Facts Views Vis Obgyn, 2016, 8 (4): 223-231 Structured view

Delayed-interval delivery can save the second twin: evidence

from a systematic review

S. Feys, Y. Jacquemyn

Department of Obstetrics and Gynaecology, Antwerp University Hospital (UZA), Edegem, Belgium
Correspondence at: [email protected]

Abstract

Background: In case of preterm birth in twins, it is not well established if the second twin benefits from a delayed-
interval delivery.
Objective: The main objective of this systematic review is to evaluate survival benefit of the second twin from
delayed-interval delivery compared to the first twin. Secondly, we will evaluate the survival benefit of the procedure
when performed equal to or after 24 weeks gestational age of the first born.
Methods: Delayed interval delivery was defined as every attempt to perform a delayed-interval delivery with at
minimum placement of a high ligature of the umbilical cord and a delay of delivery of at least 24 hours.
Based on the PRISMA method, a systematic review was performed.
Controlled and observational studies reporting at least 3 cases of delayed interval delivery in dichorionic diamniotic
twin pregnancy describing the outcome of the first and the second twin were included. Case reports and papers on
triplet or higher order pregnancies were excluded.
Primary data included gestational age and outcome of the first and second born. Metadata concern management
strategies (tocolysis, antibiotics, cerclage), neonatal data (sex, birth weight and morbidity) and maternal
complications.
The methodological quality of included studies was assessed using the “IHE quality appraisal checklist for
assessing the quality of case series”. Meta-analysis was performed by computing relative risk (RR) with its 95%
confidence interval (CI) using the random-effects model. Statistical heterogeneity was tested using the I2 and Chi2
statistics.
Since there is no control group for the secondary outcomes, these are presented by narrative synthesis.
Results: Mortality data were extracted from 13 articles, reporting a total of 128 cases of delayed-interval delivery.
In the analysis, the second born had a significantly lower mortality risk compared to the first born (relative risk =
0.44, 95% confidence interval = 0.34 – 0.57, P<0.0001, I2 =0%, P=0.70).
For the analysis of mortality of the second born foetus versus the first born when the first delivery was at ≥24
weeks of gestational age, 12 articles were included. In the analysis 4 reports were excluded since there were no
events (no mortality) in both groups (first and second born) making analysis impossible. For the 36 cases included,
the second born had a significantly lower mortality risk compared to the first born if delivery of the first born
occurred at ≥ 24 weeks gestational age (relative risk=0.37, 95% confidence interval= 0.17 – 0.82, P=0.014, I2=0%,
P=0.82).
Conclusions and implications: In carefully selected twin pregnancies the survival of the second born twin may
improve with delayed-interval delivery, also if the first was born at or after 24 weeks. Management protocols in
the studies included vary, making it difficult to propose a uniform strategy for delayed-interval delivery. Families
must be informed about the possibility that a nonviable infant would survive to a periviable gestational age with a
risk of severe sequels after birth as well as the possibility of maternal complications.
Key words : Twin, delivery, dichorionic, tocolysis, cerclage, antibiotics.

223
Introduction weeks gestational age as this is still considered a
limit of viability in most high resource countries.
Rationale Secondary outcomes are neonatal morbidity in first
and second twins and maternal morbidity.
Twin pregnancies are associated with a higher risk
of preterm delivery and this at a significantly earlier Methods
gestational age than singletons, resulting in infant
morbidity and mortality. (Gardner et al., 1995) Protocol
In pregnancies presenting with extremely preterm
labour or rupture of the membranes, significant To ensure the accuracy of this review, we reported
prolongation of gestation and hence increase in it based on the Preferred Reporting Items for
foetal weight is expected to improve foetal outcome. Systematic Reviews and Meta-analyses (PRISMA)
For this reason, it can be tried to stop labour after statement. (Liberati, Altman et al., 2009) Methods
birth of the first foetus. This procedure is defined of the analysis and inclusion criteria were specified
as a delayed-interval delivery. It was first described in advance and documented in a review protocol.
by Carson in 1880 with a wait-and-see management
(Carson, 1880). Currently, obstetrical management Eligibility criteria
has changed to active management using tocolysis,
prophylactic antibiotics and cerclage. Currently Study design
delayed interval delivery is defined as every attempt
to postpone birth of the second twin with at least Controlled and observational studies were included.
placement of a high ligature of the umbilical cord Case series were included when reporting at least
and a delay of delivery of at least 24 hours. 3 cases. Case reports were excluded. There was no
Even though a number of case reports and case limit on the year of publication.
series have been published describing attempts
to delay delivery of remaining twins, triplets and Participants
higher-order multiples after immature delivery,
intentional delayed delivery of the second foetus Only dichorionic diamniotic twin pregnancies were
in twin pregnancies is of very rare occurrence. To included. Triplet or higher order pregnancies were
our knowledge no previous systematic review was excluded.
performed evaluating the impact of the procedure on
the mortality of the second born compared with the Interventions
first born, nor on neonatal and maternal morbidity.
Furthermore there is no consensus about the Every attempt to perform a delayed-interval delivery
effectiveness of the procedure after a specific was included since there is no agreement regarding
gestational age. This depends also on procedures the best management strategy for this procedure.
relating to neonatal resuscitation on the limits At least high ligature of the umbilical cord should
of viability, e.g. in Flanders (Belgium), active have been performed and a delay of at least 24 hours
management of the new-born is offered from, but achieved.
not before, 24 weeks of pregnancy, meaning that The administration of tocolysis, the use of
delaying delivery after 24 weeks highly impacts antibiotics and the placement of cerclage were
neonatal treatment. (Jacquemyn et al., 2014) registered if mentioned in the included article.

Objectives Comparators- outcome

The main objective of this systematic review is to Mortality rate of the first born was compared with
evaluate survival benefit of the second twin from the second born. Secondly we examined if there is a
delayed-interval delivery compared to the first twin. difference in mortality between the first and second
We will review publications describing delayed- born if the first twin was born at or after 24 weeks
interval delivery in dichorionic diamniotic twins. of gestational age.
The primary outcome measure is mortality of the first
and second twin. When available, full information Language
on gestational age will be registered to complete
an additional analysis to evaluate the benefit of Only articles reported in the English and Dutch
the procedure when performed after or equal to 24 language were included.

224 Facts Views Vis Obgyn

Information sources for assessing the quality of case series”. For this
checklist ‘quality’ covers both risk of bias and
Searching an electronic database identified reports quality of reporting. (Guo et al., 2015)
and scanning reference lists of articles. The search
was applied to MEDLINE database using PubMed Summary measures and synthesis of results
up to November 2014.
Relative risk of mortality was the primary measure of
Search treatment effect. The meta-analysis was performed
by computing relative risk (RR) with its 95%
The following terms were used: (all fields) “delayed confidence interval (CI) using the random-effects
delivery”, twin AND delayed interval delivery, model as we expect the interventions in the studies
“interval delivery” AND multiple gestation, delayed to differ in ways that have impact on the results.
interval delivery AND multiple gestation. Therefore we could not assume a common effect
size. (Borenstein et al., 2009) The DerSimonian-
Study selection Laird method was used for the random effects
model.
Titles were screened first. Secondly, abstracts were Statistical heterogeneity was tested using the
read and checked if they accorded with the inclusion I2 and Chi2 statistics. Heterogeneity was found
criteria. Full reports were obtained for the titles and substantial if an I2 was greater than 30% or if the
abstracts that appeared to meet the inclusion criteria, p-value in the Chi2 test for heterogeneity was less
as well as those that were uncertain. Thereafter, than 0.10.
the full text reports were screened and decision Statistical analysis was carried out using the
was made whether these reports met the inclusion statistical software StatsDirect 3.0.
criteria.
Constraints were inaccessible full reports. When Risk of bias across studies
there were reports using the same database, only the
report providing the most applicable information Funnel plots were used to explore the presence
was used. of publication bias. The degree of funnel plot
asymmetry was assessed by Begg’s and Egger’s
Data items test, p-value with a significant level at 0.05. This
statistical analysis was performed using StatsDirect
In the excel database following data were registered: 3.0.
— author and year of publication
— interval (days) Additional analyses
— days of gestation first and second born
— route of delivery of second born (vaginally or Since there is no control group for the secondary
caesarean) outcomes, these results are presented by narrative
— outcome of first and second born (APGAR score synthesis.
at 5’)
— tocolysis: product, indication (standard or Results
therapeutic)
— antibiotics: product, indication (standard or Study selection (Fig. 1)
therapeutic)
— cerclage: performed or not after delivery of the A total of 13 studies were identified for inclusion,
first twin answering our primary research question. For
— any maternal complications during the period our secondary research question, 12 studies were
after the first delivery (all information provided) included. The MEDLINE database search using
— neonatal morbidity (all information on foetal PubMed and reference lists provided a total of 196
morbidity provided) citations. After adjusting for duplicates 103 studies
— foetal sex of the first and second born remained. Of these, 76 were discarded because
— weight of the first and second born they did not meet the inclusion criteria, based on
their abstract (23 abstracts about triplets or higher
Risk of bias in individual studies order multiples, 46 case-reports and seven had no
full-text available in English or Dutch). Full text of
The methodological quality of included studies was the remaining 27 citations was examined in detail.
assessed using the “IHE quality appraisal checklist Again, 14 reports were discarded. Five because

DELAYED-INTERVAL DELIVERY CAN SAVE THE SECOND TWIN – FEYS et al. 225
 et al., 2003) The included case series did not specify
a time period in which they registered all cases. This
can result in a source of bias where unsuccessful
delayed interval delivery intentions are not reported.
The primary outcome of the case-control study
was neonatal morbidity compared to a group
of singletons born at the same gestational age.
(Rosbergen et al., 2005)

Participants

Within the 13 studies 173 interval delivery

procedures were discussed, 132 were twin
pregnancies, 33 were triplets, 2 quadruplets and 1
quintuplet. Of the 132 discussed twin pregnancies
a total of 128 twins were included. Four were
excluded because the interval achieved was <24
hours.
Between studies, the maximum gestational age of
the first foetus when delayed-interval delivery was
performed, differed. For Kalchbrenner et al. (1998)
the delivery of the first foetus should be between
18 and 28 weeks of gestation, for Rosbergen et al.
(2005) before 32 weeks of gestation, for Petousis
et al. (2012) before 24 weeks of gestation and for
Fig. 1. — Flow diagram of study selection Farkouh et al. between 16 and 29 weeks of gestation
(Farkouh et al., 2000).
full text of the study was not available. Nine more
citations were discarded for different reasons: Intervention
one concerning a procedure other than delayed-
interval delivery, three had a database with many There is no agreement regarding the best
assumptions, one was a review of literature, two management strategy for this procedure. Inclusion
examined other outcomes and in two times articles criteria for intervention was performing an attempt
results of database were re-used. No unpublished of delayed-interval delivery with at least high
relevant studies were obtained. ligature of the umbilical cord being performed and
obtaining a delay of at least 24 hours.
Study characteristics There was registration for the use of tocolysis
and antibiotics and the placement of cerclage.
Methods Tocolysis was administered prophylactically
in 8 reports (Doger et al., 2014; Petousis et al.,
Of the 13 studies finally selected for this review nine 2012; Arias, 1994; Kalchbrenner et al., 1998;
were retrospective observational studies (Doger et van Doorn et al., 1999; Farkouh et al., 2000; Van
al., 2014; Padilla-Iserte et al., 2014; Petousis et al., der Straeten et al., 2001; Rosbergen et al., 2005),
2012; Arias 1994; Kalchbrenner et al., 1998; van variably prophylactic or therapeutic use in 3 reports
Doorn et al., 1999; Farkouh et al., 2000; Fayad et (Wittmann et al., 1992; Weemhoff et al., 2001;
al., 2003; Cristinelli et al., 2005), three were case Cristinelli et al., 2005), not discussed in one report
series (Wittmann et al., 1992; Van der Straeten et (Padilla-Iserte et al., 2014) and centre dependent
al., 2001; Weemhoff et al., 2001) and one study without discussing the modalities in one report
was a case-control study design (Rosbergen et al., (Fayad et al., 2003). Different types of tocolysis
2005). The retrospective studies described a specific were used, depending on local procedures.
period in which all of the cases with delayed In most cases antibiotics were administered
interval delivery registered were reported. One of prophylactically (Doger et al., 2014; Padilla-Iserte
these retrospective studies was a multicentric study, et al., 2014; Arias, 1994; Kalchbrenner et al., 1998;
where an information letter was sent to 20 hospitals van Doorn et al., 1999; Farkouh et al., 2000; Van der
enquiring about cases of an interval of more than Straeten et al., 2001; Weemhoff et al., 2001; Fayad
48hours. There was response of 12 hospitals. (Fayad et al., 2003; Rosbergen et al., 2005). In two reports

226 Facts Views Vis Obgyn

some patients received antibiotics prophylactically Results of individual studies
and some therapeutically (Wittmann et al., 1992;
Cristinelli et al., 2005) and in one report, antibiotics Results of individual studies, relative risk and
were only administered in case of preterm premature confidence intervals (CI) for mortality of the second
rupture of membranes (PPROM). When antibiotics born versus the first born are presented in Table I.
were started prophylactically, the duration was very Funnel plot is presented in Figure 2.
variable. In two reports they were administered for Results of individual studies, relative risk and
a few days with culture-directed switch according confidence intervals for mortality of the second born
to cervico-vaginal cultures (Padilla-Iserte et al., versus the first born when delivery of the first was
2014; Rosbergen et al., 2005). The type of antibiotic ≥ 24weeks gestational age are presented in Table II.
therapy was generally a broad-spectrum antibiotic. Funnel plot is presented in Figure 3.
In four reports cerclage placement was performed
in all cases (Petousis et al., 2012; Arias, 1994; Synthesis of results
Kalchbrenner et al., 1998; Farkouh et al., 2000).
It was dependent on preference of patient and Mortality data were studied in 13 articles, reporting
clinician in seven reports (Doger, Cakiroglu et al., a total of 128 cases of delayed-interval delivery.
2014; Wittmann et al., 1992; van Doorn et al., 1999; In the analysis, the second born is associated with
Van der Straeten et al., 2001; Fayad et al., 2003; significantly lower mortality compared to the first
Cristinelli et al., 2005). In one report it was never born (relative risk = 0.44, 95% confidence interval
placed (Rosbergen et al., 2005) and in one report it = 0.34 – 0.57, p<0.0001). There is no evidence of
was not mentioned (Padilla-Iserte et al., 2014). heterogeneity (I2 =0%, p=0.70).
For the analysis of mortality second born versus
Outcomes first born when the first delivery was ≥24 weeks
of gestational age 12 articles were included. In the
Primary analysis 4 reports were excluded since there were
no events (no mortality) in both groups (first and
In all studies mortality of the first and second born second born) and the analysis for those reports was
twin was registered. Some studies had a different not possible. For the 36 cases included, the second
primary outcome but they were included if data on born is associated with significantly lower mortality
mortality were mentioned. compared to the first born (relative risk=0.37, 95%
Fayad et al did not mention the term of delivery confidence interval= 0.17 – 0.82, p=0.014). There is
of the first and second baby, for the second analysis no evidence of heterogeneity (I2=0%, p=0.82).
this report was excluded. (Fayad et al., 2003)
Risk of bias across studies
Secondary
Funnel plot, Begg’s test and Egger’s test were used
Only Kalchbrenner et al. specified the maternal
complications they would screen for in advance so to explore the publication bias. The funnel plots
possibly there is a reporting bias. (Kalchbrenner et were symmetrical in general for both analyses.
al., 1998) Also for neonatal morbidity there was The Begg’s test and Egger’s test showed no
no unambiguous registration. Only Rosbergen evidence for publication bias in meta-analyses. For
et al. (2005) had a clear registration on neonatal the first analysis Begg’s test p=0.11 and Egger’s test
morbidity. p=0.11. For the second analysis Begg’s test p=0.88,
For the registration of maternal and neonatal Egger’s test p=0.24.
morbidities we used the same reports as for the
first analysis with that difference that higher order Additional analysis
multiples were also included. The goal of this
registration is to mention the possible morbidities Only Kalchbrenner et al. (1998) specified in
rather in a descriptive than a quantitative way, so advance the maternal complications they would
our results can function as a starting point for future screen for. The lack of such research question can
research. cause reporting bias.
There were 4 reports that did not report on
Risk of bias within studies maternal morbidity. For those we do not know if
there were no complications or if they were just not
The quality of the methodology of included studies mentioned. (Petousis et al., 2012; Wittmann et al.,
was assessed using the “IHE quality appraisal 1992; Fayad et al., 2003; Rosbergen et al., 2005) In
checklist for assessing the quality of case series”. total, 90 pregnant women were included. There were

DELAYED-INTERVAL DELIVERY CAN SAVE THE SECOND TWIN – FEYS et al. 227
 Table I. — Results of individual studies, mortality of the second born versus the first born.
95% Confidence %
Study (N, number of twin Relative
A* B* C* D* Interval (CI, Weights
pregnancie) Risk
Koopman) (random)
Wittmann 1992 (4) 0 4 4 0 0.11 0.00 062 0.96
Arias 1994 (8) 3 7 5 1 0.40 0.15 0.96 7.76
Kalchbrenner 1998 (5) 0 1 5 4 0.33 0.00 3.29 0.75
Van Doorn 1999 (7) 3 4 4 3 0.75 0.25 2.15 5.89
Farkouh 2000 (20) 8 17 12 3 0.47 0.25 0.78 20.94
Weemhoff 2001 (3) 1 2 2 1 0.50 0.08 2.45 2.11
Van der Straeten 2001 (5) 0 4 5 1 0.11 0.00 0.62 0.93
Fayad 2002 (28) 6 26 22 2 0.23 0.11 0.43 13.13
Rosbergen 2004 (24) 8 15 16 9 0.53 0.27 0.98 16.21
Cristinelli 2005 (4) 1 3 3 1 0.33 0.06 1.46 2.11
Petousis 2012 (5) 1 5 4 0 0.27 0.04 0.81 3.41
Doger 2014 (11) 6 11 5 0 0.57 0.29 0.91 23.68
Padilla-Iserte 2014 (4) 1 3 3 1 0.33 0.06 1.46 2.11
* A = exposed positive = second born twin with Apgar Score at 5 minutes (AS 5’) = 0
* B = control positive = first born twin with AS 5’ = 0
* C = exposed negative = second born twin with AS 5’ > 0
* D = control negative = first born twin with AS 5’ > 0

Table II. — Results of individual studies, mortality of the second born versus the first born, when the first delivery was at ≥24
weeks gestational age.
95% Confidence %
Study (N= number of Relative
A* B* C* D* Interval (CI, Weights
twins in study) Risk
Koopman) (random)
Wittmann 1992 (2) 0 1 1 0 0.33 0.00 3,02 9,63
Arias 1994 (1) 0 0 1 1 NA NA NA NA
Kalchbrenner 1998 (4) 0 0 4 4 NA NA NA NA
Van Doorn 1999 (5) 2 2 3 3 1.00 0.23 4.33 26.75
Farkouh 2000 (7) 0 4 7 3 0,11 0 0.72 8.14
Weemhoff 2001 (1) 0 0 1 1 NA NA NA NA
Van der Straeten 2001 (7) 0 2 3 1 0.20 0.00 1.25 8.45
Rosbergen 2004 (18) 2 7 14 9 0.29 0.07 0.99 30.99
Cristinelli 2005 (2) 0 1 2 1 0.33 0.00 2.68 8.02
Doger 2014 (1) 1 1 0 0 NA NA NA NA
Padilla-Iserte 2014 (2) 0 1 2 1 0,33 0.00 2.68 8.02
* A = exposed positive = second born twin with Apgar Score at 5 minutes (AS 5’) = 0
* B = control positive = first born twin with AS 5’ = 0
* C = exposed negative = second born twin with AS 5’ > 0
* D = control negative = first born twin with AS 5’ > 0
NA: Not Applicable

28 cases of chorioamnionitis and one was followed (2.2%) and psychological problems in one case.
by sepsis and septic shock. Placental solutio was There were no cases of maternal death reported.
mentioned four times with one leading to excessive As well for neonatal morbidity, systematic
blood loss of more than 2 litres. In total 3 patients registration was lacking in the include studies. In
were reported to have severe bleeding. More rare most reports there was no protocol for registration.
was thromboflebitis that was mentioned in 2 cases Only Rosbergen et al. (2005) and Kalchbrenner

228 Facts Views Vis Obgyn

et al. (1998) had a clearly specified registration Discussion
strategy on neonatal morbidity. Furthermore
Rosbergen et al. used a control group for the Summary of evidence
second born neonates, to evaluate morbidity due to
prematurity or the procedure, concluding the long- The main objective of this review was to evaluate
term outcome was comparable to children with the survival benefit of the second twin from delayed-
same gestational age. interval delivery compared to the first twin. A

Fig. 2. — Forest plot showing relative mortality risk of second born twin after delayed-interval
delivery compared to first born twin.

Fig. 3. — Forest plot showing relative mortality risk of second born twin after delayed-interval
delivery compared to first born twin, with first born delivery at ≥ 24 weeks of gestation.

DELAYED-INTERVAL DELIVERY CAN SAVE THE SECOND TWIN – FEYS et al. 229
literature search resulted in 13 reports, mainly language. Another important remark is the variation
retrospective case series. A meta-analysis for in year of publication. This diverges from 1992 to
mortality was performed using data of 128 twins. 2014. Since this is a long period it is conceivable
This analysis showed that the second born has a that improvement of obstetric and neonatal care
significantly lower mortality risk compared to the can cause different outcome of the included patient
first born (relative risk = 0.44, 95% CI 0.34-0.57, group nowadays. There was also an exclusion of
p<0.0001). unsuccessful procedures since the delay time was
Secondly we analysed if there was a benefit if this set at least 24 hours.
procedure was performed when the first twin was
delivered at or after 24 weeks gestational age. Thirty- Conclusion
six cases were included, showing a significantly
lower mortality risk for the second born (relative Implications for practice
risk =0.37, 95% CI 0.17-0.82, p=0.014). Although
risk of bias analysis using Begg’s and Egger’s From this review we can conclude that carefully
test showed no publication bias, results should be selected twin pregnancies may benefit from delayed-
interpreted with caution. interval delivery as this improves the survival of the
For the registration of maternal as well as second twin. Protocols however vary enormously
neonatal morbidity only respectively two and one so there is absence of agreement regarding the best
publication mentioned the reported outcomes in management of these pregnancies. Families must
advance. It is difficult to draw conclusions because be informed about the possibility that a nonviable
when a reporting protocol is lacking, the variety of infant would survive to a periviable gestational age
outcomes reported is large. Our goal was to mention with a risk of severe sequels after birth as well as the
the possible morbidities in a descriptive rather than possibility of maternal complications.
a quantitative way, facilitating future research.
Implications for research
Limitations
Large multicentre studies should be performed
Outcome level
investigating the best management strategy as well
as the neonatal and maternal morbidity. These studies
The meta-analysis reported here combines data
should randomize between attempting delayed
across studies in order to estimate treatment effects
delivery or not and in case of a delay procedure
with more precision than is possible in a single
at least for the use of tocolytics, and preferably
study. A major limitation of this meta-analysis is that
also for antibiotics and placement of cerclage. For
the management strategies did not follow a uniform
pattern, such as the performance of cerclage, morbidity investigation a registered study protocol
administration of antibiotics and tocolysis. Studies should be used as well as a control group.
also differed in reporting maternal and neonatal
morbidity not using standardized checklists. References

Study level Arias F. Delayed delivery of multifetal pregnancies with

premature rupture of membranes in the second trimester.
Am J Obstet Gynecol. 1994;170: 1233-7.
Publication bias is a potential problem in Borenstein M, Hadger LV, Higgins JPT and Rothstein HR
systematically reviewing case series. Positive (eds). Introduction to meta-analysis. John Wiley & Sons,
results tend to be published more frequently than West Sussex, United Kingdom, 2009
Carson JL. Twins born with an interval of forty-four days. Br
negative results. In the retrospective case series
Med J. 1880;1:242.
included in our review, any failed attempts at Cristinelli S, Fresson J, André M et al. Management of delayed-
performing delayed-interval delivery may not have interval delivery in multiple gestations. Fetal Diagn Ther.
been recorded in the notes, with only the successful 2005;20:285-90.
attempts recorded. In all but two reports, cases were Doger E, Cakiroglu Y, Ceylan Y et al. Obstetric and neonatal
outcomes of delayed interval delivery in cerclage and non-
registered during a period that was specified in cerclage cases: An analysis of 20 multiple pregnancies. J
advance. This should have minimized publication Obstet Gynaecol Res. 2014;40:1853-61.
bias. Farkouh LJ, Sabin ED, Heyborne KD et al. Delayed-interval
delivery: extended series from a single maternal-fetal
Review level medicine practice. Am J Obstet Gynecol. 2000;183:1499-
1503.
Fayad S, Bongain A, Holhfeld P et al. Delayed delivery of
Limitation on review level was the use of only one second twin: a multicentre study of 35 cases. Eur J Obstet
online database and restriction to English and Dutch Gynecol Reprod Biol. 2003;109:16-20.

230 Facts Views Vis Obgyn

Gardner MO, Goldenberg RL, Cliver SP et al. The origin and Petousis S, Goutzioulis A, Margioula-Siarkou C et al.
outcome of preterm twin pregnancies. Obstet Gynecol. Emergency cervical cerclage after miscarriage of the first
1995;85: 553-7. fetus in dichorionic twin pregnancies: obstetric and neonatal
Guo B, Moga C, Harstall C et al. A principal component analysis outcomes of delayed delivery interval. Arch Gynecol .
is conducted for a case series quality appraisal checklist. J 2012;286:613-7.
Clin Epidemiol 2015;69:199-207. Rosbergen M, Vogt HP, Baerts W et al. Long-term and short-
Jacquemyn Y, Beckstedde I, et al. Consensustekst perinatale term outcome after delayed-interval delivery in multi-fetal
zorgen rond levensvatbaarheid in Vlaanderen. Tijdschr pregnancies. Eur J Obstet Gynecol Reprod Biol. 2005;122:
Geneeskd. Tijdschr voor Geneeskunde 2014; 70: 1159-67. 66-72.
Kalchbrenner MA, Weisenborn EJ, Chyu JK et al. Delayed Van der Straeten FM, De Ketelaere K, Temmerman M. Delayed
interval delivery in multiple pregnancies. Eur J Obstet
delivery of multiple gestations: maternal and neonatal
Gynecol Reprod Biol. 2001;99: 85-9.
outcomes. Am J Obstet Gynecol. 1998;179: 1145-9.
van Doorn HC, van Wezel-Meijler G, van Geijn HP et al. Delayed
Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement
interval delivery in multiple pregnancies. Is optimism
for reporting systematic reviews and meta-analyses of justified? Acta Obstet Gynecol Scand. 1999;78:710-5.
studies that evaluate health care interventions: explanation Weemhoff M, van Meir CA, Walther FJ et al. Delayed birth
and elaboration. J Clin Epidemiol. 2009;62:e1-34. of the second child in multiple gestation. Ned Tijdschr
Padilla-Iserte P, Vila-Vives JM, Ferri B et al. Delayed interval Geneeskd. 2001;145: 1377-80.
delivery of the second twin: obstetric management, neonatal Wittmann BK, Farquharson D, Wong GP et al. Delayed
outcomes, and 2-year follow-up. J Obstet Gynaecol India. delivery of second twin: report of four cases and review of
2014;64:344-8. the literature. Obstet Gynecol. 1992;79: 260-3.

DELAYED-INTERVAL DELIVERY CAN SAVE THE SECOND TWIN – FEYS et al. 231