Share this article on:

Prospective Risk of Stillbirth in Multiple-Gestation Pregnancies: A Population-Based Analysis

Sairam, Shanthi MD, MRCOG; Costeloe, Kate MBBChir, FRCP; Thilaganathan, Baskaran MD, MRCOG

Original Research

OBJECTIVE To evaluate the prospective risk of stillbirth in multiple gestations.

METHODS We conducted a retrospective analysis of birth notifications and infant mortality records relating to all multiple gestations to residents in a predefined health district. The incidence of live births and stillbirths was used to calculate the prospective risk of stillbirth at each week of gestation.

RESULTS The risk of stillbirth in multiple gestations increased from 1:3333 at 28 weeks' gestation to 1:69 at 39 or more weeks' gestation. The stillbirth risk in multiple gestations at 39 weeks surpassed that of postterm singleton pregnancies (1:526).

CONCLUSION Multiple gestations at 37–38 weeks have a risk of stillbirth equivalent to that of postterm singleton pregnancy. Because multiple gestations rarely proceed beyond 39 weeks, and because stillbirth risk increases severalfold beyond this stage, elective delivery might be justified at this gestational age.

Risk of stillbirth in multiple gestations at 37–38 weeks' gestation is equivalent to that of postterm singleton pregnancy.

Feto-Maternal Medicine Unit, Academic Department of Obstetrics and Gynaecology, St George's Hospital Medical School, London, United Kingdom; and Academic Department of Child Health, St. Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary Westfield College, London, United Kingdom.

Address reprint requests to: B. Thilaganathan, MD, MRCOG, Academic Department of Obstetrics and Gynaecology, St George's Hospital Medical School, Cranmer Terrace, Tooting, London SW17 0RE England; E-mail:

This work was supported by the former NorthEast Thames Regional Health Authority, Hospitals and Regional Reproductive Review Information Project, from which the data were derived.

We are grateful for the help of the managers of the Regional Interactive Child Health System and the contribution of nurses, midwives, health visitors, pediatricians, obstetricians and clerical staff, whose record keeping contributes birth notification data. We are grateful to Ms. Lisa Hilder, Epidemiologist at St. Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary Westfield College, for help in compiling the data.

Received February 8, 2002. Received in revised form April 3, 2002. Accepted April 11, 2002.

Multiple gestation is an acknowledged cause of increased risk to the fetus, but the exact magnitude of this risk is a matter of ongoing debate.1–10 Whereas some studies found elevated fetal and neonatal mortality rates in multiple gestations,1–5 others concluded that the perinatal mortality rate is equivalent to singletons born at the same gestational age (gestation-specific data).6–10

Most previous studies have assumed incorrectly that the perinatal mortality rate reflects the risk of stillbirth and neonatal death.11,12 Because stillbirth can only occur before delivery, the risk of stillbirth is better represented as a proportion of ongoing pregnancies, rather than of total births.13–15 Gestation-specific data on the magnitude of fetal and neonatal mortality risks in multiple gestations are invaluable for clinical decision making, as gestational age is usually known before delivery. Such data are also required in order to assess the merits of elective delivery at various gestations.

The aim of this study is to evaluate the gestation-specific risks of stillbirth in multiple gestations. The calculated risk for stillbirth in multiple gestations is compared with that for singletons from the same cohort.

Back to Top | Article Outline


Information relating to 4193 multiple births to residents of the North-East Thames region of London from 1989 to 1991 was obtained from a computerized database of the Regional Interactive Child Health System. Records of death registration were linked to birth notification records in 96% of cases of stillbirth and infant death. Details of data entry, linkage, aggregation, and validation are published elsewhere.13,15 Data for comparison with singleton pregnancies were taken from the same cohort over the same time and are published elsewhere.13,15 Agreement for release of data was obtained from each of the Data Custodians of the Child Health Data and the Local Research Ethics Committee. The institutional review board permitted limited access only to the mortality data from the data custodians; therefore, maternal demographics were not available. All 18 hospitals at the time of data collection were offering routine second trimester anomaly scans, but early dating scans were available selectively. At the time of data collection, none of the maternity units involved had a routine policy of elective delivery before 40 weeks in multiple gestations. Fetal surveillance was performed in line with current evidence at the time of the study. Only pregnancies progressing beyond 41 weeks' gestation were monitored with nonstress testing (n = 47).

Assessment of gestational age, whether based on maternal history or ultrasound data, was performed during the first or early second trimester, depending on when the mother booked for antenatal care. Redating was not done when the pregnancy outcome resulted in stillbirth. The gestational age recorded on birth notifications for stillbirths is that at delivery, which is only a proxy for gestational age at death. Data on chorionicity were not available from birth registration records. Notifications with gestations recorded as more than 45 completed weeks were excluded from the analysis. Risk of stillbirth was calculated as a proportion of the ongoing gestations at the onset of each week of gestation according to the following equation: risk of stillbirth at 38 weeks = number of stillbirths at 38 to 38+6 weeks/1000 ongoing gestations at the onset of the 38th week.

The data released did not allow separation of twins from triplets. However, twins accounted for 99.8% of all multiple pregnancies in our region, and so these data were interpreted as pertaining to twin pregnancies. In addition, few, if any, triplet pregnancies progress to 38 weeks' gestation and would not influence the outcome in this study. Additionally, risk was calculated per 1000 ongoing gestations rather than pregnancies (two gestations per pregnancy in twins), as each twin pregnancy has two gestations at risk of stillbirth. The latter correction is necessary in order not to inflate the apparent mortality for twins relative to singletons. Data from singleton pregnancies from the same region collected at the same time with identical exclusion criteria are presented for comparison.13 The statistics package, Confidence Interval Analysis (CIA 1.1, 1991, Martin Gardener, BMJ Publications, London, United Kingdom) was used for calculation of relative risks and confidence intervals.

Back to Top | Article Outline


Information on 4154 multiple gestations was available for analysis. The median gestation at delivery in multiple gestations was 37 weeks, with 6% of deliveries occurring after 39 weeks and 0.4% after 41 weeks (Figure 1). In contrast, the median gestation at delivery in singletons was 40 weeks, with 6.3% of deliveries occurring after 42 weeks (Figure 1).



The risk of fetal death in multiple gestations increased from one in 3333 ongoing gestations at 28 weeks to one in 69 ongoing gestations at 39+ weeks' gestation (Table 1 and Figure 2). The risk of stillbirth in multiple gestations at 39 weeks surpassed that of postterm singletons (one in 526) (Table 2 and Figure 2).

Table 1

Table 1

Table 2

Table 2



Back to Top | Article Outline


Our data show that the risk of stillbirth in multiple gestations increases progressively to exceed, by 39 weeks, that of postterm singleton pregnancy. The pattern of increase in fetal loss in multiple gestations mimics that of singletons, but the gestational age at which the risk increases is 2–3 weeks earlier, and the magnitude of the risk is significantly higher.

Previous studies reported elevated perinatal mortality rates in multiple gestations compared with singleton pregnancy.1–5 One such study of 138,779 twin gestations in the United States reported that the perinatal mortality rate in twins was significantly higher compared with singletons.3 That and other similar studies examined the perinatal mortality data collectively, independent of gestational age,1–5 which is known to be a major contributor to perinatal death. In contrast, other authors have demonstrated that, when comparing pregnancies of similar gestational age (gestation-specific data), twins on average weighed less but had a mortality rate equivalent to that of singletons.6–10 Those authors assumed that the gestation-specific stillbirth rate reflects the risk of fetal death. However, because stillbirth can only occur before delivery, the risk of fetal death is better represented as a proportion of ongoing pregnancies rather than total births.11–15 When analyzed in this manner, our data demonstrate a gestation-related trend for the risk of stillbirth. This increase in stillbirths is likely to be a consequence of complications related to the monochorionic placentation and intrapartum and peripartum events. Although data on chorionicity were not available in the current study, previous studies have found elevated perinatal mortality rates associated with monochorionic twin gestation.16 The introduction of first-trimester ultrasound has made this information now routinely available in most multiple gestations. The results of the current study might be different if the stillbirths are assessed for monochorionic and dichorionic gestations separately. Additionally, the excess mortality might partly be accounted for by fetal malformations, which would be less affected by gestational age at delivery. That information was not available to us.

The definition of “term” in singleton pregnancy (37–42 completed weeks' gestation) is used to indicate the most likely time for onset of spontaneous labor and delivery. “Term” also signifies the gestation beyond which induction of labor is undertaken with the intention of avoiding the fetal and neonatal risks of prolonged pregnancy.11–15,18 The findings of the present study suggest that multiple gestations normally deliver at 37–38 weeks, with the risks of fetal death at 39 or more weeks exceeding those of postterm singleton pregnancy. In support of this finding, a retrospective analysis of all deliveries in the United States between 1983 and 1988 reported that the lowest fetal mortality rate for singletons was at 40–41 weeks and 3700–4000 g (0.9 stillbirths per 1000 conceptions), and for twins it was 36–37 weeks and 1900–2200 g (5.2 stillbirths per 1000 conceptions).10 Similarly, Minakami et al,9 in an analysis of 4.9 million births in Japan between 1989 and 1992, found that the risk of fetal death exceeded the risk of neonatal death at 37 weeks' gestation in twins. Our study uses an “every twin for itself” approach, in keeping with previous analyses.6–10 Hence, calculations of risk are valid estimations of fetal death risk for either twin.

Although there was no policy of elective delivery for multiple gestations before 40 weeks in any of the maternity units studied, it is likely that some pregnancies were induced before term for perceived maternal or fetal risks. It is significant, therefore, that despite possible selective early delivery in our cohort, the risks of stillbirth continued to increase exponentially with advancing gestation. In the absence of effective techniques of surveillance at term to prevent fetal loss, a policy of elective delivery of multiple-gestation pregnancies by 39 weeks might be justified. Although preterm deliveries are potentially at increased risk of neonatal mortality and morbidity mainly because of pulmonary immaturity, our data indicate that the advantages of elective delivery by 39 weeks might outweigh the fetal and neonatal risks of an expectant approach. This hypothesis should be tested in randomized controlled trials, as with singleton pregnancy.18

The gestation-specific perinatal mortality rate is commonly used as a factor on which to base clinical decisions, under the misconception that it reflects the prevailing risks to the pregnancy. We found that when the true gestation-specific risks of stillbirth are calculated, multiple gestations at 37–38 weeks' gestation have risks of fetal mortality equivalent to those of postterm singleton pregnancy. As multiple gestations rarely proceed beyond 39 weeks, and because the fetal mortality rate increases several-fold beyond this gestational age, elective delivery might be justified.

Back to Top | Article Outline


1. Fleigner JRH. When do perinatal deaths in multiple pregnancies occur? Obstet Gynaecol 1989;29:371–4.
2. Howarth GR, Pattinson RC, De Jong G. Total perinatal-related wastage in twin pregnancies. S Afr Med J 1991;80:31–3.
3. Powers, Wampler NS. Further defining the risks confronting twins. Am J Obstet Gynecol 1996;175:1522–8.
4. West CR, Adi Y, Pharoah POD. Fetal and infant death in mono- and dizygotic twins in England and Wales 1982–1991. Arch Dis Child 1999;80:F217–20.
5. Bajoria R, Kingdom J. The case for routine determination of chorionicity and zygosity in multiple pregnancy. Prenat Diagn 1997;17:1207–25.
6. Gardner MO, Goldenberg RL, Cliver SP, Tucker JM, Nelson KG, Copper RL. The origin and outcome of preterm twin pregnancies. Obstet Gynecol 1995;85:553–7.
7. Kilpatrick SJ, Jackson R, Croughan-Minihane MS. Perinatal mortality in twins and singletons matched for gestational age at delivery at 30 weeks gestation or greater. Am J Obstet Gynecol 1996;174:66–71.
8. Dunn A, Macfarlane A. Recent trends in the incidence of multiple births and associated mortality in England and Wales. Arch Dis Child 1996;75:F10–9.
9. Minakami H, Kimura H, Honma Y, Tamada T, Sato I. When is the optimal time for delivery? – The fetal perspective. Gynaecol Obstet Invest 1995;40:174–8.
10. Luke B. Reducing stillbirths in multiple births: Optimal birthweights and gestational ages for infants of twin and triplet births. Acta Genet Med Gemellol 1996;45:333–48.
11. Yudkin PL, Wood L, Redman CWG. Risk of unexplained stillbirth at different gestational ages. Lancet 1987;1:1192–4.
12. Myers SA, Ferguson R. A population study of the relationship between fetal death and altered fetal growth. Obstet Gynecol 1989;74:325–31.
13. Hilder L, Costeloe K, Thilaganathan B. Prolonged pregnancy: Evaluating gestation-specific risks of fetal and infant mortality. Br J Obstet Gynaecol 1998;105:169–73.
14. Raymond EG, Cnattingius S, Kiely JL. Effects of maternal age, parity, and smoking on the risk of stillbirth. Br J Obstet Gynaecol 1994;101:301–6.
15. Cotzias CS, Paterson-Brown S, Fisk NM. Prospective risk of unexplained stillbirth in singleton pregnancies at term: Population based analysis. BMJ 1999;319:287–8.
16. Sebire NJ, Snijders RJM, Hughes K, Sepulveda W, Nicolaides KH. The hidden mortality of monochorionic twin pregnancies. Br J Obstet Gynaecol 1997;104:1203–7.
17. Hilder L, Alberman E. Monitoring fetal and infant survival using regional birth notification data. J Epidemiol Comm Health 1998;52:253–6.
18. Hannah ME, Hannah WJ, Hellmann J, Hewson S, Milner R, Willan A. Induction of labour as compared with serial antenatal monitoring in post-term pregnancy. N Engl J Med 1992;326:1587–92.
© 2002 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.