The high-risk nature of twin pregnancy is presented in detail in the preceding chapter. Twin deliveries pose great challenges to the skill of the obstetrician and to the team of assistants that must necessarily be present at the moment of birth. These events should never be considered as two sequential singleton deliveries. Planning for delivery should begin antenatally, and these plans should consider specific events that could occur as labor commences, continues, and terminates.

Hospitals and/or obstetric units providing maternity services for twin deliveries should have a full range of personnel and equipment prepared to (1) reduce the potential sequelae of preterm delivery; (2) decrease the risk of birth trauma; and (3) optimize the quality of resuscitative efforts provided to newborns. Requisite personnel for a twin delivery should comprise the following: two qualified birth attendants (i.e., two obstetricians or an obstetrician-midwife team); an anesthesiologist with an assistant if required; two neonatal resuscitation teams; and, finally, qualified sonographic assistance. Equipment should include sufficient clamps for two umbilical cords and sufficient resuscitative equipment for two infants; ultrasound equipment; maternal and fetal monitoring equipment; a variety of oxytocic agents, including methylergonovine maleate (Methergine) and prostaglandin analogs; a complete setup for emergency operative intervention or cesarean section in the delivery room; and setups for regional and/or general anesthesia.

In the event that a hospital has inadequate facilities and/or staff, health care providers and patients necessarily face the difficult decision as to whether it is preferable to arrange an antenatal transfer for delivery, often miles away from family and support systems, or to risk the possibility of complications that may arise from inadequate facilities. Postpartum transfer of newborns to neonatal intensive-care units does not allow the same quality of resuscitation that can be obtained when mothers are transferred antenatally and delivery takes place in an optimally equipped facility.

Five areas of obstetric practice (fetal monitoring, fetal imaging, anesthesia, cesarean section and vaginal birth after cesarean section [VBAC], and immediate neonatal care) impact on labor and delivery in a twin pregnancy.

Fetal Monitoring

Monitoring of both fetal heart rates (FHRs) is essential throughout labor and delivery. Newer technology permits monitoring both twins simultaneously, by the external and/or internal route. Whereas auscultation of FHR with a stethoscope or hand-held Doppler is sometimes advocated in the absence of appropriate technology, such a system cannot detect subtle postcontraction changes in the heart rates of both fetuses, even when two examiners are present simultaneously. Moreover, from a cost-effective point of view, use of the equipment to accomplish electronic fetal monitoring is probably less labor intensive than intermittent auscultation. Moreover, in the absence of dual-channel monitoring, it is possible to elicit heart tones from one fetus for long periods of time, during which the second fetus is essentially unmonitored.

Two caveats are associated with dual-channel monitoring: (1) it is possible to monitor one twin twice; and (2) it is possible to confuse the maternal heart rate with bradycardia in one of the twins. The acute onset of bradycardia in one twin should always alert the caregiver to check the maternal pulse immediately to determine if the mother's heart rate differs from the rate of bradycardia: if this latter rate is identical to the rate of bradycardia, the transducers can then be adjusted until the two FHRs reappear. In such circumstances, ultrasound can be helpful in locating the heart and determining the FHR of the twin suspected of having bradycardia. If this does not solve the problem, immediate steps must be undertaken to determine the reason for the abrupt change in the FHR in one twin.

After the birth of the first twin, it is essential that the second twin continue to be monitored. When the second twin enters the pelvis, and if there is no contraindication to amniorrhexis, a spiral electrode can be applied to the second twin's scalp or buttock. Before that time, however, the FHR of the second twin can be assessed with a cardiotransducer and/or a hand-held Doppler. The continued use of the tocotransducer after the birth of the first twin permits simultaneous assessment of the strength and frequency of maternal contractions. A comprehensive protocol for monitoring twins in labor has been described by Eganhouse (Table 1).¹

TABLE 1. Protocol for the Use of Monitors with Twins

1. Apply EFM with mother in the lateral position or semi-Flower's.
   1. Use two different brands of monitors to improve signal quality.
      
   2. Synchronize internal clock of both monitors.
      
   3. Auscultate twin A and apply the cardiotransducer.
      
   4. Apply the tocotransducer.
      
   5. Auscultate twin B and discriminate from A's tracing. Apply cardiotransducer for twin B.
      
   6. Adjust belts to old transducers in place. Try to accomplish monitoring with two or three belts rather than four.
      
   7. Observe FHR tracings carefully to ascertain that both fetuses are being monitored.
      Document location of FHRs.
      
   
   
2. Apply a spiral electrode to twin A as soon as possible.
   1. Ascertain that US and spiral electrode produce different tracings. (It is possible to monitor twin A using two modes.)
      
   2. If only one tracing appears, reauscultate and readjust the cardiotransducer for twin B.
      
   
   
3. Observe for nonreassuring FHR patterns, including decreased variablility, late, severe variable, or prolonged decelerations.
   1. Watch for nonreassuring patterns.
      
   2. Institute appropriate interventions for nonreassuring FHR patterns as needed.
      
   
   

Chervenak and colleagues² noted that ultrasonography is also indispensable for external cephalic version of the second twin after delivery of the first. Chervenak and associates³ reported success in 10 of 14 transverse lies and 8 of 11 breech presentations, respectively. In these two small series, epidural anesthesia enhanced the likelihood of success. In two of the unsuccessful attempts at positional change, the weight of twin B exceeded that of twin A by 500 g or more. In the absence of ultrasonography to continuously monitor position and FHR, external cephalic version is considered potentially unsafe by many physicians.

Anesthesia

Scant data exist comparing general versus regional anesthesia for the delivery of multiples. Regardless of which modality is contemplated antenatally or at the onset of labor, a skilled anesthesiologist should be available throughout labor in the event that immediate operative intervention is required.

Several precautions should be observed when the patient is admitted to labor and delivery, regardless of whether her status is characterized as preterm or term:

1. Blood should be obtained and sent for type and screen. The risk of life-threatening postpartum hemorrhage for mothers of twins or higher order multiples is real, whereas cross-matching all singleton patients at the onset of labor is not cost-effective and this practice is no longer routine.
   
2. Isotonic crystalloid solution should be infused intravenously (IV) to prevent dehydration that would otherwise decrease uterine perfusion.
   
3. All IV lines should be of sufficient diameter to allow fluid resuscitation or therapy with blood or blood products.
   
4. Ingestion of solid food is contraindicated. Some obstetric units prefer a strict nil per os regimen; others permit ingestion of clear liquids. In either case, the aim is to decrease the likelihood of aspiration of solid or partially digested stomach contents.
   
5. The mother's lumbothoracic spine should be evaluated in the sitting position early in labor to facilitate insertion of an epidural needle if deemed desirable or necessary. At the same time, the patient's neck and airway should be evaluated in preparation for a rapid induction of general anesthesia should that be necessary.
   

Regardless of which type of anesthesia is planned, or even if none is administered, steps should be taken to prevent gastric acid aspiration and its sequelae. The most popular combination of medications used to reduce the risk of gastric acid aspiration is 30 mL of an oral nonparticulate antacid, such as sodium bicitrate, together with 10 mg of IV metoclopramide to enhance stomach emptying and increase the tone of the gastroesophageal sphincter. The administration of an H₂ blocker, such as ranitidine (50 mg IV), decreases the acidity of subsequent gastric secretions. No adverse maternal or neonatal side effects have been described after the use of these agents. Metoclopramide and ranitidine may be begun as soon as the patient enters the active phase of labor, and may be repeated every 6 hours.

Before 1975, it was thought that regional analgesia and anesthesia were contraindicated for multiple gestation. Practitioners were concerned that regional blocks prolonged the duration of labor as well as increased the risk of aortocaval compression, with resulting maternal hypotension and reduced placental perfusion. This opinion changed in subsequent years: currently, regional anesthesia is considered an excellent method for pain relief during labor and delivery of twins.⁴ It allows continuous or near-continuous pain relief for the mother without direct depression of the fetus, provided that the regional analgesia is properly administered and complications are avoided. To enhance patient safety and acceptability, regional blocks should be administered by those who are experienced in administration of epidural analgesia for normal deliveries.⁴ In the case of twins, analgesic blocks should not be started until obstetric conditions are appropriate (i.e., frequent, strong uterine contractions producing an adequate rate of cervical dilation).⁴ Many services place the catheter early on, but delay delivery of the anesthetic dose until conditions are deemed favorable.

Proponents of epidural anesthesia cite four main points to justify its use⁴:

1. The need for maternal narcotics and sedatives is greatly reduced, thus avoiding the fetal respiratory depressant effect associated with such drugs.
   
2. Perineal anesthesia prevents reflex bearing-down by the mother and permits a more controlled delivery of the infant over a relaxed perineum.
   
3. Operative delivery, common in twins, is facilitated by the presence of epidural anesthesia.
   
4. When required, postpartum manual removal of the placenta and/or manual exploration of the uterine cavity are facilitated in the presence of adequate anesthesia and in the absence of compromised uterine contractility.
   

Additional arguments in favor of regional anesthesia are as follows: premature infants fare better, the likelihood of gastric acid aspiration is decreased, mothers are more alert during recovery, and postoperative pain can be reduced with appropriate medication and dosage. Continuous lumbar epidural anesthesia, if used successfully for first-stage pain relief, may be renewed to provide anesthesia for delivery. In some circumstances, pudendal block or local infiltration may be used concomitantly. Regardless of the type of anesthesia selected, however, two factors are of overriding importance: the fetuses must be monitored throughout labor and delivery, and the anesthesiologist must be present and prepared to give general endotracheal anesthesia should circumstances require it.

Systemic narcotics may be administered with safety early in labor if due regard is given their potential respiratory depressant effect on the fetuses. Should delivery occur unexpectedly or at any time when drug activity may still be present, naloxone (Narcan) should be administered to the mother to reverse the narcotic effects before delivery and, in an appropriate dose, to the infant to reverse narcotic effects after delivery. The use of paracervical block is no longer favored because of its association with fetal bradycardia.

Psychoprophylactic methods of pain relief are popular in North America and can be used as easily with twins as with singletons. They are not only harmless, as far as obstetric outcome is concerned, but appear to be a benefit to many patients. Indeed, when premature labor occurs, use of psychoprophylaxis may minimize analgesic needs, as may the presence of a trained labor coach or a doula.

On occasion, rapid uterine relaxation is required. A small bolus of nitroglycerin (50 to 100 μg) is efficacious and apparently not harmful. In the double dilution technique, 5 mg of nitroglycerin is diluted in 10 mL of saline, yielding 500 μg/mL. The dilution procedure is repeated so that 1 mL of the doubly diluted solution contains 50 μg. This is an appropriate concentration for the bolus administration of nitroglycerin.

When postpartum blood loss is severe, oxytocin should be administered after the delivery of the fetuses. Some practitioners prefer to delay administration of oxytocin until after the delivery of the placenta in order to avoid its entrapment. Because oxytocin is not always effective, prostaglandin F₂ should be available in the delivery room for direct injection into the uterus (through the abdominal wall) at a concentration of 0.2 mg/mL and a maximum total dose of 1 mg.

Cesarean Section and Vaginal Birth After Cesarean Section

Patients with twin gestation should be cared for in an institution capable of providing an emergency cesarean section should circumstances warrant this. Institutions that cannot provide the means and personnel to deliver infants of high-risk mothers by cesarean section in 30 minutes or less should not attempt to provide care for multiple pregnancy.

The vaginal delivery of twins after prior cesarean section was reported as early as the 1960s, well before the present trend toward VBAC. Although the 1988 Committee Opinion (No. 64) of the American College of Obstetricians and Gynecologists stated that data were insufficient to evaluate the safety of this delivery modality in twins, subsequent reports support this practice.

Nevertheless, the risk for maternal trauma in the form of scar rupture is always present in association with VBAC. Intrauterine manipulations, such as total breech extraction or internal podalic version, theoretically enhance this risk, although data to support this contention are lacking. Further, in the absence of data from comparative randomized trials, the risk/benefit ratio of VBAC must be evaluated on a case-by-case basis. Diverse elements merit consideration, including the patient's desires and potential fears, the obstetrician's experience, and the ability of the hospital to provide a safe vaginal delivery or cesarean section with appropriate intrapartum monitoring by personnel and technologic equipment 24 hours per day, 365 days per year. Also to be considered is the indication for the prior cesarean. Permanent indications (e.g., contracted pelvis) speak against VBAC except in cases of extreme prematurity, whereas transient indications (e.g., failure to progress) favor VBAC. The use of epidural anesthesia does not mask the pain of scar dehiscence/rupture, and whereas in the past fetal heart tone decelerations were thought to be indicative of scar dehiscence/rupture, this clinical sign is no longer deemed useful by all knowledgeable parties.

Immediate Neonatal Care

Twins mirror, and at the same time magnify, the two major problems facing obstetricians in the United States today: preterm delivery and low birth weight. Under these circumstances the first few minutes of extrauterine existence are of paramount importance to twins in terms of their ability to survive and the degree of subsequent morbidity or lack thereof. Ideally, an experienced pediatrician or neonatologist should be present at the delivery. Although it may be acceptable for the obstetrician, anesthesiologist, or delivery room nurse to provide emergency neonatal stabilization for the deliveries of low-risk patients, this is less than ideal for deliveries of twins.

Figure 1 depicts the nine possible combinations of fetal positions before delivery in twin pregnancy.⁵ The frequencies of these combinations are shown in Figure 2, taken from the classic work by Chervenak and co-workers.³ Unless specific obstetric conditions preclude vaginal delivery, it is appropriate in vertex-vertex combinations. Many experienced clinicians are also comfortable with the vaginal delivery route for combinations of vertex-breech, breech-vertex, or breech-breech; others are not. The delivery mode for the remaining combinations of fetal positions depends greatly on the circumstances of the individual case, the skill and experience of the operator, and the skill of the team of assistants available to help with the delivery and resuscitation--both under the best of circumstances and at other times, such as nights, holidays, and weekends.

Physicians contemplating vaginal delivery of malpositioned twins should carefully consider that much of the morbidity and mortality associated with breech delivery is a function of physician intervention, particularly in the case of premature breech birth. Regardless of the route of delivery, the premature fetus in the breech or vertex presentation may have less respiratory reserve and less mature cardiovascular reflexes. Such fetuses are more likely to suffer long-lasting effects if subjected to the stress of hypoxia.

Either type of uterine incision, low vertical or low transverse, could be appropriate. Selection of the skin incision may be dictated by the reason for the section. For example, many clinicians who would otherwise use a low transverse incision prefer to use a midline incision in emergencies. Regardless of the external incision, the larger uterine bulk of the twin pregnancy usually leads to development of the lower uterine segment earlier than in singleton gestations. Little difficulty should attend the extraction of smaller fetuses, one at a time, from a low transverse incision. If, however, the second fetus is transverse, a low vertical incision may be preferable because it can be extended superiorly if circumstances warrant, such as in the presence of extreme prematurity.

In the past, decisions to practice cesarean section or vaginal delivery were often dogmatic and/or based on the personal preference of the physician or the clinic director in charge.⁶ More recently, this decision is based on outcome analyses. The first and most widely quoted is the pioneer work of Chervenak and associates.⁷ Other authors have also addressed this issue.^8,9 Among the major considerations in published decision schema are the weight of the twins and the desire to avoid or reduce birth trauma. A full discussion of various options is presented by Chervenak and colleagues.⁷ The types of trauma, birth presentation, birth weight, and mode of delivery in 362 consecutive twin deliveries are shown in Table 2. Figure 3 and Figure 4 are the protocols for the intrapartum management of twin gestation proposed by Chervenak and co-workers^3,10 and Warenski and Kochenauer,⁹ respectively.

TABLE 2. Significant Birth Trauma for 362 Consecutive Twin Gestations

The characteristics of labor in twins are not well studied, aside from a recent analysis of normal and abnormal labor patterns by Friedman.¹¹ As Friedman put it, “conflicting information abounds.” Gravidas with twin pregnancies are reported to have short labors or unduly long ones, depending on the source of information. Friedman further observed that this issue may be of less clinical importance than it was in the past, because so many women with twin pregnancies (44% to 52%) are delivered by cesarean section.¹¹

A major problem of prior investigations that attempted to characterize labor in twins and higher-order multiples lies in the fact that only the total duration of labor was examined, rather than its dynamic progression over time.¹¹ Labor-duration studies fall short of the mark in two important areas. First, they fail to adjust for factors known to affect the course of labor or that may have an impact on it (e.g., parity, cervical dilation at onset of labor, maternal age). Second, they deal only with the duration of labor from the onset to delivery as the measured endpoint.¹¹ As those who have become familiar with the graphic analysis of labor in the last three decades know, it is reasonable to suggest that total duration is a crude means for assessing progress because it ignores subtle and often not so subtle alterations in labor patterns.

Friedman reanalyzed previously published data of graphic analyses of labor in singletons and twins.¹¹ The labor patterns of 184 multiple pregnancies were reconstructed from a total of 24,000 gravidas studied over a 5-year period in order to provide dilation and descent patterns. As expected, multiparas were more common than nulliparas (118 vs 66), more than 50% of infant birth weights were less than 2500 g, and malpositions were common. The difference in cervical dilation at the onset in labor was obvious. On average, the percentage of labors of twins where the cervix was dilated 2, 3, or 4 cm at the onset of labor was double that of singletons (Fig. 5). When data based on the graphic analyses of labor were analyzed by parity, twins had a shortened latent phase compared to singletons (6.2 vs 8.6 hours, nulliparas; and 4.0 vs 5.3 hours, multiparas). In contrast, other phases of labor were slower, significantly more so for the deceleration phase and the second stage in the multiparous patients.

When these data were analyzed further in a subset of 50 pairs of twins compared to 50 singletons (matched by maternal age, parity, weight of the presenting fetus, fetal position, analgesic effect, uterotonic stimulation, and finally, the use of conduction anesthesia), the shorter latent phase resulting from prelabor cervical dilation and the prolonged deceleration phase and second stage were confirmed. These findings are shown in Table 3. The relative frequencies of dysfunctional labor according to phase of labor and parity are shown in Table 4. Simply stated, all identifiable labor abnormalities were more common in twins in this carefully matched study of twins and singletons.

TABLE 3. Labor Data, 50 Pairs of Multiparous Singletons and Twins Matched by Maternal Age, Parity, Fetal Weight (Presenting Fetuses), Fetal Position, Analgesic Effect, Uterotonic Stimulation, and Conduction Anesthesia

TABLE 4. Relative Frequency of Dysfunctional Labor According to Phase of Labor and Parity

To obtain further clarification of paradoxical comments in the literature, Friedman¹¹ performed a multivariate analysis of data from 976 of 1195 twin pregnancies in the National Collaborative Perinatal Program (NCPP) database. Abnormal labors were not uncommon. Among the maternal factors favoring abnormality were race, white; parity, high; Socioeconomic Index (SEI), high; weight gain, excessive; cephalopelvic disproportion (CPD), present; and gestational diabetes, present. Among the fetal factors conducive to abnormality were first twin greater than 3500 g, combined weight greater than 6000 g, breech presentation, occipitoposterior or occipitotransverse position, and polyhydramnios.

Friedman also performed logistic regression on the factors affecting labor from 176 pairs in a subset available in the NCPP database.¹¹ The factors significantly affecting labor included first twin greater than 2500 g (RR, 1.94; p = 0.004); gestational diabetes (RR, 2.15; p = 0.006); and prematurity (RR, 0.70; p = 0.018).

Friedman concluded the following: conditions that favor normal labor are preterm labor and small fetuses; conditions that predispose patients to abnormal labor are term labor and a large first twin.

After delivery of the first twin, monitoring of the second twin must continue. One obstetrician should attend to the second twin, while the birth attendant assumes primary responsibility for the mother and the neonatologist cares for the first twin. Continuous fetal monitoring, regardless of the elapsed time since the delivery of the first twin, provides reassurance about the continued intrauterine health of the second twin. Monitoring can be external in most cases. If membranes are ruptured, an internal lead can be applied. If the FHR pattern is normal, there should be no reason to intervene prematurely. Indeed, dilute oxytocin may be given with safety to effect delivery. Operative intervention is not necessary in the presence of a normal FHR tracing, and the obstetrician should await the natural processes of labor. If, however, the fetal monitor tracing indicates a sudden, serious abnormality, a fetal scalp pH determination should be made if the clinician believes this could be done without jeopardizing the health of the fetus. If sufficient time is indeed present and if the result shows fetal acidosis, a decision must be made to undertake immediate vaginal delivery or cesarean section. In other instances, this decision is made on grounds such as the type of tracing and a clinical assessment of the fetal condition. If the decision is to proceed with a vaginal delivery, in order for this approach to have an advantage over cesarean section, it should be accomplished within a reasonable amount of time and at no added risk of maternal or fetal trauma. Adequate expertise and supervision are required for the safe performance of an operative vaginal delivery. If cesarean section is to be performed, haste is required to achieve the maximum fetal benefit. Significant delay may obviate these efforts.

There is no agreement as to when membranes should be ruptured for the second twin. Some physicians advocate leaving membranes unruptured as long as possible to maintain a dilating wedge against the cervix; others favor administration of oxytocin by infusion pump to stimulate contractions and accelerate labor; still others favor artificial rupture of membranes after a specific arbitrary time (e.g., 5 or 10 minutes) has elapsed after the delivery of the first twin. Adequate prospective studies have yet to demonstrate conclusively the value of any of these maneuvers.

Similarly, there is no agreement as to which method of delivery of the second twin is optimal, especially if the second twin is nonvertex. There are three options: primary cesarean section, external cephalic version, or primary breech extraction. Whereas breech extraction was not enthusiastically endorsed some years ago, several more recent reports now describe it as appropriate, especially if the second twin is breech. An excellent review of the recent literature with outcome results has recently been published by Chauham and Roberts.¹² The case for version was reexamined in the early 1980s by Chervenak and associates, who noted an 80% success rate after version in 25 cases; 20% could not be converted.^2,3 In a later report, Gocke and Nageotte¹³ were able to document only a 66% success rate and a total cesarean section rate of 40% in those patients in whom version was attempted. It is important to remember that if the version attempt fails, breech extraction may be more difficult.⁹ A factor that strongly influences the success of version attempts is the degree to which the uterus contracts after the delivery of the first twin. These issues are discussed in detail by Chauham and Roberts.¹²

The combined method of vaginal delivery followed by a cesarean section no longer raises eyebrows. Despite this, the decision to conduct a cesarean section of the second twin should be based on documented indications that are clearly recorded in the progress notes. Most such operations are based on objective evidence of a decline in the status of the second twin. The potential benefit to the fetus should always be considered in light of the risks of cesarean delivery in terms of maternal morbidity or mortality.

Older texts and clinical aphorisms suggested a need to deliver the second twin within a relatively short interval (i.e., 20 minutes or less). Undoubtedly, these dicta contributed in some part to the use of unnecessary interventions. The basis of the obstetrician's concern no doubt relates to the risk of hypoperfusion of the placenta or abruption, when the uterine size is rapidly reduced after the delivery of the first twin. Other concerns include the possibility of uterine inertia, cord prolapse, and constriction of the cervical os, thus making rapid delivery of the second twin difficult if not nearly impossible without undue trauma in the event that fetal distress occurs.

The length of time for delivery of the second twin was reviewed at Prentice Women's Hospital.¹⁴ Intervals varied from 30 to 300 minutes without adverse fetal consequences. Although our experience was never published, other authors have described similar times.^15,16 It should be clearly understood, however, that in all instances the second twin was carefully monitored by internal or external tocodynamometry until delivery took place. Moreover, ultrasound was available to make an initial assessment of the second twin's position after the first twin had delivered and to document the fact that this position did not change. Furthermore, anesthesia personnel were present to assist in supporting the mother, and an operating room was ready in the event that circumstances changed and required an immediate delivery by cesarean section.

Pathologic Confirmation

It is standard practice to submit the placenta from all twin gestations for pathologic examination for determination of chorionicity, although this can be accomplished in the delivery room in most instances.¹⁷

Preparing for the delivery of twins should be done well in advance. Prior sections of this chapter cite specific areas of care that must be planned and provided for. Requisite attention should go into the selection of the delivery team and the setup of the room. The patient should be amply forewarned of the following four points. First, she and her support person will be surrounded by a multitude of hospital personnel in the moments preceding, during, and immediately after the delivery. Second, once it is determined that she and her babies are in stable condition postpartum, most of the staff will disappear, leaving the mother, her support person, and her babies time to become acquainted in private. Third, even if all goes well, low birth weight of the infants or any number of unforeseen possibilities may necessitate a 24-hour stay for one or both infants in the neonatal intensive-care unit. Finally, should this last eventuality take place, the bonding process will not be irrevocably disrupted.

1. Eganhouse D: Fetal monitoring of twins. J Obstet Gynecol Nurs 21: 17, 1992

2. Chervenak FA, Johnson RE, Berkowitz RL et al: Intrapartum external version of the second twin. Obstet Gynecol 62: 160, 1983

3. Chervenak FA, Johnson RE, Youcha S et al: Intrapartum management of twin gestation. Obstet Gynecol 65: 119, 1985

4. Cohen H, Brunner EA: Analgesic and anesthetic considerations. In Keith LG, Papiernik E, Keith DM, Luke B (eds): Multiple Pregnancy: Epidemiology, Gestation and Perinatal Outcome, pp 517 – 525. London, Parthenon Publishing Group, 1995

5. Keith LG, Johnson TRB, Lopez-Zeno JA, Creinin M: Labor and delivery. In Keith LG, Papiernik E, Keith DM, Luke B (eds): Multiple Pregnancy: Epidemiology, Gestation and Perinatal Outcome, pp 503 – 516. London, Parthenon Publishing Group, 1995

6. Depp R, Keith LG, Sciarra JJ: The Northwestern University twin study: VII. The mode of delivery in twin pregnancy, North American considerations. Acta Genet Med Gemellol 37: 111, 1988

7. Chervenak FA, Johnson RE, Berkowitz RL et al: Is routine cesarean section necessary for vertex-breech and vertex-transverse twin gestations? Am J Obstet Gynecol 148: 1, 1984

8. Rabinovich J, Barkai G, Reichman B et al: Randomized management of the second nonvertex twin: Vaginal delivery or cesarean section? Am J Obstet Gynecol 156: 52, 1992

9. Warenski JC, Kochenauer NK: Intrapartum management of twin gestation. Clin Perinatol 16: 4, 1989

10. Chervenak FA: The optimum route of delivery. In Keith LG, Papiernik E, Keith DM, Luke B (eds): Multiple Pregnancy: Epidemiology, Gestation and Perinatal Outcome, pp 491 – 501. London, Parthenon Publishing Group, 1995

11. Friedman EA: Normal and abnormal patterns of labor. In Keith LG, Papiernik E, Keith DM, Luke B (eds): Multiple Pregnancy: Epidemiology, Gestation and Perinatal Outcome, pp 427 – 435. London, Parthenon Publishing Group, 1995

12. Chauham SP, Roberts WE: Intrapartum management. In Gall SA (ed): Multiple Pregnancy and Delivery, pp 243 – 280. St. Louis, CV Mosby, 1996

13. Gocke S, Nageotte MP: Management of the nonvertex second twin: Primary cesarean section, external version or primary breech extraction. Eighth Annual Meeting of the Society of Perinatal Obstetricians, Las Vegas, 1988

14. Wittman R, Keith L, Method M, Depp R: The Northwestern University Multihospital Twin Study: VI. Factors relating to low Apgar scores in twin delivery. Acta Genet Med Gemellol 35: 41, 1986

15. Chitkara U, Berkowitz RL: Multiple gestations. In Multiple Gestation in Obstetrics: Normal and Problem Pregnancy, pp 881 – 920. New York, Churchill Livingstone, 1992

16. Rayburn WF, Lavin JP Jr, Miodovnik M et al: Multiple gestation: Time interval between the delivery of the first and second twins. Obstet Gynecol 63: 502, 1984

17. Derom R: Placentation. In Keith LG, Papiernik E, Keith DM, Luke B (eds): Multiple Pregnancy: Epidemiology, Gestation and Perinatal Outcome, pp 113 – 128. London, Parthenon Publishing Group, 1995