Multifetal reduction of triplets to twins: a prospective comparison of pregnancy outcome

Pierre Boulot1,3, Jacques Vignal1, Christine Vergnes2, Hervé Dechaud1, Jean-Michel Faure1 and Bernard Hedon1

1 Foetal Medicine Unit, Department of Obstetrics and Gynecology, Hopital Arnaud de Villeneuve, Avenue du Doyen Gaston Giraud, 34 000 Montpellier Cedex and 2 Department of Medical Information, Hopital Lapeyronie, Avenue du Doyen Gaston Giraud, 34 000 Montpellier Cedex, France


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The aim of this study was to compare the outcome of triplets managed expectantly or by multifetal reduction to twins to assess the potential benefit of fetal reduction. The study design was prospective, comparative and monocentric and the study was conducted in a teaching hospital. Out of 148 women with triplets mostly obtained after infertility treatment, 83 were expectantly managed while 65 chose reduction to obtain twins. Main outcome measures were fetal loss before 24 weeks, premature deliveries before 28, 32 and 34 weeks, rate of low birthweight infants and neonatal and perinatal mortality rates. The fetal loss rate before 24 weeks did not differ between the ongoing group and the reduced group (6 versus 5.4%). Reducing triplets was associated with a signicantly lower incidence of the following: prematurity before 28, 32 and 34 weeks (P < 0.001), low birthweight infants whose weights were under the third centile (P < 0.002) and infants whose weights were less than 1000, 1500 and 2000 g (P < 0.001). Neonatal (although apparently lower in the reduced group) and perinatal mortality did not significantly differ. Our results indicate that reduction of triplets to twins is effective to improve preterm birth and fetal growth.

Key words: multifetal pregnancy reduction/perinatal mortality/prematurity/triplet gestation/twins


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Over the past 10 years, the number of women carrying three fetuses has increased dramatically as a result of infertility treatments including the use of ovulation-inducing agents and assisted reproductive technology. First trimester multifetal pregnancy reduction (MFPR) can be performed in order to decrease the risk of complications associated with multiple gestations. These complications include extreme prematurity before 32 weeks, low birthweight infants, fetal death in utero, high levels of perinatal mortality, and high levels of perinatal morbidity, which includes cerebral palsy and necrotizing enterocolitis. The termination of one of the normal embryos in a triplet pregnancy is an alternative to either the abortion of all the fetuses or to the acceptance of the risk of extremely premature delivery. Although the procedure poses no ethical problems when dealing with high order multiple fetal gestations (quadruplets, quintuplets or more), the same is not true concerning triplets for whom the indications have not been clearly established. With today's perinatal and neonatal care, the outcome of triplets has improved considerably, leading to low perinatal or neonatal mortality rates (Newman et al., 1989Go; Boulot et al., 1993aGo; Lipitz et al., 1994Go; Albrecht and Tomich, 1996Go). Consequently, the issue of MFPR on triplets to obtain twins is controversial (Boulot et al., 1993aGo; Macones et al., 1993Go; Berkowitz et al., 1996Go; Evans et al., 1998Go). The aim of this study was to compare the outcome of triplet pregnancies managed expectantly or by multifetal reduction to twins in order to draw conclusions that may help both the patients and the physicians faced with this dilemma.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A total of 148 women pregnant with triplets were referred to our institution over a 12 year period. This series is an expanded series of the initial series of 80 triplets that we published as a preliminary report a few years ago (Boulot et al., 1993aGo). Eighty-three women wished to continue their pregnancies whereas 65 couples requested MFPR to obtain twins. Triplet gestations reduced to singletons were excluded from the study because it may have created a bias in calculating the rate of prematurity and the fetal growth. The mean age of patients was 30.9 ± 4.6 years for the reduced group (range: 18–42 years) and 31.1 ± 4.3 years for the non-reduced group (range: 22–42 years). There was no significant difference concerning maternal age. The origins of triplet pregnancies for the two groups are reported in Table IGo.


View this table:
[in this window]
[in a new window]
 
Table I. Origin of the 148 triplet pregnancies
 
The expectantly managed group was managed as previously reported (Boulot et al., 1993aGo): all triplets showed ultrasonographic features of trichorionic placentation. Leave from work was systematically imposed as well as a substantial reduction in maternal physical activity, once the diagnosis of triplet gestation was established. Management at home began at 20 weeks by a midwife. In cases of cervical change or increased uterine activity, women were immediately referred to our centre. The women were examined by obstetricians monthly until the fifth month and then every 2 weeks. Serial ultrasound scans were done every month by the same practitioner. Originally, ß-methasone was given to accelerate fetal lung maturation only in cases of imminent preterm labour; however, as of 1993, it was systematically given from weeks 27–33 of gestation. Delivery date was decided either by when the patient went into labour or by when tocolysis was unable to prevent preterm labour. Delivery was performed by Caesarean section under spinal anaesthesia or vaginally, at the discretion of the senior obstetrician who was present at the time of the delivery. For each delivery, three experienced neonate paediatricians were present in the delivery room.

Management of the reduced group was as follows: as of 1993, all reductions were performed as previously described (Boulot et al., 1993bGo). Before 1993, two techniques were used, both performed by the same practitioner. The technique used in the initial 16 cases of our experience with MFPR was the transcervical mini-suction at 8–11 weeks using an echo-guided Karmann cannula no. 6 to 10 inserted through the cervix under local or general anaesthesia. Transabdominal echo-guided embryo puncture was then used in the 49 following cases. Under local anaesthesia, a 22.5 gauge needle was transabdominally inserted through the uterus wall, then into the amniotic cavity, and finally pushed into the fetal thorax at a level above the diaphragm. Then, potassium chloride solution (amount ranging from 0.5 to 5 ml) was injected, leading to the cessation of the heartbeat. Antibioprophylaxy was used in all cases. The choice of technique was based on the period at which the patients entered the study. The transcervical approach was used until 1989 and then was replaced by the other technique because of a lower miscarriage rate. The procedures were performed between 8.0 and 13.5 weeks and the mean gestational age at the time of the reduction was 9.7 ± 2.5 weeks. All pregnancies obtained were twins with typical ultrasonographic features of a dichorionic placentation. Scans were then performed monthly until the end of pregnancy. After the procedure, nearly half of the patients in the reduced group were referred to their primary care providers for the remainder of the pregnancy.

Statistics
Data are presented as means ± SD. Length of gestation was established according to the date of IVF or ovulation induction or, for cases of spontaneous gestation, on the first day of the last menstrual period, which was then verified by ultrasound examination. Durations of pregnancies are presented as number of weeks gestation. Delivery at term was after 37 completed weeks gestation. Fetal weights were reported on growth curves for singletons. The perinatal mortality rate (PMR) included deaths occurring from the beginning of the 22nd week of gestation to the seventh day after birth. The neonatal mortality rate included deaths occurring from birth to the 28th day after birth. Therefore deaths occurring during the first 7 days after birth were included in both perinatal and neonatal categories. Data were analysed using the {chi}2-test, Fisher's exact test and the Kruskal–Wallis test. For uniformity, infant follow-up refers to follow-up until 6 months of age for both populations.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Results are reported in Tables II, III and IVGoGoGo. Fetal loss before 24 weeks did not significantly differ between the two groups (Table IIGo). Fifteen embryos were lost due to five complete miscarriages that occurred in the triplet cohort (6.02%) (Table IIIGo). In the reduced group, seven embryos were lost due to two complete miscarriages and three partial miscarriages resulting in singleton pregnancies (5.38%). Rates of prematurity were different in the two groups as deliveries ocurred earlier in the triplet group compared with the reduced group (33.5 ± 2.7 versus 36 ± 2.9 weeks, P < 0.001).


View this table:
[in this window]
[in a new window]
 
Table II. Obstetric outcome according to the method used for multifetal pregnancy reduction (MFPR)
 

View this table:
[in this window]
[in a new window]
 
Table III. Obstetric outcome [fetal loss before 24 weeks, prematurity and intrauterine death (IUD)]
 

View this table:
[in this window]
[in a new window]
 
Table IV. Birth and neonatal data
 
In the triplet cohort, Caesarean section was performed in 71 cases (91%) and birth was vaginally performed in seven cases (9%). In this group, 234 fetuses were delivered, 223 alive and 11 dead (Table IVGo). Among these 11 deaths, eight were spontaneous deaths in utero and three were related to second-trimester selective terminations performed on malformed fetuses (neural tube defects or hydrocephaly). The mean weight of newborns (excluding deaths in utero) was 1791 g ± 489 g (extremes from 640 to 3180 g). There were four neonatal deaths (4:223, 1.79%) that occurred within the first 7 days after birth and were therefore included in the perinatal mortality rate. The perinatal mortality rate was 6.41% (15/234). Among the reduced pregnancies, 28 patients (45.2%) underwent Caesarean section and 35 (54.8%) delivered vaginally. In all, 123 fetuses were delivered including 116 alive and seven dead (all were deaths in utero). Mean birthweight was 2362 ± 554 g (extremes from 775 to 3550 g). The comparison of birthweights based on gestational age and the distribution of birthweights according to the tenth or third percentile are reported in Table IVGo. One death occurred in the immediate neonatal period due to a cerebral haemorrhage related to extreme prematurity. The neonatal mortality rate was 0.86% (1:116) and the perinatal mortality rate was 6.5% (8:123).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The results from this expanded series confirm the results of our initial study (Boulot et al., 1993aGo). MFPR reduces extreme prematurity before 32 weeks and improves fetal growth without excessive fetal loss. Our data indicate a rate of fetal loss before 24 weeks of 6% among triplets, which was similar to the 5.4% fetal loss rate observed in the reduced group. This last rate is in agreement with those reported in large series in which MFPR was performed by a limited number of experienced operators (Evans et al., 1993Go, 1994Go; Dechaud et al., 1998Go). Furthermore, because the fetal loss rate is strongly associated with the starting and the finishing number of embryos, reducing triplets to twins is expected to lead to a lower rate of miscarriage than that observed when reducing triplets to singletons (Berkowitz et al., 1996Go; Evans et al., 1998Go). The ultrasonic prospective observation of expectantly managed triplet pregnancies reveals that fetal death is common before 24 weeks. Lipitz et al. (1994) reported a spontaneous fetal loss rate among expectantly managed triplets of nearly 21% before 24 weeks. Kol et al. (1993) reports that spontaneous fetal death ocurred in 5% of multiple pregnancies. These data suggest that the spontaneous loss rate in women carrying three fetuses may be similar to, or even higher than, the fetal loss rate observed following MFPR. Late MFPR at the mean term of 20 weeks could lead to a more favourable perinatal outcome. For Hartoov (Hartoov et al., 1998Go), delaying the procedure as late as 20 weeks may facilitate the detection of structural and chromosomal anomalies of the fetuses. However, studies are requested on larger series that should include only triplets with normal embryos in an effort to achieve homogeneity. In our study, only one out of 13 couples with spontaneously obtained triplets decided to reduce whereas half of the ovulation induction and IVF groups decided to reduce. Couples requesting infertility treatments are well informed of the risks of multiple pregnancies before starting treatments. As information of MFPR is given at this time, it probably makes couples more inclined towards reduction in case of high order multiple pregnancy. Conversely, couples with spontaneous triplets do not always have an earlier diagnosis of the multiple gestation, and are not aware of the possibility of reducing the pregnancy.

The main challenge in triplet pregnancies is the high rate of premature delivery, ranging from 87.5 to 91.6%, with a mean age at delivery of 33.9 weeks (Lipitz et al., 1989Go; Newman et al., 1989Go; Seoud et al., 1991Go; Albrecht and Tomich, 1996Go; Berkowitz et al., 1996Go; Evans et al., 1998Go). The proportion of births occurring before 32 weeks reaches ~25% according to the series in the literature, while the rate of premature deliveries occurring between 24 and 28 weeks varies from 3 to 7% (Boulot et al., 1993; Berkowitz et al., 1996Go; Evans et al., 1998Go; Torok et al., 1998Go). There was a significant decrease (P = 0.001) in prematurity in our reduced groups compared with the triplet group since pregnancies of the reduced group continued for 2.5 weeks longer on average. The mean gestational age at delivery in the expectantly managed group was 33.5 weeks, whereas for the twins resulting from reduced triplets, it was 36 weeks. Extreme premature births occurred more frequently in the non-reduced group than in the reduced group. The rate of preterm births was 2-fold higher between 24 and 28 weeks and 2.5-fold higher in the interval 29–32 weeks. In another study, the rate of prematurity among twins from reduced triplets did not significantly differ from that of dichorionic twins naturally obtained or by means of assisted reproduction procedures (Macones et al., 1993Go; Alexander et al., 1995Go; Groutz et al., 1996Go). Similar results were observed in the two other series that compare the outcome of triplets reduced to twins versus non-reduced triplets (Macones et al., 1993Go; Lipitz et al., 1994Go) and in our previous study (Boulot et al., 1993). For Lipitz et al. (1994), the rate of preterm births among triplets was respectively 23.8% between 25 and 31 weeks and 67.9% between 32 and 36 weeks versus 9.7% and 16.1% among reduced triplets (Lipitz et al., 1994Go). For Macones et al. (1993), a higher percentage of pregnancies was delivered before 32 weeks in the non-reduced triplets (43%) compared with the reduced group (7%) (Macones et al., 1993Go). In a collaborative study (Evans et al., 1998Go), the prematurity rate observed among triplets managed with MFPR, mostly to obtain twins, was 3.3% in the interval of 25–28 weeks and reached 7.5% between 29 and 32 weeks. These rates are consistent with our data. Similar data were reported by Berkowitz et al. (1996) on their monocentric series of 179 triplets reduced mostly to twins with 2% of reduced triplets being delivered between 24 and 28 weeks, 6% between 28 and 32 weeks and 35% between 24 and 28 weeks.

As for fetal growth, there was significant weight gain in the reduced group since the birthweights of the infants from that group were 571 g higher than in the triplet cohort. Two other similar studies found differences of nearly 450 to 700 g (Macones et al., 1993Go; Lipitz et al., 1994Go). In our series, the percentage of fetal weights <1500 g was significantly lower among the reduced groups (6.9 versus 27.4%). We also found that the proportion of infants whose weights were under the third centile was significantly lower (P = 0.002) in the reduced group (8.7%) when compared to the non-reduced group (18.4%). In the study of Lipitz et al. (1994) the incidence of infants weighing less than 1000 g and those whose weights were between 1000 and 1500 g were respectively 1.7 versus 6.7% and 5.1 versus 19.7% in comparison with the non-reduced group. Macones et al. (1993) reported an 83% rate of expectantly managed triplets weighing less than 2500 g compared with 73% in the reduced group. This weight gain may indicate that MFPR has no detrimental effects on the growth of the remaining twins, in agreement with the findings of Torok et al. (1998), who reported that MFPR is not associated with an increased risk of intrauterine growth restriction (IUGR) unless the starting number of embryos is five or more. In this latter study, no significant difference was found in the frequency of birthweight discordance or the incidence of IUGR when 233 twins from reduced triplets were compared with a control group of 136 dichorionic twins.

Our study fails to demonstrate a significant decrease in perinatal or neonatal mortality rates among the reduced triplets despite a larger cohort of patients although the perinatal mortality rate of the non-reduced triplets was already very low (6.41%). The study of Lipitz et al. (1994) reported a decrease in the incidence of perinatal mortality—although it was not statistically significant—in the patients undergoing MFPR (48 versus 109 per 1000 births). Macones et al. (1993) reported a perinatal mortality rate of 210 per 1000 births among expectantly managed triplets whereas it was only 30 per 1000 births in the reduced group (P < 0.001). In their series, the perinatal morbidity of twins was significantly lower when compared with the non-reduced group because the proportion of infants admitted to a neonatal intensive care unit was higher for the triplets than for the twins, as was the need of ventilatory support (respectively 84.6 versus 36.2% and 51.3 versus 14%). Similar data have been reported by others (Lipitz et al., 1994Go) and our group (Boulot et al., 1993aGo). Our present data suggest that a reduction from triplet to twin decreases the risk of extreme prematurity, and consequently that it may reduce the risk of cerebral palsy linked to intraventricular haemorrhage. Interestingly, in a retrospective series (Khadel et al., 1998Go) comparing 24 ongoing triplets and 17 triplets reduced to twins conducted over a short (3 years) and recent period, pregnancies received standardized care (births in a level III perinatal centre, fetal lung maturation). Although prematurity was lower in the reduced pregnancies group, severe prematurity (<32 weeks, and between 26 and 31 weeks) was not different. As in our study, no significant difference was observed for perinatal mortality in the two groups. Bollen et al. (1993) reported that 8% of surviving triplets suffer from neurological sequelae and this percentage may be higher because many of these infants were too young to permit full evaluation. Although MFPR decreases the number of babies going home per couple, it may increase the proportion of surviving infants without complications.

It should be noted, however, that MFPR on triplets is performed in several centres: in one study (Evans et al., 1994Go) 42% of the 1084 procedures performed were for reducing triplets. This rate is similar for a later expanded series including 1789 gestations out of which 759 were triplets (Evans et al., 1998Go). In a French national study, 41% of 262 cases of reductions performed were on triplets (Dommergues et al., 1994Go). Similar data are reported from monocentric series, with a mean rate of 40%: 44% (Wapner et al., 1990Go), 43% (Salat-Baroux et al., 1988Go), 26% (Dommergues et al., 1991Go), 57.5% (Tabsh, 1990Go), 24% (Lipitz et al., 1994Go), 50% (Trimor-Trisch et al., 1993), 44% (Berkowitz et al., 1993Go), and 49% in their recently expanded series (Berkowitz et al., 1996Go). MFPR in triplets (resulting in twins and less often in singletons) should be considered in pregnancies with serious maternal disease (e.g. heart disease, diabetes), poor uterine conditions (bicornuate uterus, exposure to diethylstilboestrol, previous Caesarean section, previous preterm delivery) or in cases where an embryo is affected by a severe disease (neural tube defect, cystic hygroma). In the absence of worrisome factors, one might consider that the decision belongs to the well-informed couple. The results shown above indicate that reduction of triplets improves some obstetric parameters. Consequently, couples with triplets should be informed of both the benefits and risks of reduction as well as of the risks of naturally managed triplet gestation. Comprehensive psychological support should be offered to parents of multiple pregnancies, regardless of their willingness or not to reduce the pregnancy. Lack of knowledge about long-term psychological consequences of MFPR on both mothers and babies is an argument to reinforce the prevention of these undesirable pregnancies, at least for the majority of cases. In a 2 year comparative prospective follow-up study between reduced and ongoing triplets (Garel et al., 1997Go), it was shown that for a majority of women, MFPR was experienced as stressful, painful and frightening in the immediate post-procedure period. However, most women had overcome these emotional difficulties 2 years later. The comparison with mothers having delivered reduced triplets indicated that psychological health and the relationship with the children were more satisfactory in the reduction group. Families with triplets had much more acute psychosocial problems than families who underwent MFPR.

Our experience and that of other teams suggest that MFPR on triplets is currently performed not only for obstetric reasons, some of which are difficult for the couples to understand despite full medical disclosure provided by obstetricians or reproductive medicine practitioners (Boulot et al., 1993aGo; Smith-Levitin et al., 1996Go; Berkowitz et al., 1996Go; Evans et al., 1998Go). Additional reasons for requesting MFPR include financial and social concerns, as well as the mother's obligation to stop working and the anticipation of a poor quality of life after the birth of triplets. It should be emphasized that the aim of assisted reproduction procedures is a viable pregnancy leading to the birth of a healthy baby. In most cases of infertility, the desire for a child is best fulfilled by a singleton or twin pregnancy rather than the birth of triplets. As the main series in the literature offer similar results to our own, we believe that MFPR applied on triplets is a reasonable parental option and should be accepted if requested.


    Notes
 
3 To whom correspondence should be addressed at: Foetal Medicine Unit, Department of Obstetrics and Gynecology, Hopital Arnaud de Villeneuve, 34 000, Montpellier, France. E-mail: p-boulot{at}chu-montpellier.fr Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Albrecht, J.L. and Tomich, P.G. (1996) The maternal and neonatal outcome of triplet gestations. Am. J. Obstet. Gynecol., 174, 1551–1556.[ISI][Medline]

Alexander, J.M., Hammond, K.R. and Steinkampf, M.P. (1995) Multifetal reduction of high-order multiple pregnancy: comparison of obstetrical outcome with non reduced twin gestations. Fertil. Steril., 64, 1201–1203.[ISI][Medline]

Berkowitz, R.L., Lynch. L., Lapinski, R. et al. (1993) First-trimester transabdominal multifetal pregnancy reduction: A report of two hundred completed cases. Am. J. Obstet. Gynecol., 169, 17–21.[ISI][Medline]

Berkowitz, R.L., Lynch, L., Stone, J. et al. (1996) The current status of multifetal pregnancy reduction. Am. J. Obstet. Gynecol., 174, 1265–1272.[ISI][Medline]

Bollen, N., Camus, M., Tournaye, H. et al. (1993) Embryo reduction in triplet pregnancies after assisted procreation: a comparative study. Fertil. Steril., 60, 504–509.[ISI][Medline]

Boulot, P., Hedon, B., Pelliccia, G. et al. (1993a) The effects of selective reduction in triplet gestations: a comparative study on 80 cases managed with or without this procedure. Fertil. Steril., 60, 497–503.[ISI][Medline]

Boulot, P., Hedon, B., Pelliccia, G. et al. (1993b) Multifetal pregnancy reduction: a consecutive series of 61 cases. Br. J. Obstet. Gynæcol., 100, 63–68.

Dechaud, H., Picot, M.C., Hedon, B. et al. (1998) First-trimester multifetal pregnancy reduction: evaluation of technical aspects and risks from 2,756 cases of the literature. Fetal. Ther. Diagn., 13, 261–265.

Dommergues, M., Nisand, I., Mandelbrot, L. et al. (1991) Embryo reduction in multifetal pregnancies following infertility therapy: obstetrical risks and perinatal benefits are related to operative strategy. Fertil. Steril., 55, 801–811.

Dommergues, M., Aknin, J., Boulot, P. et al. (1994) Réductions embryonnaires dans les grossesses multiples. Une enquête multicentrique franciaise. J. Gynecol. Obstet. Biol. Reprod., 23, 415–418.[Medline]

Evans, M.I., Dommergues, M., Wapner, R.J. et al. (1993) Efficacy of transabdominal multifetal pregnancy reduction: collaborative experience among the world's largest centers. Obstet. Gynecol., 82, 61–66.[Abstract]

Evans, M.I., Dommergues, M., Wapner, R.J. et al. (1994) Transabdominal versus transcervical and transvaginal multifetal pregnancy reduction: International collaborative experience of more than thousand cases. Am. J. Obstet. Gynecol., 170, 902–909.[ISI][Medline]

Evans, M.I., Kramer, R.L., Yaron, Y. et al. (1998) What are the ethical and technical problems associated with multifetal pregnancy gestation? Clin. Obstet. Gynecol., 41, 47–54.[ISI]

Garel, M., Stark, C., Blondel, B. et al. (1997) Psychological reactions after multifetal pregnancy reduction: a 2-year follow-up study. Hum. Reprod., 12, 617–622.[ISI][Medline]

Groutz, A., Yovel, I., Amit, A. et al. (1996) Pregnancy outcome after multifetal pregnancy reduction to twins compared with spontaneously conceived twins. Hum. Reprod., 11, 1334–1336.[Abstract]

Hartoov, J., Geva, E., Wolman, I. et al. (1998) A 3 year prospectively-designed study of late selective multifetal pregnancy reduction. Hum. Reprod., 13, 1996–1998.[Abstract]

Khadel, P., Olivennes, F., Fernandez, H. et al. (1998) Are there still obstetric and perinatal benefits for selective embryo reduction of triplet pregnancies? Hum. Reprod., 13, 3555–3559.[Abstract]

Kol, S., Levron, J., Lewit, N. et al. (1993) The natural history of multiple pregnancies after assisted reproduction: is spontaneous demise a clinically significant phenomenon? Fertil. Steril., 60, 127–130.[ISI][Medline]

Lipitz, S., Reichman, B., Paret, G. et al. (1989) The improving outcome of triplet pregnancies. Am. J. Obstet. Gynecol., 161, 1279–1284.[ISI][Medline]

Lipitz, S., Reichman, B., Uval, J. et al. (1994) A prospective comparison of the outcome of triplet pregnancies managed expectantly or by multifetal reduction to twins. Am. J. Obstet. Gynecol., 170, 874–879.[ISI][Medline]

Macones, G.A., Schemmer, G., Pritts, E. et al. (1993) Multifetal reduction of triplets to twins improves perinatal outcome. Am. J. Obstet. Gynecol., 169, 982–986.[ISI][Medline]

Newman, R., Hamer, C. and Miller, M. (1989) Outpatient triplet management: a contemporary review. Am. J. Obstet. Gynecol., 161, 547–553.[ISI][Medline]

Salat-Baroux, J., Aknin, J., Antoine, J.M. et al. (1988) The management of multiple pregnancies after induction for superovulation. Hum. Reprod., 3, 399–401.[Abstract]

Seoud, M.A., Kruithoff, C. and Muasher, S.J. (1991) Outcome of triplet and quadruplet pregnancies resulting from in vitro fertilization. Eur. J. Obstet. Gynecol. Reprod. Biol., 41, 79–84.[ISI][Medline]

Smith-Levitin, M., Kowlaik, A., Birnholz, J. et al. (1996) Selective reduction of multifetal pregnancies to twins improves outcome over nonreduced triplet gestations. Am. J. Obstet. Gynecol., 175, 878–882.[ISI][Medline]

Tabsh, K.A. (1990) Transabdominal multifetal pregnancy reduction: report of 40 cases. Obstet. Gynecol., 75, 739–741.[Abstract]

Timor-Tritsch, I.E., Peisner, B.D., Monteagudo, A. et al. (1993) Multifetal pregnancy reduction by transvaginal puncture: Evaluation of the technique used in 134 cases. Am. J. Obstet. Gynecol., 168, 799–804.[ISI][Medline]

Torok, O., Lapinski, R., Salafia, C.M. et al. (1998) Multifetal pregnancy reduction is not associated with an increased risk of intrauterine growth restriction, except for very-high-order multiples. Am. J. Obstet. Gynecol., 179, 221–225.[ISI][Medline]

Wapner, R.J., Davis, G.H., Johnson, A. et al. (1990) Selective reduction of multifetal pregnancies. Lancet, 335, 90–94.[ISI][Medline]

Submitted on April 9, 1999; accepted on March 31, 2000.