The incidence of twin pregnancy has increased worldwide over the past 10 years, largely as a consequence of the assisted reproductive technologies. Issues such as intrapartum monitoring and operative interventions, especially relating to the second twin, provide a unique challenge in labour and delivery. Epidemiological and cohort data suggest that twins have a three-fold higher mortality rate than singletons, and that the second twin might have a better outcome if delivered by lower segment caesarean section. The recently completed Twin Birth Study has found that planned vaginal lower segment caesarean section is not advantageous to the fetus. In the light of this large randomised-controlled trial, vaginal delivery if twin A presents by the vertex is recommended as long as guidelines for the conduct of such delivery are followed.
Introduction
The conduct of a twin delivery remains one of the most challenging events in the daily practice of obstetrics. In Canada, the incidence of twin delivery increased by 15% between 2007 and 2012 . Apart from the Province of Quebec, which has funded in-vitro fertilization (IVF) services and legislated the number of embryos that may be implanted during IVF cycles, the incidence of twin gestation continues to rise . Therefore, the issues addressed in this chapter will be of continuing international importance. The perennial dilemmas that apply to any singleton delivery, such as intra-partum monitoring and operative interventions, are compounded by the presence of the second fetus. The second twin presents unique challenges in labour, and much of the controversy and discussion in this chapter will focus on the delivery of the second twin.
Twin pregnancies are at a higher risk of perinatal and neonatal mortality than singleton pregnancies. Even among twin fetuses that are over 2500 g at birth , a higher risk of death occurs among singletons of the same birth weight. Neonatal seizures, respiratory morbidity, and low Apgar scores at 1 and 5 mins have also been shown to be higher for twins compared with the singleton infant at birth weights over 1500 g and over 3000 g .
The Term Breech Trial found that planned caesarean section reduced the risk of perinatal death or serious neonatal morbidity three-fold (from 5.0% to 1.6%; P < 0.001). This resulted in general agreement that twins in which the first twin is breech should be delivered by caesarean section. The findings of this trial, however, where extrapolated to all twin pregnancies in which the second twin is also often in a non-vertex presentation, and it was this increasingly considered beneficial for all twin pregnancies to be delivered by planned caesarean section.
What method of delivery should be planned for twin pregnancies? How should the delivery be conducted, and what is the optimal way to conduct a vaginal twin birth? Fortunately, many of these answers are now available to us in the form of level 1 randomised-controlled trial evidence, the Twin Birth Study .
Indications for elective caesarean section
Absolute indications for elective caesarean section are minimal, and certainly no good clinical studies on which to base strong recommendations are available. It seems that caesarean section without a trial of labour should be carried out in cases of conjoined twins and mono-amniotic twins. Timing of delivery, however, is still in question, with a recommended range of 32–36 weeks. It is my practice to deliver conjoined and mono-amniotic twins at around 32–33 weeks after a period of inpatient observation .
The other indications for elective caesarean section are not dissimilar to those of a singleton pregnancy, and include placenta praevia, and antenatal evidence of significant fetal compromise likely to worsen during labour.
Indications for elective caesarean section
Absolute indications for elective caesarean section are minimal, and certainly no good clinical studies on which to base strong recommendations are available. It seems that caesarean section without a trial of labour should be carried out in cases of conjoined twins and mono-amniotic twins. Timing of delivery, however, is still in question, with a recommended range of 32–36 weeks. It is my practice to deliver conjoined and mono-amniotic twins at around 32–33 weeks after a period of inpatient observation .
The other indications for elective caesarean section are not dissimilar to those of a singleton pregnancy, and include placenta praevia, and antenatal evidence of significant fetal compromise likely to worsen during labour.
The first twin breech
Caesarean section, when the first twin presents as a breech, has long been regarded as a relative contraindication for vaginal delivery . One of the major concerns with breech and vertex twins, which occurs in about 20% of all twins in labour , is the risk of locked twins . This complication is uncommon, with an estimated frequency of one in 645 twin births, and only 147 cases reported in the world literature between 1958 and 1987. The mortality associated with fetal entanglement, however, is extremely high: between 30– 43% . Some studies have challenged the need to carry out caesarean section for all twins with the first fetus in breech presentation . In a total of 141 twin pairs, all of whom had a non-vertex twin A, no difference was found in neonatal mortality or morbidity.
The randomised-controlled trial on singleton term breech delivery, The Term Breech Trail , reported that the chance of an infant dying as a result of a policy of planned vaginal birth is one out of 300, and the chance of significant handicap is one out of 20 . This study, not without controversy, has provided level 1 evidence that a policy of planned lower segment caesarean section will reduce morbidity and mortality, without a significant increase in immediate maternal complications. This seems to have been borne out in at least one country study, in which the implication of changing to lower segment caesarean section for breech presentation has been studied . It seems difficult, therefore, to recommend vaginal birth in twins that have the additional risks discussed above and in the remainder of this chapter.
Timing of delivery
The timing of delivery, either by induction or elective caesarean section, is an important consideration in twin delivery. The controversy, in general, addresses data that show an increase in still birth, neonatal death rates, or both, around 37–39 weeks of gestation . On the other hand, the incidence of respiratory disorders is dramatically increased in twin pregnancies, especially in weight ranges over 2500 g . Thus, the risk of transient tachypnoea of the newborn, respiratory distress syndrome, and even persistent fetal circulation associated with elective caesarean section, is likely to be higher in twin pregnancies. Most data are cohort and retrospective data ( Table 1 ), and seem to recommend that delivery of dichorionic– diamniotic twins around 37–38 weeks is associated with the lowest perinatal mortality. An under-powered randomised-controlled trial from Australia reported a similar gestation .
| Characteristic a | Planned caesarean ( n = 1393) | Planned vaginal birth ( n = 1393) |
|---|---|---|
| Maternal age (years) ≥30 | 632 (45.4%) | 632 (45.4%) |
| Parity ≥1 | 857 (61.5%) | 856 (61.5%) |
| Previous caesarean | 100 (7.2%) | 97 (7.0%) |
| Gestational age at randomisation (weeks) | ||
| Mean (+/–SD) | 34.9 (1.8) | 34.9 (1.8) |
| <32 weeks | 0 (0.0%) | 1 (0.1%) |
| 320–336 weeks | 475 (34.1%) | 477 (34.2%) |
| 340–366 weeks | 679 (48.8%) | 665 (47.7%) |
| 370–386 weeks | 239 (17.2%) | 250 (18.0%) |
| Mean estimated fetal weight (g) (+/–SD) | ||
| Twin A c | 2238 (424) | 2238 (419) |
| Twin B c | 2223 (413) | 2232 (422) |
| Chorionicity on ultrasound | ||
| Dichorionic diamniotic | 961 (69.0%) | 970 (69.6%) |
| Monochorionic diamniotic | 334 (24.0%) | 326 (23.4%) |
| Unknown | 98 (7.0%) | 97 (7.0%) |
| Not in labor at randomisation c | 1190 (85.5%) | 1159 (83.2%) |
| Membranes ruptured at randomisation | 83 (6.0%) | 76 (5.5%) |
| National perinatal mortality rate of country b | ||
| <15/1000 | 724 (52.0%) | 730 (52.4%) |
| 15–20/1000 | 596 (42.8%) | 591 (2.4%) |
| >20/1000 | 73 (5.2%) | 72 (5.2%) |
a No significant differences between groups in any of the baseline variables.
b Countries: <15/1000: Australia, Belgium, Canada, Chile, Croatia, Estonia, Germany, Greece, Hungary, Israel, The Netherlands, Oman, Poland, Qatar, Romania, Serbia, Spain, UK, USA, Uruguay; 15-20/1000: Argentina, Brazil, Jamaica; >20/1000: Egypt, Jordan.
c The numbers missing in the planned caesarean and planned vaginal birth groups, respectively, were estimated fetal weight for twin A ( n = 2, n = 2); estimated fetal weight for twin B ( n = 2, n = 1); not in labor at randomisation ( n = 1, n = 0).
This argument is even more passionate in monochorionic–diamniotic twins. Some studies have indeed placed the risk of in-utero still birth for twin pregnancies at monochorionic-diamniotic twin pregnancies at 3–4% per week, and suggested delivery at 36 weeks . This recommendation was incorporated into the Royal College of Obstetricians and Gynaecologists National Institute of Health and Clinical Excellence guidelines . More recent studies, however, have suggested that the rate of still birth was overestimated in those studies and, provided that the pregnancy is otherwise uncomplicated, then gestation can be advanced until 37 weeks . As in every case of scheduled delivery, accurate dating of the pregnancy is essential, along with parental advice about potential complications.
Twin A vertex, and over 32 weeks gestation
A systematic review of studies that compared the policies of planned vaginal birth and planned caesarean section for the delivery of twins weighing at least 1500 g or reaching at least 32 weeks gestation, did not show any difference in the rates of adverse perinatal outcome . Several cohort studies, however, have shown that the second twin is at significantly higher risk of adverse outcome compared with the first twin. In addition, three large, well-conducted cohort studies reported that this increased risk to the second twin (OR from 3.1 up top 21) was diminished or prevented completely when both babies where delivered by planned caesarean section .
Our group undertook the Twin Birth Study , a randomised-controlled trial of 2804 women with twin pregnancies, in which the first twin was vertex. The aim was to determine if planned caesarean section was safer for the baby than planned vaginal birth for the twin fetuses between 32 and 38 weeks, when the leading twin was cephalic, in centres that could assure women having a vaginal twin birth, and ensure that an experienced clinician was present at the birth.
The study enrolled 2804 women at 106 centres in 25 countries between December 13, 2003 and April 4, 2011; 1398 women were randomised to planned caesarean and 1406 to planned vaginal birth. A total of 2781 babies were born to 1392 women in the planned caesarean group, and 2782 babies were born to 1392 women in the planned vaginal birth group.
Baseline characteristics were similar in the two groups ( Table 1 ). Most women (82.4%) were randomised between 32 0/7 and 36 6/7 weeks gestation.
Labour and delivery outcomes for all women are presented in Table 2 . Of the 1392 women randomised to planned caesarean, 89.9% had a caesarean section for the delivery of both babies, 0.8% had a combined vaginal birth and caesarean delivery, and 9.3% delivered both twins vaginally. Most of the caesareans in this group were carried out before labour ( n = 748). For those randomised to planned vaginal birth, 56.2% delivered both twins vaginally, and 4.2% had a combined vaginal birth and caesarean section. The remaining proportion (39.6%) had a caesarean section for both twins. Most of the caesarean sections were carried out during labour ( n = 412).
| Characteristics | Planned caesarean ( n = 1393) | Planned vaginal birth ( n = 1393) | P -value |
|---|---|---|---|
| Mode of delivery a | |||
| Caesarean for both b | 1252 (89.9%) | 551 (39.6%) | |
| Vaginal and caesarean | 11 (0.8%) | 59 (4.2%) | |
| Vaginal for both | 129 (9.3%) | 783 (56.2%) | |
| Timing of caesarean a | |||
| Before the onset of labour | 748 (53.8%) | 196 (14.1%) | |
| During labour | 514 (37.0%) | 412 (29.6%) | |
| No caesarean | 129 (9.3%) | 783 (56.3%) | |
| Presentation at delivery for twin A/twin B a | |||
| Cephalic/cephalic | 798 (57.4%) | 845 (60.7%) | |
| Cephalic/non-cephalic | 542 (39.0%) | 507 (36.4%) | |
| Non-cephalic/cephalic or non-cephalic | 51 (3.7%) | 41 (2.9%) | |
| Gestational age at delivery of twin A (weeks) a , c | |||
| Mean (+/SD) | 36.7 (1.5) | 36.8 (1.5) | 0.01 |
| 320/7–336/7 | 88 (6.3%) | 66 (4.7%) | |
| 340/7–366/7 | 582 (41.8%) | 581 (41.7%) | |
| 370/7–386/7 | 694 (49.9%) | 696 (50.0%) | |
| ≥39 | 28 (2.0%) | 49 (3.5%) | |
| Time from randomisation to delivery of twin A (days) a Mean (+/–SD) | 12.4 (12.0) | 13.3 (12.3) | 04 |
| Twin delivery interval (min) Mean (+/–SD) a | 3.6 (9.3) | 10.0 (16.7) | <0.0001 |
| Use of antenatal corticosteroids after randomisation a | 179 (12.9%) | 152 (10.9%) | |
| Anaesthesia or analgesia a , d | 1323 (95.2%) | 996 (71.6%) | |
| Regional | 1226 | 867 | |
| General | 86 | 53 | |
| Other | 22 | 109 | |
| Chorionicity at birth a , e | |||
| Dichorionic/diamniotic | 1016 (73.1%) | 1035 (74.4%) | |
| Monochorionic/diamniotic | 346 (24.9%) | 324 (23.3%) | |
| Monochorionic/monoamniotic | 3 (0.2%) | 1 (0.1%) | |
| Unknown | 25 (1.8%) | 31 (2.2%) | |
a The missing numbers in the planned caesarean and planned vaginal birth groups, respectively, were: mode of delivery ( n = 2, n = 0); timing of caesarean ( n = 1, n = 2); presentation at delivery for twin A/twin B ( n = 2, n = 0); gestational age at delivery of twin A ( n = 1, n = 1); time from randomisation to delivery of twin A ( n = 2, n = 1); twin delivery interval ( n = 2, n = 0); use of antenatal corticosteroids after randomisation ( n = 2, n = 0); anaesthesia or analgesia ( n = 3, n = 2); chorionicity at birth ( n = 3, n = 2).
b Includes one singleton pregnancy.
c Delivery at 37 + 5 – 38 + 6 weeks occurred for 396 (28.5%) and 411 (29.5%) in the planned caesarean and planned vaginal birth groups, respectively.
d More than one response may apply.
e Chorionicity was confirmed by pathology at birth in the planned caesarean and planned vaginal birth groups, respectively: dichorionic/diamniotic ( n = 785, n = 768); monochorionic/diamniotic ( n = 235, n = 212); monochorionic/monoamniotic ( n = 3, n = 0).
The time from randomisation to delivery was shorter for the planned caesarean than the planned vaginal birth group (mean 12.4 v 13.3 days; P = 0.04), and the mean gestational age at delivery was lower in the planned caesarean group ( P = 0.01).
Fetal and neonatal mortality and serious neonatal morbidity outcomes are presented in Table 3 .
| Outcome | Planned caesarean | Planned vaginal birth | OR (95% CI) ( P value) |
|---|---|---|---|
| Total ( n = 2783) | Total ( n = 2782) | ||
| Primary composite outcome: fetal or neonatal mortality or serious neonatal morbidity a | 60 (2.2%) | 52 (1.9%) | 1.16 (0.77, 1.74) (0.49) |
| Gestational age at randomisation | |||
| 32–33 weeks b | 32 (3.4%) | 26 (2.7%) | 1.25 (0.70, 2.24) |
| 34–36 weeks | 26 (1.9%) | 19 (1.4%) | 1.34 (0.71, 2.54) |
| 37–38 weeks | 2 (0.4%) | 7 (1.4%) | 0.30 (0.06, 1.43) |
| Deaths c | 24 (0.9%) | 17 (0.6%) | |
| Fetal death | 13 | 9 | |
| Before labour | 11 | 8 | |
| During delivery | 0 | 1 | |
| Unknown | 2 | 0 | |
| Neonatal death | 11 | 8 | |
| Serious neonatal morbidity c | 36 (1.3%) | 35 (1.3%) | |
| Neonatal morbidity (excluding death of either twin) | Total (N = 2759) | Total (N = 2765) | |
| Birth trauma | |||
| Long bone fracture c | 0 | 4 | |
| Other bone fracture | 1 | 1 | |
| Facial nerve injury at 72 h of age or at discharge c | 0 | 1 | |
| Intracerebral haemorrhage c | 3 | 1 | |
| Apgar score < 4 at 5 mins c | 2 (0.1%) | 7 (0.3%) | |
| Abnormal level of consciousness | |||
| Coma c | 0 | 1 | |
| Stupor or decreased response to pain c | 2 | 0 | |
| Hyperalert, drowsy or lethargic | 9 | 7 | |
| ≥2 Neonatal seizures within 72 h of age c | 3 | 3 | |
| Ventilation ≥24 h via endotracheal tube initiated within 72 h of age c | 27 (1.0%) | 17 (0.6%) | |
| Neonatal sepsis within 72 h of age c | 1 (0.04%) | 2 (0.1%) | |
| Necrotising enterocolitis c | 1 (0.04%) | 3 (0.1%) | |
| Cystic periventricular leukomalacia c | 2 (0.1%) | 0 (0.0%) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
