereas antiphospholipid antibodiesare more strongly associated with fetal deaths (9). Finally, theprognosis in subsequent pregnancies varies for losses indifferent developmental periods. In general, later losses aremore likely to be recurrent than early losses.

It also is important to distinguish between preterm livebirths and in utero deaths before the onset of labor. At presentall losses before 20 weeks' gestation are lumped together asspontaneous abortions. However, the threshold of 20 weeks'gestation is arbitrary because the pathophysiology of sponta-neous preterm births (SPTBs) is similar in cases before and af-ter 20 weeks' gestation (10). Addionally, the recurrence rate issimilar for those with SPTB before and after 20 weeks' gesta-tion (11). Accordingly, definitions should not use 20 weeks asa criterion for defining SPTB.

The definition of stillbirth can also be problematic withregard to SPTB. Spontaneous preterm birth often leads to still-birth that occurs intrapartum. Typically, some combination ofpreterm labor, PPROM, cervical insufficiency, bleeding, andchorioamnionitis leads to preterm birth. If this occurs at a pre-viable gestation (e.g., <24 weeks' gestation), most cliniciansdo not intervene with cesarean delivery for the usual fetal in-dications. Thus, if the fetus does not tolerate the stress of la-bor, for example due to cord compression, it results in anintrapartum stillbirth before 24 weeks' gestation. In contrast,the same scenario results in a preterm live birth (with a cesar-ean delivery) if it occurs later in gestation. Hence it is impor-tant to distinguish between intrapartum and antepartumstillbirths. Table 1 shows proposed definitions for reportingpregnancy losses (and preterm births).

OBSTETRIC COMPLICATIONSMany obstetric complications have the potential to harm bothmother and fetus. Pre-eclampsia serves as a good example.However, others are limited to the mother (such as endome-tritis) or fetus/neonate (e.g., FGR). It is important to considermaternal, fetal, and neonatal consequences of obstetricdisorders.

Preterm Birth

Preterm birth is one of the most important adverse obstetricoutcomes, affecting more than 12% of pregnancies in theUnited States (12). It is typically defined as delivery before37 weeks' gestational age. Thus, precise knowledge of gesta-tional age is critical in the accurate reporting of this

VOL. 101 NO. 5 / MAY 2014

TABLE 1

Proposed nomenclature for pregnancy loss and spontaneous preterm birth.

Proposed nomenclature Definition Old nomenclature Developmental stage

Early pregnancy loss Pregnancy loss before 10 wk of gestation with no documentationof whether an embryo was visible on ultrasonogram. This mayinclude cases in which there are more than 10 wk from theLMP if there is no clear evidence of passing a formed fetus;positive urine or serum hCG (pregnancy test)

Spontaneous abortionAlso termed ‘‘positive pregnancy

test’’ or ‘‘chemical pregnancy’’

Morula, blastocyst,gastrula, embryo

Peri-implantational loss Pregnancy loss before 5 wk of gestation with no gestational sacvisible on ultrasonogram

Spontaneous abortionAlso termed ‘‘positive pregnancy

test’’ or ‘‘chemical pregnancy’’

Morula, blastocyst,gastrula

Pregnancy of unknownlocation, includingectopic pregnancy

A pregnancy that is not in the uterus; this includes definite ectopicpregnancy, probable ectopic pregnancy, pregnancy ofunknown location, and probable intrauterine pregnancy

Ectopic pregnancy N/A

Pre-embryonic loss Pregnancy loss at 5 0/7–5 6/7 wk of gestation with visiblegestational sac or yolk sac or both but no visible embryo onultrasonogram; gestational sac may measure greater than5 wk of gestation

Spontaneous abortionAlso termed ‘‘clinical pregnancy’’

Morula, blastocyst,gastrula,

Embryonic loss Pregnancy loss at 6 0/7–9 6/7 wk of gestation of an embryo withCRL less than 10 mm and no cardiac activity visible onultrasonogram

Spontaneous abortionAlso termed ‘‘clinical pregnancy’’

Embryo

Fetal death At 10 0/7–19 6/7 wk of gestation, either [1] passage of aconceptus measuring at least 30 mm in CRL, [2]ultrasonographic documentation of a dead conceptus withCRL consistent with a gestational age of at least 10 wk ofgestation (CRL at least 30 mm), or [3] loss of the conceptusafter documented fetal cardiac activity at or beyond 10 wk ofgestational age; the fetus is dead at the time of delivery, withApgar scores of 0 at 1 and 5 min

Spontaneous abortion Fetus

Early fetal death 10 0/7–15 6/7 wk of gestationLate fetal death 16 0/7–19 6/7 wk of gestationNeonatal death Death of a formed fetus alive at birth in the first 28 d of life Neonatal death NeonateStillbirth Fetus at least 20 wk of gestation that is dead at the time of

delivery, with Apgar scores of 0 at 1 and 5 minStillbirth Fetus

Voluntary terminationof pregnancy

Termination of pregnancy by induction of labor or surgicalevacuation of the uterus

Abortion N/A

Spontaneous preterm birth Preterm birth from 16 0/7 to 36 6/7 wk of gestation associatedwith one of the following: [1] classic preterm labor, [2]preterm premature rupture of membranes, [3] bleeding orabruption, [4] spontaneous premature cervical ripening, [5]chorioamnionitis

Spontaneous abortion (before 20 wk ofgestation) and spontaneous pretermbirth (after 20 wk of gestation)

Neonate

Medically indicatedpreterm birth

Preterm birth from 16 0/7 to 36 6/7 wk of gestation associatedwith induction of labor or cesarean delivery owing to amaternal or fetal complication such as pre-eclampsia orabnormal fetal heart rate tracing

Indicated abortion (before 20 wk ofgestation) and indicated preterm birth(after 20 wk of gestation)

Neonate

Late preterm birth Preterm birth from 34 0/7 to 36 6/7 wk of gestation Late preterm birth NeonateVery late preterm birth Preterm birth from 37 0/7 to 38/67 wk of gestation Near term or early term birth NeonateNote: LMP ¼ last menstrual period; N/A ¼ not applicable; CRL ¼ crown–rump length.Reproduced with permission from Silver RM, Branch DW, Goldenberg R, Iams JD, Klebanoff MA. Nomenclature for pregnancy outcomes. Time for a change. Obstet Gynecol 2011;118:1402–8.

Silver. Pregnancy outcomes in infertility trials. Fertil Steril 2014.

VOL.101

NO.5

/MAY2014

1211

Fertilityand

Sterility®

VIEWS AND REVIEWS

complication. The condition accounts for considerablemorbidity and mortality. For example, PTB contributes to35% of neonatal deaths and is the secondmost common causeof death in children aged <5 years (13). Long-term morbidityin surviving children is common, accounting for at least $26.2billion in costs in the United States in 2005 (14).

It is important to also provide details regarding the rea-sons for the preterm birth. There are several ways to catalogthe various causes of preterm birth. One of the most commoninvolves distinguishing between SPTB and medically indi-cated preterm birth. Spontaneous preterm birth includes de-liveries at %37 weeks' gestation due to the spontaneousonset of uterine contractions leading to preterm birth and/or PPROM leading to preterm birth. Preterm prematurerupture of membranes may lead to spontaneous preterm la-bor, induction of labor, or cesarean delivery without labor,but if the primary reason for preterm birth is PPROM it isconsidered an SPTB. Spontaneous preterm birth typically ac-counts for approximately two-thirds of all preterm births.

In contrast, medically indicated preterm birth occurs inthe setting of iatrogenic delivery (either after induction of la-bor or cesarean delivery) due to medical, obstetric, or fetalcomplications. In these cases expectant management of thepregnancy poses an increased risk of adverse maternal, fetal,and/or neonatal outcomes. The clinician has determined thatthe risks of expectant management exceed the risks of prema-turity at the gestational age at delivery. Examples include ob-stetric problems such as pre-eclampsia or placenta accreta,maternal medical conditions such as worsening cardiac dis-ease, or fetal conditions such as abnormal antenatal testingor FGR. A (partial) list of conditions wherein medically indi-cated preterm birth is considered appropriate is outlined inTable 2 (15).

There is no doubt that treatments for infertility havecontributed to many cases of preterm birth, primarily throughthe generation of multiple gestations. This occurs after med-ical therapy to stimulate ovulation, as well as with IVF.Indeed, it is estimated that in 2011 in the United States,36% of twin births and 77% of triplet or higher-order multiplegestations were due to fertility treatments (16).

Hypertensive Disorders

Hypertensive disorders of pregnancy affect 5–10% of gesta-tions and are a major cause of both maternal and fetal/neonatal morbidity and even mortality. The main fetal risksare due to prematurity, although FGR, abruption, stillbirth,and placental insufficiency also contribute to adverse fetaloutcomes. In cases of severe disease, mothers may sufferstroke due to, for example, hypertension, hemorrhage, pul-monary edema, or ruptured liver. There also is an increasedrisk of cesarean delivery as well as an increased risk of subse-quent cardiovascular and metabolic disease. Pre-eclampsiacauses an estimated 50,000–60,000 deaths per yearthroughout the world.

There are new guidelines for the definition of hyperten-sive disorders in pregnancy (17). These were developed by alarge taskforce of experts using an evidence-based approachand best available data. The guidelines are endorsed

1212

by numerous professional organizations and serve as anexcellent guideline for reporting guidelines for hypertensivedisorders in pregnancy. The diagnostic criteria for pre-eclampsia advised by the task force are shown in Table 3 (17).

Fetal Growth Restriction

Fetal growth restriction is associated with an increased riskfor adverse outcome such as stillbirth, fetal/neonatal braininjury, and preterm birth (18). In addition, because of in uterofetal programming it increases the risk for subsequent cogni-tive delay and adult disorders such as diabetes and cardiovas-cular disease (19). As stated previously, the risk of FGRmay beincreased with IVF or certain types of IVF (4, 7). Thus, detailsregarding FGR are of great interest in fertility trials.Additionally, among the many causes of FGR is fetalexposure to numerous medications (18).

Definitions of FGR vary, and the optimal definition iscontroversial. Even the term FGR is not universally accepted,and it is often used interchangeably with ‘‘intrauterine growthrestriction’’ and ‘‘small for gestational age’’ fetus or infant.Obstetricians tend to focus on identifying FGR in utero, tooptimize perinatal outcome. Conversely, it makes sense tofocus on infant birth weight as the outcome for infertilitytrials.

The most commonly used threshold to define FGR orsmall for gestational age is birth weight less than the 10thpercentile for gestational age. However, this inherently labels10% of the ‘‘normal’’ population as having the condition.Such individuals are termed constitutionally small for gesta-tional age. As expected, using a threshold of birth weight<5th% or<3rd% is more specifically associated with adverseoutcomes. Numerous population-based growth charts areavailable to identify FGR. Ideally these should be derivedfrom populations that are similar to the one studied. Manyfactors influence birth weight, including race/ethnicity andaltitude. Thus, growth charts from Denver are different fromthose in a city at sea level. Taken to an extreme, some author-ities advocate using customized birth weight percentiles thatreflect numerous characteristics that influence birth weight innormal individuals. This allows better distinction betweenconstitutionally small infants and pathologically small in-fants, and such growth charts report stronger correlations be-tween FGR and adverse outcomes relative to standard growthcharts (20). For most infertility trials, simply reporting birthweight and an accurate gestational age can allow for subse-quent calculation of FGR.

Placental Abnormalities

The most common placental abnormalities are placenta pre-via and placental abruption. These conditions also contributeto maternal and perinatal morbidity and mortality. Previa af-fects approximately 1 in 200 pregnancies and is associatedwith a 2.3% perinatal mortality rate (21). The condition isdefined as the placenta covering the cervical os (eitherpartially or completely). Diagnosis is usually made by sono-gram, although it can be made clinically or, in cases of hyster-ectomy, histologically. The condition has been linked to IVF,

VOL. 101 NO. 5 / MAY 2014

TABLE 2

Guidance regarding timing of delivery when conditions complicate pregnancy at or after 34 weeks of gestation.

Condition Gestational agea at deliveryGrade of

recommendationb

Placental and uterine issuesPlacenta previac 36–37 wk BSuspected placenta accreta, increta,

or percreta with placenta previac34–35 wk B

Prior classic cesarean (upper segmentuterine incision)c

36–37 wk B

Prior myomectomy necessitatingcesarean deliveryc

37–38 wk (may require earlier delivery, similar toprior classic cesarean, in situations with moreextensive or complicated myomectomy)

B

Fetal issuesFetal growth restriction—singleton 38–39 wk

Otherwise uncomplicated, no concurrent findings34–37 wk

Concurrent conditions (oligohydramnios, abnormalDoppler studies, maternal risk factors, comorbidity)Expeditious delivery regardless ofgestational age:Persistent abnormal fetal surveillance suggestingimminent fetal jeopardy

B

B

Fetal growth restriction—twin gestation 36–37 wkDichorionic-diamniotic twins with isolated fetalgrowth restriction

32–34 wkMonochorionic-diamniotic twins with isolated fetalgrowth restrictionConcurrent conditions (oligohydramnios, abnormalDoppler studies, maternal risk factors, comorbidity)Expeditious delivery regardless of gestational age:Persistent abnormal fetal surveillance suggestingimminent fetal jeopardy

B

B

B

Fetal congenital malformationsc 34–39 wkSuspect worsening of fetal organ damagePotential for fetal intracranial hemorrhage (e.g., veinof Galen aneurysm, neonatal alloimmunethrombocytopenia)When delivery before labor is preferred (eg, EXITprocedure)Previous fetal interventionConcurrent maternal disease (e.g., pre-eclampsia,chronic hypertension)Potential for adverse maternal effect from fetalconditionExpeditious delivery regardless of gestational age:When intervention is expected to be beneficialFetal complications develop (abnormal fetalsurveillance, new-onset hydrops fetalis, progressive ornew-onset organ injury)Maternal complications develop (mirror syndrome)

B

B

Multiple gestations—dichorionic-diamnioticc 38 wk BMultiple gestations—monochorionic-diamnioticc 34–37 wk BMultiple gestations—dichorionic-diamniotic or

monochorionic-diamniotic with single fetal deathcIf occurs at or after 34 wk, consider delivery

(recommendation limited to pregnancies at or after34 wk; if occurs before 34 wk,individualize according to concurrent maternal orfetal conditions)

B

Multiple gestations—monochorionic-monoamnioticc 32–34 wk BMultiple gestations—monochorionic-monoamniotic

with single fetal deathcConsider delivery; individualized according to

gestational age and concurrent complicationsB

Oligohydramnios—isolated and persistentc 36–37 wk BMaternal issues

Chronic hypertension—no medicationsc 38–89 wk BChronic hypertension—controlled on medicationc 37–39 wk BChronic hypertension—difficult to control (requiring

frequent medication adjustments)c36–37 wk B

Gestational hypertensiond 37–38 wk BSilver. Pregnancy outcomes in infertility trials. Fertil Steril 2014.

VOL. 101 NO. 5 / MAY 2014 1213

Fertility and Sterility®

TABLE 2

Continued.

Condition Gestational agea at deliveryGrade of

recommendationb

Preeclampsia—severec At diagnosis (recommendation limited to pregnanciesat or after 34 wk)

C

Preeclampsia—mildc 37 wk BDiabetes—pregestational, well controlledc LPTB or ETB not recommended BDiabetes—pregestational, with vascular diseasec 37–39 wk BDiabetes—pregestational, poorly controlled 34–39 wk (individualized to situation) BDiabetes—gestational, well controlled on dietc LPTB or ETB not recommended BDiabetes—gestational, well controlled on

medicationcLPTB or ETB not recommended B

Diabetes—gestational, poorly controlled onmedicationc

34–39 wk (individualized to situation) B

Obstetric issuesPrior stillbirth—unexplainedc LPTB or ETB not recommended

Consider amniocentesis for fetal pulmonary maturityif delivery planned at less than 39 wk

BC

Spontaneous preterm birth: preterm prematurerupture of membranesc

34 wk (recommendation limited to pregnancies at or after34 wk)

B

Spontaneous preterm birth: active preterm laborc Delivery if progressive labor or additional maternal or fetalindication

B

Note: LPTB ¼ late-preterm birth at 34 0/7 weeks through 36 6/7 weeks; ETB ¼ early-term birth at 37 0/7 weeks through 38 6/7 weeks.a Gestational age is in completed weeks; thus, 34 weeks includes 34 0/7 weeks through 34 6/7 weeks.b Grade of recommendations are based on the following: recommendations or conclusions or both are based on good and consistent scientific evidence (A); limited or inconsistent scientific ev-idence (B); primarily consensus and expert opinion (C). The recommendations regarding expeditious delivery for imminent fetal jeopardy were not given a grade. The recommendation regardingsevere pre-eclampsia is based largely on expert opinion; however, higher-level evidence is not likely to be forthcoming because this condition is believed to carry significant maternal risk with limitedpotential fetal benefit from expectant management after 34 weeks.c Uncomplicated, thus no fetal growth restriction, superimposed pre-eclampsia, etc. If these are present, then the complicating conditions take precedence and earlier delivery may be indicated.d Maintenance antihypertensive therapy should not be used to treat gestational hypertension.

Silver. Pregnancy outcomes in infertility trials. Fertil Steril 2014.

VIEWS AND REVIEWS

with one population-based study noting a six-fold increasedrisk (22). It is possible that placing the embryo in the uterus ismore likely to lead to implantation lower in the uterine cavitythan in spontaneous conceptions.

Abruption is slightly more common than previa and alsohas a considerable rate of perinatal mortality (23). The defini-tion of abruption is inconsistent and somewhat subjective. Itis classically defined as a premature separation of the placentabefore delivery. Diagnosis is made by clinical features(bleeding and/or abdominal pain), obstetric sonogram, bloodclot adherent to the placenta, and placental histology.

A less common but even more serious placental compli-cation is placenta accreta. This occurs when the placenta isabnormally adherent to the uterus, usually in the setting ofplacenta previa in women with prior cesarean deliveries.The condition is extremely morbid for the mother, leadingto massive hemorrhage that can be life-threatening. It is note-worthy that a recent Japanese study reported an increased riskof placenta accreta after frozen–thawed embryo transfer (24).

Fetal Abnormalities

Some of the scariest complications for parents are birth de-fects and genetic abnormalities. Some associations betweentreatments for infertility and birth defects and genetic con-ditions are highlighted in a previous section. It is beyondthe scope of most infertility trials to comprehensivelyassess and report all minor fetal abnormalities. However,all major malformations and genetic abnormalities shouldbe reported.

1214

Gestational Diabetes

Gestational diabetes (GDM) is another important obstetriccomplication that is associated with both short-term andlong-term neonatal and maternal morbidity. Adversematernal outcomes associated with GDM include cesarean de-livery, pre-eclampsia, and subsequent development of type 2diabetes (25). Gestational diabetes also is associated withfetal/neonatal complications, such as macrosomia, shoulderdystocia, hyperbilirubinemia, and hypoglycemia (24). Medi-cations used to treat some causes of infertility, such as metfor-min, may influence the risk of GDM.

The definition of GDM is somewhat controversial. In theUnited States most clinicians use a two-step screening pro-cess. Typically this involves a 1-hour glucose challenge test(measuring serum glucose 1 hour after a woman drinks a50-g oral glucose solution). If the result from this is abnormal,it is followed by a 3-hour glucose tolerance test (measuringfasting blood glucose, followed by an oral 100-g glucoseload, followed by blood glucose monitoring at 1, 2, and3 hours after the load). Additionally, various thresholdswith different sensitivities and sensitivities are used to defineabnormal. In contrast, the International Association of theDiabetes and Pregnancy Study Groups advocates using aone-step screening process involving a fasting blood glucose,75-g glucose load, and 1- and 2-hour postload blood glucosemonitoring. This has been adopted in many parts of the worldand has the advantage of being evidence-based, using datafrom the Hyperglycemia and Adverse Pregnancy Outcomestrial (26). However, it may lead to a dramatic increase in the

VOL. 101 NO. 5 / MAY 2014

TABLE 3

Diagnostic criteria for pre-eclampsia.

Blood pressure R140 mm Hg systolic or R90 mm Hg diastolic on two occasions at least 4 h apart after20 wk gestation in a woman with a previously normal blood pressure

R160 mm Hg systolic or R110 mm Hg diastolic, hypertension can be confirmed within ashort interval (minutes) to facilitate timely antihypertensive therapy

andProteinuria R300 mg per 24-h urine collection (or this amount extrapolated from a timed collection)

orProtein/creatinine ratio R0.3a

Dipstick reading of 1þ (used only if other quantitative methods not available)Or in the absence of proteinuria, new-onset

hypertension with the new onset of anyof the following:

Thrombocytopenia Platelet count <100,000/mLRenal insufficiency Serum creatinine concentrations >1.1 mg/dL or a doubling of the serum creatinine

concentration in the absence of other renal diseaseImpaired liver function Elevated blood concentrations of liver transaminases to twice normal concentrationPulmonary edemaCerebral or visual symptomsNote: Reproduced with permission from Hypertension in pregnancy. Report of the American College of Obstetricians and Gynecologists' Task Force on Hypertension in Pregnancy. Obstet Gynecol2013;122:1122–31.a Each measured as mg/dL.

Silver. Pregnancy outcomes in infertility trials. Fertil Steril 2014.

Fertility and Sterility®

proportion of women diagnosed with GDM, and the cost-benefit ratio for identifying these additional cases has notbeen established (25). For now it is adequate for infertility tri-als to simply report whether the patient had GDM and whichdefinition was used. Ideally, specific parameters (e.g., resultsof glucose tolerance testing) would be collected and availablefor further in-depth analysis.

Other Complications and Conditions

There are numerous other obstetric outcomes that warrantconsideration when reporting results of infertility treatmenttrials. Examples include cesarean delivery, vaginal bleeding,hyperemesis gravidarum, and ovarian hyperstimulation syn-drome. Some are straightforward, such as the rate of cesareandelivery. Many of the other, less-common conditions can becaptured by systematically assessing adverse events. It maynot be necessary to report them in detail unless an effect ofthe treatment is clearly apparent.

PRACTICAL CONSIDERATIONSThere are many practical measures that can be taken to facil-itate the optimal reporting of obstetric outcomes in infertilitytrials. A major logistical problem is that most fertility clinicsare not hospital based. Consequently, treated patients ulti-mately deliver at a variety of hospitals that the infertility phy-sicians may or may not work in. In turn, access to medicalrecords and subsequent pregnancy outcomes may be difficultto come by. Several approaches can be helpful. One option isto obtain institutional review board (IRB) approval at all of thehospitals that potential participants may deliver in. Manyhospitals now honor national, regional, or local universityIRBs. Our group routinely does this for populations-based ob-stetric studies. This allows research staff to directly obtain re-cords on site or through direct access to electronic medicalrecords. This may prove difficult in some locales, but thereis a trend toward more centralization of IRBs, which should

VOL. 101 NO. 5 / MAY 2014

ease the burden on investigators. Second, hospitals and phy-sicians' offices will routinely send medical records to researchstaff if the patient giveswritten informed consent. This also hasbeen very successful and cost-efficient for obstetric studies.Third, information regarding some outcomes can be reliablyobtained from the patients themselves. This can be accom-plished with in-person visits, telephone calls, protected healthinformation protected e-mails, and questionnaires electroni-cally filled out by participants that are directly downloaded.

The most cost- and time-efficient window to obtainrecords and/or contact patients is 6 weeks after delivery. Bythat time almost all major adverse outcomes (e.g., birth seriousbirth defects) will be apparent, and inpatient and outpatientmedical records should be complete. Noncoercive amounts ofremuneration may be useful in increasing compliance withfollow-up visits or phone calls, although most patients infertility trials willingly undergo postpartum follow-up. Smallmementos such as ultrasound or infant pictures or charmsalso are helpful with encouraging compliance and follow-up.

It also is important to use stringent, generally accepteddefinitions for outcomes that are delineated before beginningthe trial. Often it is necessary to obtain objective data so as toallow for verification of obstetric complications. Examplesinclude blood pressure and platelet counts for the determina-tion of pre-eclampsia or glucose tolerance tests results incases of gestational diabetes. Such details should be obtainedduring chart abstraction so that the subset of cases withimportant obstetric complications can be vetted.

There are several aspects of the medical record that arepertinent to obstetric complications that may be unfamiliarto individuals not actively practicing obstetrics. These includeprenatal records, labor and delivery records, delivery notes,antenatal sonograms, antenatal testing, and antenatal labo-ratory studies, such as genetic screening or screening forgestational diabetes. Unfortunately, many of these data arecontained in outpatient or inpatient records, but not both.Thus it is almost always necessary to obtain both outpatient

1215

VIEWS AND REVIEWS

and inpatient medical records. Numerous examples of datacollection fields for obstetric outcomes are available. Mostare listed on clinical trials Web sites, and many investigatorsare happy to share data collection instruments for obstetrictrials.

Another good resource for appropriate information tocollect regarding obstetric outcomes is summarized in anarticle detailing modified Consolidated Standards of Report-ing Trials guidelines for obstetric research (26). This article de-lineates details that are pertinent to obstetric complicationsthat may not be applicable to general medical trials. Thereis a focus on pertinent data fields that should be reportedfor obstetric trials, such as parity, plurality (singletoncompared with multiple gestations), gestational weight gain,obstetric history, gestational age at delivery, and neonatalgender (27). Many if not most of these also should be reportedfor infertility treatment trials.

REFERENCES1. Wen J, Jiang J, Ding C, Dai J, Liu Y, Xia Y, et al. Birth defects in children

conceived by in vitro fertilization and intracytoplasmic sperm injection: ameta-analysis. Fertil Steril 2012;97:1331–7.

2. DaviesMJ, Moore VM,Willson KJ, Van Essen P, Priest K, Scott H, et al. Repro-ductive technologies and the risk of birth defects. N Engl J Med 2012;366:1803–13.

3. Vermeiden JP, Bernardus RE. Are imprinting disorders more prevalent afterhuman in vitro fertilization or intracytoplasmic sperm injection? Fertil Steril2013;99:642–51.

4. Pandey S, Shetty A, Hamilton M, Bhattacharya S, Maheshwari A. Obstetricand perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: asystematic review and meta-analysis. Hum Reprod Update 2012;18:485–503.

5. Messerlian C, Maclagan L, Basso O. Infertility and the risk of adverse preg-nancy outcomes: a systematic review and meta-analysis. Hum Reprod2013;28:125–37.

6. Maheshwari A, Pandey S, Shetty A, Hamilton M, Battacharya S. Obstetricand perinatal outcomes in singleton pregnancies resulting from the transferof frozen thawed versus fresh embryos generated through in vitro fertiliza-tion treatment: a systematic review and meta-analysis. Fertil Steril 2012;98:368–77.

7. Maheshwari A, Kalampokas T, Davidson J, Bhattacharya S. Obstetric andperinatal outcomes in singleton pregnancies resulting from the transfer ofblastocyst-stage versus cleavage-stage embryos generated through in vitrofertilization treatment: a systematic review and meta-analysis. Fertil Steril2013;100:1615–21.

8. Silver RM, Branch DW, Goldenberg R, Iams JD, Klebanoff MA. Nomencla-ture for pregnancy outcomes. Time for a change. Obstet Gynecol 2011;118:1402–8.

9. Oshiro BT, Silver RM, Scott JR, Yu H, Branch DW. Antiphospholipid anti-bodies and fetal death. Obstet Gynecol 1996;87:489–93.

1216

10. Creinin MD, Simhan HN. Can we communicate gravidity and parity better?Obstet Gynecol 2009;113:709–11.

11. Edlow AG, Srinivas SK, Elovitz MA. Second trimester loss and subsequentpregnancy outcomes: what is the real risk? Am J Obstet Gynecol 2007;197:581.e1–6.

12. Hamilton B, Martin J, Ventura S. Births: preliminary data for 2007. Natl VitalStat Rep 2009;57:1.

13. Blencowe H, Cousens S, Oestergaard MZ, Chou D, Moller AB, Narwal R,et al. National, regional, and worldwide estimates of preterm birth rates inthe year 2010 with time trends since 1990 for selected countries: a system-atic analysis and implications. Lancet 2012;379:2162–72.

14. Behrman RE, Butler AS, editors. Preterm birth: causes, consequences, andprevention. Washington, DC: National Academies Press; 2007.

15. Spong CY, Mercer BM, D’alton M, Kilpatrick S, Blackwell S, Saade G. Timingof indicated late-preterm and early term birth. Obstet Gynecol 2011;118:323–33.

16. Kulkami AD, Jamieson DJ, Jones HW Jr, Kissin DM, Gallo MF, Macaluso M,et al. Fertility treatments and multiple births in the United States. N Engl JMed 2013;369:2218–25.

17. Hypertension in pregnancy. Report of the American College of Obstetriciansand Gynecologists’ Task Force on Hypertension in Pregnancy. Obstet Gyne-col 2013;122:1122–31.

18. American College of Obstetricians and Gynecologists Committee on Prac-tice Bulletins. Fetal growth restriction. Number 134, May 2013. Obstet Gy-necol 2013;121:1122–33.

19. Pallotto EK, Kilbride HW. Perinatal outcome and later implications of intra-uterine growth restriction. Clin Obstet Gynecol 2006;49:257–69.

20. Figueras F, Gardosi J. Intrauterine growth restriction: new concepts in ante-natal surveillance, diagnosis, and management. Am J Obstet Gynecol 2011;204:288–300.

21. Rao KP, Belogolovkin V, Yankowitz J, Spinnato JA. Abnormal placentation:evidence-based diagnosis and management of placenta previa, placenta ac-creta, and vasa previa. Obstet Gynecol Survey 2012;67:503–19.

22. Romundstad LB, Romundstad PR, Sunde A, von D€uring V, Skjaerven R,Vatten LJ. Increased risk of placenta previa in pregnancies following IVF/ICSI; a comparison of ART and non-ART pregnancies in the same mother.Hum Reprod 2006;21:2353–8.

23. Tikkanen M. Placental abruption: epidemiology, risk factors and conse-quences. Acta Obstet Gynecol Scand 2011;90:140–9.

24. Ishihara O, Araki R, Kuwahara A, Itakura A, Saito H, Adamson GD. Impact offrozen-thawed single-blastocyst transfer on maternal and neonataloutcome: an analysis of 277,042 single-embryo transfer cycles from 2008to 2010 in Japan. Fertil Steril 2014;101:128–33.

25. National Institutes of Health Consensus Development Conference State-ment. Diagnosing gestational diabetes mellitus, March 4–6, 2013. ObstetGynecol 2013;122:358–69.

26. HAPO Study Cooperative Research Group,Metzger BE, Lowe LP, Dyer AR,Trimble ER, Chaovarindr U, et al. Hyperglycemia and adverse pregnancy out-comes. N Engl J Med 2008;358:1991–2002.

27. Chauhan SP, Blackwell SC, Saade GR, for the Society for Maternal FetalMedicine Health Policy Committee. A suggested approach for implementingCONSORT guidelines specific to obstetric research. Obstet Gynecol 2013;122:952–6.

VOL. 101 NO. 5 / MAY 2014