Study population

This was a retrospective cohort study of twins born at 24 ^0/7 to 27 ^6/7 weeks’ gestation and admitted to level III neonatal intensive care units (NICUs) participating in the Canadian Neonatal Network (CNN) between 2010 and 2017.

The following cases were excluded: infants with major congenital anomalies, infants born outside a level III NICU (ie, transferred to a level III NICU after birth, designated as “outborn”), cases with an interval of more than 2 hours between the birth times of the first and second twins, twins with no matching cotwin, cases with incomplete or missing information, and cases in which mode of birth was not consistent with current protocols (eg, VD of a first twin in non-Vx presentation). We included only pregnancies in which both cotwins met the study criteria.

Data source

The CNN maintains a national database of outcomes, risk factors, and practices for infants admitted to level III NICUs in Canada. At each site, data are collected from patient charts by trained abstractors according to standard definitions and electronically entered into a customized data entry program with built-in error checks. The CNN database has been shown to have very high accuracy and internal consistency. Data collection for this study was approved by the local research ethics boards. We obtained ethics approval for retrospective investigation of this dataset from the Sunnybrook Health Sciences Centre Research Ethics Board (approval number 444-2018) and from the executive committee of the CNN.

Exposure

The primary exposure was a trial of vaginal twin delivery in pregnancies with the first twin in Vx presentation and the second twin in non-Vx presentation (Vx and non-Vx pairs). Because information on whether a trial of VD occurred is not coded in the CNN database, we considered women to be exposed to a trial of VD if the first twin was born vaginally, irrespective of the mode of delivery of the second twin. Therefore, the exposure group (study group) included Vx and non-Vx twin pairs for which the mode of delivery was “VD-VD” or “VD-CD” ( Figure 1 ).

Selection of the study groups

CD , cesarean delivery; VD , vaginal delivery; Vx , vertex.

Hiersch et al. Mode of delivery and outcome of twins born at <28 weeks’ gestation. Am J Obstet Gynecol 2021.

Neonatal outcomes of the study group were compared with those in 2 separate nonexposed control groups. The first control group (control group 1) included Vx and non-Vx pairs in which both twins were born by primary CD (CD-CD). We recognized that this control group may include cases that actually had a trial of VD but underwent CD before the birth of the first twin. However, because the main concerns associated with a trial of vaginal twin delivery relate to complications occurring at the time of birth, especially the birth of the second twin, we felt that this control group would reflect, from a pragmatic perspective, the outcomes of twin pairs not exposed to a trial of VD. In addition, to overcome this potential limitation, we included a second control group (control group 2) consisting of pairs in which both twins were born by CD owing to a non-Vx presentation of the first twin, thereby representing twin pairs who definitively had no exposure to a trial of VD ( Figure 1 ). Of note, for the purpose of the current study, the presentation considered was the actual presentation at which the fetus was born, because we were mainly interested in the birth trauma associated with the delivery of the non-Vx second twin.

Outcomes

The primary outcome was a composite neonatal outcome defined as 1 or more of the following: (1) neonatal death before discharge; (2) severe neurologic injury, defined as grade 3 or 4 intraventricular hemorrhage (according to the criteria of Papile et al ) or periventricular leukomalacia, diagnosed by cranial ultrasound; or (3) birth trauma, defined as any of the following: intracranial hemorrhage, subarachnoid hemorrhage, or other traumatic injury (coded as “birth trauma otherwise unspecified” in the CNN database).

Secondary outcomes included (1) neonatal death before discharge; (2) severe neurologic injury (as defined above); (3) bronchopulmonary dysplasia, defined as requirement for oxygen at postmenstrual age of 36 weeks’ gestation or at the time of transfer to a level II facility ; (4) necrotizing enterocolitis stage 2 or 3, defined according to the Bell criteria ; (5) 5-minute Apgar score of <7; and (6) need for mechanical ventilation.

Data analysis

Baseline characteristics and neonatal outcomes of the study group were compared with those of the 2 control groups. Outcomes were analyzed using the following 2 approaches:

• 1.
  

  Using pregnancy (or twin pair) as the unit of analysis. With this approach, adverse outcomes were assigned to a pregnancy if 1 or both twins within a pair experienced the adverse outcome.

  
  
• 2.
  

  Using neonate as the unit of analysis. With this approach, each neonate was considered individually, and outcomes were analyzed separately for the first and second twins in a pair. Based on CNN policy, for patient privacy purposes, outcomes with fewer than 5 entries in the database were not reported.

The Student t test or nonparametric tests were used for continuous variables, and the chi-square or Fisher exact tests were used for categorical variables, as appropriate. Multivariable log-binomial regression analysis was used to determine the association between exposure and the primary and secondary outcomes after adjusting for gestational age at birth and birthweight below the 10th percentile (based on a national reference chart) and using each of the 2 control groups as reference. Models were fitted with generalized estimating equations to account for the correlation within twin pairs. Associations were expressed as crude and adjusted relative risk (aRR) with 95% confidence intervals (CI).

The significance of statistical tests was evaluated using 2-sided P values at the 5% significance level. Data were analyzed using the statistical package Statistical Analysis System (version 9.3; SAS Institute, Cary, NC).

Patient and public involvement

Neither patients nor members of the public were involved in the research described in this article.

Characteristics of the study population

A total of 8972 infants born at 24 ^0/7 to 27 ^6/7 weeks’ gestation were admitted to the CNN during the study period; 1082 of these were twin infants (541 twin pairs) who met the study criteria ( Figure 1 ). Of the 220 Vx and non-Vx pairs, 112 pairs had a trial of labor (study group), and 108 pairs had both twins born by CD (control group 1). Of the 170 pairs with the first twin in non-Vx presentation, all twins were born by CD (control group 2) ( Figure 1 ).

Among the 112 Vx and non-Vx pairs that had a trial of labor, the rate of CD for the second twin was 30%, compared with 7% for the 71 Vx and Vx pairs that had a trial of labor ( P <.001) ( Figure 1 ).

The baseline characteristics of the study group and the 2 control groups were similar ( Table 1 ). Specifically, there was no marked difference between the groups concerning exposure to antenatal corticosteroids and magnesium sulfate, gestational age at birth, infant sex, and birthweight below the 10th percentile for gestational age ( Table 1 ).

Neonatal outcomes by mode of delivery: per pregnancy analysis

We first determined the association between a trial of labor and neonatal outcomes using pregnancy as the unit of analysis, whereby an event was defined as adverse outcome for 1 or both infants within each twin pair ( Table 2 ). The unadjusted rates of neonatal outcomes were also visually presented to simplify the comparison between groups ( Figure 2 ). There was no difference between the groups in the rates of the primary composite outcome (42% in the study group, 37% in the control group 1, 34% in the control group 2) or any of the secondary outcomes. Findings remained similar after adjustment for gestational age at birth and birthweight <10th percentile ( Table 2 ).

Table 2

Neonatal outcomes by mode of delivery: per pregnancy analysis

Outcome	Study group: Vx and non-Vx pairs; trial of labor; mode of birth, VD-VD or VD-CD (n=224; 112 pairs) Rate	Control group 1: Vx and non-Vx pairs,mode of birth, CD-CD (n=216; 108 pairs)			Control group 2: first twin non-Vx; mode of birth, CD-CD (n=340; 170 pairs)
		Rate	Crude RR (95% CI, study vs control 1)	Adjusted RR (95% CI, study vs control 1) a	Rate	Crude RR (95% CI, study vs control 2)	Adjusted RR (95% CI, study vs control 2) a
Composite outcome b	47 (42)	40 (37)	1.13 (0.82–1.57)	0.93 (0.71–1.22)	58 (34)	1.23(0.91–1.66)	1.20 (0.92–1.58)
Neonatal death	26 (23)	26 (24)	0.96 (0.60–1.55)	0.90 (0.58–1.39)	34 (20)	1.16 (0.74–1.82)	1.22 (0.81–1.86)
Severe neurologic injury	29 (26)	24 (22)	1.17 (0.73–1.87)	1.01 (0.63–1.62)	35 (21)	1.26 (0.82–1.93)	1.18 (0.77–1.82)
Bronchopulmonary dysplasia	78 (72)	72 (69)	1.04 (0.88–1.24)	1.01 (0.90–1.13)	115 (69)	1.04 (0.89–1.21)	0.94 (0.84–1.06)
Necrotizing enterocolitis	17 (15)	16 (15)	1.02 (0.55–1.92)	0.99 (0.53–1.86)	27 (16)	0.96 (0.55–1.67)	0.92 (0.53–1.60)
5-min Apgar score <7	23 (21)	24 (22)	0.98 (0.85–1.12)	0.94 (0.84–1.05)	43 (25)	0.94 (0.83–1.07)	0.89 (0.79–1.01)
Mechanical ventilation	103 (92)	101 (94)	0.98 (0.91–1.06)	N/A	161 (95)	0.97 (0.91–1.04)	N/A

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Long-term effectiveness and safety of open Burch colposuspension vs retropubic midurethral sling for stress urinary incontinence—results from a large comparative study
2. Dichorionic twin-specific vs singleton growth references for diagnosis of fetal growth restriction
3. Services and payer mix of Black-serving hospitals and related severe maternal morbidity
4. Intracervical balloon catheter for labor induction after rupture of membranes: a systematic review and meta-analysis

Stay updated, free articles. Join our Telegram channel

Join