ArticlePDF Available

Abstract and Figures

To estimate the risk of stillbirth in a second pregnancy when previous stillbirth, preterm, and small-for-gestational age (SGA) births occurred in the previous pregnancy. This was a population-based cohort study in New South Wales Australia from 2002 to 2006. Singleton births in a first pregnancy were linked to a second pregnancy using data from the New South Wales Midwives Data Collection and the New South Wales Perinatal Death Database. Deaths were classified according to the Perinatal Mortality Classifications of the Perinatal Society of Australia and New Zealand. Crude and adjusted hazard ratios were estimated using a proportional hazards model. Delivery of an SGA newborn in the first pregnancy was associated with increased risks of stillbirth in a second pregnancy (hazard ratio 1.73, 95% confidence interval [CI] 1.15-2.60) and risk was further increased with prematurity (hazard ratio 5.65, 95% CI 1.76-18.12). Stillbirth in a first pregnancy had a nonsignificant association with stillbirth in the second pregnancy (hazard ratio 2.03, 95% CI 0.60-6.90). For women aged 30-34 years, the absolute risk of stillbirth up to 40 completed weeks of gestation was 4.84 per 1,000 among women whose first pregnancy was a stillbirth and 7.19 per 1,000 among women whose first pregnancy was preterm and SGA. Delivering an SGA and preterm neonate in a first pregnancy is associated with greater risks for stillbirth in a second pregnancy than delivering a previous stillbirth. All factors merit improved surveillance in a subsequent pregnancy, and research should address underlying factors common to all three outcomes.
Content may be subject to copyright.
Stillbirth Risk in a Second Pregnancy
Adrienne Gordon,
MBChB
,
MPH
, Camille Raynes-Greenow,
MPH
,
PhD
, Kevin McGeechan,
MBiostat
,
PhD
,
Jonathan Morris,
MBChB
,
PhD
, and Heather Jeffery,
MBBS
,
MRCP
OBJECTIVE: To estimate the risk of stillbirth in a second
pregnancy when previous stillbirth, preterm, and small-
for-gestational age (SGA) births occurred in the previous
pregnancy.
METHODS: This was a population-based cohort study in
New South Wales Australia from 2002 to 2006. Singleton
births in a first pregnancy were linked to a second preg-
nancy using data from the New South Wales Midwives Data
Collection and the New South Wales Perinatal Death Da-
tabase. Deaths were classified according to the Perinatal
Mortality Classifications of the Perinatal Society of Australia
and New Zealand. Crude and adjusted hazard ratios were
estimated using a proportional hazards model.
RESULTS: Delivery of an SGA newborn in the first preg-
nancy was associated with increased risks of stillbirth in a
second pregnancy (hazard ratio 1.73, 95% confidence
interval [CI] 1.15–2.60) and risk was further increased
with prematurity (hazard ratio 5.65, 95% CI 1.76–18.12).
Stillbirth in a first pregnancy had a nonsignificant associ-
ation with stillbirth in the second pregnancy (hazard ratio
2.03, 95% CI 0.606.90). For women aged 30–34 years,
the absolute risk of stillbirth up to 40 completed weeks of
gestation was 4.84 per 1,000 among women whose first
pregnancy was a stillbirth and 7.19 per 1,000 among
women whose first pregnancy was preterm and SGA.
CONCLUSION: Delivering an SGA and preterm neonate
in a first pregnancy is associated with greater risks for
stillbirth in a second pregnancy than delivering a previous
stillbirth. All factors merit improved surveillance in a
subsequent pregnancy, and research should address un-
derlying factors common to all three outcomes.
(Obstet Gynecol 2012;119:509–17)
DOI: 10.1097/AOG.0b013e31824781f8
LEVEL OF EVIDENCE: II
Stillbirth is estimated to occur in 2.65 million births
globally each year.1It is a tragedy both for the
family and their maternity health care provider. One
of the greatest challenges is how to counsel regarding
recurrence risk of stillbirth and management of a
subsequent pregnancy. Reported recurrence risks
vary widely with estimates ranging from a two to a
10-fold increase.2However, these data arise from
studies performed in the 1990s in selected or case–
control populations that did not perform adjusted
analyses and therefore their clinical usefulness is
limited.3,4 It is apparent that recurrence risks are
dependent on whether an initial cause of death was
established2,5 with some conditions such as those with
genetic causes associated with well-defined recur-
rence.6However, with much of recent data on still-
birth coming from large established population-based
data collections, it is unusual to be able to interpret
recurrence risks based on previous cause because
such data are often not available. The three largest
population-based studies to date do not contain any
information on cause of death.7–9
From being relatively understudied, stillbirth re-
search is increasing globally.10 There remains a rela-
tive paucity of data on recurrence. A recent meta-
analysis of population-based studies in high-income
countries pooled data from five studies (three cohorts
and two case–control studies) and demonstrated in-
creased risk of stillbirth after previous stillbirth with
an adjusted odds ratio (OR) of 2.61 (95% confidence
See related editorial on page 495.
From the Department of Neonatal Medicine, Royal Prince Alfred Hospital,
Sydney School of Public Health and Sydney Medical School, the University of
Sydney, and Perinatal Research, Kolling Institute of Medical Research, Sydney
University, Royal North Shore Hospital, St Leonards, New South Wales,
Australia.
Camille Raynes-Greenow was supported by a National Health and Medical
Research Council postdoctoral fellowship.
The authors thank Lee Taylor and the Centre for Epidemiology and Research at
the New South Wales Department of Health for providing the MDC and
perinatal death data. They also thank the Centre for Health Record Linkage for
performing probabilistic record linkage of the data sets.
Corresponding author: Adrienne Gordon, MBChB, MRCP, FRACP, MPH
(Hons), Department of Neonatal Medicine, Royal Prince Alfred Hospital,
Missenden Road, Camperdown, Sydney, NSW 2050, Australia; e-mail:
adrienne.gordon@email.cs.nsw.gov.au.
Financial Disclosure
The authors did not report any potential conflicts of interest.
© 2012 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins.
ISSN: 0029-7844/12
VOL. 119, NO. 3, MARCH 2012 OBSTETRICS & GYNECOLOGY 509
interval [CI] 1.504.55)11 There was variation in
results resulting in important statistical heterogeneity.
One large population-based study from the United
States demonstrated a large effect (adjusted OR 5.8,
95% CI 3.70–9.00)12 and two other studies showed no
difference.13,14
One major issue in interpreting available data on
recurrence risk is how factors known to be associated
with stillbirth are adjusted for. Recently exposures in
a previous pregnancy and how they relate to subse-
quent stillbirth have also been examined. Exposures
studied include: infant mortality, preterm birth,
growth restriction, and cesarean delivery. A summary
of the relevant literature, confounders adjusted for,
and whether these are from the first or second preg-
nancy is presented in Appendix 1 (available online
at http://links.lww.com/AOG/A280).3,4,7–9,12,13,15–27
For studies in which underlying causal mechanisms
are the focus, it seems appropriate to adjust for factors
known to be associated a priori with stillbirth and
present in a subsequent pregnancy. This, however,
includes factors known to be present only at or after
the time of delivery such as placental abruption or the
neonate being small for gestational age. For a family
who attends antenatal counseling in a pregnancy after
a previous stillbirth, this information is not yet known
when assessing their risk of a subsequent stillbirth.
The purpose of this study was to estimate clinically
useful appropriate risks for counseling families regarding
subsequent stillbirth using information available from
the previous pregnancy and delivery and known current
pregnancy risk factors.
MATERIALS AND METHODS
This was a statewide population-based cohort study
using deidentified linked data from two New South
Wales data sets: the New South Wales Midwives Data
Collection and the Perinatal Death Data from the
New South Wales Maternal and Perinatal Committee.
Ethical approval was obtained from the New South
Wales Department of Health Ethics Committee Ref-
erence No. DoHEC 2005-06-11. New South Wales is
the most populous state in Australia and its 90,000
births per annum account for approximately 30% of
the nation’s births.
The New South Wales Midwives Data Collection is
a mandated population-based surveillance system cov-
ering all births in New South Wales public and private
hospitals as well as home births.28 It encompasses all live
births and stillbirths of at least 20 weeks of gestation or
at least 400 g birth weight and performs well in valida-
tion studies.29 The Midwives Data Collection requires
the attending midwife or doctor to complete a notifica-
tion form when a birth occurs and collects demographic,
maternal health, pregnancy, labor, delivery, and perina-
tal outcomes data28 (Appendix 2, available online at
http://links.lww.com/AOG/A281). The majority of
these data is submitted electronically to the Depart-
ment of Health. The New South Wales Maternal and
Perinatal Committee is a quality assurance committee
established under the New South Wales Health Ad-
ministration Act 1982 and is privileged under this Act
to carry out confidential reviews of both maternal and
perinatal deaths.28 Members are appointed by the
Minister for Health.28 A subcommittee called the
Perinatal Outcomes Working Party reviews and classi-
fies perinatal deaths using the Perinatal Society of Aus-
tralia and New Zealand–Perinatal Mortality Classifica-
tion System, which has been documented to have a high
interobserver reliability with a
value of 0.83–0.95.30
Information available to the working party includes a
confidential report on perinatal death (Appendix 3,
available online at http://links.lww.com/AOG/A282),
postmortem, and placental pathology reports. These
reports are forwarded by local hospital perinatal
death review committees plus other information that
may be relevant, eg, microbiologic results. The clas-
sification system has been used throughout all New
South Wales since 2002 and formed part of a policy
directive, which mandated hospital procedures for
review and reporting of perinatal deaths.31 The direc-
tive included recommended investigations for still-
birth, a requirement for autopsy to be discussed and
offered to every family, and for the placenta to be
examined pathologically. The classifications assigned
by the Committee after review of the hospital infor-
mation comprise the final cause of death reported in
the population data.
Records for neonates born between 2002 and
2006 from the Midwives Data Collection and Perina-
tal Deaths data compiled by the New South Wales
Maternal and Perinatal Committee subgroup Perina-
tal Outcomes Working Party were linked using prob-
abilistic record linkage methods in Centre for Health
Record Linkage. Neonates were included if they were
at least 22 weeks of gestation. Cause of death classi-
fication was extracted from the Perinatal Death Data-
base on all stillborn fetuses that had been reviewed by
the Perinatal Outcomes Working Party. Perinatal
deaths classified as being the result of congenital
abnormality by the Perinatal Society of Australia and
New Zealand Perinatal Death Classification System
(see Appendix 3, http://links.lww.com/AOG/A282)
were excluded. This category includes deaths in
which a congenital abnormality (structural, func-
tional, or chromosomal) is considered to have made a
510 Gordon et al Stillbirth Risk in a Second Pregnancy OBSTETRICS & GYNECOLOGY
major contribution to the death and includes termi-
nations of pregnancy for congenital abnormality. De-
mographic data were collected from both data sets.
Gestational age was determined by certain dates
confirmed by ultrasonography before 20 weeks of
gestation or, if dates uncertain, by ultrasonography
alone before 20 weeks of gestation or, rarely, if both
are unavailable, by examination of the newborn.
Stillbirth was defined as a neonate born of at least 20
weeks of gestation or 400 g birth weight who did not,
at any time after delivery, breathe or show any
evidence of life such as a heartbeat. Cause of death
was defined using the Perinatal Society of Australia
and New Zealand Perinatal Death Classification (Ap-
pendix 3, http://links.lww.com/AOG/A282).32 Unex-
plained stillbirth was defined as the “death of a
normally formed fetus before the onset of labor where
no predisposing factors are considered likely to have
caused the death“ (Perinatal Society of Australia and
New Zealand Perinatal Death Classification of 10 or
11). Explained stillbirth was defined as a Perinatal
Society of Australia and New Zealand Perinatal Death
Classification of 2–9. Small for gestational age (SGA)
was defined as a birth weight less than the 10th
percentile using Australian population-based percen-
tile charts.33 Premature birth was defined as delivery
before 37 weeks of gestation. Further analysis of
prematurity and SGA in the first pregnancy was
performed after subclassifying births into four catego-
ries: appropriate-for-gestational-age term; appropri-
ate-for-gestational-age preterm; SGA term; and SGA
preterm.
Crude rates for stillbirth were computed as num-
ber of events divided by total number of births and
multiplied by 1,000. Univariable comparisons of di-
chotomous data for frequency distributions of second
pregnancy characteristics were made by use of the
2
test, the
2test for trend, or Fisher’s exact as appro-
priate.34 The Pvalues for all hypotheses tests were
two-sided, and statistical significance was set at P.05.
The risk of stillbirth was compared between groups
by time-to-event analyses using gestation as the time
scale and stillbirth as the event. All other births were
censored. Crude and adjusted hazard ratios were
estimated using a Cox proportional hazards model.
The proportional hazards assumption was tested us-
ing Schoenfeld residuals, which were plotted against
each covariate and the graphs inspected for any trend
in the residuals.35,36 Interactions between the log of
gestational age and each covariate was created and
added to the model individually to test for any departure
from the proportional hazards assumption.37 Records
with missing data for the variables analyzed were ex-
cluded. Statistical analysis was performed using SAS 9.2.
RESULTS
Of 176,260 women who had a first pregnancy in New
South Wales between 2002 and 2006, 53,607 had a
second pregnancy. After exclusion of missing data,
the study group for adjusted analysis of stillbirth risk
consisted of 52,110 second pregnancies (Fig. 1). Still-
birth rates for first, second, and third consecutive
pregnancies are shown in Figure 2. The first preg-
nancy stillbirth rate was 4.8 per 1,000. For women
whose first pregnancy resulted in a stillbirth, the rate
of stillbirth in their subsequent pregnancy was ap-
proximately doubled to 8.6 per 1,000. Conversely, for
women in whom the first neonate was liveborn, the
stillbirth rate in their second pregnancy was 2.8 per
1,000. There were very few stillbirths related to third
pregnancies in this cohort; however, increased risks
were seen in the third pregnancy relative to a first
pregnancy stillbirth (30 per 1,000) even when the
second neonate was liveborn.
Table 1 compares the frequency distribution of
second pregnancy characteristics between women
who had a stillbirth or not in their first pregnancy
using univariable analyses. There were significant
differences in maternal age with more women
younger than 20 years and 40 years or older in the
second pregnancy who had experienced a stillbirth in
the first pregnancy. There were also higher incidences
of both pre-existing and gestational diabetes, pre-
eclampsia, and smoking as well as a higher rate of
stillbirth in the subsequent pregnancy. Table 2 shows
the adjusted hazard ratios (HRs) for stillbirth in the
second pregnancy based on first pregnancy outcome
and characteristics of the second pregnancy. Stillbirth
in the first pregnancy did not significantly increase the
risk for the subsequent pregnancy when the other
confounders were adjusted for HR 2.03 (95% CI
0.606.90). Significant association remained for sub-
sequent stillbirth if the previous birth was SGA (HR
1.73, 95% CI 1.15–2.60) and a similar increase in risk,
although not reaching statistical significance for pre-
vious prematurity (HR 1.75, 95% CI 0.97–3.15).
There was some evidence (P.03) that the propor-
tional hazards assumption was not met for preterm
birth in the first pregnancy with the risk not constant
through gestation and higher before 32 weeks of
gestation. Allowing for this nonproportional hazards
with the inclusion of an interaction term did not affect
the estimates of the other HRs. We did not find an
increase in subsequent stillbirth with previous cesar-
ean delivery.
VOL. 119, NO. 3, MARCH 2012 Gordon et al Stillbirth Risk in a Second Pregnancy 511
The outcome of the first pregnancy was grouped
into five categories based on premature birth and fetal
growth restriction to further explore their association
with stillbirth (Table 3). Although numbers of still-
births in the second pregnancy were small, significant
risks were observed for previous SGA term births
(HR 1.66, 95% CI 1.01–2.74) and the risk increased if
the SGA neonate was also premature (HR 5.65, 95%
CI 1.76–18.12). The risk of subsequent stillbirth for
women who previously delivered a liveborn SGA
neonate at less than 37 weeks of gestation was greater
than the risk conferred by previous stillbirth. For
women aged 30–34 years (median age group) without
any of the risk factors listed in Table 2, the estimated
absolute risk of stillbirth at 40 weeks of gestation or
before for the second pregnancy was 2.39 per 1,000.
Absolute risks of stillbirth in the second pregnancy
related to first pregnancy outcomes were: 4.84 per
1,000 among women whose first pregnancy was a
stillbirth; 4.17 per 1,000 among women whose first
pregnancy was preterm but not SGA; and 4.12 per
1,000 among women whose first pregnancy was SGA
but not preterm. The highest risk was 7.19 per 1,000
among women whose first pregnancy was both pre-
term and SGA.
Perinatal Society of Australia and New Zealand
Classification of cause of death was available for 85%
of the first pregnancy stillbirths and 81% of the second
First pregnancies
n=176,260
Excluded: n=748
Gestational age missing or
outside range of 22–44
weeks: 369
Deaths from congenital
abnormality: 236
Fourth pregnancies: 143
Singleton, consecutive
pregnancies at or beyond 22
weeks of gestation; first birth
occurred in New South Wales
between 2002 and 2006
N=234,785
Second pregnancies
n=53,607
Included in study cohort
n=174,396
Third pregnancies
n=4,170
Included in study cohort*
n=52,110
Included in study cohort
n=3,993
Missing data on any variable
n=1,864
Missing data on any variable
n=1,497
Missing data on any variable
n=177
Fig. 1. Selection of study cohort from all singleton consecutive pregnancies at or beyond 22 weeks of gestation where first
birth within the study period (2002–2006) was in New South Wales. Total exclusions: 4,286 (1.8%). *Second pregnancy
used for calculation of hazard ratios for subsequent stillbirth.
Gordon. Stillbirth Risk in a Second Pregnancy. Obstet Gynecol 2012.
Singleton, consecutive pregnancies at or beyond 22 weeks gestation;
first birth occurred in New South Wales between 2002 and 2006
N=234,785
First pregnancy:
Stillbirths
n=737; 4.2/1,000
First pregnancy:
Live births
n=173,659
Second pregnancy:
Stillbirths
n=3; 8.6/1,000
Second pregnancy:
Live births
n=345
Second pregnancy:
Stillbirths
n=145; 2.8/1,000
Second pregnancy:
Live births
n=51,617
No births;
rate not calculable
Third pregnancy:
Stillbirths
n=2; 30/1,000
Third pregnancy:
Live births
n=63
Third pregnancy:
Stillbirths
n=5; 1.3/1,000
Third pregnancy:
Live births
n=43
Third pregnancy:
Live births
n=3,880
No stillbirths;
rate not calculable
Fig. 2. Occurrence and recurrence of stillbirth for consecutive singleton pregnancies in New South Wales, 2002–2006. For
first births, n174,396; second births, n52,110; and third births, n3,993.
Gordon. Stillbirth Risk in a Second Pregnancy. Obstet Gynecol 2012.
512 Gordon et al Stillbirth Risk in a Second Pregnancy OBSTETRICS & GYNECOLOGY
pregnancy stillbirths (Table 4); the remaining 15% and
19% could not be classified based on incomplete or not
forwarded confidential reports. Placental histopathology
was performed in 90% of the classified initial stillbirths
(560/624) and postmortem examination in 51% (317/
624). There remained substantial numbers of stillbirths
classified as unexplained despite these two investiga-
tions: 49% of the placental examinations (273 of 560)
and 52% of the postmortems (166 of 317). The risk of
stillbirth in the second pregnancy did not differ based on
whether the first stillbirth was explained or not; however
it must be noted that there were only three second
pregnancy stillbirths to women whose first fetus was
stillborn. The adjusted risk for previous explained
stillbirth was HR 1.82 (95% CI 0.24–13.75) com-
pared with previous unexplained stillbirth (HR
3.11, 95% CI 0.72–13.50).
DISCUSSION
This is a large population-based study from Australia
examining stillbirth recurrence in consecutive preg-
nancies. Although absolute risks of stillbirth are ap-
proximately doubled in a subsequent pregnancy, the
finding is not statistically significant after adjusting for
current pregnancy characteristics and previous preg-
nancy outcome. It confirms the findings of prior
research that previous pregnancy outcomes of SGA
and preterm birth are particularly important when
counseling regarding subsequent risks of stillbirth
and, in this cohort, conferred greater risk than a
previous stillbirth alone.
The strengths of this study include the large
record-linked population-based data set and the abil-
ity to assess consecutive pregnancies. New South
Wales is representative of national data comprising
one third of the country’s births and is generalizable
to other high-income countries. There were few miss-
ing data (1.8%) and the data collection has previously
been validated.29 The only other published popula-
tion-based Australian study to assess recurrence risk
for stillbirth was published in 2001 and was likely
underpowered.24 The standardized Perinatal Society
of Australia and New Zealand Perinatal Death Clas-
sification System has been used routinely in New
South Wales since 2002 and access to these classifica-
tions has enabled this analysis to have some signifi-
cant advantages. The first is the ability to exclude
deaths as a result of congenital abnormality from a
recurrence risk analysis. This exclusion is important
because 1) many of the diagnosed abnormalities may
result in a termination of pregnancy beyond 20 weeks
of gestation; 2) genetic abnormalities may have high
recurrence risks; and 3) many abnormalities are asso-
ciated with both prematurity and growth restriction.
All of these issues could lead to an overestimate of
recurrence risk of stillbirth when these deaths remain
in the data set. The second advantage is the ability to
assess future risk of stillbirth based on the initial cause
of death. We did not find any difference in subse-
quent risk based on whether the initial stillbirth was
explained or not; however, there were only three
second pregnancy stillbirths for women in whom the
first fetus was stillborn, so a definitive answer to this
question remains unanswered. There was also a high
proportion of unexplained deaths in the first preg-
nancy. The contribution of unexplained deaths in
Table 1. Comparison of Characteristics in the Second Pregnancy for Women With Stillbirth or Live Birth
in the First Pregnancy
Characteristics of Second Pregnancy
Stillbirth in First
Pregnancy (n348)
Livebirth in First
Pregnancy (n51,762) P*
Indigenous 8 (2.3) 1,074 (2.1) .77
Maternal age group (y)
Younger than 20 19 (5.5) 1,425 (2.7) .002
20–24 62 (17.8) 8,075 (15.6)
25–29 84 (24.1) 14,086 (27.2)
30–34 120 (34.5) 18,916 (36.5)
35–39 48 (13.8) 8,140 (15.7)
40 or older 15 (4.3) 1,120 (2.2)
Pre-existing diabetes 7 (2.0) 237 (0.46) .001
Pre-existing hypertension 4 (1.1) 426 (0.82) .50
Smoking in pregnancy 50 (14.4) 5,456 (10.5) .02
Gestational diabetes 26 (7.5) 1,838 (3.6) .001
Preeclampsia 30 (8.6) 1,531 (3.0) .001
Stillbirth in second pregnancy 3 (0.90) 145 (0.28) .042
Data are n (%) unless otherwise specified.
*
2
test,
2test for trend, or Fisher exact test where appropriate.
VOL. 119, NO. 3, MARCH 2012 Gordon et al Stillbirth Risk in a Second Pregnancy 513
studies of stillbirth is variable ranging from 10% to
70%38–40 and depends on the classification system
used.41 Recent classification systems that are focused
toward particular underlying causes or risks for still-
birth such as placental pathologies or growth restric-
tion yield low proportions of unexplained deaths.39,42
Earlier classification systems (eg, the extended Wiggles-
worth or modified Aberdeen classifications) were usu-
ally applied after minimal investigation of perinatal
deaths and resulted in high proportions of deaths classed
as unexplained.43–45 The Perinatal Society of Australia &
New Zealand Perinatal Death Classification definition of
unexplained antepartum deaths is “deaths of normally
formed fetuses before the onset of labor where no
predisposing factors are considered likely to have
caused the death.”32 Unlike unexplained sudden infant
death syndrome, this definition does not require autopsy
findings or a mandated set of investigations. The pro-
portion of stillbirths classified as unexplained is lower in
studies with an extensive set of investigations and higher
autopsy rates.38,46 New South Wales has a low autopsy
rate for stillbirths with the most recent report stating that
an autopsy was performed in only 30.8% of unexplained
antepartum deaths.28 Although the majority of the still-
births in this cohort had placental pathology performed
(90%), only half (51%) had a postmortem. We are
unable to assess the extent of other recommended
investigations for stillbirth because those data are not
routinely collected on a population basis. It is likely,
however, that many of the unexplained deaths in this
cohort were underinvestigated. The Perinatal Society of
Australia & New Zealand Perinatal Death Classification
forms are part of a clinical practice guideline on Perina-
tal Mortality published in 2005.32 Although there has
been national consensus to use the classification system,
the guideline also recommends a set of core investiga-
tions for stillbirth; however, uptake of the full guideline
is variable.47 When the recommended investigations are
performed in a tertiary hospital setting, the proportion of
deaths classified as unexplained using Perinatal Society
of Australia & New Zealand Perinatal Death Classifica-
tion has been shown to decrease from 34% to 13%.48
Because our cohort only extends until 2006, it is likely
that uptake of all aspects of the guideline at the time of
the data collection was inadequate.
There are inherent limitations in using popula-
tion-based data. Analysis is limited to what routine
variables are collected and we have not been able to
assess the effect of maternal obesity because prepreg-
nancy weight or body mass index are not routinely
collected in these population-based data sets. It is of
note, however, that the most recently published pop-
ulation-based study that assessed subsequent stillbirth
and previous infant mortality showed no increase in
stillbirth risk with maternal prepregnancy obesity
(HR 1.05, 95% CI 0.9–1.20).16 We also were unable to
assess the effect of interpregnancy interval because
Table 2. Adjusted Hazard Ratios for Stillbirth in
the Second Pregnancy by Second
Pregnancy Characteristics and First
Pregnancy Outcome in New South
Wales, 2002–2006
Maternal and
Obstetric
Characteristics
Multivariate
Hazard Ratio
95%
Confidence
Interval P
First pregnancy
outcome
Stillbirth
No Referent
Yes 2.03 0.60–6.90 .26
Small for gestational
age
No Referent
Yes 1.73 1.15–2.60 .009
Preterm
No Referent
Yes 1.75 0.97–3.15 .06
Cesarean delivery
No Referent
Yes 0.88 0.60–1.31 .53
Second pregnancy
predictors
Aboriginal or Torres
Strait Islander
No Referent
Yes 1.39 0.58–3.30 .46
Maternal age group
(y)
Younger than 20 0.86 0.35–2.10 .09
20–24 Referent
25–29 0.75 0.47–1.21
30–34 0.71 0.44–1.13
35–39 0.54 0.29–1.01
40 or older 1.85 0.80–4.26
Pre-existing diabetes
No Referent
Yes 3.53 0.85–14.70 .08
Pre-existing
hypertension
No Referent
Yes 2.82 0.88–9.04 .08
Smoking in
pregnancy
No Referent
Yes 1.44 0.91–2.29 .12
Gestational diabetes
No Referent
Yes 0.56 0.18–1.76 .32
Pre-eclampsia
No Referent
Yes 0.41 0.10–1.66 .21
514 Gordon et al Stillbirth Risk in a Second Pregnancy OBSTETRICS & GYNECOLOGY
the deidentification of the data set did not include
date of birth. We used a definition of SGA as a birth
weight less than the 10th percentile using Australian
population-based birth weight percentiles rather than
customized percentiles because data on maternal
weight and height are unavailable. This definition will
not identify all neonates who have failed to reach
their growth potential and will include both patholog-
ically and constitutionally small neonates; however, it
would be consistent with definitions used in other
population-based data. We were only able to use time
of delivery, not estimated time of death, which could
also affect classification of neonates as SGA. A further
limitation is that this cohort only comprises 5 years of
data and therefore there are relatively small numbers
of stillbirths seen for second pregnancies. A priori, we
aimed to perform this analysis with the ability to
analyze stillbirths classified using the Perinatal Society
of Australia & New Zealand classification system and
therefore could only start the analysis from 2002
onward when it was in routine use throughout the
state. The precision of the recurrence risk estimates of
stillbirth in second pregnancies are likely to be af-
fected by the small numbers as indicated by the wide
CIs. A database of over a million births would be
required to provide a more definitive answer. There
are potential benefits of analyzing recent data, which
is more representative of current obstetric practice. It
is possible that smaller recurrence risks than previ-
ously reported are being seen because these women
are already being managed differently in a subsequent
pregnancy. Findings within our data suggest that this
might be the case because although the proportions of
women diagnosed with gestational diabetes and pre-
eclampsia are greater in the subsequent pregnancy,
these factors are then not associated with the subsequent
second pregnancy stillbirth risk. It is also known that
many women with a previous stillbirth will be electively
Table 3. Stillbirth, Growth, and Prematurity in First Pregnancy and Adjusted Risk of Stillbirth During the
Second Pregnancy Among Women With Two Consecutive Singleton Births in New South
Wales 2002–2006
Outcome of First
Pregnancy
Second Pregnancy Births
P
Total
n
No. of Stillbirths
(Rate per 1,000)
Multivariate
Hazard Ratio*
95% Confidence
Interval
Liveborn and AGA
Term 44,917 116 (2.6) Referent
Preterm 2,446 8 (3.26) 1.41 0.69–2.90 .003
Liveborn and SGA
Term 4,024 18 (4.7) 1.66 1.01–2.74
Preterm 227 3 (13.0) 5.65 1.76–18.12
Stillbirth 348 3 (8.6) 3.74 1.18–11.81
AGA, average for gestational age; SGA, small for gestational age.
* Also adjusted for cesarean delivery in first pregnancy, indigenous status, maternal age, pre-existing diabetes, pre-existing
hypertension, gestational diabetes, pre-eclampsia, and smoking.
Table 4. Distribution of Cause of Death for Stillbirth in First and Second Pregnancies
Perinatal Society of Australia and
New Zealand Classification
First Pregnancy Second Pregnancy
No. of
Stillbirths
%of
Total
No. of
Stillbirths
%of
Total
Unexplained antepartum death 306 49 60 50
Spontaneous preterm 73 12 13 11
Hypertension 54 8.7 5 4.2
Fetal growth restriction 46 7.4 7 5.9
Antepartum hemorrhage 38 6.1 12 10
Maternal conditions 30 4.8 9 7.6
Specific perinatal conditions 27 4.3 4 3.4
Perinatal infection 27 4.3 4 3.4
Hypoxic peripartum death 21 3.4 5 4.2
No obstetric antecedent 2 0.32 0 0
Total 624 classified from 737
total stillbirths (85%)
119 classified from 148
total stillbirths (81%)
VOL. 119, NO. 3, MARCH 2012 Gordon et al Stillbirth Risk in a Second Pregnancy 515
induced or delivered in a subsequent pregnancy poten-
tially reducing second pregnancy risks.49
The finding of greater risk of subsequent stillbirth
with a previous SGA neonate adds to the increasing
literature that supports common underlying mecha-
nisms for stillbirth related to placental dysfunction.20,23,50
Mandating placental examination for every stillbirth
and improving adequate investigations of stillbirth will
enable population-based data to play a greater role in
assisting with further research into potential underlying
mechanisms but many of these research questions will
need to be answered in a basic science setting. The
finding that previous preterm birth increases the associ-
ation with subsequent stillbirth in the SGA neonates and
conveys stronger risks at less than 32 weeks of gestation
supports other literature that suggests recurrence of
stillbirth occurs within the same “gestational age win-
dow.”9,20 Our data suggest that in the absence of growth
restriction and preterm, birth women are at low risk of a
subsequent stillbirth.
In summary, we found that the delivery of an
SGA neonate in a first pregnancy increases the risk of
stillbirth in a second pregnancy, especially if the SGA
neonate was also premature. These previous preg-
nancy outcomes conferred greater risk than previous
stillbirth. Although absolute risks remain small, all
these factors merit improved surveillance and coun-
seling in a subsequent pregnancy. Future population-
based studies on stillbirth should include standardized
cause of death classifications to assess future risks
based on previous cause and also need to address how
women are managed in subsequent pregnancies and
whether different strategies affect outcome.
REFERENCES
1. Lawn JE, Blencowe H, Pattinson R, Cousens S, Kumar R,
Ibiebele I, et al. Stillbirths: Where? When? Why? How to
make the data count? Lancet 2011;377:144863.
2. Reddy UM. Prediction and prevention of recurrent stillbirth.
Obstet Gynecol 2007;110:1151–64.
3. Samueloff A, Xenakis EM, Berkus MD, Huff RW, Langer O.
Recurrent stillbirth. Significance and characteristics. J Reprod
Med 1993;38:883–6.
4. Greenwood R, Samms-Vaughan M, Golding J, Ashley D. Past
obstetric history and risk of perinatal death in Jamaica. Paedi-
atr Perinat Epidemiol 1994;8(suppl 1):40–53.
5. Management of stillbirth. ACOG Practice Bulletin No. 102.
American College of Obstetricians and Gynecologists. Obstet
Gynecol 2009;113:74861.
6. Wapner RJ, Lewis D. Genetics and metabolic causes of
stillbirth. Semin Perinatol 2002;26:704.
7. Bhattacharya S, Prescott GJ, Black M, Shetty A. Recurrence
risk of stillbirth in a second pregnancy. BJOG 2010;117:
1243–7.
8. Lykke JA, Paidas MJ, Langhoff-Roos J. Recurring complica-
tions in second pregnancy. Obstet Gynecol 2009;113:1217–24.
9. Melve KK, Skjaerven R, Rasmussen S, Irgens LM. Recurrence
of stillbirth in sibships: population-based cohort study. Am J
Epidemiol 2010;172:1123–30.
10. The Lancet.com. Stillbirths. Available at: www.thelancet.com/
series/stillbirth. Retrieved April 14, 2011.
11. Flenady V, Koopmans L, Middleton P, Froen JF, Smith GC,
Gibbons K, et al. Major risk factors for stillbirth in high-income
countries: a systematic review and meta-analysis. Lancet 2011;
377:1331–40.
12. Sharma PP, Salihu HM, Kirby RS. Stillbirth recurrence in a
population of relatively low-risk mothers. Paediatr Perinat
Epidemiol 2007;21(suppl 1):24–30.
13. Black M, Shetty A, Bhattacharya S. Obstetric outcomes subse-
quent to intrauterine death in the first pregnancy. BJOG
2008;115:269–74.
14. Flenady V. Causes and risk factors of stillbirth in Australia.
Perinatal Society of Australia and New Zealand Congress.
Gold Coast: Journal of Paediatrics and Child Health; 2008. p.
A2–10.
15. Salihu HM, Duan J, Nabukera SK, Mbah AK, Alio AP.
Younger maternal age (at initiation of childbearing) and recur-
rent perinatal mortality. Eur J Obstet Gynecol Reprod Biol
2011;154:31–6.
16. August EM, Salihu HM, Weldeselasse H, Biroscak B, Mbah A,
Alio AP. Infant mortality and subsequent risk of stillbirth: a
retrospective cohort study. BJOG 2011;118:163645.
17. Measey MA, Tursan d’Espaignet E, Charles A, Douglass C.
Unexplained fetal death: are women with a history of fetal loss
at higher risk? AustNZJObstet Gynaecol 2009;49:151–7.
18. Goy J, Dodds L, Rosenberg MW, King WD. Health-risk
behaviours: examining social disparities in the occurrence of
stillbirth. Paediatr Perinat Epidemiol 2008;22:314–20.
19. Lurie S, Eldar I, Glezerman M, Sadan O. Pregnancy outcome
after stillbirth. J Reprod Med 2007;52:289–92.
20. Smith GS, Shah I, White IR, Pell JP, Dobbie R. Previous
preeclampsia, preterm delivery, and delivery of a small for
gestational age infant and the risk of unexplained stillbirth in
the second pregnancy: a retrospective cohort study, Scotland,
1992–2001. Am J Epidemiol 2007;165:194–202.
21. Sharma PP, Salihu HM, Oyelese Y, Ananth CV, Kirby RS. Is
race a determinant of stillbirth recurrence? Obstet Gynecol
2006;107:391–7.
22. Onwude JL, Eisman V, Selo-Ojeme DO. Recurrent stillbirths:
a matched case–control study of unexplained stillbirths at
term. J Obstet Gynaecol 2006;26:205–7.
23. Surkan PJ, Stephansson O, Dickman PW, Cnattingius S.
Previous preterm and small-for-gestational-age births and the
subsequent risk of stillbirth. N Engl J Med 2004;350:777–85.
24. Robson S, Chan A, Keane RJ, Luke CG. Subsequent birth
outcomes after an unexplained stillbirth: preliminary popula-
tion-based retrospective cohort study. AustNZJObstet
Gynaecol 2001;41:29–35.
25. Heinonen S, Kirkinen P. Pregnancy outcome after previous
stillbirth resulting from causes other than maternal conditions
and fetal abnormalities. Birth 2000;27:33–7.
26. Weeks JW, Asrat T, Morgan MA, Nageotte M, Thomas SJ,
Freeman RK. Antepartum surveillance for a history of still-
birth: when to begin? Am J Obstet Gynecol 1995;172:486–92.
27. Freeman RK, Dorchester W, Anderson G, Garite TJ. The
significance of a previous stillbirth. Am J Obstet Gynecol
1985;151:7–13.
28. Centre for Epidemiology and Research. New South Wales
mothers and babies 2008. New South Wales (Australia): NSW
Department of Health; 2010.
516 Gordon et al Stillbirth Risk in a Second Pregnancy OBSTETRICS & GYNECOLOGY
29. NSW Health Department. Validation study: NSW midwives
data collection 1998. NSW Public Health Bull 2000;9:97–9.
30. Chan A, King JF, Flenady V, Haslam RH, Tudehope DI.
Classification of perinatal deaths: development of the Austra-
lian and New Zealand classifications. J Paediatr Child Health
2004;40:340–7.
31. NSW Ministry of Health Policy Directive (2006). Hospital proce-
dures for review and reporting of perinatal deaths. Available at
www.health.nsw.gov.au/archive/policies/pd/2006/PD2006_006.
html. Retrieved October 1, 2011.
32. Flenady V, King J, Charles A, Gardener G, Ellwood D, Day K,
et al. PSANZ clinical practice guideline for perinatal mortality.
Version 2.2. Available at: www.psanzpnmsig.org, 2009.
Retrieved October 6, 2011.
33. Roberts CL, Lancaster PA Australian national birthweight
percentiles by gestational age. Med J Aust 1999;170:1148.
34. Armitage P, Berry G, Matthews JNS. Trends in proportions.
In: Malden M, editor. Statistical methods in medical research.
4th ed. Oxford (UK): Blackwell Science; 2002.
35. Grambsch PM, Therneau TM. Proportional hazards tests and
diagnostics based on weighted residuals. Biometrika 1994;81:
515–26.
36. Platt RW, Joseph KS, Ananth CV, Grondines J, Abrahamow-
icz M, Kramer MS. A proportional hazards model with
time-dependent covariates and time-varying effects for analysis
of fetal and infant death. Am J Epidemiol 2004;160:199–206.
37. Hosmer DS, Lemeshow SM. Applied survival analysis: regres-
sion modeling of time-to-event data.2nd ed. Chichester (UK):
Wiley; 2008.
38. Petersson K, Bremme K, Bottinga R, Hofsjo¨ A, Hulthe´ n-Varli
I, Kublickas M, et al. Diagnostic evaluation of intrauterine fetal
deaths in Stockholm 1998–99. Acta Obstet Gynecol Scand
2002;81:284–92.
39. Korteweg FJ, Gordijn SJ, Timmer A, Erwich JJ, Bergman KA,
Bouman K, et al. The Tulip classification of perinatal mortality:
introduction and multidisciplinary inter-rater agreement.
BJOG 2006;113:393–401.
40. Maternal and Child Health Research Consortium. Confiden-
tial enquiry into stillbirths and deaths in infancy (CESDI). 8th
annual report. London (UK): Maternal and Child Health
Research Consortium; 2001.
41. Gordon A, Jeffery HE. Classification and description of still-
births in New South Wales, 2002–2004. Med J Aust 2008;188:
645–8.
42. Gardosi J, Kady SM, McGeown P, Francis A, Tonks A. Classifi-
cation of stillbirth by relevant condition at death (ReCoDe):
population based cohort study. BMJ 2005;331:1113–7.
43. Wigglesworth J. Monitoring perinatal mortality. A pathophys-
iological approach. Lancet 1980;2:6846.
44. Cole SK, Hey EN, Thomson AM. Classifying perinatal death:
an obstetric approach. Br J Obstet Gynaecol 1986;93:1204–12.
45. Hey EN, Lloyd DJ, Wigglesworth JS. Classifying perinatal
death: fetal and neonatal factors. Br J Obstet Gynaecol 1986;
93:1213–23.
46. Pauli RM, Reiser CA. Wisconsin Stillbirth Service Program: II.
Analysis of diagnoses and diagnostic categories in the first
1,000 referrals. Am J Med Genet 1994;50:135–53.
47. Uptake of the PSANZ perinatal mortality audit guidelines: a
survey of midwives and obstetricians in Australia and New
Zealand. Melbourne (Australia): Perinatal Society of Australia
and New Zealand Congress; 2007.
48. Headley E, Gordon A, Jeffery H. Reclassification of unex-
plained stillbirths using clinical practice guidelines. AustNZJ
Obstet Gynaecol 2009;49:285–9.
49. Robson SJ, Leader LR. Management of subsequent pregnancy
after an unexplained stillbirth. J Perinatol 2010;30:305–10.
50. Wikstro¨m A-K, Svensson T, Kieler H, Cnattingius S. Recur-
rence of placental dysfunction disorders across generations.
Am J Obstetr Gynecol 2011;205:454.e1–8.
VOL. 119, NO. 3, MARCH 2012 Gordon et al Stillbirth Risk in a Second Pregnancy 517
... Increased maternal age, non-Western ethnic origin, use of alcohol or tobacco and maternal comorbidities as obesity, hypertension, diabetes, preterm birth and small for gestational age fetus (SGA) are risk factors for stillbirth [2]. A history of stillbirth may independently be associated with an increased risk of recurrent stillbirth; previous studies showed an increased risk of fetal death varying from zero to tenfold [3][4][5][6][7][8]. ...
Article
Full-text available
Background A history of stillbirth is a risk factor for recurrent fetal death in a subsequent pregnancy. Reported risks of recurrent fetal death are often not stratified by gestational age. In subsequent pregnancies increased rates of medical interventions are reported without evidence of perinatal benefit. The aim of this study was to estimate gestational-age specific risks of recurrent stillbirth and to evaluate the effect of obstetrical management on perinatal outcome after previous stillbirth. Methods A retrospective cohort study in the Netherlands was designed that included 252.827 women with two consecutive singleton pregnancies (1st and 2nd delivery) between 1999 and 2007. Data was obtained from the national Perinatal Registry and analyzed for pregnancy outcomes. Fetal deaths associated with a congenital anomaly were excluded. The primary outcome was the occurrence of stillbirth in the second pregnancy stratified by gestational age. Secondary outcome was the influence of obstetrical management on perinatal outcome in a subsequent pregnancy. Results Of 252.827 first pregnancies, 2.058 pregnancies ended in a stillbirth (8.1 per 1000). After adjusting for confounding factors, women with a prior stillbirth have a two-fold higher risk of recurrence (aOR 1.96, 95% CI 1.07–3.60) compared to women with a live birth in their first pregnancy. The highest risk of recurrence occurred in the group of women with a stillbirth in early gestation between 22 and 28 weeks of gestation (a OR 2.25, 95% CI 0.62–8.15), while after 32 weeks the risk decreased. The risk of neonatal death after 34 weeks of gestation is higher in women with a history of stillbirth (aOR 6.48, 95% CI 2.61–16.1) and the risk of neonatal death increases with expectant obstetric management (aOR 10.0, 95% CI 2.43–41.1). Conclusions A history of stillbirth remains an important risk for recurrent stillbirth especially in early gestation (22–28 weeks). Women with a previous stillbirth should be counselled for elective induction in the subsequent pregnancy at 37–38 weeks of gestation to decrease the risk of perinatal death.
... Likewise, analysis of the Norwegian birth register between 1996 and 2013 containing >700,000 women from their first to second pregnancy showed the risk of preterm preeclampsia is 4-7 times higher in women with prior preterm birth without preeclampsia compared with women with prior term birth (19). Other studies report that the risk of stillbirth is significantly increased in women with a history of preeclampsia or premature birth in prior pregnancy (20,21), whereas preterm birth risk is also increased in women with a history of stillbirth in a prior pregnancy (22,23). ...
Article
Full-text available
In the fifteen minutes it takes to read this short commentary, more than 400 babies will have been born too early, another 300 expecting mothers will develop preeclampsia, and 75 unborn third trimester fetuses will have died in utero (stillbirth). Given the lack of meaningful progress in understanding the physiological changes that occur to allow a healthy, full term pregnancy, it is perhaps not surprising that effective therapies against these great obstetrical syndromes that include prematurity, preeclampsia, and stillbirth remain elusive. Meanwhile, pregnancy complications remain the leading cause of infant and childhood mortality under age five. Does it have to be this way? What more can we collectively, as a biomedical community, or individually, as clinicians who care for women and newborn babies at high risk for pregnancy complications, do to protect individuals in these extremely vulnerable developmental windows? The problem of pregnancy complications and neonatal mortality is extraordinarily complex, with multiple unique, but complementary perspectives from scientific, epidemiological and public health viewpoints. Herein, we discuss the epidemiology of pregnancy complications, focusing on how the outcome of prior pregnancy impacts the risk of complication in the next pregnancy — and how the fundamental immunological principle of memory may promote this adaptive response.
... Lastly, subsequent pregnancy outcomes depend heavily on the outcome of previous pregnancies where each birth is not independent of births [52][53][54]. An anticipated complication of our analysis that will impact on the interpretation of results is the absence of a unique identifier for mothers to account for potential clustering. ...
Article
Full-text available
Background Despite advances in the care of women and their babies in the past century, an estimated 1.7 million babies are born still each year throughout the world. A robust method to estimate a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform decision-making around the timing of birth to reduce the risk of stillbirth from 35 weeks of gestation in Australia, a high-resource setting. Methods This is a protocol for a cross-sectional study of all late-pregnancy births in Australia (2005–2015) from 35 weeks of gestation including 5188 stillbirths among 3.1 million births at an estimated rate of 1.7 stillbirths per 1000 births. A multivariable logistic regression model will be developed in line with current TransparentReporting of a multivariable prediction model forIndividualPrognosis orDiagnosis (TRIPOD) guidelines to estimate the gestation-specific probability of stillbirth with prediction intervals. Candidate predictors were identified from systematic reviews and clinical consultation and will be described through univariable regression analysis. To generate a final model, elimination by backward stepwise multivariable logistic regression will be performed. The model will be internally validated using bootstrapping with 1000 repetitions and externally validated using a temporally unique dataset. Overall model performance will be assessed with R², calibration, and discrimination. Calibration will be reported using a calibration plot with 95% confidence intervals (α = 0.05). Discrimination will be measured by the C-statistic and area underneath the receiver-operator curves. Clinical usefulness will be reported as positive and negative predictive values, and a decision curve analysis will be considered. Discussion A robust method to predict a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform timely, appropriate care to reduce stillbirth. Among existing prediction models designed for obstetric use, few have been subject to internal and external validation and many fail to meet recommended reporting standards. In developing a risk prediction model for late-gestation stillbirth with both providers and pregnant women in mind, we endeavor to develop a validated model for clinical use in Australia that meets current reporting standards.
... Lastly, subsequent pregnancy outcomes depend heavily on the outcome of previous pregnancies where each birth is not independent of births (52)(53)(54). An anticipated complication of our analysis that will impact interpretation of results is the absence of a unique identi er for mothers to account for potential clustering. ...
Preprint
Full-text available
Background: Despite advances in the care of women and their babies in the past century, an estimated 1.7 million babies are born still each year throughout the world. A robust method to estimate a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform decision-making around timing of birth to reduce the risk of stillbirth from 35 weeks gestation in Australia, a high-resource setting. Methods: This is a protocol for a cross-sectional study of all late-pregnancy births in Australia (2005-2015) from 35 weeks gestation including 5,188 stillbirths among 3.1 million births at an estimated rate of 1.7 stillbirths per 1000 births. A multivariable logistic regression model will be developed in line with current Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) guidelines to estimate the gestation-specific probability of stillbirth with prediction intervals. Candidate predictors were identified from systematic reviews and clinical consultation and will be described through univariable regression analysis. To generate a final model, elimination by backward stepwise multivariable logistic regression will be performed. The model will be internally validated using bootstrapping with 1000 repetitions and externally validated using a temporally unique dataset. Overall model performance will be assessed with R², calibration, and discrimination. Calibration will be reported using a calibration plot with 95% confidence intervals (α=0.05). Discrimination will be measured by the C-statistic and area underneath the receiver-operator curves. Clinical usefulness will be reported as positive and negative predictive values and a decision curve analysis will be considered. Discussion: A robust method to predict a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform timely, appropriate care to reduce stillbirth. Among existing prediction models designed for obstetric use, few have been subject to internal and external validation and many fail to meet recommended reporting standards. In developing a risk prediction model for late-gestation stillbirth with both providers and pregnant women in mind, we endeavor to develop a validated model for clinical use in Australia that meets current reporting standards.
... Lastly, subsequent pregnancy outcomes depend heavily on the outcome of previous pregnancies where each birth is not independent of births (51)(52)(53). An anticipated complication of our analysis that will impact interpretation of results is the absence of a unique identi er for mothers to account for potential clustering. ...
Preprint
Full-text available
Background: Despite advances in the care of women and their babies in the past century, an estimated 1.7 million babies are born still each year throughout the world. A robust method to estimate a pregnant woman’s individualized risk of late-pregnancy stillbirth may inform decision-making around timing of birth to reduce the risk of stillbirth from 35 weeks gestation in Australia, a high-resource setting. Methods: This is a protocol for a cross-sectional study of all late-pregnancy births in Australia (2005-2015) from 35 weeks gestation including 5,188 stillbirths among 3.1 million births at an estimated rate of 1.7 stillbirths per 1000 births. A multivariable logistic regression model will be developed in line with current Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) guidelines to estimate the gestation-specific probability of stillbirth with prediction intervals. Candidate predictors were identified from systematic reviews and clinical consultation and will be assessed through univariable regression analysis. To generate a final model, elimination by backward stepwise multivariable logistic regression will be performed. The model will be internally validated using bootstrapping with 1000 repetitions and externally validated using a temporally unique dataset. Overall model performance will be assessed with R², calibration, and discrimination. Calibration will be reported using a calibration plot with 95% confidence intervals (α=0.05). Discrimination will be measured by the C-statistic and area underneath the receiver-operator curves. Clinical usefulness will be reported as positive and negative predictive values and a decision curve analysis will be considered. Discussion: A robust method to predict a pregnant woman’s individualized risk of late-pregnancy stillbirth may inform timely, appropriate care to reduce stillbirth. Among existing prediction models designed for maternity use, few have been subject to internal and external validation and many fail to meet recommended reporting standards. In developing a risk prediction model for late-gestation stillbirth with both providers and pregnant women in mind, we endeavor to develop a validated model for clinical use in Australia that meets current reporting standards.
... Lastly, subsequent pregnancy outcomes depend heavily on the outcome of previous pregnancies where each birth is not independent of births (44)(45)(46). An anticipated complication of our analysis that will impact interpretation of results is the absence of a unique identifier for mothers to account for potential clustering. ...
Preprint
Full-text available
Background Despite advances in the care of women and their babies in the past century, an estimated 1.7 million babies are born still each year throughout the world. A robust method to estimate a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform decision-making around timing of birth to reduce the risk of stillbirth from 35 weeks gestation. Methods This is a protocol for a retrospective cohort study of all late-pregnancy births in Australia (1998-2015) from 35 weeks gestation including 7,200 stillbirths among 4.9 million births at an estimated rate of 1.47 stillbirths per 1000 live births. A multivariable logistic regression model will be developed in line with current T ransparent R eporting of a multivariable prediction model for I ndividual P rognosis or D iagnosis (TRIPOD) guidelines to estimate the gestation-specific probability of stillbirth with prediction intervals. Candidate predictors were identified from systematic reviews and clinical consultation and will be assessed through univariate regression analysis. To generate a final model, elimination by backward stepwise logistic regression will be performed. The model will be internally validated using K-fold cross-validation and externally validated using a geographically unique dataset. Overall model performance will be assessed with R 2 in addition calibration and discrimination. Calibration will be visualized using a calibration plot. Discrimination will be measured by the C- statistic and visualized using area underneath the receiver-operator curves (AUROC). Clinical usefulness will be reported as positive and negative predictive values and a decision curve analysis will be considered. Discussion A robust method to predict a pregnant woman’s individualized risk of late-pregnancy stillbirth is needed to inform timely, appropriate care to reduce stillbirth. Among existing prediction models designed for obstetric use, few have been subject to internal and external validation and many fail to meet recommended reporting standards. In developing a risk prediction model for late-gestation stillbirth with both providers and pregnant women in mind, we endeavor to develop a validated model for clinical use in Australia that meets current reporting standards.
Article
Full-term deliveries are defined as occurring between 39 weeks and 40 weeks and 6 days. Because contemporary research suggests improved outcomes with delivery in the term period compared with the early term period, nonindicated delivery should be pursued no earlier than 39 weeks. There are, however, multiple medical, obstetric, and fetal indications for delivery before 39 weeks, and the obstetric provider must weigh the risks and benefits of delivery versus expectant management on both the mother and fetus. This review serves to provide a basic framework of evidentiary support toward optimizing the term delivery.
Article
Stillbirth is one of the most common adverse pregnancy outcomes, occurring in 1 in 160 deliveries in the United States. In developed countries, the most prevalent risk factors associated with stillbirth are non-Hispanic black race, nulliparity, advanced maternal age, obesity, preexisting diabetes, chronic hypertension, smoking, alcohol use, having a pregnancy using assisted reproductive technology, multiple gestation, male fetal sex, unmarried status, and past obstetric history. Although some of these factors may be modifiable (such as smoking), many are not. The study of specific causes of stillbirth has been hampered by the lack of uniform protocols to evaluate and classify stillbirths and by decreasing autopsy rates. In any specific case, it may be difficult to assign a definite cause to a stillbirth. A significant proportion of stillbirths remains unexplained even after a thorough evaluation. Evaluation of a stillbirth should include fetal autopsy; gross and histologic examination of the placenta, umbilical cord, and membranes; and genetic evaluation. The method and timing of delivery after a stillbirth depends on the gestational age at which the death occurred, maternal obstetric history (eg, previous hysterotomy), and maternal preference. Health care providers should weigh the risks and benefits of each strategy in a given clinical scenario and consider available institutional expertise. Patient support should include emotional support and clear communication of test results. Referral to a bereavement counselor, peer support group, or mental health professional may be advisable for management of grief and depression.
Article
Objective: In attempt to shed new light on the etiopathogenesis of fetal growth restriction (FGR) we aimed to compare pregnancy outcomes and placental histopathology in cases of first vs. subsequent FGR occurrence. Study design: Pregnancy and placental reports of FGR pregnancies (defined by birth weight <10th percentile), born between 2008 and 2018 were reviewed. Included only cases with recurrent FGR, in two consecutive pregnancies, thus each subject served as her own control in two FGRs consecutive pregnancies. Neonatal outcome and placental histopathology were compared between the first FGR delivery (first FGR group) and the subsequent FGR delivery (subsequent FGR group). Composite adverse neonatal outcome was defined as one or more early neonatal complications. Results: Included in the study a total of 96 cases with recurrence of FGR pregnancies. Placentas from the first FGR group were characterized by higher rate of maternal vascular malperfusion (MVM) lesions as compared with the subsequent FGR group (71.8% versus 55.2%, respectively, p = 0 .02). Adverse neonatal outcome was more prevalent in the first FGR group as compared to the recurrent FGR group (41.6% versus 25%, respectively, p = 0.02). After controlling for confounders, using multivariate regression analysis, placental MVM lesions (aOR = 1.36, 95% CI = 1.12-1.45) and composite adverse neonatal outcome (aOR = 1.18 95% CI = 1.09-1.55) were found to be independently associated with the first FGR group. Conclusion: First event of FGR is associated with a higher rate of placental MVM lesions and adverse neonatal outcome as compared to FGR in subsequent pregnancies.
Article
SUMMARY Nonproportional hazards can often be expressed by extending the Cox model to include time varying coefficients; e.g., for a single covariate, the hazard function for subject i is modelled as exp {β(t)Zi(t)}. A common example is a treatment effect that decreases with time. We show that the function βi(t) can be directly visualized by smoothing an appropriate residual plot. Also, many tests of proportional hazards, including those of Cox (1972), Gill & Schumacher (1987), Harrell (1986), Lin (1991), Moreau, O'Quigley & Mesbah (1985), Nagelkerke, Oosting & Hart (1984), O'Quigley & Pessione (1989), Schoenfeld (1980) and Wei (1984) are related to time-weighted score tests of the proportional hazards hypothesis, and can be visualized as a weighted least-squares line fitted to the residual plot.
Article
THE MOST PRACTICAL, UP-TO-DATE GUIDE TO MODELLING AND ANALYZING TIME-TO-EVENT DATANOW IN A VALUABLE NEW EDITION Since publication of the first edition nearly a decade ago, analyses using time-to-event methods have increase considerably in all areas of scientific inquiry mainly as a result of model-building methods available in modern statistical software packages. However, there has been minimal coverage in the available literature to9 guide researchers, practitioners, and students who wish to apply these methods to health-related areas of study. Applied Survival Analysis, Second Edition provides a comprehensive and up-to-date introduction to regression modeling for time-to-event data in medical, epidemiological, biostatistical, and other health-related research. This book places a unique emphasis on the practical and contemporary applications of regression modeling rather than the mathematical theory. It offers a clear and accessible presentation of modern modeling techniques supplemented with real-world examples and case studies. Key topics covered include: variable selection, identification of the scale of continuous covariates, the role of interactions in the model, assessment of fit and model assumptions, regression diagnostics, recurrent event models, frailty models, additive models, competing risk models, and missing data. Features of the Second Edition include: Expanded coverage of interactions and the covariate-adjusted survival functions The use of the Worchester Heart Attack Study as the main modeling data set for illustrating discussed concepts and techniques New discussion of variable selection with multivariable fractional polynomials Further exploration of time-varying covariates, complex with examples Additional treatment of the exponential, Weibull, and log-logistic parametric regression models Increased emphasis on interpreting and using results as well as utilizing multiple imputation methods to analyze data with missing values New examples and exercises at the end of each chapter Analyses throughout the text are performed using Stata Version 9, and an accompanying FTP site contains the data sets used in the book. Applied Survival Analysis, Second Edition is an ideal book for graduate-level courses in biostatistics, statistics, and epidemiologic methods. It also serves as a valuable reference for practitioners and researchers in any health-related field or for professionals in insurance and government.
Article
Singleton survivors born to multigravidae in the whole island of Jamaica in 2 months (September-October 1986) were compared with singleton perinatal deaths occurring to multigravidae throughout the island in the 12-month period September 1986 to August 1987. Past obstetric history was obtained from the mothers using a structured questionnaire. Deaths were categorised using the Wigglesworth classification. Logistic regression was used to compare current outcomes in women who had had at least one previous pregnancy.
Article
An adequate fetomaternal circulatory system may be compromised by a variety of disturbances leading to stillbirth. The purpose of this study was to assess subsequent pregnancy outcome in women with a history of stillbirth as a result of causes other than maternal conditions and fetal abnormalities. Ninety-two deliveries after stillbirth were identified among 11,910 deliveries of parous women recorded in the birth registry at Kuopio, Finland. Using logistic regression, pregnancy outcome measures were compared with those of a parous healthy obstetric population (n = 11,818). Women with a history of stillbirth as a result of causes other than maternal conditions and fetal abnormalities were older than their unaffected controls (32.4 yr vs 30.3 yr). Stillbirth in an earlier pregnancy was associated with a significantly higher (p < 0.001) frequency of placental abruption in subsequent pregnancy (5.4% vs 0.7%). A history of stillbirth was predictive of preterm delivery (OR = 2.25) and low-birthweight infants (OR = 2.70). No recurrence was reported. Pregnancy with a history of stillbirth as a result of causes other than maternal conditions and fetal abnormalities is a moderate risk state, with prematurity and low-birthweight rates somewhat higher than those in the general population. The overall probability of a favorable outcome is good. These findings may be useful in counseling pregnant women with a history of stillbirth.
Article
To examine the association between infant mortality in a first pregnancy and risk for stillbirth in a second pregnancy. Population-based, retrospective cohort study. Maternally linked cohort data files for the state of Missouri. Women who had two singleton pregnancies in Missouri during the period 1989-2005 (n = 320 350). Women whose first pregnancy resulted in infant death were compared with those whose infant from the first pregnancy survived the first year of life. The Kaplan-Meier product limit estimator was employed to compare probabilities for stillbirth in the second pregnancy between both groups of women. Adjusted hazard ratios (AHRs) and 95% confidence intervals (95% CIs) were generated to assess the association between infant mortality in the first pregnancy and stillbirth in the second pregnancy. Exposure was defined as infant mortality in the first pregnancy, and the outcome was defined as stillbirth in the second pregnancy. Women with prior infant deaths were about three times as likely to experience stillbirth in their subsequent pregnancy (AHR 2.91; 95% CI 2.02-4.18). White women with a previous infant death were nearly twice as likely to experience a subsequent stillbirth, compared with white women with a surviving infant (AHR 1.96; 95% CI 1.13-3.39). Black women with a previous infant death were more than four times as likely to experience subsequent stillbirth, compared with black women with a surviving infant (AHR 4.28; 95% CI 2.61-6.99). Previous infant mortality results in an elevated risk for subsequent stillbirth, with the most profound increase observed among black women. Interconception care should consider prior childbearing experiences to avert subsequent fetal loss.