INTRODUCTION — Multifetal pregnancies may be complicated by spontaneous extremely preterm birth, thereby placing the neonates at high risk for prolonged hospitalization, serious morbidity, and mortality. Typically, all fetuses of a multiple gestation deliver within a short interval; however, in selected cases, the preterm birth of one fetus may not require delivery of the other fetus(es). An extended time interval between births of fetuses at a critical gestational age may improve neonatal survival and reduce morbidity from preterm birth.
This topic will discuss selection of candidates for delayed-interval delivery and management of these pregnancies. Interpretation of available information is challenging because the evidence base consists of case reports, small case series, and literature reviews, but no randomized trials. Ascertainment bias is a significant problem as successful outcomes often are reported, while many unsuccessful cases go unreported.
The routine management of twin and triplet births are reviewed separately.
●(See "Twin pregnancy: Labor and delivery".)
●(See "Triplet pregnancy".)
KEY PRINCIPLES — The best approach for achieving a delayed-interval delivery is unclear. The authors' approach is based upon our early experience with over 50 delayed-interval deliveries [1,2], our ongoing experience with delayed-interval delivery, and our insights from case series reported in the literature [3-14]. The key principles of our approach are:
●Appropriate selection of candidates
●Informed consent
●Exclusion of intraamniotic infection in undelivered fetus(es)
●Drug-induced uterine relaxation
●Antibiotic prophylaxis
●Placement of a cerclage
●Administration of antenatal glucocorticoids
SELECTING CANDIDATES — We discuss the option of delayed-interval delivery with patients when delivery of some fetuses of a multiple gestation is indicated because of maternal, obstetric, or fetal factors and concurrent delivery of fetuses unaffected by these factors would likely result in their death or severe morbidity. This is a rare event and a complex clinical decision as it is difficult to objectively quantify and compare the potential risks and benefits for the various stakeholders. There are no high-quality data on which to select or exclude candidates for delayed-interval delivery and no generally accepted, published guidelines for selection of appropriate pregnancies.
Optimal candidates — We believe the best candidates for delayed-interval delivery are patients with a multiple gestation at an early gestational age (<24 weeks) in which only the first (presenting) fetus spontaneously delivers vaginally due to preterm labor, cervical insufficiency, prelabor rupture of membranes, or intrauterine demise.
Contraindications — We do not offer delayed-interval delivery in the following scenarios:
●Pregnancies ≥28 weeks of gestation. Neonatal outcome at our institution, and similar institutions with an appropriate level of neonatal care, is generally good at this gestational age.
●Pregnancy complications associated with a high risk of serious maternal or fetal morbidity/mortality in ongoing pregnancies, such as preeclampsia with severe features, placental abruption, or intraamniotic infection of the nonpresenting fetus(es). In our practice, the finding of intraamniotic infection involving the first fetus or the requirement for oxytocin augmentation to facilitate its delivery does not exclude the patient as a candidate for a delayed-interval delivery; however, laboratory evidence of amniotic fluid infection (obtained by amniocentesis) of the undelivered fetus(es) would be a contraindication to this approach and would require delivery of the entire pregnancy. (See 'Evaluation for and management of infection' below.)
●Monochorionicity with fetus that has delivered. Vascular anastomoses in the retained placenta may cause complications of the undelivered fetus(es), although successful delayed-interval delivery has been reported in this setting [15,16].
●Operative birth of the first fetus is a relative contraindication; however, single case reports of selective hysterotomy [17] or operative vaginal evacuation [18] for delivery of one fetus with successful outcomes for the undelivered fetus(es) have been described. There is also a case report of cesarean birth of the first twin (birth weight 530 g) at 22+4 weeks followed 10 days later (24+0 weeks) by cesarean birth of the second twin (birth weight 600 g); unexpectedly, the first twin survived, and the second twin died on day 2 of life of cardiovascular and respiratory failure [14].
In our practice, if the first fetus is breech and at a gestational age compatible with newborn survival, we generally perform a cesarean birth and do not attempt delayed-interval delivery of the remaining fetuses.
INFORMED CONSENT — The potential benefits and risks of delayed-interval delivery and the range of potential outcomes are the focus of informed consent.
Key points of informed consent include:
●Delayed-interval delivery is one option, but not the only option. Regardless of gestational age, parents may choose to deliver the entire pregnancy when one fetus of a multiple gestation delivers.
●Delaying delivery increases the risk for maternal, fetal, and/or neonatal infection and, in turn, the potential sequelae of infection (eg, maternal: sepsis, localized infection, infertility; pediatric: periviable or preterm birth, neurodevelopmental impairment, death). (See 'Outcome' below and "Intraamniotic infection (clinical chorioamnionitis)", section on 'Fetal and neonatal outcome' and "Intraamniotic infection (clinical chorioamnionitis)", section on 'Potential maternal sequelae'.)
●The potential benefit of delaying delivery depends on both the gestational age at birth of the first fetus and latency duration. For instance, a predicted median three-week latency interval is of considerable value if the first fetus is born at 23 weeks but not if born at 17 weeks.
●The possibility of extending the pregnancy from a previable to a periviable gestational age is a concern that should be raised in the counseling and consent process. Potentially appropriate candidates for delayed-interval delivery who are at a previable gestational age may elect not to attempt an approach that increases the possibility of birth of an extremely preterm infant who survives with significant long-term morbidity. These patients can reasonably choose to deliver the entire pregnancy at the time of birth of the firstborn infant. A neonatology consultation may assist in providing the best information for this issue. (See "Periviable birth (limit of viability)" and "Long-term outcome of the preterm infant".)
●Since we generally perform a cerclage, we review the risks of cerclage. (See "Transvaginal cervical cerclage".)
EVALUATION FOR AND MANAGEMENT OF INFECTION — We have avoided serious infectious morbidity over the years by an active and preemptive approach to infection.
●We always perform amniocentesis on the undelivered fetus(es) to analyze amniotic fluid for evidence of subclinical intraamniotic infection, which would exclude the pregnancy from consideration of delayed-interval delivery. Intraamniotic infection is a common cause of failure to prolong pregnancy [7,19,20].
●The amniocentesis is almost always performed before delayed-interval delivery is attempted, generally before the presenting fetus delivers. Rarely, we are forced to complete delivery of the first fetus (including cerclage placement) and then perform the amniocentesis immediately thereafter. We have not had to remove the cerclage and deliver an unexpectedly infected retained sibling in these rare instances.
●The amniotic fluid is sent to the laboratory for Gram stain, culture, leukocyte count, and glucose concentration. One or more of the following findings is suggestive of intraamniotic infection: amniotic fluid leukocyte count >50 cells/mm3, glucose concentration <20 mg/dL (1.1 mmol/L), or bacteria on Gram stain. (See "Intraamniotic infection (clinical chorioamnionitis)", section on 'When to perform amniocentesis to test amniotic fluid'.)
•If the initial laboratory results strongly suggest infection (they are not just "borderline"), we do not attempt to delay delivery.
•If initial laboratory results do not suggest infection but the culture (which is checked daily) returns positive, then the attempt at delaying delivery is terminated, and the undelivered fetus is delivered.
●If clinical chorioamnionitis in the delivered fetus's sac is suspected but there is no laboratory evidence of intraamniotic infection in the undelivered fetus(es), we administer broad-spectrum antibiotics with anaerobic coverage and proceed with delayed-interval delivery (in parents who have chosen this approach) (see "Intraamniotic infection (clinical chorioamnionitis)", section on 'Antibiotic therapy'). Importantly, if maternal signs of clinical chorioamnionitis do not resolve within 24 hours after birth of the first fetus and initiation of antibiotic therapy, we repeat the amniocentesis and testing for intraamniotic infection, and deliver if results suggest infection.
PROCEDURE
Delivery of the presenting fetus — The presenting fetus usually delivers spontaneously.
Rarely, the pregnancy is induced. This clinical scenario could occur in a multiple gestation with prelabor rupture of membranes at 22 weeks with evidence of intraamniotic infection of the presenting fetus but no evidence of infection in the other fetus(es). Waiting for spontaneous active labor may increase the risk of development of infection in the other fetus(es) and maternal systemic signs of chorioamnionitis. Therefore, we would induce/augment labor with oxytocin, as needed, and then inhibit uterine activity after the delivery of the first twin, as described.
If oxytocin was administered for induction or augmentation of labor for the presenting fetus, it is discontinued as soon as this fetus delivers. Nitroglycerin is administered intravenously to promptly initiate uterine quiescence [21]. A tocolytic, such as indomethacin, is administered and continued for 24 to 48 hours postoperatively, depending on uterine activity. (See "Inhibition of acute preterm labor", section on 'Fetal side effects'.)
Management of the cord and placenta — After the first fetus has delivered, its umbilical cord is clamped, cut, and ligated with absorbable suture as close to the placental insertion site as possible. Cord traction is avoided and the placenta is left in situ. We do not obtain cord cultures.
Local and systemic antibiotic prophylaxis — After delivery, the lower uterine segment is irrigated with 500 to 1000 mL of an antibiotic solution (eg, 1 gram of cefazolin per liter of saline) using an irrigating syringe. The cervix is grasped with ring forceps (in preparation for cerclage), and the solution is directed purposefully into the lower uterine segment while placing the patulous cervix on gentle traction.
Broad-spectrum antibiotics (eg, gentamicin and clindamycin or ampicillin-sulbactam) are administered intravenously for a three-day course, followed by a cephalosporin (eg, cephradine 500 mg orally every six hours) plus metronidazole 500 mg orally every eight hours orally for four days.
Cerclage — Because it is usually impossible to exclude the possibility that cervical insufficiency was a contributing factor to the early delivery of the first fetus, we generally perform a cerclage in all patients attempting delayed-interval delivery. Cerclage may forestall the birth of the undelivered fetus(es) and is a common intervention in this setting [15,19,20,22-27]. Even if cervical insufficiency was not initially a contributing factor, reapproximation of the cervix after delivery of the first fetus may prevent future prolapse of fetal membranes and decrease the risk of development of intraamniotic infection. However, there are numerous reports of successful delayed-interval deliveries without placement of a cerclage [7,18,28,29].
If membranes from an undelivered fetus are prolapsed, they are gently retracted cephalad using standard maneuvers (see "Transvaginal cervical cerclage", section on 'Replace prolapsed membranes, if present'). We place a McDonald cerclage using #2 monofilament nonabsorbable suture swaged onto a medium taper needle. Two sutures are initially placed; the second is offset by 45 degrees so as not to be placed directly on top of the first suture. Additionally, we may place a "box suture" from 12 o'clock to 6 o'clock to close any distal patulous cervix (Wϋrm cerclage).
Repeat cerclage is unlikely to be successful in pregnancies where a cerclage had been placed earlier in pregnancy, before delivery of the first fetus [2]. This group of patients has a more guarded prognosis due to a particularly foreshortened latency.
Antenatal glucocorticoids — A course of antenatal glucocorticoids is administered to pregnancies ≥23+0 weeks of gestation since only a few primitive alveoli are present below this gestational age. Earlier administration in the 22nd week is reasonable if delivery at 23 weeks is anticipated and aggressive neonatal intervention is planned after thorough counseling by the neonatology service. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)
In determining when to administer the first course of steroids, we consider the gestational age and try to be reasonably sure of the absence of intraamniotic infection.
We administer a rescue (salvage) course after a two-week latency interval from the first course, if clinical factors suggest delivery may be imminent; otherwise, we wait until approximately 31 weeks of gestation.
RhD-negative patients — Anti-D immune globulin is administered to RhD-negative patients after delivery of the first fetus. (See "RhD alloimmunization: Prevention in pregnant and postpartum patients".)
Maternal and fetal monitoring
●The patient is kept in the hospital for a minimum of seven days as many unsuccessful attempts at delayed delivery will declare themselves within this seven-day period (eg, occurrence of maternal infection, preterm prelabor rupture of membranes, preterm labor, or nonreassuring fetal status) [5].
●In fetuses in whom urgent delivery for an abnormal fetal heart rate pattern would be considered, we perform electronic fetal heart rate monitoring for at least 30 minutes, three times daily, while the patient is in the hospital. Abnormal findings (such as recurrent decelerations, persistent loss of previous heart rate variability, or tachycardia) should prompt assessment and/or intervention, such as ultrasound, delivery, or amniocentesis, and retesting for intraamniotic infection.
OUTPATIENT CARE AND MONITORING — Outpatient management is considered after seven days. Ideal candidates are those with minimal uterine activity or bleeding, proven adherence to medical care, and minimal barriers to seeking appropriate care for questions or emergencies.
Appropriate candidates for outpatient care are discharged home with instructions to limit physical activity to light activities in the home (eg, self-care) and to avoid coitus. They are also counseled to be aware of contractions, discharge, fever, or abdominal tenderness and to call if any of these symptoms occur.
We schedule office visits weekly and perform transvaginal ultrasound evaluation approximately every two weeks to evaluate the cervix and the integrity of the cerclage. If the plane of the cerclage is broken by prolapse of the membranes, especially in patients with some uterine activity, then readmission for closer surveillance would be indicated if the pregnancy is still remote from term. We remove the cerclage if preterm prelabor rupture of membranes or labor occurs.
OUTCOME — Key outcomes of delayed delivery of multiple gestations are latency interval, perinatal mortality, neonatal morbidity, and maternal morbidity.
Latency — Latency is important since achieving an extended time interval between births of fetuses at critical gestational ages is the basis for improved neonatal survival and reduced morbidity in the undelivered fetus. In two large series, mean latency was 18 to 19 days [7] and 36 days [2]; in two other large series, median latency was 6 days [5] and 16 days [8]. However, in individual patients, the duration of latency can be very short or quite long: The range reported in the literature is extremely wide, from 1 to 152 days [2,5,7,8,29].
In one study, gestational age ≤24 weeks at delivery of the first fetus was associated with longer latency (≥49 days) [2], but not in another study (latency ≥22 days in 9 percent of first deliveries at 17 to 23 weeks versus 16 percent of first deliveries at 24 to 29 weeks) [5]. Although pregnancies <24 weeks at presentation derive the greatest benefit from delayed-interval delivery rather than contemporaneous delivery [6], survival rates are poor at this gestational age since latency is often not long enough to reach viability [6,7,30].
The most common reasons for the end of latency are preterm labor (with or without ruptured membranes) and clinical evidence of intraamniotic infection (eg, maternal fever) [2]. (See "Intraamniotic infection (clinical chorioamnionitis)".)
No data on latency and pregnancy/neonatal outcome are available for the rare scenario of delayed-interval delivery when the delivered fetus had clinical evidence of infection.
Perinatal mortality — In a 2020 meta-analysis of the effect of delayed-interval delivery on survival of the remaining fetuses (16 studies, 492 pregnancies), the intervention was associated with significantly improved perinatal survival compared with the first born (odds ratio [OR] 5.22, 95% CI 2.95-9.25), particularly when the first born delivered before 20+0 weeks (OR 6.32, 95% CI 1.99-20.13) versus between 20+0 and 23+6 weeks (OR 3.31, 95% CI 1.95-5.63) or after 24+0 weeks (OR 1.92, 95% CI 1.21-3.05) [31].
Neonatal morbidity — Neonatal morbidity is primarily dependent on gestational age at birth; sparse information is available. In the meta-analysis described above, among the survivors, there were no significant differences in short-term and long-term neonatal morbidities between the first born and the remaining fetus(es) [31].
Maternal morbidity — In the meta-analysis described above, serious maternal morbidity was reported in 39 percent of pregnancies after delayed-interval delivery (71 of 183), but there were no maternal deaths [31]. Maternal morbidity included one or more of the following: local infection/sepsis (56), postpartum hemorrhage (12), placental abruption (8), and hysterectomy (2; 1 because of sepsis and 1 because of hemorrhage).
Amniocentesis to evaluate for infection before attempting delayed-interval delivery was not always successful in preventing future maternal morbidity. This highlights the small, but potentially life-threatening, risk of delaying delivery after the birth of one fetus.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Multiple gestation".)
SUMMARY AND RECOMMENDATIONS
●Candidates
•Delayed-interval delivery may be an option when some fetuses of a multiple gestation deliver because of maternal, obstetric, or fetal factors and concurrent delivery of fetuses unaffected by these factors would likely result in their death or severe morbidity. Modest intervals between the birth of fetuses at critical gestational ages can improve neonatal survival and decrease neonatal morbidity among the second-born survivors. (See 'Selecting candidates' above and 'Outcome' above.)
•We believe the best candidates for delayed-interval delivery are pregnancies at an early gestational age (<24 weeks) in which only the first (presenting) fetus delivers due to preterm labor, cervical insufficiency, preterm prelabor rupture of membranes, or intrauterine demise. (See 'Optimal candidates' above.)
•We do not offer delayed-interval delivery to patients with pregnancy complications associated with a high risk of serious maternal or fetal injury (eg, severe preeclampsia, abruptio placentae, intraamniotic infection of the nonpresenting fetus) or gestational age ≥28 weeks. We consider monochorionicity and the need for cesarean birth of the first fetus relative contraindications, but do not consider intraamniotic infection limited to the first fetus an absolute contraindication. (See 'Contraindications' above.)
●Procedure –The optimal strategy to achieve a successful delayed-interval delivery has not been validated by randomized studies, nor have there been adequate sample sizes in observational studies to draw statistically meaningful conclusions. There is no consensus as to optimum management of potential candidates for delayed-interval delivery. (See 'Introduction' above.)
•Informed consent – The counseling and consent process should address the risks and benefits of delayed-interval delivery, including the possibility and consequences of extending the pregnancy from a previable to a periviable gestational age; the risk of maternal, fetal, and/or neonatal infection, as well as potential sequelae of infection; and the option of not attempting delayed-interval delivery. (See 'Informed consent' above.)
•Evaluating for infection – Prior to an attempt at delayed-interval delivery, we perform amniocentesis to exclude subclinical microbial invasion of the amniotic fluid of the proposed undelivered fetus(es). (See 'Evaluation for and management of infection' above.)
•Management of delivery – For patients undergoing delayed-interval delivery, we perform high ligation of the umbilical cord with absorbable suture and leave the placenta in situ, administer local and systemic antibiotics, and place a cerclage. If uterine relaxation is necessary, we use nitroglycerin followed by a tocolytic, such as indomethacin. (See 'Procedure' above.)
•Antenatal corticosteroids – We administer a first course of antenatal glucocorticoids at viability and a rescue (salvage) course after a two-week latency interval from the first course, if clinical factors suggest delivery may be imminent; otherwise, we wait until approximately 31 weeks of gestation. (See 'Antenatal glucocorticoids' above.)
●Postdelivery maternal and fetal monitoring – Ongoing maternal and fetal surveillance during latency involves assessment for signs of infection, preterm labor, preterm prelabor rupture of membranes, cervical change, fetal heart rate abnormalities, and other complications. (See 'Maternal and fetal monitoring' above and 'Outpatient care and monitoring' above.)
●Outcome
•Latency – Length of latency is important since achieving an extended time interval between births of fetuses at critical gestational ages is the basis for improved neonatal survival and reduced morbidity in undelivered fetuses. (See 'Latency' above.)
•Perinatal survival – In a meta-analysis of the effect of delayed-interval delivery on survival of the remaining fetuses, the intervention was associated with significantly improved perinatal survival compared with the first born, particularly when the first born delivered before 20+0 weeks versus between 20+0 and 23+6 weeks or after 24+0 weeks. (See 'Perinatal mortality' above.)
•Neonatal morbidity – Neonatal morbidity is primarily dependent on gestational age at birth. In the meta-analysis described above, among the survivors, there were no significant differences in short-term and long-term neonatal morbidities between the first born and the remaining fetus(es). (See 'Neonatal morbidity' above.)
•Maternal morbidity – In the meta-analysis described above, serious maternal morbidity was reported in 39 percent of pregnancies after delayed-interval delivery, but there were no maternal deaths. Maternal morbidity included one or more of the following: local infection/sepsis, postpartum hemorrhage, placental abruption, and hysterectomy because of sepsis or hemorrhage. (See 'Maternal morbidity' above.)
