Uterine rupture, perioperative and perinatal morbidity after

single-layer and double-layer closure at cesarean delivery

Celeste Durnwald, MD, and Brian Mercer, MD
Cleveland, Ohio

OBJECTIVE: This study was undertaken to evaluate the risks and benefits of single-layer uterine closure at
cesarean delivery on the index and subsequent pregnancy.
STUDY DESIGN: A retrospective study of women delivered of their first live-born infants by primary low
transverse cesarean delivery (1989-2001) and their subsequent pregnancy at our institution was performed.
RESULTS: Of 768 women studied, 267 had single-layer and 501 had double-layer uterine closures in the index
pregnancy. Single-layer closure was associated with slightly decreased blood loss (646 vs 690 mL, P < .01),
operative time (46 vs 52 minutes, P < .001), endometritis (13.5% vs 25.5%, P < .001), and postoperative stay
(3.5 vs 4.1 days, P < .001). In the second pregnancy, prior single-layer closure was not associated with uterine
rupture after a trial of labor (0% vs 1.2%, P = .30), or other maternal or infant morbidities. Prior single-layer
closure was associated with increased uterine windows (3.5% vs 0.7%, P = .046) at subsequent cesarean
delivery.
CONCLUSION: Single-layer uterine closure is associated with decreased infectious morbidity in the index
surgery, but not uterine rupture or other adverse outcomes in the subsequent gestation. (Am J Obstet
Gynecol 2003;189:925-9.)

Key words: Uterine rupture, cesarean delivery, vaginal birth after cesarean section

In 1926, Kerr1 introduced the low transverse uterine
incision for cesarean delivery. This incision had the
advantage of being on a more passive part of the uterus
allowing better wound healing and less stretch of the scar
during subsequent pregnancies. Traditionally, the uterine
incision has been closed in two layers, but single-layer
closure has been recognized as an alternative if satisfac-
tory approximation of the lower uterine segment can be
achieved.2 Radiographic study has revealed fewer lower
uterine segment abnormalities with single-layer closure,
suggesting that this closure might reduce local ischemia,
hematoma formation, and infection.3
Recent studies have shown slightly shorter operative
times, less need for hemostatic sutures, and similar
infectious complications with single-layer closure.4-7 Few
studies have evaluated the impact of single-layer closure
on subsequent pregnancies.8-10 Although two studies
found no increased risk of subsequent uterine rupture,8,9
a third analysis has suggested a 4- to 6-fold increase in
uterine rupture after single layer closure.10 Prior studies
are confounded by inclusion of women with prior vaginal
From the Department of Obstetrics and Gynecology, MetroHealth
Medical Center, Case Western Reserve University School of Medicine.
Supported in part by a grant from the National Center for Research,
MO1-RR-00080.
Presented at the Twenty-Third Annual Meeting of the Society for
Maternal-Fetal Medicine, San Francisco, Calif, February 3-8, 2003.
Reprints not available from the authors.
Ó 2003, Mosby, Inc. All rights reserved.
0002-9378/2003 $30.00 + 0
doi:10.1067/S0002-9378(03)01056-1
deliveries and/or intervening successful trials of labor.
These factors could potentially alter the likelihood of
successful trial of labor, perioperative complications, and
uterine rupture.
Because of the conflicting results and limitations of
prior studies, we undertook this study to evaluate the risks
and benefits of single-layer closure of primary low trans-
verse cesarean delivery on the index and subsequent
pregnancy in women without a prior vaginal delivery.
Methods
With institutional review board approval, we performed
a retrospective study of all nulliparous women delivered of
their first and second live-born singleton infants at
MetroHealth Medical Center between 1989 and 2001,
with the first delivered by low transverse cesarean delivery.
We excluded women with an upper uterine segment
extension or T-incision, prior uterine myometrial surgery,
and those delivered of either pregnancy before 24 weeks.
Individual chart and operative report review of all eligible
women was performed for maternal clinical characteris-
tics, labor course, operative findings, and postpartum
outcomes for both pregnancies. Neonatal outcomes,
including premature birth, birth weight, Apgar scores,
and neonatal intensive care unit (NICU) admission, were
recorded.
For the subsequent pregnancy, antenatal hemorrhage,
including placenta previa and accreta, was recorded.
Labor characteristics, including labor induction, duration

925
926 Durnwald and Mercer October 2003
Am J Obstet Gynecol

Table I. Maternal clinical characteristics and perioperative complications in the first (index) pregnancy

Single layer (n = 267) Double layer (n = 501) P value

Maternal age (y)* 23.2 (7.3) 22.7 (5.8) .31

Race (% white) 137 (51.3) 277 (55.3) .32
Government insurance (%) 211 (79.0) 375 (74.9) .21
Diabetes (%) 17 (6.4) 25 (5.0) .41
Operative time (min)* 46.4 (15.7) 51.9 (16.2) < .001
Delivery to closure time* 36.7 (13.0) 42.0 (13.9) < .0001
Additional sutures (%) 91 (34.1%) 121 (24.2%) .004
Number addedy 1 (1–3) 1 (1-4) .63
Estimated blood loss (mL)* 646 (190) 690 (197) .003
Endometritis (%) 36 (13.5) 128 (25.5) < .0001
Postoperative stay (d)* 3.5 (1.8) 4.1 (2.0) < .0001
Postpartum hemorrhage (%) 8 (3.0) 10 (2.0) .45
Blood transfusion (%) 2 (0.8) 4 (0.8) .99

*Mean (SD).
ymedian (range).

of labor and membrane rupture, oxytocin use, anesthesia,
meconium, internal uterine monitors, amnionitis, non-
reassuring fetal heart rate, and mode of delivery, were
obtained. Evaluated postpartum outcomes included post-
partum hemorrhage, blood transfusion, and endomet-
ritis. Operative reports were reviewed for indication,
operative times, anesthesia type, estimated blood loss,
bladder/bowel trauma, adhesions, uterine window, and
uterine rupture. Evaluated neonatal outcomes included
gestational age, birth weight, Apgar scores, NICU admis-
sion, positive pressure ventilation, intraventricular hem-
orrhage, seizures, birth asphyxia, and neonatal death
(within 28 days of life).
Single-layer closure was defined as one continuous
running suture applied to the lower uterine segment with
additional interrupted hemostatic sutures placed as
needed. Double-layer closure was defined as a second
continuous imbricating layer applied over the first layer.
Uterine rupture was defined as a full-thickness defect
through myometrium and peritoneum. A uterine window
was defined as an asymptomatic uterine defect with intact
overlying peritoneum.
Statistical analysis was conducted using Statview (SAS
Institute Inc, Cary, NC). Fisher exact and Student t tests
were used where appropriate. The log rank test was
performed for nonnormally distributed temporal data.
Logistic regression was performed to evaluate endome-
tritis after the first delivery, controlling for the potentially
confounding influence of perioperative antibiotics. A P
value less than .05 was considered significant.
Results
Over 12 years, 768 women met inclusion criteria and
1536 maternal and infant charts were reviewed. In the
index pregnancy, there were 267 single-layer and 501
double-layer closures. Of uterine closures, 99.2% were
accomplished with polyglactin 910 suture (0-Vicryl,
Ethicon, Somerville, NJ). In the second pregnancy,
68.2% and 67.9% of women with single- and double-layer
closures underwent a trial of labor, respectively. Of these,
68.1% and 64.7% of women with prior single- and double-
layer closures, respectively, had a successful vaginal birth
after cesarean section (VBAC).
Maternal clinical characteristics in the index pregnancy
were similar for both groups (Table I). Women with
single-layer closure had slightly shorter operative times
and a lower estimated blood loss (P # .01 for each).
Double-layer closure was associated with a 1.9-fold higher
incidence of endometritis and longer postoperative stay
(P < .0001 for each). Perioperative antibiotics were not
significantly less common with double-layer closure
(81.6% vs 87.3%, P = .052); nor were women with endo-
metritis after double-layer closure less likely to receive
perioperative antibiotics (85.2% vs 80.6%, P = .61). The
increase in endometritis with double-layer closure persis-
ted after controlling for perioperative antibiotic exposure
by logistic regression (P < .0001). Postpartum hemor-
rhage and blood transfusions were similar between
groups, as were infant outcomes, including premature
birth (12.4% vs 11.4%, P = .72), 5-minute Apgar score less
than 7 (4.1% vs 2.4%, P = .19), and NICU admission
(12.7% vs 16.2%, P = .24) in single- and double-layer
closure groups, respectively. Infants born to women with
single-layer closure were slightly smaller (3230 ± 713g vs
3337 ± 702 g, P = .047).
For all subsequent pregnancies, including trials of labor
and elective repeat cesarean deliveries, prior single- and
double-layer closures had similar rates of antepartum
bleeding (0.8% vs 1.6%, P = .51) and placenta previa
(0.4% vs 1.0%, P = .67). The interdelivery interval was 4.5
months longer, on average, for those with double-layer
closure in the first pregnancy (2.97 ± 1.8 years vs
2.60 ± 1.42 years, P = .004).
There were no uterine ruptures with prior single-layer
closure and four uterine ruptures with prior double-layer
closure (0% vs 0.8%, respectively, P = .30). Neither group
Volume 189, Number 4 Durnwald and Mercer 927
Am J Obstet Gynecol

Table II. Labor characteristics and pregnancy outcomes of women undergoing a trial of labor in the
second pregnancy

Single layer (n = 182) Double layer (n = 340) P value

Gestational age (wks)* 39.1 (2.2) 39.4 (2.1) .21

Induction (%) 38 (20.9) 55 (16.2) .19
Oxytocin use (%) 111 (61.0) 186 (54.7) .19
Labor duration (h)* 9.4 (6.5) 9.9 (7.0) .44
Rupture of membranes duration (h)* 14.1 (56.9) 12.3 (30.9) .91
Protracted labor (%) 14 (7.7) 26 (7.6) .99
Epidural (%) 159 (87.4) 298 (87.6) .99
Internal uterine monitor (%) 131 (72.0) 241 (70.9) .84
Meconium (%) 31 (17.0) 59 (17.4) .99
Amnionitis (%) 7 (3.8) 24 (7.1) .17
Uterine rupture (%) 0 4 (1.2) .30
Cesarean section for fetal distress (%) 15 (8.2) 38 (11.2) .36
Cesarean section for labor arrest (%) 36 (19.8) 72 (21.2) .74
Endometritis (%) 5 (2.7) 19 (5.6) .19
Postpartum hemorrhage (%) 0 5 (1.5) .17
Blood transfusion (%) 0 2 (0.6) .54

*Mean (SD).

had placenta accreta nor maternal death. Infants born
after a prior single- or double-layer closure, respectively,
were similar in gestational age (39.1 ± 2.2 weeks vs
39.3 ± 2.0 weeks, P = .16), birth weight (3306 ± 609 g vs
3347 ± 547 g, P = .33), 5-minute Apgar score less than 7
(1.1% vs 1.8%, P = .56), NICU admission (5.2% vs 8.2%,
P = .14), positive pressure ventilation (3.0% vs 3.2%, P =
.99), intraventricular hemorrhage (0% vs 0.4%, P = .55),
seizures (0% vs 0.2%, P = .99), and birth asphyxia (0% vs
0.2%, P = .99). There were no neonatal deaths.
Of women undergoing a trial of labor in the second
pregnancy, the rates of labor induction, oxytocin use,
protracted labor, regional anesthesia, internal uterine
monitors, meconium, and amnionitis were similar be-
tween the groups (Table II). All four uterine ruptures oc-
curred during a trial of labor after a double-layer closure.
In each case, cesarean delivery was performed for a non-
reassuring fetal heart rate. Single-layer closure was not
associated with increased cesarean delivery for fetal dis-
tress or labor arrest. No cesarean hysterectomies were re-
quired. Postpartum hemorrhage, endometritis, and infant
outcomes were similar between the groups (Table III).
Those undergoing a cesarean delivery in the second
pregnancy had similar incidences of adhesions, incision
extensions, bowel/bladder trauma, as well as operative
times and estimated blood loss (Table IV). The quality of
adhesions was not recorded. Uterine windows were more
commonly diagnosed with prior single-layer closure
(3.5% vs 0.7%, P = .046). Data regarding uterine windows
were not available for the entire cohort because only
30.9% of women with a successful VBAC had a docu-
mented examination of the previous uterine scar. No
uterine ruptures or windows were discovered in those
examined. The incidence of uterine windows for the 530
evaluated women was 2.8% (5/181) after single-layer
closure and 0.6% (2/349) after double-layer closure,
P = .049. Two windows in the single-layer closure group
were complicated by a ‘‘J’’ incision extension described
to be within the lower uterine segment in the index
pregnancy. One of these, and one of the two women with
a uterine window after double-layer closure had post-
operative endometritis. However, all women with a uterine
window and three of four women with uterine rupture
had received perioperative antibiotics at the first delivery.
Because perinatal outcomes may be potentially worse
after a failed trial of labor, we evaluated this group
separately. At cesarean delivery after failed trial of labor,
there were no differences in adhesions (46.6% vs 36.7%,
P = .25), incision extension (17.2% vs 17.5%, P = .99),
estimated blood loss (686 ± 190 mL vs 750 ± 710 mL,
P = .50) as well as operative times, including total (56.9 vs
59.4 minutes, P = .55) and skin-to-infant delivery time
(14.1 vs 12.9 minutes, P = .61) with prior single- or
double-layer closure, respectively. Although the incidence
of uterine rupture (0% vs 3.3%, P = .31) was similar,
a statistically insignificant trend to increased uterine
windows with prior single-layer closure persisted (5.2% vs
0.8%, P = .10). Infants in this subgroup were similar in
birth weight (3385 ± 611 g vs 3432 ± 512 g, P = .59),
5-minute Apgar score less than 7 (0% vs 4.2%, P = .17),
NICU admission (5.2% vs 9.2%, P = .55), positive pres-
sure ventilation (1.7% vs 5.0%, P = .43), and intraven-
tricular hemorrhage (0% vs 0.8%, P = .99) after a prior
single- or double-layer closure, respectively. There were
no cases of birth asphyxia or seizures.
Comment
Our findings are consistent with previous studies
showing a slight decrease in operative times and estimated
blood loss with single-layer closure.4,5,7 Although statisti-
cally significant, these findings may not be clinically
relevant and do not in isolation justify single-layer closure.
928 Durnwald and Mercer October 2003
Am J Obstet Gynecol

Table III. Infant outcomes after a trial of labor in the second pregnancy

Single layer (n = 182) Double layer (n = 340) P value

Birth weight (g)* 3241 (565) 3294 (527) .29

Male (%) 99 (54.4) 170 (50.0) .36
NICU admission (%) 10 (5.5) 29 (8.5) .23
1-min Apgar score <5 (%) 10 (5.5) 22 (6.5) .71
5-min Apgar score <7(%) 3 (1.6) 9 (2.6) .56
Pressure ventilation (%) 4 (2.2) 14 (4.1) .32
Intraventricular hemorrhage (%) 0 2 (0.6) .54
Seizures (%) 0 1 (0.3) .99
Birth asphyxia (%) 0 1 (0.3) .99

*Mean (SD).

Table IV. Operative findings for women undergoing cesarean delivery in the second pregnancy with or without a
trial of labor

Single layer (n = 143) Double layer (n = 281) P value

Adhesions (%) 53 (37.1) 92 (32.7) .39

Trauma (%) 0 1 (0.4) .99
Extension (%) 17 (11.9) 26 (9.3) .40
Operative time (min)* 56.8 (18.9) 56.3 (25.2) .84
Incision to delivery (min)* 15.3 (8.5) 13.1 (12.2) .06
Delivery to closure (min)* 41.6 (14.4) 43.2 (20.7) .39
Estimated blood loss (mL)* 660 (179) 683 (490) .59
Uterine rupture (%) 0 4 (1.4) .31
Uterine window (%) 5 (3.5) 2 (0.7) .046

*Mean (SD).

However, in contrast to previous studies that found similar
rates of postoperative infectious morbidity regardless of
the type of uterine closure, we found a significant reduc-
tion in endometritis and shorter postoperative stay with
single-layer closure. This effect persisted after control-
ling for perioperative antibiotic exposure.
Our findings are consistent with those of Chapman
et al8 and Tucker et al9 who found no increased risk of
uterine rupture with prior single layer closure during
a trial of labor. In our study, all four uterine ruptures with
prior double-layer closure occurred during a trial of labor
but were not associated with prostaglandin use. Only one
of these women had endometritis that may have affected
incisional healing after the first cesarean delivery. Our
findings are in contrast to those of Bujold et al10 in which
a 6-fold increase in uterine rupture was found with a trial
of labor after prior single-layer closure (4-fold after
multivariate analysis). The major differences between
our practice and that of the institution of Bujold et al10
appear to be the suture type and method used for uterine
closure. In our institution, polyglactin 910 (0-Vicryl,
Ethicon) was used almost uniformly, whereas chromic
catgut was used predominantly in the study by Bujold et al.
Polyglactin 910, a synthetic absorbable suture, is degraded
by hydrolysis and maintains all of its tensile strength for at
least 7 to 10 days. Chromic catgut is degraded by
proteolytic enzymes and loses half of its tensile strength
within 7 to 10 days. This degradation occurs more rapidly
with infection such as endometritis. It is plausible that
using a suture that maintains its tensile strength longer
and is less susceptible to degradation in infected tissue
may account for the lower rate of uterine rupture with
single-layer closure in our study. Although Bujold et al
reported applying a continuous locked stitch, the general
practice at our institution is to perform a continuous
unlocked closure. It is plausible that a running locked
closure leads to vascular occlusion and poorer wound
healing.11 This, combined with the lower incidence of
postoperative infection, could also explain the discrep-
ancy between the two studies. Single-layer uterine closure
may also contribute to less vascular occlusion and tissue
ischemia than double-layer closure in which there is
theoretically more tissue strangulation. Furthermore, in
our study, double-layer closures were more likely to need
additional hemostatic sutures that could also lead to more
tissue strangulation. Bujold et al10 found an association of
uterine rupture with an interdelivery interval of 2 years or
less. In our study, the mean interdelivery interval for the
uterine ruptures was 2.6 years, with only one case
occurring with an interdelivery interval less than 2 years.
It is likely that interval between deliveries is a less
important factor than perioperative incisional healing.
Although our findings regarding perioperative com-
plications in the index pregnancy and uterine rupture in
the subsequent pregnancy with single-layer closure are
reassuring, the apparent increased risk of uterine
Volume 189, Number 4
Am J Obstet Gynecol
windows requires consideration. All uterine windows were
found at cesarean delivery and the majority was found
after a failed trial of labor. It is unknown if these windows
would progress to symptomatic uterine rupture with ad-
vancing gestation, prolonged labor, or in a subsequent
pregnancy. Only 31% of women with a successful VBAC
had palpation of the uterine scar documented. Although
no uterine windows were identified in these women, it is
possible that we may not have detected all uterine windows.
Our findings are consistent with previous reports
showing no increase in adverse infant outcomes with
single-layer closure, including those born after a failed
trial of labor.8 In addition, we found no increase in other
infant morbidities, including intraventricular hemor-
rhage, seizures, and birth asphyxia.
This retrospective cohort study depends on availability
and completeness of charts used in data collection and
cannot control for other less evident changes in surgical
practice that may influence perioperative and subsequent
pregnancy outcomes. Factors influencing decisions for
uterine closure at the time of cesarean delivery such as
bleeding, myometrial thickness, and surgeon preference
cannot be accounted for. Double-layer closure was used
predominantly until the mid 1990s after which single-
layer closure became the predominant practice. Although
our study has adequate power (81%) to evaluate the 6-fold
increase in uterine rupture with single-layer closure
suggested by Bujold et al,10 our power to detect more
modest differences is low.
To adequately evaluate a 2-fold increased risk of uterine
rupture after single-layer closure from a 1.2% with double-
layer closure with 80% power (a = .05 two-sided), 3848
women would need to be studied.
A running single-layer closure of the uterine incision
with polyglactin 910 suture at cesarean delivery appears to
offer benefits, including reduced perioperative infection,
Durnwald and Mercer 929
in addition to a slight decrease in operative time and
estimated blood loss. Although we have not found an
increased risk of uterine rupture with a trial of labor after
a single-layer closure, our finding regarding increased
uterine windows with single-layer closure is notable. We
recommend using a running unlocked suture of poly-
glactin 910 for single-layer closure at low transverse
cesarean delivery. For women planning a trial of labor in
a future pregnancy, a double-layer closure may be
appropriate.
REFERENCES
1. Kerr JMM. The technique of cesarean section, with special refer-
ence to the lower uterine segment. Am J Obstet Gynecol 1926;12:
729-34.
2. Pritchard JA, MacDonald PC. Cesarean section and cesarean hyster-
ectomy. In: Pritchard JA, MacDonald PC, editors. Williams’ obstetrics.
15th ed. New York: Appleton-Century-Crofts; 1976. p. 903-23.
3. Lal K, Tsomo P. Comparative study of single layer and conventional
closure of uterine incision in cesarean section. Int J Gynecol Obstet
1988;27:349-52.
4. Hauth JC, Owen J, Davis R. Transverse uterine incision closure: one
versus two layers. Am J Obstet Gynecol 1992;167:1108-11.
5. 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.
6. Malkamy H, Ardekany MS. Is it safe to repair low transverse cesarean
section in one layer? Int J Gynecol Obstet 1993;41:93-5.
7. Ohel G, Younis S, Lang N, Levit A. Double-layer closure of uterine
incision with visceral and parietal peritoneal closure: are they obli-
gatory steps of routine cesarean sections? J Matern Fetal Med 1996;
5:366-9.
8. 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.
9. 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.
10. 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.
11. Cruikshank DP. The impact of single- or double-layer closure on
uterine rupture. Am J Obstet Gynecol 2003;188:295-6.