Background
The Twin Birth Study randomized women with uncomplicated pregnancies, between 32 0/7 -38 6/7 weeks’ gestation where the first twin was in cephalic presentation, to a policy of either a planned cesarean or planned vaginal delivery. The primary analysis showed that planned cesarean delivery did not increase or decrease the risk of fetal/neonatal death or serious neonatal morbidity as compared with planned vaginal delivery.
Objective
This study presents the secondary outcome of death or neurodevelopmental delay at 2 years of age.
Study Design
A total of 4603 children from the initial cohort of 5565 fetuses/infants (83%) contributed to the outcome of death or neurodevelopmental delay. Surviving children were screened using the Ages and Stages Questionnaire with abnormal scores validated by a clinical neurodevelopmental assessment. The effect of planned cesarean vs planned vaginal delivery on death or neurodevelopmental delay was quantified using a logistic model to control for stratification variables and using generalized estimating equations to account for the nonindependence of twin births.
Results
Baseline maternal, pregnancy, and infant characteristics were similar. Mean age at assessment was 26 months. There was no significant difference in the outcome of death or neurodevelopmental delay: 5.99% in the planned cesarean vs 5.83% in the planned vaginal delivery group (odds ratio, 1.04; 95% confidence interval, 0.77–1.41; P = .79).
Conclusion
A policy of planned cesarean delivery provides no benefit to children at 2 years of age compared with a policy of planned vaginal delivery in uncomplicated twin pregnancies between 32 0/7 -38 6/7 weeks’ gestation where the first twin is in cephalic presentation.
Introduction
Twin pregnancies occur more frequently now than in the past and are reported to complicate 2-3% of all births. Cohort studies have shown a reduced risk of adverse perinatal outcomes for both twins, or for the second twin, when twins at or near term were delivered by means of an elective cesarean delivery. This approach has led to increased rates of elective cesarean delivery for twins in North America and worldwide.
The Twin Birth Study was a multicenter, international randomized controlled trial that enrolled and randomized women with uncomplicated twin pregnancies, between 32 0/7 -38 6/7 weeks’ gestation, to a policy of planned cesarean or planned vaginal delivery. The primary analysis showed that planned cesarean delivery did not increase or decrease the risk of fetal or neonatal death or serious neonatal morbidity as compared with a planned vaginal delivery. A secondary outcome for the trial was a composite of death or neurodevelopmental delay of the children at 2 years of age. This report presents the 2-year outcomes of the children in the Twin Birth Study.
Materials and Methods
Initial study
Women were enrolled in the Twin Birth Study if they were between 32 0/7 -38 6/7 weeks of gestation, the first twin was in the cephalic presentation, and both twins were alive with an estimated weight between 1500-4000 g, confirmed by ultrasonography within 7 days of randomization. Exclusion criteria included monoamniotic twins, fetal reduction at ≥13 weeks of gestation, the presence of a lethal fetal anomaly, contraindication to labor or vaginal delivery (eg, fetal compromise, second twin substantially larger than the first twin, fetal anomaly or condition that might cause mechanical problems at delivery, and previous vertical uterine incision or >1 previous lower-segment cesarean delivery), and prior participation in the Twin Birth Study.
Randomization took place from Dec. 13, 2003, through April 4, 2011; women were randomly assigned to planned cesarean delivery or planned vaginal delivery. Randomization was centrally controlled at the Center for Mother, Infant, and Child Research at the Sunnybrook Health Sciences Center in Toronto with the use of a computerized randomization program stratified according to parity (0 vs ≥1) and gestational age (32 0/7 -33 6/7 weeks, 34 0/7 -36 6/7 weeks, or 37 0/7 -38 6/7 weeks), and used random blocks of varying sizes.
Data were abstracted from the medical records at participating centers by trained study staff and were recorded, after delivery, on standardized data-collection forms. Participating centers assessed fetal growth and well-being with the use of ultrasonography at least every 4 weeks and with the use of nonstress or biophysical profile tests twice weekly if needed; were prepared to perform a cesarean delivery within 30 minutes if necessary; and had anesthetic, obstetrical, and nursing staff available in the hospital at the time of planned vaginal delivery.
Elective delivery by means of either cesarean delivery (for women in the planned cesarean delivery group) or labor induction (for women in the planned vaginal delivery group) was planned between 37 5/7 -38 6/7 weeks of gestation.
Two-year neurodevelopmental follow-up
When the children reached 23-25 months of age, corrected to estimated date of confinement if born <37 weeks’ gestation, the parental caregivers were asked to complete an Ages and Stages Questionnaire (ASQ). The ASQ is a parent/parental caregiver-administered structured questionnaire that includes questions on 5 domains of development: communication, gross motor skills, fine motor skills, problem-solving skills, and personal-social skills. The ASQ has become widely administered and has been utilized and validated in a variety of neurodevelopment monitoring settings worldwide. If the ASQ could not be completed in the 23- to 25-month window, the caregiver was asked to complete a 30-month ASQ at 29-32 months of age.
The scores of each domain were summed up centrally at the coordinating center, and if the score for any 1 of the 5 domains was below the specific cutoff for the domain, the ASQ was considered abnormal. If the ASQ was deemed abnormal, a clinical neurodevelopmental assessment (CNA) was undertaken by an individual trained to assess children for developmental delay at the center. If an ASQ could not be completed in either window described above, a CNA was then undertaken.
The primary outcome for the 2-year neurodevelopmental follow-up of the children was death at any time or a neurodevelopmental delay. Neurodevelopmental delay was defined as a motor or cognitive delay of >3 months (age at time of assessment compared with developmental age as determined by the clinician completing the CNA) or the presence of cerebral palsy on clinical assessment.
Study oversight
The research ethics committee at each participating center approved the study, including the 2-year neurodevelopmental follow-up of the children. The first and last authors take responsibility for the accuracy and completeness of the reported data and for the fidelity of the report to the study protocol. All the women provided written informed consent for enrollment and for the follow-up phase for their infants/children. This trial was registered at www.clinicaltrials.gov (identifier NCT 00187369) and the International Standard Randomized Controlled Trial Register (identifier ISRCTN74420086).
Statistical analysis
The results were analyzed according to intention to treat. Children with Down syndrome, fragile X syndrome, or other chromosomal disorders known to contribute to a neurodevelopmental delay were excluded from the analyses. We calculated that a sample of 2200 pregnancies (4400 children, 2200 per group) would be required to detect a reduction in the risk of death or abnormal neurodevelopmental outcome from 2% with a planned vaginal delivery to 0.05% with a planned cesarean delivery, with 80% power and 2-sided type 1 error probability of .05, assuming a 20% lost-to-follow-up rate. Since the expected outcome rate for twins was not known at the start of the trial, a rate of 2% was used, based on the rates of adverse outcomes found in singletons in the planned vaginal delivery group of the Term Breech Trial.
Odds ratios (OR) and 95% confidence intervals were used to contrast the relative frequency of the composite outcome among those with planned cesarean delivery as compared with planned vaginal delivery. This was done using a logistic model with the fetus/infant or child as the unit of analysis and generalized estimating equations to account for the correlation between 2 fetuses/infants or children from the same pregnancy. The model included the stratification variables of parity and gestational age at randomization. A 2-sided P value ≤.05 was considered to indicate statistical significance for the composite outcome.
Planned subgroups analyses for the outcomes were conducted by testing the interaction term between the treatment group and the following baseline variables: parity (0 vs ≥1), gestational age at randomization (32 0/7 -33 6/7 weeks, 34 0/7 -36 6/7 weeks, or 37 0/7 -38 6/7 weeks), and maternal age (<30 vs ≥30 years). A post hoc analysis was conducted to test the interaction between treatment group and birth order.
Materials and Methods
Initial study
Women were enrolled in the Twin Birth Study if they were between 32 0/7 -38 6/7 weeks of gestation, the first twin was in the cephalic presentation, and both twins were alive with an estimated weight between 1500-4000 g, confirmed by ultrasonography within 7 days of randomization. Exclusion criteria included monoamniotic twins, fetal reduction at ≥13 weeks of gestation, the presence of a lethal fetal anomaly, contraindication to labor or vaginal delivery (eg, fetal compromise, second twin substantially larger than the first twin, fetal anomaly or condition that might cause mechanical problems at delivery, and previous vertical uterine incision or >1 previous lower-segment cesarean delivery), and prior participation in the Twin Birth Study.
Randomization took place from Dec. 13, 2003, through April 4, 2011; women were randomly assigned to planned cesarean delivery or planned vaginal delivery. Randomization was centrally controlled at the Center for Mother, Infant, and Child Research at the Sunnybrook Health Sciences Center in Toronto with the use of a computerized randomization program stratified according to parity (0 vs ≥1) and gestational age (32 0/7 -33 6/7 weeks, 34 0/7 -36 6/7 weeks, or 37 0/7 -38 6/7 weeks), and used random blocks of varying sizes.
Data were abstracted from the medical records at participating centers by trained study staff and were recorded, after delivery, on standardized data-collection forms. Participating centers assessed fetal growth and well-being with the use of ultrasonography at least every 4 weeks and with the use of nonstress or biophysical profile tests twice weekly if needed; were prepared to perform a cesarean delivery within 30 minutes if necessary; and had anesthetic, obstetrical, and nursing staff available in the hospital at the time of planned vaginal delivery.
Elective delivery by means of either cesarean delivery (for women in the planned cesarean delivery group) or labor induction (for women in the planned vaginal delivery group) was planned between 37 5/7 -38 6/7 weeks of gestation.
Two-year neurodevelopmental follow-up
When the children reached 23-25 months of age, corrected to estimated date of confinement if born <37 weeks’ gestation, the parental caregivers were asked to complete an Ages and Stages Questionnaire (ASQ). The ASQ is a parent/parental caregiver-administered structured questionnaire that includes questions on 5 domains of development: communication, gross motor skills, fine motor skills, problem-solving skills, and personal-social skills. The ASQ has become widely administered and has been utilized and validated in a variety of neurodevelopment monitoring settings worldwide. If the ASQ could not be completed in the 23- to 25-month window, the caregiver was asked to complete a 30-month ASQ at 29-32 months of age.
The scores of each domain were summed up centrally at the coordinating center, and if the score for any 1 of the 5 domains was below the specific cutoff for the domain, the ASQ was considered abnormal. If the ASQ was deemed abnormal, a clinical neurodevelopmental assessment (CNA) was undertaken by an individual trained to assess children for developmental delay at the center. If an ASQ could not be completed in either window described above, a CNA was then undertaken.
The primary outcome for the 2-year neurodevelopmental follow-up of the children was death at any time or a neurodevelopmental delay. Neurodevelopmental delay was defined as a motor or cognitive delay of >3 months (age at time of assessment compared with developmental age as determined by the clinician completing the CNA) or the presence of cerebral palsy on clinical assessment.
Study oversight
The research ethics committee at each participating center approved the study, including the 2-year neurodevelopmental follow-up of the children. The first and last authors take responsibility for the accuracy and completeness of the reported data and for the fidelity of the report to the study protocol. All the women provided written informed consent for enrollment and for the follow-up phase for their infants/children. This trial was registered at www.clinicaltrials.gov (identifier NCT 00187369) and the International Standard Randomized Controlled Trial Register (identifier ISRCTN74420086).
Statistical analysis
The results were analyzed according to intention to treat. Children with Down syndrome, fragile X syndrome, or other chromosomal disorders known to contribute to a neurodevelopmental delay were excluded from the analyses. We calculated that a sample of 2200 pregnancies (4400 children, 2200 per group) would be required to detect a reduction in the risk of death or abnormal neurodevelopmental outcome from 2% with a planned vaginal delivery to 0.05% with a planned cesarean delivery, with 80% power and 2-sided type 1 error probability of .05, assuming a 20% lost-to-follow-up rate. Since the expected outcome rate for twins was not known at the start of the trial, a rate of 2% was used, based on the rates of adverse outcomes found in singletons in the planned vaginal delivery group of the Term Breech Trial.
Odds ratios (OR) and 95% confidence intervals were used to contrast the relative frequency of the composite outcome among those with planned cesarean delivery as compared with planned vaginal delivery. This was done using a logistic model with the fetus/infant or child as the unit of analysis and generalized estimating equations to account for the correlation between 2 fetuses/infants or children from the same pregnancy. The model included the stratification variables of parity and gestational age at randomization. A 2-sided P value ≤.05 was considered to indicate statistical significance for the composite outcome.
Planned subgroups analyses for the outcomes were conducted by testing the interaction term between the treatment group and the following baseline variables: parity (0 vs ≥1), gestational age at randomization (32 0/7 -33 6/7 weeks, 34 0/7 -36 6/7 weeks, or 37 0/7 -38 6/7 weeks), and maternal age (<30 vs ≥30 years). A post hoc analysis was conducted to test the interaction between treatment group and birth order.
Results
Characteristics of the participants
Outcome data were available for 4603 children of the initial cohort of 5565 fetuses/infants (83%) from 97 centers in 25 countries ( Figure ): 2320 children in the planned cesarean delivery group (83%) and 2283 children in the planned vaginal delivery group (82%).
The numbers of children from each country are provided in the Supplementary Table A1 . The baseline characteristics of the mothers and the infants were similar in the 2 study groups both at the primary analysis of the trial and for this 2-year follow-up ( Table 1 ). The mean age at assessment was similar between the 2 groups: 26.1 months in the planned cesarean delivery group (5th, 95th percentile 22-40 months) compared with 25.8 months in the planned vaginal delivery group (5th, 95th percentile 23-37 months). One third of the children required a CNA to confirm or dispute the presence of a neurodevelopmental delay or cerebral palsy: 30.3% for the planned cesarean delivery vs 28.0% for the planned vaginal delivery group. The incidence of the primary outcome of the Twin Birth Study (ie, composite fetal or neonatal death and serious neonatal morbidity) and individual neonatal morbidities were similar between the 2 groups of children included in the analyses at 2 years of age ( Table 2 ).
| Women in primary analysis N = 2786 | Women whose children are in 2-y analysis N = 2323 | |||
|---|---|---|---|---|
| Planned CS N = 1393 | Planned VB N = 1393 | Planned CS N = 1172 | Planned VB N = 1151 | |
| Characteristic | ||||
| Maternal age, y, mean ± SD | 29.3 ± 6.1 | 29.30 ± 6.1 | 29.5 ± 6.1 | 29.7 ± 6.1 |
| Maternal age ≥30 y | 632 (45.4%) | 632 (45.4%) | 552 (47.1%) | 558 (48.5%) |
| Parity ≥1 | 857 (61.5%) | 856 (61.5%) | 700 (59.7%) | 701 (60.9%) |
| Gestational age at randomization, wk | 34.9 ± 1.8 | 34.9 ± 1.8 | 34.9 ± 1.8 | 34.9 ± 1.8 |
| Gestational age at randomization | ||||
| <32 0/7 wk | 0 (0.0%) | 1 (0.1%) | 0 (0.0%) | 1 (0.1%) |
| 32 0/7 –33 6/7 wk | 475 (34.1%) | 477 (34.2%) | 402 (34.3%) | 405 (35.2%) |
| 34 0/7 –36 6/7 wk | 679 (48.7%) | 665 (47.7%) | 577 (49.2%) | 547 (47.5%) |
| 37 0/7 –38 6/7 wk | 239 (17.2%) | 250 (17.9%) | 193 (16.5%) | 198 (17.2%) |
| Presentation of twin B at randomization | ||||
| Cephalic | 792 (56.9%) | 783 (56.2%) | 668 (57.0%) | 652 (56.7%) |
| Noncephalic | 601 (43.1%) | 610 (43.8%) | 504 (43.0%) | 499 (43.4%) |
| Chorionicity | ||||
| Dichorionic and diamnionic | 961 (68.9%) | 970 (69.6%) | 812 (69.3%) | 809 (70.3%) |
| Monochorionic and diamnionic | 334 (23.9%) | 326 (23.4%) | 287 (24.5%) | 281 (24.4%) |
| Unknown | 98 (7.0%) | 97 (6.9%) | 73 (6.2%) | 61 (5.3%) |
| National perinatal mortality rate of country | ||||
| <15/1000 | 723 (51.9%) | 730 (52.4%) | 632 (53.9%) | 32 (54.9%) |
| 15–20/1000 | 597 (42.9%) | 591 (42.4%) | 475 (40.5%) | 458 (39.8%) |
| >20/1000 | 73 (5.2%) | 72 (5.2%) | 65 (5.6%) | 61 (5.3%) |
| Mode of delivery | ||||
| Cesarean for both | 1252 (89.9%) | 551 (39.6%) | 1051 (89.8%) | 444 (38.6%) |
| Vaginal and cesarean | 11 (0.8%) | 59 (4.2%) | 7 (0.6%) | 43 (3.7%) |
| Vaginal for both | 129 (9.3%) | 783 (56.2%) | 113 (9.7%) | 664 (57.7%) |
| Missing | 1 | 1 | ||
| Time from randomization to delivery of first twin, d | 12.4 ± 12.0 | 13.3 ± 12.3 | 12.5 ± 12.0 | 13.6 ± 12.4 |
| Missing | 2 | 1 | 1 | 1 |
| Interval between deliveries, min | 5 (2–10) | 2 (1–3) | ||
| Median (interquartile range) | 5 (2–11) | 2 (1–3) | 1 | 0 |
| Missing | 2 | 0 | ||
| Use of antenatal corticosteroids after randomization | 179 (12.9%) | 152 (10.9%) | 151 (12.9%) | 132 (11.5%) |
| Missing | 2 | 0 | 2 | 0 |
| Gestational age at delivery of first twin | 36.7 ± 1.5 | 36.8 ± 1.5 | 36.7 ± 1.5 | 36.8 ± 1.5 |
| 32 0/7 –33 6/7 wk | 88 (6.3%) | 66 (4.7%) | 71 (6.1%) | 48 (4.2%) |
| 34 0/7 –36 6/7 wk | 582 (41.8%) | 581 (41.7%) | 497 (42.4%) | 495 (43.0%) |
| 37 0/7 –38 6/7 wk | 694 (49.9%) | 696 (50.0%) | 577 (49.3%) | 566 (49.2%) |
| ≥39 0/7 wk | 28 (2.0%) | 49 (3.5%) | 26 (2.2%) | 41 (3.6%) |
| Missing | 1 | 1 | 1 | 1 |
| Chorionicity at birth | ||||
| Dichorionic and diamnionic | 1016 (73.1%) | 1035 (74.4%) | 859 (73.5%) | 865 (75.2%) |
| Monochorionic and diamnionic | 346 (24.9%) | 324 (23.3%) | 289 (24.7%) | 266 (23.1%) |
| Monochorionic and monoamniotic | 3 (0.2%) | 1 (0.1%) | 2 (0.2%) | 1 (0.1%) |
| Unknown | 25 (1.8%) | 31 (2.2%) | 19 (1.6%) | 19 (1.7%) |
| Missing | 3 | 2 | 3 | 0 |
| Outcome | Infants in primary analysis N = 5565 | Infants who are in 2-y analysis N = 4603 | ||
|---|---|---|---|---|
| Planned CS N = 2783 | Planned VB N = 2782 | Planned CS N = 2320 a | Planned VB N = 2283 | |
| Composite primary outcome of Twin Birth Study | 60 (2.2%) | 52 (1.9%) | 57 (2.5%) | 46 (2.0%) |
| Missing | 2 a | 0 | ||
| Death contributing to primary outcome | 24 (0.9%) | 17 (0.6%) | 24 (1.0%) | 17 (0.7%) |
| Fetal death | 13 (0.5%) | 9 (0.3%) | 13 (0.6%) | 9 (0.4%) |
| Neonatal death | 11 (0.4%) | 8 (0.3%) | 11 (0.5%) | 8 (0.4%) |
| Serious neonatal morbidity | 36 (1.3%) | 35 (1.3%) | 33 (1.4%) | 29 (1.3%) |
| Missing | 2 a | 0 | ||
| Neonatal morbidity, excluding death of either twin | Planned CS N = 2759 | Planned VB N = 2765 | Planned CS N = 2296 a | Planned VB N = 2266 |
|---|---|---|---|---|
| Birth trauma | 4 (0.1%) | 6 (0.2%) | 4 (0.2%) | 5 (0.2%) |
| Missing | 2 a | 0 | ||
| Apgar score <4 at 5 min | 2 (0.1%) | 7 (0.2%) | 2 (0.1%) | 6 (0.3%) |
| Missing | 2 a | 0 | ||
| Abnormal level of consciousness | 2 (0.1%) | 1 (0.04%) | 2 (0.1%) | 1 (0.04%) |
| Missing | 2 a | 0 | ||
| Seizures (≥2 within 72 h after birth) | 3 (0.1%) | 3 (0.1%) | 3 (0.1%) | 3 (0.1%) |
| Missing | 2 a | 0 | ||
| Assisted ventilation ≥24 h and within 72 h after birth | 24 (1.1%) | 16 (0.7%) | ||
| Missing | 27 (0.1%) | 17 (0.6%) | 2 a | 0 |
| Neonatal sepsis within 72 h of age | 1 (0.04%) | 2 (0.1%) | 1 (0.04%) | 0 (0.0%) |
| Missing | 2 a | 0 | ||
| Necrotizing enterocolitis | 1 (0.04%) | 3 (0.1%) | 1 (0.04%) | 2 (0.1%) |
| Missing | 2 a | 0 | ||
| Cystic periventricular leukomalacia | 2 (0.1%) | 0 (0.0%) | 2 (0.19%) | 0 (0.0%) |
| Missing | 2 a | 0 |
a Received 2-y follow-up for set of twins but no additional primary information was available (removed from neonatal tables in primary).
Death or neurodevelopmental delay
Table 3 shows the 2-year neurodevelopmental outcomes. The frequency of the composite outcome did not differ significantly between the planned cesarean delivery group and the planned vaginal delivery group (5.99% and 5.83%, respectively; OR with planned cesarean delivery vs planned vaginal delivery, 1.04; 95% confidence interval, 0.77–1.41; P = .79). There were no significant differences between the individual components of the composite outcome.
| Outcome | Planned CS N = 2320 | Planned VB N = 2283 | Odds ratio (95% confidence intervals) | P value |
|---|---|---|---|---|
| Composite outcome (death or neurodevelopmental delay at age 2 y) | 139 (5.99%) | 133 (5.83%) | 1.04 (0.77–1.41) | .79 |
| 32 0/7 –33 6/7 wk | 51/796 (6.4%) | 52/803 (6.5%) | ||
| 34 0/7 –36 6/7 wk | 68/1138 (6.0%) | 59/1087 (5.4%) | ||
| 37 0/7 –38 6/7 wk | 20/386 (5.2%) | 22/393 (5.6%) | ||
| Death | 35 (1.5%) | 23 (1.0%) | 1.48 (0.84–2.64) | .18 |
| Fetal or neonatal death | 24 (1.0%) | 17 (0.7%) | ||
| Death up to follow-up a | 11 (0.5%) | 6 (0.3%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
