- Email: [email protected]
Lower perinatal mortality in preterm born twins than in singletons: a nationwide study from The Netherlands
Accepted Manuscript Lower perinatal mortality in preterm born twins than in singletons; a nationwide study from the Netherlands Blanka Vasak, MD, Jessica J. Verhagen, MD, Steven V. Koenen, PhD, MD, Maria P.H. Koster, PhD, MD, P.A.O.M. de Reu, PhD, Arie Franx, PhD, MD, Jan G. Nijhuis, PhD, MD, Gouke J. Bonsel, PhD, MD, Gerard H.A. Visser, PhD, MD PII:
S0002-9378(16)30868-7
DOI:
10.1016/j.ajog.2016.10.005
Reference:
YMOB 11335
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 13 July 2016 Revised Date:
16 September 2016
Accepted Date: 3 October 2016
Please cite this article as: Vasak B, Verhagen JJ, Koenen SV, Koster MPH, de Reu PAOM, Franx A, Nijhuis JG, Bonsel GJ, Visser GHA, Lower perinatal mortality in preterm born twins than in singletons; a nationwide study from the Netherlands, American Journal of Obstetrics and Gynecology (2016), doi: 10.1016/j.ajog.2016.10.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 1
TITLE PAGE
2
Lower perinatal mortality in preterm born twins than in singletons; a nationwide study from the Netherlands
3
RI PT
4
Blanka Vasak1 (MD), Jessica J. Verhagen1 (MD), Steven V. Koenen1 (PhD, MD), Maria P.H. Koster1
6
(PhD, MD), P.A.O.M. de Reu2 (PhD), Arie Franx 1 (PhD, MD), Jan G. Nijhuis3 (PhD, MD), Gouke J.
7
Bonsel1 (PhD, MD), Gerard H.A. Visser1 (PhD, MD)
SC
5
8 1
Division Woman and Baby, University Medical Center Utrecht, The Netherlands
10
2
Independent midwife, retired, no current affiliation
11 12
3
M AN U
9
Department of Obstetrics and Gynaecology, GROW-School for Oncology and Developmental Biology, Maastricht University Medical Centre, Maastricht, The Netherlands
13
TE D
Conflict of interest: The authors report no conflict of interest.
Acknowledgment of financial support: Vrienden UMC Utrecht provided financial support.
EP
Paper presentation This study has been presented at the 49th Annual Gynaecongres, Eindhoven, the Netherlands, May 20, 2016.
AC C
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Table or Figure that needs to be included in the Print Version Figure 1
Corresponding author: B. Vasak, MD.
Work address: University Medical Center Departement Obstetrics, Huispostnummer KE 04.123.1 PO Box 85090
ACCEPTED MANUSCRIPT 2 35 36 37 38 39 40 41 42 43 44 45
Lundlaan 6, 3584 EA Utrecht The Netherlands
46
Word count main text: 2673
47
Word count abstract: 278
48 49 50 51 52 53 54 55
Condensation of the paper Overall perinatal mortality is higher in twin pregnancies than in singleton pregnancies, but during the preterm period, mortality is considerably lower in twins.
RI PT SC
+31 88 75 539 23 +31 6 41 38 52 31 +31 88 75 553 20
M AN U
Telephone: Work: Private: Fax:
EP
TE D
Short version article title "Lower perinatal mortality in preterm born twins than in singletons."
AC C
56
Email: [email protected]
ACCEPTED MANUSCRIPT 3
Abstract
58
Background: Twin pregnancies are at increased risk for perinatal morbidity and mortality, due to many factors including a high incidence of preterm delivery. Compared to singleton pregnancies overall perinatal mortality risk is higher in twin pregnancies, however for the preterm period perinatal mortality has been reported to be lower in twins.
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
Objective: To compare perinatal mortality in relation to gestational age at birth between singleton and twin pregnancies, taken into account socioeconomic status, fetal sex and parity. Study design: We studied perinatal mortality according to gestational age at birth in 1.502.120 singletons and 51.658 twins without congenital malformations who were delivered between 2002 and 2010 after 28 weeks of gestation. Data were collected from the nationwide Netherlands Perinatal Registry.
SC
62
Results: Overall perinatal mortality in twins (6.6/1000 infants) was higher than in singletons (4.1/1000 infants). However, in the preterm period perinatal mortality in twins was substantially lower than in singleton pregnancies, 10.4 per 1000 infants as compared to 34.5 per 1000 infants, respectively for infants born < 37 weeks; this held especially for antepartum deaths. After 39 weeks of pregnancy perinatal mortality was higher in twins. Differences in parity, fetal sex and socioeconomic status did not explain the observed differences in outcome.
M AN U
61
Conclusion: Overall perinatal mortality was higher in twins than in singletons, which is most likely caused by the high preterm birth rate in twins and not by a higher mortality rate for gestation, apart from term pregnancies. During the preterm period, antepartum mortality was much lower in twins than in singletons. We suggest that this might partially be due to a closer monitoring of twins, which indirectly suggests a need for closer surveillance of singleton pregnancies.
TE D
60
EP
59
RI PT
57
80
Key words
82
Gestational age, Perinatal mortality, Preterm mortality, Singleton pregnancies, Twin pregnancies
83 84 85 86 87
AC C
81
ACCEPTED MANUSCRIPT 4
Introduction
89
Twin pregnancies are at increased risk for preterm birth, intrauterine growth restriction
90
and a number of multiplet related conditions. Monochorionic twins have additional risks
91
for mortality and morbidity primarily due to the twin-to-twin transfusion syndrome
92
(TTTS) and congenital abnormalities.1-5
93
While registry data in Western countries (PERISTAT) confirm a crude 2-3 fold increased
94
perinatal mortality risk of a twin pregnancy,6 only few reports have compared mortality
95
and morbidity between twin and singleton pregnancies taking into account differences
96
in gestational age at delivery. They all found a lower perinatal mortality in twins born
97
before 37 weeks.7-9 These studies were all small, with the exception of one (Kahn) and
98
did not include separation between zygosity and chorionicity in twins, nor did they take
99
into account known prognostic important factors such as parity and socioeconomic
M AN U
SC
RI PT
88
status (SES) of the mother and sex of the child.
101
The objective of this study was to compare perinatal mortality in relation to gestational
102
age at birth between singleton and twin pregnancies, taken into account parity, SES,
103
fetal sex and mode of delivery. For this study we used a large national anonymized data
104
set from the Netherlands Perinatal Registry (Perined).10
EP
TE D
100
AC C
105 106
Materials and methods
107
General
108
This was a retrospective population-based cohort study. We obtained, after prior
109
permission from the Perinatal Registry (Perined),11 aggregated data and analyses on all
110
births in the Netherlands between 2002 and 2010. This registry contains information on
ACCEPTED MANUSCRIPT 5 111
approximately 95% of all births from 16.0 weeks gestational age onwards, where
112
coverage of twins is even higher. Details of this dataset can be found elsewhere.10
113
Variables
115
The Perined dataset includes routine maternal data on age, parity, socioeconomic
116
background (zip-code based), and ethnic background. Information is available on
117
intervention (induction of labor, mode of delivery (instrumental, elective and emergency
118
cesarean delivery), NICU admission of the child and specific features of the delivery.
119
Data on the child comprise of: singleton or multiplet, sex of the child, presence of
120
congenital anomalies, gestational age at birth, birth weight, and Apgar score at 5
121
minutes. From birth weight, sex and gestational age, the birth weight is computed as
122
centile score.12
123
The size of the multiplet is recorded and for each child its rank number during birth. No
124
information is available on zygosity/chorionicity. Congenital abnormalities were
125
recorded if they were present and recognized, either at birth or at first NICU admission
126
by the neonatologist. This includes any abnormality already noticed during antenatal
127
ultrasound scanning or, occasionally, with genetic tests. All children with congenital
128
abnormalities were excluded. Gestational age was calculated from the first day of the
129
last menstrual period or from an ultrasound dating scan by measurement of the Crown-
130
Rump Length in the first trimester. In this period 91% of the pregnancies had a reliable
131
dating, obtained by a first trimester dating scan.10 For twins conceived spontaneously
132
dating was based on the mean CRL of both fetuses. Pregnancies achieved by assisted
133
reproductive technology (ART) were dated based on the ART-derived gestational age.
134
In this report a gestational age of 37 weeks means: 37 weeks + 0 days - 37 weeks + 6
135
days containing whole week groups. For this analysis births between 28+0 and 43+6
AC C
EP
TE D
M AN U
SC
RI PT
114
ACCEPTED MANUSCRIPT 6
gestation were included, to allow comparison with data from other countries in which
137
registration practices may differ at earlier gestational ages.6 Children with missing
138
gestational age at birth were excluded from the analysis. A total of 1839 children from a
139
twin pregnancy were excluded because a twin pair could not be linked or information
140
on congenital malformations or gestational age of one of the twins was missing. In this
141
group there were 42 perinatal deaths. If one of the twins died before 28 weeks gestation
142
both children were excluded from analysis. If one of the twins died after 28 weeks
143
gestation and the pregnancy continued for the benefit of the other twin, the gestational
144
age at birth was registered for both twins, the exact gestational age of death for the
145
demised twin was not available in this database.
146
Perinatal mortality was defined as fetal or neonatal death up to 7 days after birth:
147
antepartum mortality was defined as death occurring before labor, intrapartum mortality
148
as death during labor, and neonatal mortality as death occurring within 0–7 days
149
following live birth. This information is always available in all cases.
150
Statistical analysis
151
Perinatal mortality was expressed as rate (number of deaths per 1000 infants, including
152
stillbirths). Mortality rate was subdivided to time of occurrence: antepartum, intrapartum
153
and neonatal. Perinatal mortality was compared between singleton and twin, according
154
to gestational age. The mortality rates at 42 and 43 weeks gestation were not displayed
155
owing to the extremely small number of cases. Comparisons were made between
156
groups by using multivariate logistic regression with gestational age in the model or chi
157
square test.
158
The mode of delivery in singletons and twins was studied. Vaginal delivery was divided
159
in spontaneous vaginal delivery and instrumental delivery defined as delivery by forceps
160
or a vacuum device. Cesarean delivery was divided in planned cesarean section, defined
AC C
EP
TE D
M AN U
SC
RI PT
136
ACCEPTED MANUSCRIPT 7
as a scheduled or elective cesarean, and an emergency cesarean defined as any
162
cesarean that was not planned or scheduled.
163
To investigate the influence of parity, maternal SES, and child sex, we calculated
164
perinatal mortality rates for each parameter for twins and singletons: for parity (0 vs.
165
1+), SES (p<20 poor vs. p>20) and child sex (male/female). For classification of SES, the
166
"socio-economic status score" was used as a proxy. This score takes into account the
167
average income per household in a given four-digit postcode area and the percentage
168
of households with low income, without paid job and with low education level.13,14 The
169
level of SES was categorized as low if the score was under the 20th centile.
170
No information on zygosity was available in the data set. The only information we could
171
obtain on zygosity indirectly were the twin pairs with unequal gender; male-female
172
twins, these twins are per definition dizygotic and dichorial.
M AN U
SC
RI PT
161
TE D
173
Results
175
We studied 1.502.120 singletons and 51.658 twins born between 2002 and 2010 after 28
176
weeks of gestation. In singleton pregnancies there were 6.087 perinatal deaths
177
(4.1/1000). For twins a total of 340 deaths occurred resulting in a perinatal mortality
178
rate of 6.6 per 1000 infants. Perinatal mortality according to gestational age is displayed
179
in table 1.
180
Of all twins 46% were born before 37 weeks of gestation, in contrast to only 5% in the
181
singletons. For deliveries before 37 weeks perinatal mortality rate in twins was 10.4 per
182
1000 infants compared to 34.5 per 1000 infants in singletons. For infants born before 34
183
weeks mortality rates were 17.2 per 1000 infants in twins and 93.6 per 1000 infants in
184
singletons.
AC C
EP
174
ACCEPTED MANUSCRIPT 8
In singleton pregnancies, perinatal mortality decreased with increasing gestational age,
186
with the lowest mortality rate at 40 weeks gestational age. In twin pregnancies mortality
187
also decreased with increasing gestational age with the lowest mortality rate at 38
188
weeks and an increase thereafter (Fig 1). In the preterm period, perinatal mortality in
189
twins was substantially lower than in singleton pregnancies. After 39+0 weeks mortality
190
was higher in twins.
191
In figure 2 perinatal mortality is displayed according to time period of occurrence of
192
death. In both singletons and twins the majority of perinatal deaths occurred in the
193
antepartum period, followed by the neonatal period and the intrapartum period,
194
respectively. In singletons, the distribution of the time period of death was, 72 %
195
antepartum, 9% intra partum and 19% neonatal; for twins this distribution was 59%
196
antepartum, 6% intra partum and 35% neonatal. Antepartum mortality was significantly
197
lower in twins as compared to singletons during the whole preterm period (p<0.001).
198
After 40 weeks gestation antepartum mortality was significantly higher for twins (p 0.02).
199
Intrapartum mortality was significantly lower between 31-33 weeks of gestation but did
200
not differ in the term period. Neonatal mortality was more or less similar during the
201
whole preterm period, mortality rates were higher for twins in the term period. Twins
202
were more often delivered by cesarean section than singletons (Table 2).
203
In both singletons and twins, nulliparity was related to higher overall mortality rates.
204
Poor maternal socioeconomic status (SES
205
overall perinatal mortality rates in singletons, in twins rates were also higher in these
206
groups but this was not significant (Table 3). Figure 3 shows twin and singleton mortality
207
rates per parameter for gestational age. Within the singletons and twin groups mortality
208
rates did not differ for parity, fetal sex and socioeconomic status.
209
Overall perinatal mortality was slightly lower in dizygotic twins of unequal gender as
210
compared to all twins and this held for most gestational ages (Table 1 & Fig 4). For
AC C
EP
TE D
M AN U
SC
RI PT
185
ACCEPTED MANUSCRIPT 9 211
infants born preterm mortality rates were lower in dizygotic unequal gender twins than
212
in singletons.
RI PT
213
Comment
215
This study shows that, in a large contemporary population of infants without congenital
216
abnormalities born after 28 weeks' gestation, overall perinatal mortality was higher in
217
twins compared to singletons, which is in accordance with literature.5
218
perinatal mortality in preterm twins was considerably lower than in singletons. For
219
antepartum mortality this held for the whole preterm period and for intrapartum
220
mortality between 31 and 33 weeks, but not for neonatal mortality. Data did not change
221
when parity, sex, socio-economic status were taken into account.
222
An explanation for the lower preterm perinatal death rates in twins as compared to
223
singletons might be a difference in the etiology of preterm delivery. In the latter
224
population many of the preterm deliveries are associated with pathologies such as
225
preeclampsia and fetal growth restriction whereas in twins preterm contractions due to
226
uterine overdistention is the most likely causative mechanism in the majority of
227
cases.15,16 However, it is unlikely that this may explain the lower antepartum mortality
228
rate especially since we included both mono- and dizygotic twins in our study. An
229
explanation for the lower antenatal mortality rate may be that in twin gestations with
230
one fetal death, pregnancy might have been continued for the benefit of the other twin,
231
resulting in a later (term) delivery and inaccuracy in determination of the timing of the
232
fetal death. This is unlikely to occur in a singleton pregnancy. This may have lowered the
233
preterm antenatal mortality rate and increased the term mortality rate in twins.
234
However, in this study 73 % of deaths in twins were recorded as having occurred during
235
the preterm period (Table 1), i.e. this will not account for the differences found. Another
However,
AC C
EP
TE D
M AN U
SC
214
ACCEPTED MANUSCRIPT 10 236
important reason might be that twin pregnancies are considered to be at high risk and
237
are therefore more closely monitored than singleton pregnancies in the same period.
238
According
239
measurements of fetal biometry, screening for signs of TTTS in monochorionic twins and
240
care given by an obstetrician17 Such a monitoring does not occur in case of singleton
241
pregnancies, which may be cared for by midwives, general practitioners or obstetricians.
242
Uncomplicated dichorial twins are advised to be induced into labor before 40 weeks
243
gestational age. For uncomplicated monochorial diamniotic and monoamniotic twins
244
induction of labor is advised at 36-37 and 32-34 weeks gestation respectively. For
245
uncomplicated singletons pregnancies, induction of labor is advised at 42+0 weeks of
246
pregnancy, induction may be considered between 41-42 weeks if the parents prefer
247
this.17
248
Our observations regarding mortality in relation to gestational age at birth are
249
supported by some previous studies. Kahn et al. evaluated the prospective risk of fetal
250
mortality in singleton, twin, and triplet pregnancies and found lower antepartum
251
mortality rates in twins compared to singletons until week 37 of gestation.7 In another
252
study from Korea, 609.643 singletons and 9805 twins were studied, the neonatal
253
mortality rate was lower in twins than in singletons with a gestational age > 29 weeks.8
254
A third study on fetal and neonatal mortality rates for singleton and multiple births in
255
the United States over the years 1985-1988 and 1995-1998 showed that for deliveries
256
before 37 weeks’ gestation fetal mortality rates were lower in twins than in singletons.9
257
Finally a large study from Canada,18
258
malformed twins was less than 1 per 1000 for infants born preterm, which is
259
considerably lower than antenatal stillbirth rates reported by us and others in
260
singletons.19,20
the
national
guidelines
this
includes
longitudinal
ultrasound
AC C
EP
TE D
M AN U
SC
RI PT
to
has shown that the risk of stillbirth in non-
ACCEPTED MANUSCRIPT 11
The lower antenatal mortality rate in twins as compared to singletons may well imply
262
that monitoring of singletons should be intensified to lower the stillbirth rate during the
263
preterm period. However, it remains uncertain which assessment tools should be used
264
(e.g., Doppler uterine artery, umbilical artery, middle cerebral artery; longitudinal growth
265
assessment, markers of placentation, maternal risk factors). Moreover, identification of
266
singletons at risk of stillbirth may prove difficult, given the fact that the majority will
267
have a fetal weight within the normal range. 20
268
In dizygotic twins of unequal gender there was no clear increase in mortality until 40
269
weeks of gestation, which suggests that in dizygotic twins modern surveillance
270
expectant management may last until that age, although nowadays most authors
271
suggest induction around 38 weeks.
272
delivery at 37 weeks’ gestation compared to ongoing expectant management for
273
women with an uncomplicated twin pregnancy does not appear to be associated with
274
an increased risk of morbidity or mortality, however specific data on chorionicity could
275
not be provided.21
276
A strength of this study is its large population based sample size. However, some
277
limitations of this study need to be mentioned. Twin pregnancies with missing data of
278
one or both children were excluded, as were singletons with missing data. In case all 42
279
deaths that were excluded in the twin group had occurred in the preterm period,
280
mortality in twins would have been about 17% higher, which would still be considerably
281
lower than that in singletons during the preterm period. The overall perinatal mortality
282
rate in twins born before 37weeks of gestation was 10.4/1000 and in singletons
283
34.5/1000 (Table 1); addition of 42 deaths to the twin cohort born <37 weeks, would
284
increase perinatal mortality till 12.1/1000.
285
information on the causes of death and possible inaccuracies in the registration of the
286
time of occurrence of death, especially for distinction between antepartum and
M AN U
SC
RI PT
261
Data from the Cochrane library indicate that
AC C
EP
TE D
18,21-23
Other limitations include the lack of
ACCEPTED MANUSCRIPT 12
intrapartum death.24 Another limitation is the lack of data on zygosity and placental
288
chorionicity. Perinatal mortality is higher in monozygotic/monochorial twins.2 To
289
estimate perinatal mortality rates in monozygotic twins the Weinberg formula may be
290
used.25 With this formula it is assumed that a) the number of dizygotic twins with
291
unequal gender is the same as that of dizygotic twins of equal gender, b) within
292
dizygotic twins the mortality of one member of the twin is independent from the gender
293
of the other member of the twin and c) within dizygotic twins a perfect balance exists
294
between the number of male-male and female-female twins. With this formula mortality
295
in monozygotic twins may be calculated as follows using absolute numbers: A. total twin
296
mortality – B. dizygotic twin mortality (mortality in dizygotic twins of different gender x
297
2) = C. mortality in monozygotic twins.
298
perinatal mortality was higher in monozygotic twins than in dizygotic DZ twins and this
299
held for most gestational ages (supplementary figure 1). However, preterm mortality
300
rates were lower in monozygotic twins than in singletons. This method has frequently
301
been used in previous literature on twin pregnancies. The reliability of this rule’s
302
assumptions has never been conclusively verified, nor rejected,26 We have chosen not to
303
include these data in the results section.
304
Conclusion
305
Overall perinatal mortality was higher in twins than in singletons. This is most likely
306
caused by the high preterm birth rate in twins and not by a higher mortality rate for
307
gestation, apart from term pregnancies. During the preterm period perinatal mortality
308
and especially stillbirth was much lower in twin deliveries than in singleton deliveries.
309
We hypothesize that this might, among others, be due to the closer monitoring of twins,
310
which indirectly suggests a need for closer monitoring of singleton pregnancies. Future
311
research should concentrate on data on zygosity and causes of death, to develop
312
optimal risk reduction strategies.
M AN U
SC
RI PT
287
AC C
EP
TE D
By using this formula we found that overall
ACCEPTED MANUSCRIPT 13 313
Acknowledgments
315
This project was supported by funding from Vrienden UMC Utrecht.
316
We would like to thank C. Hukkelhoven (Perined) for providing the aggregated data and
317
analyses.
RI PT
314
SC
318
References
320
1. Ortibus E, Lopriore E, Deprest J, et al. The pregnancy and long-term neurodevelopmental
321
outcome of monochorionic diamniotic twin gestations: A multicenter prospective cohort study
322
from the first trimester onward. Am J Obstet Gynecol. 2009;200(5):494.e1-494.e8. doi:
323
10.1016/j.ajog.2009.01.048 [doi].
324
2. Hack KE, Derks JB, Elias SG, et al. Increased perinatal mortality and morbidity in
325
monochorionic versus dichorionic twin pregnancies: Clinical implications of a large dutch cohort
326
study. BJOG. 2008;115(1):58-67. doi: BJO1556 [pii].
327
3. Lopriore E, Nagel HT, Vandenbussche FP, Walther FJ. Long-term neurodevelopmental
328
outcome in twin-to-twin transfusion syndrome. Am J Obstet Gynecol. 2003;189(5):1314-1319.
329
doi: S0002937803007609 [pii].
330
4. Sherer DM. Adverse perinatal outcome of twin pregnancies according to chorionicity: Review
331
of the literature. Am J Perinatol. 2001;18(1):23-37.
332
5. Powers WF, Kiely JL. The risks confronting twins: A national perspective. Am J Obstet Gynecol.
333
1994;170(2):456-461. doi: S000293789400476X [pii].
AC C
EP
TE D
M AN U
319
ACCEPTED MANUSCRIPT 14
6. EURO-PERISTAT Project with SCPE and EUROCAT. European perinatal health report. the
335
health and care of pregnant women and babies in europe in 2010.May 2013.
336
7. Kahn B, Lumey LH, Zybert PA, et al. Prospective risk of fetal death in singleton, twin, and
337
triplet gestations: Implications for practice. Obstet Gynecol. 2003;102(4):685-692. doi:
338
S0029784403006161 [pii].
339
8. Hong JS, Yi SW, Han YJ, et al. Fetal growth and neonatal mortality in korea. Paediatr Perinat
340
Epidemiol. 2007;21(5):397-410. doi: PPE850 [pii].
341
9. Alexander GR, Slay Wingate M, Salihu H, Kirby RS. Fetal and neonatal mortality risks of
342
multiple births. Obstet Gynecol Clin North Am. 2005;32(1):1-16, vii. doi: S0889-8545(04)00124-X
343
[pii].
344
10. Utrecht, perined 2016. www.perined.nl.
345
11. Perined. Perinatale zorg in Nederland 2014. Utrecht 2015.
346
12. Visser GH, Eilers PH, Elferink-Stinkens PM, Merkus HM, Wit JM. New dutch reference curves
347
for birthweight by gestational age. Early Hum Dev. 2009;85(12):737-744. doi:
348
10.1016/j.earlhumdev.2009.09.008; 10.1016/j.earlhumdev.2009.09.008.
349
13. Posthumus AG, Borsboom GJ, Poeran J, Steegers EA, Bonsel GJ. Geographical, ethnic and
350
socio-economic differences in utilization of obstetric care in the netherlands. PLoS One.
351
2016;11(6):e0156621. doi: 10.1371/journal.pone.0156621 [doi].
AC C
EP
TE D
M AN U
SC
RI PT
334
ACCEPTED MANUSCRIPT 15
14. Knol F. Status development in neighbourhoods in the netherlands 1998–2010 (in
353
dutch: statusontwikkeling van wijken in nederland 1998–2010). the netherlands institute for
354
social research, 2012 . .
355
15. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth.
356
Lancet. 2008;371(9606):75-84. doi: 10.1016/S0140-6736(08)60074-4 [doi].
357
16. Romero R, Espinoza J, Kusanovic JP, et al. The preterm parturition syndrome. BJOG.
358
2006;113 Suppl 3:17-42. doi: BJO1120 [pii].
359
17. Nederlandse Vereniging voor obstetrie en gynaecologie. Richtlijn meerlingzwangerschap.
360
http://nvog documenten.nl/index.php?pagina=/richtlijn/pagina.php&fSelectTG_62=75&fSelect
361
edSub=62&fSelectedParent=75. Accessed July,2016.
362
18. Wood S, Tang S, Ross S, Sauve R. Stillbirth in twins, exploring the optimal gestational age for
363
delivery: A retrospective cohort study. BJOG. 2014;121(10):1284-90; discussion 1291. doi:
364
10.1111/1471-0528.12866 [doi].
365
19. Gardosi J, Madurasinghe V, Williams M, Malik A, Francis A. Maternal and fetal risk factors
366
for stillbirth: Population based study. BMJ. 2013;346:f108. doi: 10.1136/bmj.f108.
367
20. Vasak B, Koenen SV, Koster MP, et al. Human fetal growth is constrained below optimal for
368
perinatal survival. Ultrasound Obstet Gynecol. 2015;45(2):162-167. doi: 10.1002/uog.14644
369
[doi].
AC C
EP
TE D
M AN U
SC
RI PT
352
ACCEPTED MANUSCRIPT 16
21. Dodd JM, Deussen AR, Grivell RM, Crowther CA. Elective birth at 37 weeks' gestation for
371
women with an uncomplicated twin pregnancy. Cochrane Database Syst Rev.
372
2014;2:CD003582. doi: 10.1002/14651858.CD003582.pub2 [doi].
373
22. Doss AE, Mancuso MS, Cliver SP, Jauk VC, Jenkins SM. Gestational age at delivery and
374
perinatal outcomes of twin gestations. Am J Obstet Gynecol. 2012;207(5):410.e1-410.e6. doi:
375
10.1016/j.ajog.2012.08.012 [doi].
376
23. Kilpatrick SJ, Jackson R, Croughan-Minihane MS. Perinatal mortality in twins and singletons
377
matched for gestational age at delivery at > or = 30 weeks. Am J Obstet Gynecol. 1996;174(1 Pt
378
1):66-71. doi: S0002-9378(96)70375-7 [pii].
379
24. Evers AC, Brouwers HA, Hukkelhoven CW, et al. Perinatal mortality and severe morbidity in
380
low and high risk term pregnancies in the netherlands: Prospective cohort study. BMJ.
381
2010;341:c5639. doi: 10.1136/bmj.c5639.
382
25. Weinberg W. Beitrage zur physiologie und pathologie der mehrlingsbeburten beim
383
menschen. Arch Ges Physiol. 1902;88:346–430.
384
26. Fellman J, Eriksson AW. Weinberg's differential rule reconsidered. Hum Biol.
385
2006;78(3):253-275. doi: 10.1353/hub.2006.0044 [doi].
SC
M AN U
TE D
EP
AC C
386
RI PT
370
ACCEPTED MANUSCRIPT
17
Tables
388
Table 1: Perinatal mortality according to gestational age in singleton and twin pregnancies
389
15 23 22 16 21 22 41 42 47 42 25 15 8 1 0 0 340
406 592 730 1048 1666 2496 3900 5088 8088 13588 9374 3502 1039 129 10 2 51658
Dizygotic twins unequal gender
PMR within week
36.95 38.85 30.14 15.27 12.61 8.81 10.51 8.25 5.81 3.09 2.67 4.28 7.70 7.75 0.00 0.00 6.58
M AN U
225.92 179.28 121.66 94.86 69.01 45.61 28.79 20.21 10.91 6.20 3.01 2.02 1.66 2.07 1.98 1.12 4.05
TE D
1443 1718 2285 2899 4130 6337 10732 17319 33999 85521 215586 357042 420017 274563 67638 891 1502120
EP
326 308 278 275 285 289 309 350 371 530 648 720 696 567 134 1 6087
Total deaths Total infants (n) (n)
AC C
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Total
All twins
Total deaths Total infants PMR within (n) (n) week
Total deaths Total infants (n) (n)
SC
Singletons Gestational age (weeks)
RI PT
387
1 5 3 4 11 4 9 14 15 19 8 4 1 0 0 0 98
96 174 208 362 510 773 1270 1783 2684 5102 3621 1377 415 59 4 0 18438
PMR within week
10.42 28.74 14.42 11.05 21.57 5.17 7.09 7.85 5.59 3.72 2.21 2.90 2.41 0.00 0.00 0.00 5.32
n= number, PMR= perinatal mortality rate *Per 1000 infants
Multivariate logistic regression analysis: Singletons versus all twins p <0.001. Singletons versus dizygotic twins unequal gender p <0.001. All twins versus dizygotic twins unequal gender p 0.18.
ACCEPTED MANUSCRIPT 18 390
Table 2: Mode of delivery in singleton and twin pregnancies Singletons n
%
n
1.129.542
75.20
26.628
51.55
<0.001
Instrumental delivery
153.354
10.21
5.006
9.69
0.23
Planned cesarean delivery
95.792
6.38
10.720
20.75
<0.001
Emergency cesarean delivery
120.141
8.00
9.171
17.75
<0.001
3.291
0.22
0.26
0.92
Spontaneous
Unknown 392
n=number, %=percentage
394
396
Parity
Socioeconomic status
400 401 402 403 404 405
PMR
4.53
3.65
SES > p20
p value
PMR
PMR
<0.001
4.80
3.76
EP
Twins 7.56 5.72 0.01 6.55 6.52 PMR: perinatal mortality rate per 1000 infants, SES: socioeconomic status
AC C
399
PMR
SES < p20
TE D
Singletons
398
133
Table 3: Perinatal mortality in singleton and twin pregnancies according to parity, sex, and socioeconomic status.
Nulliparous Multiparous
397
%
M AN U
393
395
p
SC
Delivery mode
All twins
RI PT
391
Child sex Boys
Girls
p value
PMR
PMR
p value
<0.001
4.16
3.92
<0.001
0.97
7.15
6.02
0.11
ACCEPTED MANUSCRIPT 19
Figures
407
Figure 1.
408
Figure legend: Perinatal mortality in singleton and twin pregnancies according to gestational age at birth. Perinatal mortality rates per 1000 infants with 95% confidence intervals.
409
RI PT
406
410 411
Figure 2.
413
Figure legend: Perinatal mortality in singleton and twin pregnancies according to time of occurrence of death and gestational age at birth. Perinatal mortality rates per 1000 infants with 95% confidence intervals.
415
M AN U
414
SC
412
416 417
Figure 3.
419
Figure legend: Perinatal mortality in singleton and twin pregnancies according to parity, sex, and socioeconomic status (SES) at birth. Perinatal mortality rates per 1000 infants. P20= 20th centile.
421 422 423
EP
420
TE D
418
Figure 4.
425
Figure legend: Perinatal mortality in singleton and dizygotic twin pregnancies with unequal gender (female-male twins) according to gestational age at birth. Perinatal mortality rate per 1000 infants with 95% confidence intervals. The mortality rates after 41 weeks gestation were not displayed owing to the extremely small number of cases.
426 427 428 429 430 431
432
AC C
424
ACCEPTED MANUSCRIPT 20
Supplementary
434
Supplementary Figure S1.
435
Figure legend: Perinatal mortality in singleton and twin pregnancies according to zygosity and gestational age at birth. Data on twin zygosity were estimated using the Weinberg formula. The mortality rates after 41 weeks gestation were not displayed owing to the extremely small number of cases.
436 437 438
RI PT
433
SC
439
440
M AN U
441
AC C
EP
TE D
442
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0002-9378(16)30868-7
DOI:
10.1016/j.ajog.2016.10.005
Reference:
YMOB 11335
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 13 July 2016 Revised Date:
16 September 2016
Accepted Date: 3 October 2016
Please cite this article as: Vasak B, Verhagen JJ, Koenen SV, Koster MPH, de Reu PAOM, Franx A, Nijhuis JG, Bonsel GJ, Visser GHA, Lower perinatal mortality in preterm born twins than in singletons; a nationwide study from the Netherlands, American Journal of Obstetrics and Gynecology (2016), doi: 10.1016/j.ajog.2016.10.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 1
TITLE PAGE
2
Lower perinatal mortality in preterm born twins than in singletons; a nationwide study from the Netherlands
3
RI PT
4
Blanka Vasak1 (MD), Jessica J. Verhagen1 (MD), Steven V. Koenen1 (PhD, MD), Maria P.H. Koster1
6
(PhD, MD), P.A.O.M. de Reu2 (PhD), Arie Franx 1 (PhD, MD), Jan G. Nijhuis3 (PhD, MD), Gouke J.
7
Bonsel1 (PhD, MD), Gerard H.A. Visser1 (PhD, MD)
SC
5
8 1
Division Woman and Baby, University Medical Center Utrecht, The Netherlands
10
2
Independent midwife, retired, no current affiliation
11 12
3
M AN U
9
Department of Obstetrics and Gynaecology, GROW-School for Oncology and Developmental Biology, Maastricht University Medical Centre, Maastricht, The Netherlands
13
TE D
Conflict of interest: The authors report no conflict of interest.
Acknowledgment of financial support: Vrienden UMC Utrecht provided financial support.
EP
Paper presentation This study has been presented at the 49th Annual Gynaecongres, Eindhoven, the Netherlands, May 20, 2016.
AC C
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Table or Figure that needs to be included in the Print Version Figure 1
Corresponding author: B. Vasak, MD.
Work address: University Medical Center Departement Obstetrics, Huispostnummer KE 04.123.1 PO Box 85090
ACCEPTED MANUSCRIPT 2 35 36 37 38 39 40 41 42 43 44 45
Lundlaan 6, 3584 EA Utrecht The Netherlands
46
Word count main text: 2673
47
Word count abstract: 278
48 49 50 51 52 53 54 55
Condensation of the paper Overall perinatal mortality is higher in twin pregnancies than in singleton pregnancies, but during the preterm period, mortality is considerably lower in twins.
RI PT SC
+31 88 75 539 23 +31 6 41 38 52 31 +31 88 75 553 20
M AN U
Telephone: Work: Private: Fax:
EP
TE D
Short version article title "Lower perinatal mortality in preterm born twins than in singletons."
AC C
56
Email: [email protected]
ACCEPTED MANUSCRIPT 3
Abstract
58
Background: Twin pregnancies are at increased risk for perinatal morbidity and mortality, due to many factors including a high incidence of preterm delivery. Compared to singleton pregnancies overall perinatal mortality risk is higher in twin pregnancies, however for the preterm period perinatal mortality has been reported to be lower in twins.
63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
Objective: To compare perinatal mortality in relation to gestational age at birth between singleton and twin pregnancies, taken into account socioeconomic status, fetal sex and parity. Study design: We studied perinatal mortality according to gestational age at birth in 1.502.120 singletons and 51.658 twins without congenital malformations who were delivered between 2002 and 2010 after 28 weeks of gestation. Data were collected from the nationwide Netherlands Perinatal Registry.
SC
62
Results: Overall perinatal mortality in twins (6.6/1000 infants) was higher than in singletons (4.1/1000 infants). However, in the preterm period perinatal mortality in twins was substantially lower than in singleton pregnancies, 10.4 per 1000 infants as compared to 34.5 per 1000 infants, respectively for infants born < 37 weeks; this held especially for antepartum deaths. After 39 weeks of pregnancy perinatal mortality was higher in twins. Differences in parity, fetal sex and socioeconomic status did not explain the observed differences in outcome.
M AN U
61
Conclusion: Overall perinatal mortality was higher in twins than in singletons, which is most likely caused by the high preterm birth rate in twins and not by a higher mortality rate for gestation, apart from term pregnancies. During the preterm period, antepartum mortality was much lower in twins than in singletons. We suggest that this might partially be due to a closer monitoring of twins, which indirectly suggests a need for closer surveillance of singleton pregnancies.
TE D
60
EP
59
RI PT
57
80
Key words
82
Gestational age, Perinatal mortality, Preterm mortality, Singleton pregnancies, Twin pregnancies
83 84 85 86 87
AC C
81
ACCEPTED MANUSCRIPT 4
Introduction
89
Twin pregnancies are at increased risk for preterm birth, intrauterine growth restriction
90
and a number of multiplet related conditions. Monochorionic twins have additional risks
91
for mortality and morbidity primarily due to the twin-to-twin transfusion syndrome
92
(TTTS) and congenital abnormalities.1-5
93
While registry data in Western countries (PERISTAT) confirm a crude 2-3 fold increased
94
perinatal mortality risk of a twin pregnancy,6 only few reports have compared mortality
95
and morbidity between twin and singleton pregnancies taking into account differences
96
in gestational age at delivery. They all found a lower perinatal mortality in twins born
97
before 37 weeks.7-9 These studies were all small, with the exception of one (Kahn) and
98
did not include separation between zygosity and chorionicity in twins, nor did they take
99
into account known prognostic important factors such as parity and socioeconomic
M AN U
SC
RI PT
88
status (SES) of the mother and sex of the child.
101
The objective of this study was to compare perinatal mortality in relation to gestational
102
age at birth between singleton and twin pregnancies, taken into account parity, SES,
103
fetal sex and mode of delivery. For this study we used a large national anonymized data
104
set from the Netherlands Perinatal Registry (Perined).10
EP
TE D
100
AC C
105 106
Materials and methods
107
General
108
This was a retrospective population-based cohort study. We obtained, after prior
109
permission from the Perinatal Registry (Perined),11 aggregated data and analyses on all
110
births in the Netherlands between 2002 and 2010. This registry contains information on
ACCEPTED MANUSCRIPT 5 111
approximately 95% of all births from 16.0 weeks gestational age onwards, where
112
coverage of twins is even higher. Details of this dataset can be found elsewhere.10
113
Variables
115
The Perined dataset includes routine maternal data on age, parity, socioeconomic
116
background (zip-code based), and ethnic background. Information is available on
117
intervention (induction of labor, mode of delivery (instrumental, elective and emergency
118
cesarean delivery), NICU admission of the child and specific features of the delivery.
119
Data on the child comprise of: singleton or multiplet, sex of the child, presence of
120
congenital anomalies, gestational age at birth, birth weight, and Apgar score at 5
121
minutes. From birth weight, sex and gestational age, the birth weight is computed as
122
centile score.12
123
The size of the multiplet is recorded and for each child its rank number during birth. No
124
information is available on zygosity/chorionicity. Congenital abnormalities were
125
recorded if they were present and recognized, either at birth or at first NICU admission
126
by the neonatologist. This includes any abnormality already noticed during antenatal
127
ultrasound scanning or, occasionally, with genetic tests. All children with congenital
128
abnormalities were excluded. Gestational age was calculated from the first day of the
129
last menstrual period or from an ultrasound dating scan by measurement of the Crown-
130
Rump Length in the first trimester. In this period 91% of the pregnancies had a reliable
131
dating, obtained by a first trimester dating scan.10 For twins conceived spontaneously
132
dating was based on the mean CRL of both fetuses. Pregnancies achieved by assisted
133
reproductive technology (ART) were dated based on the ART-derived gestational age.
134
In this report a gestational age of 37 weeks means: 37 weeks + 0 days - 37 weeks + 6
135
days containing whole week groups. For this analysis births between 28+0 and 43+6
AC C
EP
TE D
M AN U
SC
RI PT
114
ACCEPTED MANUSCRIPT 6
gestation were included, to allow comparison with data from other countries in which
137
registration practices may differ at earlier gestational ages.6 Children with missing
138
gestational age at birth were excluded from the analysis. A total of 1839 children from a
139
twin pregnancy were excluded because a twin pair could not be linked or information
140
on congenital malformations or gestational age of one of the twins was missing. In this
141
group there were 42 perinatal deaths. If one of the twins died before 28 weeks gestation
142
both children were excluded from analysis. If one of the twins died after 28 weeks
143
gestation and the pregnancy continued for the benefit of the other twin, the gestational
144
age at birth was registered for both twins, the exact gestational age of death for the
145
demised twin was not available in this database.
146
Perinatal mortality was defined as fetal or neonatal death up to 7 days after birth:
147
antepartum mortality was defined as death occurring before labor, intrapartum mortality
148
as death during labor, and neonatal mortality as death occurring within 0–7 days
149
following live birth. This information is always available in all cases.
150
Statistical analysis
151
Perinatal mortality was expressed as rate (number of deaths per 1000 infants, including
152
stillbirths). Mortality rate was subdivided to time of occurrence: antepartum, intrapartum
153
and neonatal. Perinatal mortality was compared between singleton and twin, according
154
to gestational age. The mortality rates at 42 and 43 weeks gestation were not displayed
155
owing to the extremely small number of cases. Comparisons were made between
156
groups by using multivariate logistic regression with gestational age in the model or chi
157
square test.
158
The mode of delivery in singletons and twins was studied. Vaginal delivery was divided
159
in spontaneous vaginal delivery and instrumental delivery defined as delivery by forceps
160
or a vacuum device. Cesarean delivery was divided in planned cesarean section, defined
AC C
EP
TE D
M AN U
SC
RI PT
136
ACCEPTED MANUSCRIPT 7
as a scheduled or elective cesarean, and an emergency cesarean defined as any
162
cesarean that was not planned or scheduled.
163
To investigate the influence of parity, maternal SES, and child sex, we calculated
164
perinatal mortality rates for each parameter for twins and singletons: for parity (0 vs.
165
1+), SES (p<20 poor vs. p>20) and child sex (male/female). For classification of SES, the
166
"socio-economic status score" was used as a proxy. This score takes into account the
167
average income per household in a given four-digit postcode area and the percentage
168
of households with low income, without paid job and with low education level.13,14 The
169
level of SES was categorized as low if the score was under the 20th centile.
170
No information on zygosity was available in the data set. The only information we could
171
obtain on zygosity indirectly were the twin pairs with unequal gender; male-female
172
twins, these twins are per definition dizygotic and dichorial.
M AN U
SC
RI PT
161
TE D
173
Results
175
We studied 1.502.120 singletons and 51.658 twins born between 2002 and 2010 after 28
176
weeks of gestation. In singleton pregnancies there were 6.087 perinatal deaths
177
(4.1/1000). For twins a total of 340 deaths occurred resulting in a perinatal mortality
178
rate of 6.6 per 1000 infants. Perinatal mortality according to gestational age is displayed
179
in table 1.
180
Of all twins 46% were born before 37 weeks of gestation, in contrast to only 5% in the
181
singletons. For deliveries before 37 weeks perinatal mortality rate in twins was 10.4 per
182
1000 infants compared to 34.5 per 1000 infants in singletons. For infants born before 34
183
weeks mortality rates were 17.2 per 1000 infants in twins and 93.6 per 1000 infants in
184
singletons.
AC C
EP
174
ACCEPTED MANUSCRIPT 8
In singleton pregnancies, perinatal mortality decreased with increasing gestational age,
186
with the lowest mortality rate at 40 weeks gestational age. In twin pregnancies mortality
187
also decreased with increasing gestational age with the lowest mortality rate at 38
188
weeks and an increase thereafter (Fig 1). In the preterm period, perinatal mortality in
189
twins was substantially lower than in singleton pregnancies. After 39+0 weeks mortality
190
was higher in twins.
191
In figure 2 perinatal mortality is displayed according to time period of occurrence of
192
death. In both singletons and twins the majority of perinatal deaths occurred in the
193
antepartum period, followed by the neonatal period and the intrapartum period,
194
respectively. In singletons, the distribution of the time period of death was, 72 %
195
antepartum, 9% intra partum and 19% neonatal; for twins this distribution was 59%
196
antepartum, 6% intra partum and 35% neonatal. Antepartum mortality was significantly
197
lower in twins as compared to singletons during the whole preterm period (p<0.001).
198
After 40 weeks gestation antepartum mortality was significantly higher for twins (p 0.02).
199
Intrapartum mortality was significantly lower between 31-33 weeks of gestation but did
200
not differ in the term period. Neonatal mortality was more or less similar during the
201
whole preterm period, mortality rates were higher for twins in the term period. Twins
202
were more often delivered by cesarean section than singletons (Table 2).
203
In both singletons and twins, nulliparity was related to higher overall mortality rates.
204
Poor maternal socioeconomic status (SES
205
overall perinatal mortality rates in singletons, in twins rates were also higher in these
206
groups but this was not significant (Table 3). Figure 3 shows twin and singleton mortality
207
rates per parameter for gestational age. Within the singletons and twin groups mortality
208
rates did not differ for parity, fetal sex and socioeconomic status.
209
Overall perinatal mortality was slightly lower in dizygotic twins of unequal gender as
210
compared to all twins and this held for most gestational ages (Table 1 & Fig 4). For
AC C
EP
TE D
M AN U
SC
RI PT
185
ACCEPTED MANUSCRIPT 9 211
infants born preterm mortality rates were lower in dizygotic unequal gender twins than
212
in singletons.
RI PT
213
Comment
215
This study shows that, in a large contemporary population of infants without congenital
216
abnormalities born after 28 weeks' gestation, overall perinatal mortality was higher in
217
twins compared to singletons, which is in accordance with literature.5
218
perinatal mortality in preterm twins was considerably lower than in singletons. For
219
antepartum mortality this held for the whole preterm period and for intrapartum
220
mortality between 31 and 33 weeks, but not for neonatal mortality. Data did not change
221
when parity, sex, socio-economic status were taken into account.
222
An explanation for the lower preterm perinatal death rates in twins as compared to
223
singletons might be a difference in the etiology of preterm delivery. In the latter
224
population many of the preterm deliveries are associated with pathologies such as
225
preeclampsia and fetal growth restriction whereas in twins preterm contractions due to
226
uterine overdistention is the most likely causative mechanism in the majority of
227
cases.15,16 However, it is unlikely that this may explain the lower antepartum mortality
228
rate especially since we included both mono- and dizygotic twins in our study. An
229
explanation for the lower antenatal mortality rate may be that in twin gestations with
230
one fetal death, pregnancy might have been continued for the benefit of the other twin,
231
resulting in a later (term) delivery and inaccuracy in determination of the timing of the
232
fetal death. This is unlikely to occur in a singleton pregnancy. This may have lowered the
233
preterm antenatal mortality rate and increased the term mortality rate in twins.
234
However, in this study 73 % of deaths in twins were recorded as having occurred during
235
the preterm period (Table 1), i.e. this will not account for the differences found. Another
However,
AC C
EP
TE D
M AN U
SC
214
ACCEPTED MANUSCRIPT 10 236
important reason might be that twin pregnancies are considered to be at high risk and
237
are therefore more closely monitored than singleton pregnancies in the same period.
238
According
239
measurements of fetal biometry, screening for signs of TTTS in monochorionic twins and
240
care given by an obstetrician17 Such a monitoring does not occur in case of singleton
241
pregnancies, which may be cared for by midwives, general practitioners or obstetricians.
242
Uncomplicated dichorial twins are advised to be induced into labor before 40 weeks
243
gestational age. For uncomplicated monochorial diamniotic and monoamniotic twins
244
induction of labor is advised at 36-37 and 32-34 weeks gestation respectively. For
245
uncomplicated singletons pregnancies, induction of labor is advised at 42+0 weeks of
246
pregnancy, induction may be considered between 41-42 weeks if the parents prefer
247
this.17
248
Our observations regarding mortality in relation to gestational age at birth are
249
supported by some previous studies. Kahn et al. evaluated the prospective risk of fetal
250
mortality in singleton, twin, and triplet pregnancies and found lower antepartum
251
mortality rates in twins compared to singletons until week 37 of gestation.7 In another
252
study from Korea, 609.643 singletons and 9805 twins were studied, the neonatal
253
mortality rate was lower in twins than in singletons with a gestational age > 29 weeks.8
254
A third study on fetal and neonatal mortality rates for singleton and multiple births in
255
the United States over the years 1985-1988 and 1995-1998 showed that for deliveries
256
before 37 weeks’ gestation fetal mortality rates were lower in twins than in singletons.9
257
Finally a large study from Canada,18
258
malformed twins was less than 1 per 1000 for infants born preterm, which is
259
considerably lower than antenatal stillbirth rates reported by us and others in
260
singletons.19,20
the
national
guidelines
this
includes
longitudinal
ultrasound
AC C
EP
TE D
M AN U
SC
RI PT
to
has shown that the risk of stillbirth in non-
ACCEPTED MANUSCRIPT 11
The lower antenatal mortality rate in twins as compared to singletons may well imply
262
that monitoring of singletons should be intensified to lower the stillbirth rate during the
263
preterm period. However, it remains uncertain which assessment tools should be used
264
(e.g., Doppler uterine artery, umbilical artery, middle cerebral artery; longitudinal growth
265
assessment, markers of placentation, maternal risk factors). Moreover, identification of
266
singletons at risk of stillbirth may prove difficult, given the fact that the majority will
267
have a fetal weight within the normal range. 20
268
In dizygotic twins of unequal gender there was no clear increase in mortality until 40
269
weeks of gestation, which suggests that in dizygotic twins modern surveillance
270
expectant management may last until that age, although nowadays most authors
271
suggest induction around 38 weeks.
272
delivery at 37 weeks’ gestation compared to ongoing expectant management for
273
women with an uncomplicated twin pregnancy does not appear to be associated with
274
an increased risk of morbidity or mortality, however specific data on chorionicity could
275
not be provided.21
276
A strength of this study is its large population based sample size. However, some
277
limitations of this study need to be mentioned. Twin pregnancies with missing data of
278
one or both children were excluded, as were singletons with missing data. In case all 42
279
deaths that were excluded in the twin group had occurred in the preterm period,
280
mortality in twins would have been about 17% higher, which would still be considerably
281
lower than that in singletons during the preterm period. The overall perinatal mortality
282
rate in twins born before 37weeks of gestation was 10.4/1000 and in singletons
283
34.5/1000 (Table 1); addition of 42 deaths to the twin cohort born <37 weeks, would
284
increase perinatal mortality till 12.1/1000.
285
information on the causes of death and possible inaccuracies in the registration of the
286
time of occurrence of death, especially for distinction between antepartum and
M AN U
SC
RI PT
261
Data from the Cochrane library indicate that
AC C
EP
TE D
18,21-23
Other limitations include the lack of
ACCEPTED MANUSCRIPT 12
intrapartum death.24 Another limitation is the lack of data on zygosity and placental
288
chorionicity. Perinatal mortality is higher in monozygotic/monochorial twins.2 To
289
estimate perinatal mortality rates in monozygotic twins the Weinberg formula may be
290
used.25 With this formula it is assumed that a) the number of dizygotic twins with
291
unequal gender is the same as that of dizygotic twins of equal gender, b) within
292
dizygotic twins the mortality of one member of the twin is independent from the gender
293
of the other member of the twin and c) within dizygotic twins a perfect balance exists
294
between the number of male-male and female-female twins. With this formula mortality
295
in monozygotic twins may be calculated as follows using absolute numbers: A. total twin
296
mortality – B. dizygotic twin mortality (mortality in dizygotic twins of different gender x
297
2) = C. mortality in monozygotic twins.
298
perinatal mortality was higher in monozygotic twins than in dizygotic DZ twins and this
299
held for most gestational ages (supplementary figure 1). However, preterm mortality
300
rates were lower in monozygotic twins than in singletons. This method has frequently
301
been used in previous literature on twin pregnancies. The reliability of this rule’s
302
assumptions has never been conclusively verified, nor rejected,26 We have chosen not to
303
include these data in the results section.
304
Conclusion
305
Overall perinatal mortality was higher in twins than in singletons. This is most likely
306
caused by the high preterm birth rate in twins and not by a higher mortality rate for
307
gestation, apart from term pregnancies. During the preterm period perinatal mortality
308
and especially stillbirth was much lower in twin deliveries than in singleton deliveries.
309
We hypothesize that this might, among others, be due to the closer monitoring of twins,
310
which indirectly suggests a need for closer monitoring of singleton pregnancies. Future
311
research should concentrate on data on zygosity and causes of death, to develop
312
optimal risk reduction strategies.
M AN U
SC
RI PT
287
AC C
EP
TE D
By using this formula we found that overall
ACCEPTED MANUSCRIPT 13 313
Acknowledgments
315
This project was supported by funding from Vrienden UMC Utrecht.
316
We would like to thank C. Hukkelhoven (Perined) for providing the aggregated data and
317
analyses.
RI PT
314
SC
318
References
320
1. Ortibus E, Lopriore E, Deprest J, et al. The pregnancy and long-term neurodevelopmental
321
outcome of monochorionic diamniotic twin gestations: A multicenter prospective cohort study
322
from the first trimester onward. Am J Obstet Gynecol. 2009;200(5):494.e1-494.e8. doi:
323
10.1016/j.ajog.2009.01.048 [doi].
324
2. Hack KE, Derks JB, Elias SG, et al. Increased perinatal mortality and morbidity in
325
monochorionic versus dichorionic twin pregnancies: Clinical implications of a large dutch cohort
326
study. BJOG. 2008;115(1):58-67. doi: BJO1556 [pii].
327
3. Lopriore E, Nagel HT, Vandenbussche FP, Walther FJ. Long-term neurodevelopmental
328
outcome in twin-to-twin transfusion syndrome. Am J Obstet Gynecol. 2003;189(5):1314-1319.
329
doi: S0002937803007609 [pii].
330
4. Sherer DM. Adverse perinatal outcome of twin pregnancies according to chorionicity: Review
331
of the literature. Am J Perinatol. 2001;18(1):23-37.
332
5. Powers WF, Kiely JL. The risks confronting twins: A national perspective. Am J Obstet Gynecol.
333
1994;170(2):456-461. doi: S000293789400476X [pii].
AC C
EP
TE D
M AN U
319
ACCEPTED MANUSCRIPT 14
6. EURO-PERISTAT Project with SCPE and EUROCAT. European perinatal health report. the
335
health and care of pregnant women and babies in europe in 2010.May 2013.
336
7. Kahn B, Lumey LH, Zybert PA, et al. Prospective risk of fetal death in singleton, twin, and
337
triplet gestations: Implications for practice. Obstet Gynecol. 2003;102(4):685-692. doi:
338
S0029784403006161 [pii].
339
8. Hong JS, Yi SW, Han YJ, et al. Fetal growth and neonatal mortality in korea. Paediatr Perinat
340
Epidemiol. 2007;21(5):397-410. doi: PPE850 [pii].
341
9. Alexander GR, Slay Wingate M, Salihu H, Kirby RS. Fetal and neonatal mortality risks of
342
multiple births. Obstet Gynecol Clin North Am. 2005;32(1):1-16, vii. doi: S0889-8545(04)00124-X
343
[pii].
344
10. Utrecht, perined 2016. www.perined.nl.
345
11. Perined. Perinatale zorg in Nederland 2014. Utrecht 2015.
346
12. Visser GH, Eilers PH, Elferink-Stinkens PM, Merkus HM, Wit JM. New dutch reference curves
347
for birthweight by gestational age. Early Hum Dev. 2009;85(12):737-744. doi:
348
10.1016/j.earlhumdev.2009.09.008; 10.1016/j.earlhumdev.2009.09.008.
349
13. Posthumus AG, Borsboom GJ, Poeran J, Steegers EA, Bonsel GJ. Geographical, ethnic and
350
socio-economic differences in utilization of obstetric care in the netherlands. PLoS One.
351
2016;11(6):e0156621. doi: 10.1371/journal.pone.0156621 [doi].
AC C
EP
TE D
M AN U
SC
RI PT
334
ACCEPTED MANUSCRIPT 15
14. Knol F. Status development in neighbourhoods in the netherlands 1998–2010 (in
353
dutch: statusontwikkeling van wijken in nederland 1998–2010). the netherlands institute for
354
social research, 2012 . .
355
15. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth.
356
Lancet. 2008;371(9606):75-84. doi: 10.1016/S0140-6736(08)60074-4 [doi].
357
16. Romero R, Espinoza J, Kusanovic JP, et al. The preterm parturition syndrome. BJOG.
358
2006;113 Suppl 3:17-42. doi: BJO1120 [pii].
359
17. Nederlandse Vereniging voor obstetrie en gynaecologie. Richtlijn meerlingzwangerschap.
360
http://nvog documenten.nl/index.php?pagina=/richtlijn/pagina.php&fSelectTG_62=75&fSelect
361
edSub=62&fSelectedParent=75. Accessed July,2016.
362
18. Wood S, Tang S, Ross S, Sauve R. Stillbirth in twins, exploring the optimal gestational age for
363
delivery: A retrospective cohort study. BJOG. 2014;121(10):1284-90; discussion 1291. doi:
364
10.1111/1471-0528.12866 [doi].
365
19. Gardosi J, Madurasinghe V, Williams M, Malik A, Francis A. Maternal and fetal risk factors
366
for stillbirth: Population based study. BMJ. 2013;346:f108. doi: 10.1136/bmj.f108.
367
20. Vasak B, Koenen SV, Koster MP, et al. Human fetal growth is constrained below optimal for
368
perinatal survival. Ultrasound Obstet Gynecol. 2015;45(2):162-167. doi: 10.1002/uog.14644
369
[doi].
AC C
EP
TE D
M AN U
SC
RI PT
352
ACCEPTED MANUSCRIPT 16
21. Dodd JM, Deussen AR, Grivell RM, Crowther CA. Elective birth at 37 weeks' gestation for
371
women with an uncomplicated twin pregnancy. Cochrane Database Syst Rev.
372
2014;2:CD003582. doi: 10.1002/14651858.CD003582.pub2 [doi].
373
22. Doss AE, Mancuso MS, Cliver SP, Jauk VC, Jenkins SM. Gestational age at delivery and
374
perinatal outcomes of twin gestations. Am J Obstet Gynecol. 2012;207(5):410.e1-410.e6. doi:
375
10.1016/j.ajog.2012.08.012 [doi].
376
23. Kilpatrick SJ, Jackson R, Croughan-Minihane MS. Perinatal mortality in twins and singletons
377
matched for gestational age at delivery at > or = 30 weeks. Am J Obstet Gynecol. 1996;174(1 Pt
378
1):66-71. doi: S0002-9378(96)70375-7 [pii].
379
24. Evers AC, Brouwers HA, Hukkelhoven CW, et al. Perinatal mortality and severe morbidity in
380
low and high risk term pregnancies in the netherlands: Prospective cohort study. BMJ.
381
2010;341:c5639. doi: 10.1136/bmj.c5639.
382
25. Weinberg W. Beitrage zur physiologie und pathologie der mehrlingsbeburten beim
383
menschen. Arch Ges Physiol. 1902;88:346–430.
384
26. Fellman J, Eriksson AW. Weinberg's differential rule reconsidered. Hum Biol.
385
2006;78(3):253-275. doi: 10.1353/hub.2006.0044 [doi].
SC
M AN U
TE D
EP
AC C
386
RI PT
370
ACCEPTED MANUSCRIPT
17
Tables
388
Table 1: Perinatal mortality according to gestational age in singleton and twin pregnancies
389
15 23 22 16 21 22 41 42 47 42 25 15 8 1 0 0 340
406 592 730 1048 1666 2496 3900 5088 8088 13588 9374 3502 1039 129 10 2 51658
Dizygotic twins unequal gender
PMR within week
36.95 38.85 30.14 15.27 12.61 8.81 10.51 8.25 5.81 3.09 2.67 4.28 7.70 7.75 0.00 0.00 6.58
M AN U
225.92 179.28 121.66 94.86 69.01 45.61 28.79 20.21 10.91 6.20 3.01 2.02 1.66 2.07 1.98 1.12 4.05
TE D
1443 1718 2285 2899 4130 6337 10732 17319 33999 85521 215586 357042 420017 274563 67638 891 1502120
EP
326 308 278 275 285 289 309 350 371 530 648 720 696 567 134 1 6087
Total deaths Total infants (n) (n)
AC C
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Total
All twins
Total deaths Total infants PMR within (n) (n) week
Total deaths Total infants (n) (n)
SC
Singletons Gestational age (weeks)
RI PT
387
1 5 3 4 11 4 9 14 15 19 8 4 1 0 0 0 98
96 174 208 362 510 773 1270 1783 2684 5102 3621 1377 415 59 4 0 18438
PMR within week
10.42 28.74 14.42 11.05 21.57 5.17 7.09 7.85 5.59 3.72 2.21 2.90 2.41 0.00 0.00 0.00 5.32
n= number, PMR= perinatal mortality rate *Per 1000 infants
Multivariate logistic regression analysis: Singletons versus all twins p <0.001. Singletons versus dizygotic twins unequal gender p <0.001. All twins versus dizygotic twins unequal gender p 0.18.
ACCEPTED MANUSCRIPT 18 390
Table 2: Mode of delivery in singleton and twin pregnancies Singletons n
%
n
1.129.542
75.20
26.628
51.55
<0.001
Instrumental delivery
153.354
10.21
5.006
9.69
0.23
Planned cesarean delivery
95.792
6.38
10.720
20.75
<0.001
Emergency cesarean delivery
120.141
8.00
9.171
17.75
<0.001
3.291
0.22
0.26
0.92
Spontaneous
Unknown 392
n=number, %=percentage
394
396
Parity
Socioeconomic status
400 401 402 403 404 405
PMR
4.53
3.65
SES > p20
p value
PMR
PMR
<0.001
4.80
3.76
EP
Twins 7.56 5.72 0.01 6.55 6.52 PMR: perinatal mortality rate per 1000 infants, SES: socioeconomic status
AC C
399
PMR
SES < p20
TE D
Singletons
398
133
Table 3: Perinatal mortality in singleton and twin pregnancies according to parity, sex, and socioeconomic status.
Nulliparous Multiparous
397
%
M AN U
393
395
p
SC
Delivery mode
All twins
RI PT
391
Child sex Boys
Girls
p value
PMR
PMR
p value
<0.001
4.16
3.92
<0.001
0.97
7.15
6.02
0.11
ACCEPTED MANUSCRIPT 19
Figures
407
Figure 1.
408
Figure legend: Perinatal mortality in singleton and twin pregnancies according to gestational age at birth. Perinatal mortality rates per 1000 infants with 95% confidence intervals.
409
RI PT
406
410 411
Figure 2.
413
Figure legend: Perinatal mortality in singleton and twin pregnancies according to time of occurrence of death and gestational age at birth. Perinatal mortality rates per 1000 infants with 95% confidence intervals.
415
M AN U
414
SC
412
416 417
Figure 3.
419
Figure legend: Perinatal mortality in singleton and twin pregnancies according to parity, sex, and socioeconomic status (SES) at birth. Perinatal mortality rates per 1000 infants. P20= 20th centile.
421 422 423
EP
420
TE D
418
Figure 4.
425
Figure legend: Perinatal mortality in singleton and dizygotic twin pregnancies with unequal gender (female-male twins) according to gestational age at birth. Perinatal mortality rate per 1000 infants with 95% confidence intervals. The mortality rates after 41 weeks gestation were not displayed owing to the extremely small number of cases.
426 427 428 429 430 431
432
AC C
424
ACCEPTED MANUSCRIPT 20
Supplementary
434
Supplementary Figure S1.
435
Figure legend: Perinatal mortality in singleton and twin pregnancies according to zygosity and gestational age at birth. Data on twin zygosity were estimated using the Weinberg formula. The mortality rates after 41 weeks gestation were not displayed owing to the extremely small number of cases.
436 437 438
RI PT
433
SC
439
440
M AN U
441
AC C
EP
TE D
442
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT