The Role of Uterine Closure in the Risk of
Uterine Rupture
Emmanuel Bujold, MD, MSc, FRCSC, Martine Goyet, MD, FRCSC, Sylvie Marcoux, MD, PhD,
Normand Brassard, MD, MBA, FRCSC, Béatrice Cormier, MD, Emily Hamilton, MD, FRCSC,
Belkacem Abdous, PhD, Elhadji A. Laouan Sidi, MSc, Robert Kinch, MD, FRCSC,
Louise Miner, MD, FRCSC, André Masse, MD, FRCSC, Claude Fortin, MD, FRCSC,
Guy-Paul Gagné, MD, FRCSC, André Fortier, MD, FRCSC, Gilles Bastien, MD, PhD, CCFP,
Robert Sabbah, MD, FRCSC, Pierre Guimond, MD, FRCSC, Stéphanie Roberge, MSc,
and Robert J. Gauthier, MD, FRCSC
OBJECTIVE: To evaluate the effects of prior single-layer
compared with double-layer closure on the risk of uter-
ine rupture.
From the Department of Obstetrics & Gynaecology, Faculty of Medicine, Centre
Hospitalier Universitaire de Québec, Université Laval, Québec, Canada, the
Department of Social and Preventive Medicine, Faculty of Medicine, Université
Laval, Québec, Canada, and the Department of Obstetrics & Gynaecology, Faculty
of Medicine, Hôpital Sainte-Justine, Université de Montréal, Montréal, Quebec,
Canada; the Department of Obstetrics & Gynaecology, Faculty of Medicine, Hôpital
Sainte-Justine, Université de Montréal, Montréal, Quebec Canada, and the Depart-
ment of Obstetrics & Gynaecology, Hôpital Lasalle, Lasalle, Quebec, Canada; the
Department of Social and Preventive Medicine, Faculty of Medicine, Université
Laval, Québec, Canada; the Department of Obstetrics & Gynaecology, Faculty of
Medicine, Centre Hospitalier Universitaire de Québec, Université Laval, Québec,
Canada, and the Department of Obstetrics & Gynaecology, Faculty of Medicine, St
Mary’s Hospital, McGill University, Montréal, Quebec, Canada; the Department of
Obstetrics & Gynaecology, Faculty of Medicine, Royal Victoria Hospital, McGill
University, Montréal, Quebec, Canada; the Department of Obstetrics & Gynaecol-
ogy, Faculty of Medicine, Royal Victoria Hospital, McGill University, Montréal,
Quebec, Canada; Department of Social and Preventive Medicine, Faculty of
Medicine, Université Laval, Québec, Canada; Department of Social and Preventive
Medicine, Faculty of Medicine, Université Laval, Québec, Canada; the Department
of Obstetrics & Gynaecology, Faculty of Medicine, Royal Victoria Hospital, McGill
University, Montréal, Quebec, Canada; the Department of Obstetrics & Gynaecol-
ogy, Faculty of Medicine, Jewish General Hospital, McGill University, Montréal,
Quebec, Canada; the Department of Obstetrics & Gynaecology, Faculty of Medicine,
Centre Hospitalier de l’Université de Montréal (CHUM), Université de Montréal,
Montréal, Quebec, Canada; the Department of Obstetrics & Gynaecology, Hôpital
Lasalle, Lasalle, Quebec, Canada; the Department of Obstetrics & Gynaecology,
Hôpital Lasalle, Lasalle, Quebec, Canada; the Department of Obstetrics &
Gynaecology, Hôpital Pierre-Boucher, Longueuil, Quebec, Canada; the Department
of Obstetrics & Gynaecology, Centre Hospitalier Cité de la Santé, Laval, Quebec,
Canada; Department of Obstetrics & Gynaecology, Hôpital Sacré-Coeur, Université
de Montréal, Montréal, Quebec, Canada; the Department of Obstetrics & Gynae-
cology, Hôpital Maisonneuve-Rosemont, Université de Montréal, Montréal, Quebec,
Canada; the Department of Obstetrics & Gynaecology, Faculty of Medicine, Centre
Hospitalier Universitaire de Québec, Université Laval, Québec, Canada, and the
Department of Social and Preventive Medicine, Faculty of Medicine, Université
Laval, Québec, Quebec, Canada; and the Department of Obstetrics & Gynaecology,
Faculty of Medicine, Hôpital Sainte-Justine, Université de Montréal, Montréal,
Quebec, Canada.
VOL. 116, NO. 1, JULY 2010
METHODS: A multicenter, case– control study was per-
formed on women with a single, prior, low-transverse
cesarean who experienced complete uterine rupture
during a trial of labor. For each case, three women who
underwent a trial of labor without uterine rupture after a
prior low-transverse cesarean delivery were selected as
control participants. Risk factors such as prior uterine
closure, suture material, diabetes, prior vaginal delivery,
labor induction, cervical ripening, birth weight, prosta-
glandin use, maternal age, gestational age, and interde-
livery interval were compared between groups. Condi-
tional logistic regression analyses were conducted.
RESULTS: Ninety-six cases of uterine rupture, including
28 with adverse neonatal outcome, and 288 control
participants were assessed. The rate of single-layer clo-
sure was 36% (35 of 96) in the case group and 20% (58 of
288) in the control group (P<.01). In multivariable anal-
ysis, single-layer closure (odds ratio [OR] 2.69; 95%
confidence interval [CI] 1.37–5.28) and birth weight
greater than 3,500 g (OR 2.03; 95% CI 1.21–3.38) were
Dr Emmanuel Bujold holds a Clinician Scientist Award from the Canadian
Institutes of Health Research and the Jeanne and Jean-Louis Lévesque Perinatal
Research Chair at Université Laval. This study was funded by the Canadian
Foundation for Women’s Health and the Jeanne and Jean-Louis Lévesque
Perinatal Research Chair at Université Laval and by the Fonds de Recherche des
Gynécologues-Obstétriciens de l’Hôpital Sainte-Justine (FOREGO).
Ovid DaSilva provided English editing assistance.
Corresponding author: Emmanuel Bujold, MD, MSc, FRCSC, Associate Pro-
fessor, Department of Obstetrics and Gynaecology, Faculty of Medicine, Univer-
sité Laval, 2705, boulevard Laurier, Québec, QC, Canada G1V 4G2; e-mail:
emmanuel.bujold@crchul.ulaval.ca.
Financial Disclosure
The authors did not report any potential conflicts of interest.
© 2010 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins.
ISSN: 0029-7844/10
OBSTETRICS & GYNECOLOGY 43
linked with increased rates of uterine rupture, whereas
prior vaginal birth was a protective factor (OR 0.47; 95%
CI 0.24 – 0.93). Single-layer closure was also related to
uterine rupture associated with adverse neonatal out-
come (OR 2.89; 95% CI 1.01– 8.27).
CONCLUSION: Prior single-layer closure carries more
than twice the risk of uterine rupture compared with
double-layer closure. Single-layer closure should be
avoided in women who could contemplate future vaginal
birth after cesarean delivery.
(Obstet Gynecol 2010;116:43–50)
LEVEL OF EVIDENCE: II
A lthough rare, the catastrophic complications of
uterine rupture make it one of the worst obstetric
emergencies. Moreover, its rapid evolution can ham-
per the best rescue efforts.1–3 Obstetricians are well
acquainted with unforeseeable labor complications
that require rapid intervention to avoid injury such as
cord prolapse or amniotic fluid embolism. Uterine
rupture is different because the decision to undergo a
trial of labor, and hence accepting exposure to this
known risk, usually results from a deliberate process
involving both the obstetrician and mother. It follows
that good estimates are required of the independent
effects of various risk factors. Some risk factors are not
modifiable, but some are potentially changeable such
as medical interventions related to the management
of labor and delivery.
The current study aims to examine the indepen-
dent contribution of several risk factors of uterine
rupture to help clinicians more accurately assess and
communicate the risks to mothers and they can make
informed clinical choices about risk factors. One of
the key potential risk factors we examine that is
modifiable is the technique of uterine closure in prior
cesarean delivery.4
More than one million cesarean deliveries are
performed every year in North America.5 Although
the technique of low-transverse cesarean delivery is
by far the most commonly performed, some technical
aspects of the procedure remain controversial. Clo-
sure of the uterus is one step in the operative proce-
dure for which there is no consensus.6 Although
two-layer closure was almost exclusively common
practice two decades ago, single-layer closure, associ-
ated with shorter operative time and fewer hemostatic
sutures, gained popularity in the 1990s and is in
common use today.7,8 However, in 2002, a retrospec-
tive study showed a fourfold increase in the risk of
uterine rupture with prior single-layer closure com-
pared with double-layer closure in women undergo-
ing a trial of labor after prior cesarean delivery.4,9
44 Bujold et al Previous Uterine Closure and Uterine Rupture
With more than 500,000 women with previous cesar-
ean delivery in North America and more than a
million around the world deciding about possibly
undergoing a trial of labor every year, it is imperative
to confirm whether a simple change in the cesarean
surgical technique, such as the number of layers for
uterus closure, could affect the incidence of uterine
rupture.5 Therefore, we have evaluated the indepen-
dent contributions of several risk factors for uterine
rupture, including the method of uterine closure
(single layer compared with double layer) at previous
cesarean delivery.
METHODS
We performed a multicenter, case– control study
comparing women who experienced complete uter-
ine rupture during a trial of labor after prior, single,
low-transverse cesarean delivery with a group of
control participants who underwent a trial of labor
after a prior, single, low-transverse cesarean delivery
without uterine rupture of any degree. Patients for the
case and control groups were selected from deliveries
at 10 centers in the Montreal metropolitan area
between 1992 and 2002. This period was predeter-
mined to avoid potential obstetric practice changes in
regard to type of uterine closure that could have
occurred after the publication of studies by Bujold et
al.4,9 All women with a documented low-transverse
cesarean delivery who underwent a trial of labor after
24 weeks of gestation were eligible irrespective of the
final mode of delivery. Exclusion criteria included all
the standard contraindications to attempting a vaginal
birth after cesarean (prior classical or J-shaped uterine
scar, previous extension above the low transverse
incision, prior transmural myomectomy, placenta
praevia) as well as the presence of more than one
prior cesarean delivery or birth weight under 500 g.
Uterine rupture was defined as complete opening of
the uterus, including the visceral serosa or vesical
wall, confirmed at the time of cesarean delivery or
during immediate postpartum laparotomy.10 Women
with incomplete uterine rupture or “uterine dehis-
cence,” defined as an opening of the uterus with intact
visceral serosa, were excluded and were not eligible
to be control participants. Patients for the case group
were identified by International Classification of Dis-
eases, 9th Revision code (665.0, 665.1, 674.1) in the
medical records combined with local perinatal data-
bases when available. All charts were reviewed by a
physician (M.G. or B.C.) for eligibility and data
collection. In case of uncertainty regarding eligibility,
the charts were reviewed independently by two other
physicians (E.B., R.J.G.) who were blinded to the
OBSTETRICS & GYNECOLOGY
prior type of closure to reach a consensus. Control
participants were selected from the list of patients with
prior cesarean delivery (International Classification of
Diseases, 9th Revision, Clinical Modification code
654.2) who delivered before the case of uterine rup-
ture in the same center. Medical charts were reviewed
in reverse chronological sequence starting from each
case of uterine rupture to find the first three control
participants that met the eligibility criteria. Control
participants were selected from among patients
treated before each index case to avoid bias from a
potential obstetric practice change that might occur in
the days or weeks after a case of uterine rupture.
Therefore, patients in the case and control groups
were matched by time period and hospital. After the
case and control groups had been established for each
center, all charts were reviewed to record maternal
age and maternal weight at the time of delivery,
gravidity, parity, gestational or pregestational diabe-
tes, indication and date of the prior cesarean delivery,
birth weight of the neonate from prior cesarean
delivery, date of the current delivery, induction of
labor, cervical maturation (use of prostaglandins or
intracervical Foley catheter), oxytocin use, birth
weight, Apgar score, umbilical cord pH, intrapartum
or neonatal death, and maternal morbidity, including
blood transfusion and peripartum hysterectomy.
When the prior operative report was available in the
maternal chart, we noted the method of uterine
closure and type of suture material. The interdelivery
interval was calculated in months from the prior
cesarean delivery to the current delivery.
Univariable and multivariable conditional regres-
sion analyses were performed for predetermined po-
tential risk factors of uterine rupture. Odds ratios
(ORs) and 95% confidence intervals (95% CIs) were
calculated. Proportions were compared using chi
square and Fisher’s exact tests when appropriate. All
statistical tests were two-sided; P⬍.05 was considered
significant. The primary independent variable was
type of closure (single layer, double layer, unknown).
Multivariable conditional logistic regression analysis
was conducted to calculate the adjusted OR for
adjusting for the effect of each variable associated
with uterine rupture with a P⬍.10 in univariable
analysis as well as the birth center (that was used for
matching). The following variables were considered:
maternal age, maternal weight at delivery (or the
closest available weight if taken after 28 weeks of
gestation), prior vaginal delivery, prior vaginal birth
after cesarean delivery, gestational or pregestational
diabetes, indication for the previous cesarean delivery
(arrest disorder or not), induction of labor, with
VOL. 116, NO. 1, JULY 2010 Bujold et al
favorable or unfavorable cervix (defined as Bishop’s
score less than 7 or cervical dilatation less than 2 cm
if Bishop’s score was not available), cervical matura-
tion (use of prostaglandins or intracervical Foley
catheter), oxytocin use, birth weight, birth weight at
the previous cesarean delivery, method of uterine
closure at the previous cesarean delivery, and type of
suture material.
A secondary analysis was performed, excluding
those patients from Sainte-Justine Hospital who were
included in the prior publication of Bujold et al.4 Finally,
we conducted a third analysis among women who
experienced uterine rupture associated with adverse
neonatal outcome and their respective control partici-
pants to determine whether any of the risk factors
independently increased the likelihood of adverse neo-
natal outcome because this event was of great concern
for the majority of women contemplating vaginal birth
after cesarean. For this analysis, adverse neonatal out-
come was defined as arterial cord blood pH less than 7.0
or 5-minute Apgar score less than 4 or intrapartum fetal
death. The stepwise logistic regression analysis included
all significant variables.
Sample size calculation during the planning
phase was based on our prior experience. We esti-
mated that approximately 25% of those in the control
group would have had single-layer closure, as would
at least 50% of the those in the case group. To obtain
a significant difference, with an ␣ error of .05 and a ␤
error of .10, with three control participants for each
case of uterine rupture and an OR of 2 or more, a
minimum of 55 cases of uterine rupture and 165
control participants would be necessary. We esti-
mated that 3–5% of deliveries during the study period
would have been patients undergoing a trial of labor
after a single prior cesarean delivery. Assuming a rate
of uterine rupture of 0.8%, we estimated that we
would need a cohort of at least 137,500 deliveries and
possibly up to 230,000 deliveries to find sufficient
patients for the case and control groups. Study partic-
ipation was, therefore, proposed to all departments of
obstetrics and gynecology of Montreal hospitals that
counted more than 2,000 deliveries per year. The
research protocol was approved by the department
head and the Ethics Review Board of each center.
RESULTS
Altogether, approximately 288,000 deliveries were
recorded in the 10 participating centers between 1992
and 2002. One center did not have a single case of
complete uterine rupture during the study period.
Ninety-six cases of uterine rupture and 288 control
participants were identified in the other nine centers.
Previous Uterine Closure and Uterine Rupture 45
Table 1. Unadjusted and Adjusted Odds Ratios for Maternal and Obstetric Characteristics Associated
With Uterine Rupture
Uterine No Uterine
Rupture Rupture Unadjusted Adjusted
Variable (nⴝ96) (nⴝ288) OR 95% CI OR* 95% CI

Maternal age at delivery (y)

Younger than 30 40 (42) 97 (34) 1.00
30–34 34 (35) 116 (40) 0.71 0.42–1.21
35 or older 22 (23) 75 (26) 0.71 0.39–1.30
‡
Maternal weight at delivery (kg)
Less than 60 6/70 (9) 23/247 (9) 1.00 0.28–2.31
60–79 44/70 (63) 139/247 (56) 1.30 0.49–3.46
80 or greater 20/70 (29) 85/247 (34) 1.01 0.35–2.90
Prior vaginal delivery
No 83 (86) 203 (70) 1.00 1.00
Yes 13 (14) 85 (30) 0.38† 0.20–0.71 0.47† 0.24–0.93
Gestational age at delivery (wk)
Three-category variable
Less than 37 3 (3) 13 (5) 0.75 0.21–2.73
37–40 68 (71) 227 (79) 1.00
41 or greater 25 (26) 48 (17) 1.74 1.00–3.02
Dichotomic variable
Less than 41 71 (74) 240 (83) 1.00
41 or greater 25 (26) 48 (17) 1.76† 1.01–3.05 1.06 0.56–2.00
Diabetes§
No 75 (86) 249 (86) 1.00
Yes 13 (14) 39 (14) 1.00 0.51–1.97
Interdelivery interval (mo)
Three-category variable
Less than 18 12 (12) 19 (7) 2.38† 1.08–5.24
18–23 19 (20) 32 (11) 2.15† 1.15–4.03
24 or greater 65 (68) 237 (82) 1.00
Dichotomic variable
Less than 24 31 (32) 51 (18) 2.23† 1.32–3.78 1.40 0.78–2.52
24 or greater 65 (68) 237 (82) 1.00 1.00
Induction of labor, cervical status, and maturation
Three-category variable
No induction (spontaneous labor) 56 (58) 197 (68) 1.00
Induction with favorable cervix 13 (14) 45 (16) 1.01 0.51–2.01
Induction with unfavorable cervix 27 (28) 46 (16) 2.10† 1.19–3.70
Using PG 8 (8) 13 (5) 2.27 0.87–5.92
Using Foley catheter 8 (8) 10 (3) 2.18 0.80–5.90
No Foley catheter or PG 11 (11) 23 (8) 1.96 0.90–4.25
Dichotomic variable
Induction with unfavorable cervix
No 69 (72) 242 (84) 1.00 1.00
Yes 27 (28) 46 (16) 2.09† 1.20–3.64 1.66 0.88–3.14
Oxytocin during labor
Four-category variable
Spontaneous labor, no augmentation 21 (25) 105 (39) 1.00
Induction of labor, no augmentation 3 (3) 8 (2.8) 1.89 0.46–7.69
Spontaneous labor followed by oxytocin 35 (37) 92 (32) 1.92† 1.04–3.55
augmentation of labor
Induction of labor with oxytocin 37 (39) 83 (29) 2.26† 1.22–4.18
Dichotomic variable
No oxytocin during labor 24 (25) 113 (39) 1.00 1.00
Oxytocin for augmentation or induction of 72 (75) 175 (61) 1.96† 1.16–3.31 1.50 0.84–2.66
labor
(continued)

46 Bujold et al Previous Uterine Closure and Uterine Rupture OBSTETRICS & GYNECOLOGY
Table 1. Unadjusted and Adjusted Odds Ratios for Maternal and Obstetric Characteristics Associated
With Uterine Rupture (continued)
Uterine No Uterine
Rupture Rupture Unadjusted Adjusted
Variable (nⴝ96) (nⴝ288) OR 95% CI OR* 95% CI

Birth weight (g)

Five-category variable
Less than 2,500 2 (2) 11 (4) 0.78 0.16–3.70
2,500–2,999 6 (6) 41 (14) 0.66 0.26–1.69
3,000–3,499 28 (29) 124 (43) 1.00
3,500–3,999 38 (40) 76 (26) 2.22† 1.27–3.90
4,000 or greater 22 (23) 36 (13) 2.84† 1.43–5.65
Dichotomic variable
Less than 3,500 36 (37) 176 (61) 1.00 1.00
3,500 or greater 60 (63) 112 (39) 2.60† 1.61–4.20 2.03† 1.21–3.38
Prior cesarean for arrest disorders
No 52 (54) 176 (61) 1.00
Yes 44 (46) 112 (39) 1.33 0.83–2.12
Birth weight at first cesarean (g)
Less than 2,500 7/88 (8) 34/272 (13) 0.63 0.26–1.57
2,500–2,999 11/88 (13) 51/272 (19) 0.66 0.31–1.43
3,000–3,499 30/88 (34) 93/272 (34) 1.00
3,500–3,999 23/88 (26) 57/272 (21) 1.25 0.67–2.35
4,000 or greater 17/88 (19) 37/272 (14) 1.42 0.70–2.87
Prior type of uterine closure
Single-layer closure 35 (36) 58 (20) 3.12† 1.69–5.74 2.69† 1.37–5.28
Double-layer closure 39 (41) 167 (58) 1.00 1.00
Unknown 22 (23) 63 (22) 1.46 0.79–2.71 1.38 0.72–2.62
Type of suture used
Chromic catgut 67 (70) 205 (71) 1.00
Vicryl or Polysorb 4 (4) 12 (4) 1.03 0.32–3.33
Unknown 25 (26) 71 (25) 1.09 0.62–1.93
OR, odds ratio; CI, confidence interval; PG, prostaglandins.
Data are n (%) or n/N (%) unless otherwise specified.
* ORs were adjusted by multivariable conditional logistic regression, including all significant variables in univariable analysis as well as
birth center (used for matching). Regression analysis was repeated with induction of labor and cervical maturation instead of
induction of labor with an unfavorable cervix, but there were no significant changes in the results (data not shown).
†
P⬍.05.
‡
Maternal weight at delivery or after 28 weeks of gestation, the closest to delivery, when not available.
§
Gestational or pregestational diabetes during the current pregnancy.

Among the 96 cases, rupture was diagnosed intra-
partum in 89 (93%) and postpartum by laparotomy in
seven (7%). Of these seven, three occurred in women
with a previous single-layer closure and four in women
with a double-layer closure. Six (6%) of the cases were
accompanied by intrapartum fetal death, 29 (30%) had
neonates with 5-minute Apgar scores less than 7, and 16
(17%) had neonates with 5-minute Apgar scores less
than 4. In the 76 with arterial blood gas measurements,
24 (32%) had pH less than 7.0. Twenty-eight (29%)
neonates had 5-minute Apgar scores less than 4, arterial
umbilical cord pH less than 7.0, or intrapartum fetal
death. Thirteen (14%) mothers received blood transfu-
sions. There were no intrapartum or postpartum mater-
nal deaths before discharge. Among the 288 control
participants, 210 (73%) succeeded in achieving a vaginal
delivery.
Maternal and obstetric characteristics for patients
in the case and control groups are reported in Table 1.
Among the 13 risk factors, univariable analysis iden-
tified seven with a statistically significant association
with uterine rupture. Uterine rupture was linked to
single-layer closure of the uterus with an OR of 2.51
(95% CI 1.45– 4.36). Other factors related to uterine
rupture included birth weight less than 3,500 g,
interdelivery interval less than 24 months, labor in-
duction with an unfavorable cervix, oxytocin use for
labor induction or augmentation, and gestational age
at delivery 41 weeks of gestation or greater. In
keeping with numerous other studies, we found that
prior vaginal birth was associated with diminished
risk of uterine rupture.11,12 The suture material used
for uterine closure was comparable for those in the
case and control groups.

VOL. 116, NO. 1, JULY 2010 Bujold et al Previous Uterine Closure and Uterine Rupture 47
Table 2. Unadjusted and Adjusted Odds Ratios for Maternal and Obstetric Characteristics Related to
Uterine Rupture Associated With Adverse Neonatal Outcomes (Arterial Umbilical Cord pH Less
Than 7.0, 5-Minute Apgar Score Less Than 4, or Both)
Uterine Rupture
With Adverse No Uterine
Neonatal Outcome Rupture Unadjusted Adjusted
Variable (nⴝ28) (nⴝ84) OR 95% CI OR* 95% CI

Prior vaginal delivery

No 24 (86) 57 (68) 1.00 1.00
Yes 4 (14) 27 (32) 0.35 0.11–1.11 0.34 0.10–1.15
Gestational age at delivery (wk)
Less than 41 22 (79) 69 (82) 1.00
41 or greater 6 (21) 15 (18) 1.24 0.44–3.50
Interdelivery interval (mo)
Less than 24 6 (21) 13 (15) 1.51 0.50–4.55
24 or greater 22 (79) 71 (85) 1.00
Induction of labor with unfavorable
cervix (dilatation less than 2 cm)
No 22 (79) 72 (86) 1.00
Yes 6 (21) 12 (14) 1.69 0.55–5.21
Oxytocin during labor
No oxytocin during labor 8 (29) 35 (42) 1.00 1.00
Oxytocin for augmentation or 20 (71) 49 (58) 1.88 0.71–4.99 2.11 0.76–5.83
induction of labor
Birth weight (g)
Less than 3,500 13 (46) 50 (60) 1.00
3,500 or greater 15 (54) 34 (40) 1.79 0.73–4.40
Prior type of uterine closure
Single-layer closure 10 (36) 15 (18) 2.69 0.99–7.32 2.89† 1.01–8.27
Double-layer closure 12 (43) 50 (60) 1.00 1.00
Unknown 6 (21) 19 (23) 1.27 0.38–4.21 1.77 0.50–6.21
OR, odds ratio; CI, confidence interval.
Data are n (%) unless otherwise specified.
* ORs were adjusted by multivariable conditional logistic regression using stepwise selection of variables with all of them included at
the first step. The final model contained the following variables: prior vaginal delivery, oxytocin use, and prior uterine closure
(categorical variable).
†
P⬍.05.

Mutivariate conditional regression analysis was
performed according to the seven factors identified
through univariable analysis as well as the birth center
(Table 1). After adjustment, single-layer closure and
birth weight greater than 3,500 g remained associated
with uterine rupture, whereas prior vaginal birth
stayed as a protective factor. Multivariable regression
analysis was repeated after exclusion of the 23 cases of
uterine rupture reported previously from Sainte-Jus-
tine Hospital and their 69 appropriate control partic-
ipants.4 In that secondary analysis, single-layer clo-
sure of the uterus remained associated with uterine
rupture (OR 2.36; 95% CI 1.10 –5.07) (data not
shown). We evaluated the effect of center type accord-
ing to the presence or the absence of an obstetrics–
gynecology residency program. We found that sur-
geons from centers with obstetrics– gynecology
residency program were more likely to use single-
layer closure (30% [48 of 159] of the control partici-
pants) and the association between single-layer clo-
sure and uterine rupture was very significant (OR
3.59; 95% CI 1.62–7.99). On the other hand, centers
without a obstetrics– gynecology residency program
were less likely to use single-layer closure (8% [10 of
129] of the control participants) and the association
was not significant (OR 0.77; 95% CI 0.17–3.57).
Finally, the analysis was repeated in only the 28
cases associated with adverse neonatal outcomes (pH
less than 7.0, 5-minute Apgar scores less than 4,
intrapartum fetal death) and their respective control
participants (Table 2). With stepwise logistic regres-
sion analysis, we observed that single-layer closure
remained the only significant factor related to uterine
rupture with adverse neonatal outcome.
DISCUSSION
The study has demonstrated that among a number of
purported risk factors for uterine rupture, only three

48 Bujold et al Previous Uterine Closure and Uterine Rupture OBSTETRICS & GYNECOLOGY
were independently associated and just one was poten-
tially modifiable. A single-layer closure in the previous
cesarean delivery carried over a twofold increase in risk
followed closely by a birth weight in the current preg-
nancy greater than 3,500 g, whereas a prior vaginal
delivery decreased the risk by nearly half. Of these three
factors, only a prior single-layer closure affected the risk
of uterine rupture associated with an adverse neonatal
outcome. We further established that it was presence of
the single-layer closure itself, not the suture material, that
was associated with the rupture.
Our study has some limitations. We could not
examine every possible variable such as clinical com-
plications or level of the surgeon’s expertise at the
time of the first cesarean delivery or aspects of
management during the labor complicated by uterine
rupture such as the amount and duration of oxytocin
use. Nevertheless, among the factors examined, three
stood apart from all the others. On the other hand, the
strengths of our study are numerous and include its
multicenter design, the large number of cases, an
unequivocal definition of uterine rupture, and thor-
ough review of the actual charts resulting in high
levels of data accuracy. Control participants were
selected to account for local practice patterns and
changes over the years. The study was performed in a
period when the method of previous uterine closure
was not believed to be related to uterine rupture and
thus did not affect decisions about a trial of labor.
The literature provides inconsistent conclusions
regarding the role of the uterine closure technique
and subsequent uterine rupture. Most studies have
compared rates of uterine rupture in two groups
defined by previous closure technique. The sample
size necessary to study a relatively rare complication
such as uterine rupture with an expected incidence of
approximately 1% and anticipating a difference be-
tween groups as large as 50% would be approximately
10,000 women undergoing a trial of labor. To date,
most cohort studies examining the role of single-
compared with double-layer closure involved fewer
than 1,000 patients. Guise et al recently reviewed the
literature regarding closure type and uterine rupture
risk and concluded that there were serious method-
ological deficiencies precluding a clear answer on the
role of uterine closure.4,13–17 The only randomized
trial reported by Chapman et al involved 906 women
who were randomized to single- compared with dou-
ble-layer uterus closure. Only 164 of them had sub-
sequent pregnancies and no uterine rupture was re-
ported. In short, these previous studies were greatly
underpowered to examine rates of uterine rupture
according to uterine closure technique. The current
VOL. 116, NO. 1, JULY 2010 Bujold et al
study applies a different technique, namely condi-
tional multiple regression, applied to a large retro-
spective sample of women with uterine rupture and
matched control participants to measure the indepen-
dent effect of interrelated risk factors. This study
examined 96 uterine ruptures, which is more than the
62 cases in all the other studies combined.14 –18
We do not have a precise understanding of the
pathophysiology involved in the association of the sin-
gle-layer closure technique and subsequent uterine rup-
ture; however, some observations and speculations sup-
port our findings. It is possible that a single, continuous
suture technique does not approximate the tissues pre-
cisely together because decidua can be included in the
scar. Poidevin19 demonstrated, in several animal mod-
els, that decidua inclusion in the scar or eversion or both
of the hysterotomy edge are important factors related to
subsequent uterine scar defect. Scar thickness measure-
ments, which were not available to us, would have been
an interesting way of further examining this mecha-
nism.20 –22 Uterine rupture is thought to result from a
biomechanical process, in which there is an imbalance
between the tensile strength of the scar that maintains its
integrity and the forces causing disruption.23 None of
these forces is directly measurable today. However, it is
plausible that a thinner or weaker scar would be more
prone to rupture. Further studies are definitely required
to explain the association between single-layer closure
and uterine rupture.
The magnitude of the association we observed
between single-layer closure and uterine rupture
makes this information crucial to all women and
healthcare providers who are making choices about
the mode of delivery and all surgeons performing
cesarean deliveries. Moreover, it has been shown that
the double-layer technique is as easy to perform as the
single-layer technique in most cases.7 Given the infor-
mation available today, we agree with Jelsema et al24
that single-layer closure be reserved for women un-
dergoing tubal ligation or those who require very
expedient closure of the uterus.
REFERENCES
1. Bujold E, Gauthier RJ. Neonatal morbidity associated with
uterine rupture: what are the risk factors? Am J Obstet Gynecol
2002;186:311– 4.
2. Guise JM, McDonagh MS, Osterweil P, Nygren P, Chan BK,
Helfand M. Systematic review of the incidence and conse-
quences of uterine rupture in women with previous caesarean
section. BMJ 2004;329:19 –25.
3. Sachs BP. A 38-year-old woman with fetal loss and hysterec-
tomy. JAMA 2005;294:833– 40.
4. Bujold E, Bujold C, Hamilton EF, Harel F, Gauthier RJ. The
impact of a single-layer or double-layer closure on uterine
rupture. Am J Obstet Gynecol 2002;186:1326 –30.
Previous Uterine Closure and Uterine Rupture 49
 5. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F,
Kirmeyer S, et al. Births: final data for 2005. Natl Vital Stat
Rep 2007;56:1–103.
6. Guise JM, Hashima J, Osterweil P. Evidence-based vaginal
birth after Caesarean section. Best Pract Res Clin Obstet
Gynaecol 2005;19:117–30.
7. Hauth JC, Owen J, Davis RO. Transverse uterine incision
closure: one versus two layers. Am J Obstet Gynecol 1992;
167:1108 –11.
8. Dandolu V, Raj J, Harmanli O, Lorico A, Chatwani AJ.
Resident education regarding technical aspects of cesarean
section. J Reprod Med 2006;51:49 –54.
9. Bujold E, Mehta SH, Bujold C, Gauthier RJ. Interdelivery
interval and uterine rupture. Am J Obstet Gynecol 2002;187:
1199 –202.
10. Pridjian G. Labor after prior cesarean section. Clin Obstet
Gynecol 1992;35:445–56.
11. Mercer BM, Gilbert S, Landon MB, Spong CY, Leveno KJ,
Rouse DJ, et al. Labor outcomes with increasing number of
prior vaginal births after cesarean delivery. Obstet Gynecol
2008;111:285–91.
12. Hendler I, Bujold E. Effect of prior vaginal delivery or prior
vaginal birth after cesarean delivery on obstetric outcomes in
women undergoing trial of labor. Obstet Gynecol 2004;104:
273–7.
13. Chapman SJ, Owen J, Hauth JC. One- versus two-layer closure
of a low transverse cesarean: the next pregnancy. Obstet
Gynecol 1997;89:16 – 8.
14. Pruett KM, Kirshon B, Cotton DB, Poindexter AN 3rd. Is
vaginal birth after two or more cesarean sections safe? Obstet
Gynecol 1988;72:163–5.
50 Bujold et al Previous Uterine Closure and Uterine Rupture
15. Tucker JM, Hauth JC, Hodgkins P, Owen J, Winkler CL. Trial
of labor after a one- or two-layer closure of a low transverse
uterine incision. Am J Obstet Gynecol 1993;168:545– 6.
16. Durnwald C, Mercer B. Uterine rupture, perioperative and
perinatal morbidity after single-layer and double-layer clo-
sure at cesarean delivery. Am J Obstet Gynecol 2003;189:
925–9.
17. Shipp TD, Zelop C, Cohen A, Repke JT, Lieberman E.
Post-cesarean delivery fever and uterine rupture in a subse-
quent trial of labor. Obstet Gynecol 2003;101:136 –9.
18. Gyamfi C, Juhasz G, Gyamfi P, Blumenfeld Y, Stone JL.
Single- versus double-layer uterine incision closure and uterine
rupture. J Matern Fetal Neonatal Med 2006;19:639 – 43.
19. Poidevin LOS. Cesarean section scars. Springfield (IL):
Charles C. Thomas Publisher; 1965.
20. Rozenberg P, Goffinet F, Phillippe HJ, Nisand I. Ultrasono-
graphic measurement of lower uterine segment to assess risk of
defects of scarred uterus. Lancet 1996;347:281– 4.
21. Bujold E, Jastrow N, Simoneau J, Brunet S, Gauthier RJ.
Prediction of complete uterine rupture by sonographic evalu-
ation of the lower uterine segment. Am J Obstet Gynecol
2009;201:320.e1– 6.
22. Jastrow N, Chaillet N, Roberge S, Morency AM, Lacasse Y,
Bujold E. Sonographic lower uterine segment thickness and
risk of uterine scar defect: a systematic review. J Obstet
Gynaecol Can 2010;32:321–7.
23. Hamilton EF, Bujold E, McNamara H, Gauthier R, Platt RW.
Dystocia among women with symptomatic uterine rupture.
Am J Obstet Gynecol 2001;184:620 – 4.
24. Jelsema RD, Wittingen JA, Vander Kolk KJ. Continuous,
nonlocking, single-layer repair of the low transverse uterine
incision. J Reprod Med 1993;38:393– 6.
OBSTETRICS & GYNECOLOGY